University of Southern California Dissertations and Theses

Heart disease among middle -aged and elderly persons in the United States: Trends and a multistate model

(USC Thesis Other)

Heart disease among middle -aged and elderly persons in the United States: Trends and a multistate model

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Heart Disease among Middle-Aged and Elderly Persons in the United States: Trends and a Multistate Model by Hiroshi Ueda A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Gerontology and Public Policy) August 2000 Copyright 2000 Hiroshi Ueda Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3018138 Copyright 2000 by Ueda, Hiroshi All rights reserved. _ _ ® UMI UMI Microform 3018138 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE S C H O O L UNIVERSITY PA R K LOS AN0ELE5, CALIFORNI A 9000 7 This dissertation, written by H iro s h i Ueda under the direction of fcia— Dissertation Committee, and approved fry all its m e m b e rs * has be en presented to and a c c e p te d by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHI LOSOPHY D m o f Gr arf uai r Studiet Date J flQ Q . DISSERTATION COMMITTEE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENT I would like to express my appreciation to the following persons for their support and guidance with writing this dissertation. First of all, I would like to thank all of my dissertation committee members for their inspiration, enthusiasm and advice. I would especially like to thank Dr. Eileen Crimmins, the chairperson of my committee, who motivated and guided me through my study. Her vigor, compassion, and erudition always kept me on the right track. Next, I would like to thank Drs. David M. Heer and Merril Silverstein for their remarks on my study. Both professors inspired my empirical writing inside and outside of the classroom. I would also like to thank Dr. Yasuhiko Saito for teaching me invaluable research skills and techniques. The faculty and staff of the Andrus Gerontology Center, the USC Sociology Department, and the Population Research Laboratory were all very supportive. In addition, I would like to acknowledge the support I have received from friends whc have remained close to me throughout my studies. Their support has given me the encouragement I needed to complete my degree. Finally, I would like to dedicate this dissertation to my parents, Masuji and Hiroko Ueda, who have supported me for many years from abroad. I have always found inspiration in the musical groups, BOSTON and PSEUDO ECHO. The lyrics of their songs sen/e as a metaphor for the motivation I have received from all those listed above. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS PAGE ACKNOWLEDGEMENTS ii LIST OF TABLES vi LIST OF FIGURES xi ABSTRACT .................................................................................................. xv CHAPTER I INTRODUCTION AND OVERVIEW 1 A. Introduction and overview of the study .............................................. 1 B. Contributions to the literature ............................................................ 4 C. Outline of the dissertation .................................................................. 5 CHAPTER II LITERATURE REVIEW 7 A. Epidemiology theory........................................................................... 7 B. Diseases of the heart ........................................................................ 8 C. Mortality trends .................................................................................. 1 1 D. Survival trends .................................................................................. 13 E. Incidence trends ................................................................................ 13 F. Prevalence trends .............................................................................. 20 G. Trends in risk factors, diagnoses, and treatments.............................. 23 H. Comprehensive views of mortality, survival, incidence, and prevalence ........................................................................................ 24 I. Research questions .......................................................................... 27 CHAPTER III MATERIALS AND METHODS 29 A. Data Sources, Samples, and Measures ............................................ 29 1. Vital Statistics of the United Status - Mortality .............................. 29 iii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS (Continued) PAGE CHAPTER III (Continued) A. 2. National Health Interview Survey (NHIS)..................................... 31 3. Health and Retirement Study (HRS) and Assets and Health Dynamics among the Oldest Old (AHEAD)................................... 36 B. Analytical Methods ............................................................................ 42 1. Analytical methods for trends in heart disease with population (non-individual) data .................................................................... 42 2. Analytical methods for trends in heart disease with individual d a ta ............................................................................................... 44 3. Event history analysis ................................................................. 45 4. Multistate life table ..................................................................... 48 a. Estimated years of life with and without heart disease ............... 50 b. Estimated prevalence rate ........................................................ 52 CHAPTER IV RESULTS - TRENDS IN HEART DISEASE 55 A. Mortality trends .................................................................................. 55 1. Trends in overall heart disease mortality rates ........................... 55 2. Trends in acute myocardial infarction (AMI) mortality rates ......... 66 B. Prevalence trends.............................................................................. 76 1. Overall heart disease prevalence trends ..................................... 76 2. Ischemic heart disease prevalence trends................................... 94 C. Other trends - Onset, activity/work limitations, and survival trends • • 111 1. Time since the onset of heart disease......................................... 112 2. Limitations of activity/work for those who have heart disease 120 3. Survival trends ........................................................................... 136 D. Summary.............................................................................................. 143 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS (Continued) PAGE CHAPTER V RESULTS - HAZARD MODELS, TRANSITION SCHEDULES, AND MULTISTATE LIFE TABLES 145 A. Transition in heart disease status ..................................................... 146 B. Hazard models and annual transition rates ....................................... 153 C. Multistate life tables ......................................................................... 155 1. Expected years of life with and without heart diseases ................ 158 2. Estimated implied heart disease prevalence rates........................ 167 D. Summary........................................................................................... 177 CHAPTER VI SUMMARY AND CONCLUSIONS 179 REFERENCES.............................................................................................. 190 APPENDIX A ................................................................................................ 199 APPENDIX B ................................................................................................ 217 v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UST OF TABLES TABLE PAGE 2.1 Average Annual Number of Ischemic Heart Disease (IHD) Per 1,000 Persons Aged 55 Years and Over, by Sex and Age .............. 22 3.1 The Criteria and Code Number of Heart Disease in the International Classification of Diseases: 9th Revision (ICD-9) in the Vital and Health Statistics of the Department of Health and Human Services, Public Health Sen/ice, Centers for Disease Control and Prevention, National Center for Health Statistics ............. 30 3.2 The Criteria and Code Number for Diseases of Heart in the National Health Interview Survey (NHIS) - Diagnostic Recode (B) 32 3.3 Eligible Sample Size for Heart Disease Study from the National Health Interview Survey (NHIS) Among Those Aged 50 and Over, by Sex, by Y ear.................................................................................. 33 3.4 Heart Disease Related Variables in the National Health Interview Survey (NHIS).................................................................................... 34 3.5 Heart Disease Study Sample for the Merged Data from the Health and Retirement Study (HRS) and the Assets and Health Dynamics of the Oldest Old (AHEAD) - Numbers of Sample in Each Step towards the Final Sample Selection.................................. 38 3.6 Heart Disease Questions in the Health and Retirement Study (HRS) and the Assets and Health Dynamics of the Oldest Old (AHEAD), Wave I and Wave II .......................................................... 39 3.7 Weight Variables for the Health and Retirement Study (HRS) and the Assets and Health Dynamics of the Oldest Old (AHEAD) Data S ets.................................................................................................... 41 4.1 Death Rates for Diseases of Heart for Men, by 5-Year Age Group, Vital Statistics 1968-1996 .................................................................. 56 4.2 Parameter Estimates of the Annual Change in Death from Overall Heart Problems by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 .......................................................................................... 59 vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES (Continued) TABLE PAGE 4.3 Average Annual Percentage Changes (and 95% Confidence Intervals) in Overall Heart Disease Mortality by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 ................................................ 61 4.4 Death Rates for Diseases of Heart for Women, by 5-Year Age Group, Vital Statistics 1968-1996 ...................................................... 64 4.5 Death Rates for Acute Myocardial Infarction (AMI) for Men, by 5- Year Age Group, Vital Statistics 1968-1996 ...................................... 67 4.6 Parameter Estimates of the Annual Change in Death from Acute Myocardial Infarction (AMI) by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 .......................................................................... 69 4.7 Average Annual Percentage Changes (95% Confidence Intervals) in Acute Myocardial Infarction (AMI) Mortality by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 ................................................ 71 4.8 Death Rates for Acute Myocardial Infarction (AMI) for Women, by 5-year Age Groups, Vital Statistics 1968-1996 .................................. 74 4.9 Annual Prevalence of Overall Heart Disease, by Sex, by 5-Year Age Group, NHIS 1982-1996 ............................................................ 77 4.10 Prevalence of Overall Heart Disease, by Sex, by 5-Year Age Group, NHIS 1982-1996 (Average of a 3-Year Range)...................... 80 4.11 Annual Prevalence of Overall Heat Disease, by Sex, by 10-Year Age Group, NHIS 1982-1996 ............................................................ 82 4.12 Parameter Estimates of the Annual Change in Overall Heart Disease Prevalence, by Sex, by 5-Year Age Group, NHIS 1982- 1996 .................................................................................................. 83 4.13 Average Annual Percentage Changes (with 95% Confidence Intervals) in Overall Heart Disease Prevalence by Sex, by 5-Year Age Group, NHIS 1982-1996 ............................................................ 84 4.14 Average Annual Percentage Changes (with 95% Confidence Intervals) in Overall Heart Disease Prevalence, by Sex, by 10- Year Age Group, NHIS 1982-1996 ..................................................... 87 vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES (Continued) TABLE PAGE 4.15 Odds Ratios from the Logistic regression Analyses of the Prevalence of Overall Heart Disease by Sex, NHIS 1982-1996 89 4.16 Annual Prevalence of Ischemic Heart Disease (IHD), by Sex, by 5-Year Age Group, NHIS 1982-1996 ................................................ 95 4.17 Prevalence of Ischemic Heart Disease (IHD), by Sex, by 5-Year Age Group, NHIS 1982-1996 (Average of 3-Year Range).................. 98 4.18 Annual Prevalence of Ischemic Heart Disease (IHD), by Sex, by 10-Year Age Group, NHIS 1982-1996 ................................................ 100 4.19 Parameter Estimates of the Annual Change in Ischemic Heart Disease (IHD) Prevalence, by Sex, by 5-Year Age Group, NHIS 1982-1996 .......................................................................................... 101 4.20 Average Annual Percentage Changes (with 95% Confidence Intervals) in Ischemic Heart Disease (IHD) Prevalence by Sex, by 5-Year Age Group, NHIS 1982-1996 .................................................. 102 4.21 Average Annual Percentage Changes (with 95% Confidence Intervals) in Ischemic Heart Disease (IHD) Prevalence, by Sex, by 10-Year Age Group, NHIS 1982-1996 ................................................ 104 4.22 Odds Ratios from the Logistic Regression Analyses of the Prevalence of Ischemic Heart Disease (IHD) by Sex, NHIS 1982- 1996 .................................................................................................... 107 4.23 Percentage of Interviewees who Experienced Heart Disease Onset within 1 and 5 Years before Interview, by Sex, NHIS 1982- 1996 .................................................................................................... 113 4.24 Odds Ratios of Experiencing the Onset of Heart Disease within 1 and 5 Years before Interview, by Sex, NHIS 1982-1996 .................... 115 4.25 Percentages in Specified Disability States for Those Aged 50 and Over with Heart Disease, by Sex, NHIS 1982-1996 ........................... 121 4.26 Odds Ratios of Activity Limitation Status for Individuals with Heart Disease Aged 50 and Over, by Sex, NHIS 1982-1996 ...................... 125 viii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES (Continued) TABLE PAGE 4.27 Odds Ratios of Work Limitation Status for Individuals with Heart Disease Aged 50-69, by Sex, NHIS 1983-1996 .................................. 129 4.28 Percentages of People who Deceased by Time since Interview, by Sex, NHIS 1986-1994 .......................................................................... 137 4.29 Odds Ratios for Deaths During Specific Time After Interview for Men, by Heart Disease Conditions, by Period, NHIS 1986-1994 .......... 140 4.30 Odds Ratios for Deaths During Specific Time After Interview for Women, by Heart Disease Conditions, by Period, NHIS 1986- 1994 .................................................................................................... 141 5.1 Observed Numbers and Percentages for Heart Disease Transition, HRS/AHEAD Wave I - Wave II, by Sex .............................. 147 5.2 Observed Prevalence of Heart Disease at Wave I, by Sex, by Age Group, HRS/AHEAD .......................................................................... 148 5.3 Observed Incidence of Non-Fatal Heart Disease for Men and Women, Between Wave I and Wave II, HRS/AHEAD........................ 151 5.4 Summary of Equations of Hazard Analyses for Heart Disease Transitions, by Sex, by Transitions, HRS/AHEAD.............................. 154 5.5 Annual Transition Rates for Multi-State Heart Disease for Those Aged 50-97, by Sex, HRS/AHEAD...................................................... 157 5.6 Expectations of Years of Life With and Without Heart Disease at Certain Age for Aged 50 and Over, Population-Based, by Sex ....... 159 5.7 Expectations of Years of Life by Heart Disease Status at Certain Age for Aged 50 and Over, Status-Based (Initial Status: Without Heart Disease), by Sex, HRS/AHEAD................................................ 162 5.8 Expectations of Years of Life by Heart Disease Status at Certain Age for Aged 50 and Over, Status-Based (Initial Status: With Heart Disease), by Sex, HRS/AHEAD ............................................... 164 5.9 Expectations of Years of Life by Heart Disease Status at Certain Ages for Aged 50 and Over Status-Based (With Heart Attack), by Sex, HRS/AHEAD.............................................................................. 165 ix Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES (Continued) TABLE PAGE 5.10 Implied Population-Based Age-Specific Prevalence Rate, by Sex • 170 5.11 Implied Status-Based Age-Specific Prevalence Rate (Initial Status: without Heart Disease), by S e x .............................................. 174 5.12 Implied Status-Based Age-Specific Prevalence Rate (Initial Status: With Heart Disease), by Sex................................................... 175 x Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES FIGURE PAGE 3.1 Status of Disease in the Heart Disease Model..................................... 47 4.1 Death Rates for Diseases of the Heart for Men, Vital Statistics, 1968-1996 .......................................................................................... 57 4.2 Average Annual Change for Overall Heart Disease Mortality Rates, by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 60 4.3 Comparison of the Average Annual Percentage Change for Overall Heart Disease Mortality Change, by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 ...................................................... 62 4.4 Death Rates for Diseases of the Heart for Women, Vital Statistics 1968-1996 .......................................................................................... 65 4.5 Death Rates for Acute Myocardial Infarction (AMI) for Men, Vital Statistics 1968-1996 ........................................................................... 68 4.6 Average Annual Changes for Acute Myocardial Infarction (AMI) Mortality Rates by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 ......................................................................................... 70 4.7 Comparison of the Average Annual Percentage Change for Acute Myocardial Infarction (AMI) Mortality Rates by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 ...................................................... 73 4.8 Death Rates for Acute Myocardial Infarction (AMI) for Women, Vital Statistics 1968-1996 .................................................................. 75 4.9 Annual Prevalence of Heart Disease for Men, NHIS 1982-1996 ......... 79 4.10 Prevalence of Heart Disease for Men, NHIS 1982-1996 (3-Year Average).............................................................................................. 81 4.11 Comparison of the Average Annual Percentage Change for Overall Heart Disease Prevalence by Sex, by 5-Year Age Group, NHIS 1982-1996 ................................................................................ 86 xi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES (Continued) FIGURE PAGE 4.12 Comparison of the Average Percentage Change for Overall Heart Disease Prevalence by Sex, by 10-Year Age Group, NHIS 1968- 1996 ..................................................................................................... 88 4.13 Parameter Estimates for Overall Heart Disease Prevalence with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 ............................................................................................. 91 4.14 Annual Prevalence of Heart Diseases for Women, NHIS 1982- 1996 ..................................................................................................... 92 4.15 Prevalence of Heart Disease for Women, NHIS 1982-1996 (3- Year Average)....................................................................................... 93 4.16 Annual Prevalence of Ischemic Heat Disease (IHD) for Men, NHIS 1982-1996 ......................................................................................... 96 4.17 Prevalence of Ischemic Heart Disease (IHD) for Men, NHIS 1982- 1996 (3-Year Average) ...................................................................... 99 4.18 Comparison of the Average Annual Percentage Change for Ischemic Heart Disease (IHD) Prevalence by Sex, by 10-Year Age Group, NHIS 1968-1996 ...................................................................... 103 4.19 Comparison of the Average Percentage Change for Ischemic Heart Disease (IHD) Prevalence by Sex, by 5-Year Age Group, NHIS 1968-1996 .................................................................................. 105 4.20 Parameter Estimates for Ischemic Heart Disease Prevalence with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 .......................................................................................... 108 4.21 Annual Prevalence of Ischemic Heart Disease (IHD) for Women, NHIS 1982-1996 ................................................................................ 109 4.22 Prevalence of Ischemic Heart Disease (IHD) for Women, NHIS 1982-1996 (3-Year Average)................................................................ 110 xii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES (Continued) FIGURE PAGE 4.23 Percentage of Interviewees who Experienced Heart Disease Onset within 1 and 5 Years before Interview, by Sex, NHIS 1982- 1996 .................................................................................................... 114 4.24 Parameter Estimates for Experiencing the Overall Heart Disease within 1 Year Before Interview with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 ................................. 117 4.25 Parameter Estimates for Experiencing the Overall Heart Disease within 5 Years Before Interview with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 ................................. 118 4.26 Activity Limitation Among Men with Heart Disease Aged 50 and Over, NHIS 1982-1996 ...................................................................... 122 4.27 Work Limitation Among Men with Heart Disease Aged 50-69, NHIS 1983-1996 ................................................................................ 124 4.28 Parameter Estimates for the Status of No Activity Limitation Among Those with Heart Disease with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 ................................ 127 4.29 Parameter Estimates for the Status of Unable to Perform A Major Activity Among Those with Heart Disease with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 128 4.30 Parameter Estimates for the Status of No Work Limitation Among Those with Heart Disease with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 .............................................. 131 4.31 Parameter Estimates for the Status of Unable to Work Among Those with Heart Disease with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 .............................................. 132 4.32 Activity Limitation Among Women with Heart Disease Aged 50 and Over, NHIS 1982-1996 ................................................................ 133 4.33 Work Limitation Among Women with Heart Disease Aged 50-69, NHIS 1983-1996 ................................................................................ 134 xiii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES (Continued) FIGURE PAGE 4.34 Percentage Deceased among Men Aged 50 and Over by Time since NHIS Interview, by Interview Y ear............................................. 138 4.35 Percentage Deceased among Women Aged 50 and Over by Time since NHIS Interview, by Interview Year ............................................ 142 5.1 Observed Prevalence of Heart Disease at Wave I, by Sex, HRS/AHEAD...................................................................................... 150 5.2 Observed Incidence of Non-Fatal Heart Disease, between Wave I and Wave II, by Sex, HRS/AHEAD .................................................... 152 5.3 Annual Transition Rates for Multi-State Heart Disease, by Transition, by Sex, HRS/AHEAD........................................................ 156 5.4 Expected Age of Onset of Heart Disease and Heart Attack, by Age and Sex, Population-Based ........................................................ 160 5.5 Expected Age of Onset of Heart Disease and Heart Attack for Those who are Free of Heart Disease at Specific Age, by Sex, Status-Based .................................................................................... 163 5.6 Expected Age of Onset of Heart Attack for Those with Heart Disease at Specific Age, by Sex, Status-Based.................................. 168 5.7 Expected Years of Life after Heart Attack who have had Heart Attack at Specific Age, by Sex, Status-Based.................................... 169 5.8 Population-Based Age-Specific Prevalence Rate (Initial Status: Without Heart Disease), by Sex.......................................................... 171 5.9 Status-Based Age-Specific Prevalence Rate (Initial Status: Without Heart Disease), by Sex.......................................................... 173 5.10 Status-Based Age-Specific Prevalence Rate (Initial Status: With Heart Disease), by Sex ...................................................................... 176 xiv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT This dissertation investigates heart disease in the middle-aged and older US population with two major concerns: time trends and length of life with and without heart disease. First, it examines trends in mortality, prevalence, onset, related disability, and death rates among those with and without heart disease. Second, it estimates the average length of an individual's life with and without heart disease and estimated implied prevalence from multistate models. Results indicate that heart disease mortality rates have continuously declined for three decades for both sexes. Older persons show faster mortality declines in overall heart disease and myocardial infarction (Ml). Men show greater annual percent declines in heart disease mortality rates than women, except for overall heart disease among older age groups. While the overall heart disease prevalence rate has increased, the ischemic heart disease (IHD) prevalence rate has decreased among men. Women have opposite results: a decrease in overall heart disease prevalence and an increase in ischemic heart disease. Women tend to have more recent onset of heart disease than do men. Both men and women with heart disease have less disability in recent years. Death rates from all causes and death rates due to heart disease among those with and without heart disease have decreased for both sexes. Estimated years of life with and without heart disease and the implied prevalence of heart disease differ by sex due to different transition rates into and out of states of health defined by heart disease. Men and women are expected to xv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. live different years with and without heart disease. The likelihood of having had a heart attack is higher for men than women. These research findings benefit researchers and policymakers by providing information on time trends of heart disease and related issues as well as changes in a population’s health status and information on life with and without heart disease. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER I INTRODUCTION AND OVERVIEW This chapter begins with an introduction and overview of the dissertation. It outlines the rationale for and purposes of the study. The contribution of the dissertation to the literature is then addressed. An outline of the dissertation is presented in the last part of this chapter. A. Introduction and Overview of the Study Heart disease is one of the most important public health concerns in the United States. The death rate from heart disease increases with age among the middle-aged and older population. Heart disease has been the leading cause of death among the elderly for many years. The category of ischemic heart disease (IHD) includes myocardial infarction (Ml, or a heart attack), which is a major cause of death among diseases of the heart. People who survive heart disease may be affected for the rest of their lives in their daily activities because of the disease. These people may have a limitation in their ability to work or may even need to withdraw from the labor force because they are not able to perform their work. Limitation in daily activities because of heart disease may result in an inability to perform tasks required to maintain an independent household or in impairment in instrumental activities of daily living (IADL) such as cooking, using the telephone, shopping, doing light housework, managing money, and getting around the community. In severe cases of physical 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. limitation, an individual with heart disease may have difficulty performing basic self- care activities often referred to as activities of daily living (ADL), which include eating, using the toilet, dressing, transferring, walking, and bathing. Because the population of the U.S. is aging and will continue to do so in the future, the number of people with heart disease is expected to increase. If persons live more years with heart disease, the number will increase even more. Disability due to heart disease burdens both individuals and society in terms of health care costs and need for health services. One of the objectives of this study is to describe and analyze time trends in a variety of aspects of heart disease. This study will provide a comprehensive understanding of trends in heart disease. While many studies have examined heart disease trends in the past, researchers have focused on the trend in overall mortality, or in the prevalence, incidence, survival, or risk factors for heart disease. Outcomes from past research differ partially due to the focus on different aspects of heart disease, the examination of different age groups, or the period of study. Moreover, most studies are not well integrated in presenting a comprehensive view of heart disease. Generally, heart disease mortality has declined over recent decades among both men and women, but time trends linked to mortality such as prevalence and disability linked to heart disease are not well documented. To overcome these issues, different aspects of short-term and long-term heart disease trends are examined in this study using two national data sets: the Vital Statistics and the National Health Interview Survey (NHIS). These data sets provide annual time trends based information on heart disease mortality, prevalence, time since onset, related disability, and death rates among those with and without heart 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. disease for men and for women. Heart disease trends are examined separately for overall heart disease and myocardial infarction (Ml) or ischemic heart disease (IHD). Another objective in this study is to build a model relating heart disease mortality and morbidity. This will allow estimates of years of life with and without heart disease as well as the prevalence of heart disease implied in the population through the use of heart disease multistate life tables. The current health status of the US population is related to the past trends in heart disease death rates and to the length of life with and without heart disease. Despite the recognition of heart disease as a serious cause of death and morbidity, little evidence is available on trajectories of heart disease over the life cycle. Details of heart disease trajectories for men and for women are not well known. The interplay of heart disease mortality and onset of heart disease determines the duration of life with and without heart disease over the life cycle. Information on time trends in heart disease, such as in mortality, prevalence, and onset can provide comprehensive views and better understanding of the heart disease process and of population health for both men and women. This study estimates the average length of an individual's life cycle with and without heart disease for men and for women in the middle-aged and older population. For the detailed investigation of heart disease, heart disease status is categorized into three statuses: "without heart disease," "with heart disease but no heart attack,” and “with heart attack." The merged data set from the Health and Retirement Survey (HRS) and the Asset and Health Dynamics of the Oldest Old (AHEAD) expands the age range and provides transition rates between different disease statuses. A multistate life table approach will be used to simulate the heart disease experience by linking the mortality change to change in population morbidity. 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The life table models estimate life with and without heart disease for men and for women. Moreover, through the use of the multistate model, the expected length of life with heart disease, and the implied health status of the middle-aged and elder U.S. population, and heart disease prevalence rates are generated. B. Contributions to the Literature This dissertation makes two contributions. First, it provides a comprehensive investigation of time trends in mortality and the prevalence of heart disease among the middle-aged and the elderly in the United States. These heart disease analyses inform us on changes in mortality, prevalence, time since onset, disability among those with heart disease, and survival among those with and without heart disease. Prevalence, for instance, is a very important indicator of population health, but the trend using annual change over a number of years has never been reported in the past. Most studies have reported the prevalence change between two years. This study provides information on the short-term and long-term time trends of heart disease on an annual basis. Second, the dissertation applies the multistate approach to building a life table to clarify how morbidity and mortality from heart disease are related and to estimate life with and without overall heart disease and life after a heart attack. Although multistate life table models have been used in analyses of disability, the models have not been used frequently in studies of heart disease. Application of multistate life tables will help us to understand how past changes in mortality and possible future changes in mortality and transition schedules affect estimated life expectancy in various states of heart disease and affect the prevalence of heart disease in a life Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. table population. This multistate life table combines information on age specific heart disease morbidity states with mortality rates into a single summary indicator. We thereby gain a clearer understanding of the dynamics of both population health and individual health experiences with heart disease over the life cycle. Results highlight the usefulness of researchers examining measures of population health as well as life cycle indicators of health related to heart disease. For the same reason, the results may help policymakers have a better understanding of how changes in a population's health status come about. By identifying trends in heart disease, researchers and policymakers would be able to relate past health policies and services, as well as risk factors for heart disease to health trends. They would also be able to understand how these trends change the average life cycles of individuals in terms of years with and without heart diseases. In sum, this dissertation helps us to understand the trend in heart disease and the process of life changes with and without diseases of the heart among middle-aged and older American men and women. C. Outline of the Dissertation In Chapter Two, the literature on heart disease trends in mortality, survival, incidence, and prevalence is described. Hypotheses for this study are proposed. In Chapter Three, the study design is described. This includes the description of data sources, sampling methods and selection, measures, and analytical methods. In Chapter Four, the trends in different aspects of heart disease are examined. The trends for heart disease include mortality, prevalence, time since onset, and related disability or resulting activity/work limitations. In Chapter Five, based on hazard 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. analyses, transition schedules are used to develop multistate life tables. Multistate models of two types are produced: population-based and state-based. In Chapter Six, the concluding chapter, the findings from the analyses are reviewed and related to the initial hypotheses. Theoretical implications are drawn and future research suggestions outlined. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER II LITERATURE REVIEW In this chapter, an epidemiologic theory is briefly introduced. Then, literature on diseases of the heart is reviewed and information on heart diseases prevalence is provided. Next, literature on trends in mortality rates, survival, incidence rates, and the prevalence of diseases of the heart is explored. In addition to the trends in heart disease itself, the trends in risk factors, diagnoses, and treatments are explored. Then, literature integrating heart disease mortality, survival, incidence, and prevalence is presented. In the final section of this chapter, research questions are presented and the literature is summarized. A. Epidemiology theory Knowledge of the prevalence, incidence, and duration of a condition or illness in a population is necessary for understanding trends in morbidity from that condition or illness. Mortality, incidence, and prevalence of heart disease are related. A basic epidemiological equation is expressed as: Prevalence (Pn ) = Incidence (lr) x Duration (E,(D)). (Eq. 2.1) This equation means that “the expected number of diseased individuals is equal to the number of new cases per time unit multiplied by the expected duration of disease” (Freeman & Hutchinson, 1980; Hennekens & Buring, 1987; Keiding, 1991). In this equation, information on two of these factors can be used to calculate the third in a defined steady state. However, this formula does not work when there is Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. population growth. In other words, only if age-specific birth rates and death rates are constant over time (in a stationary population), is this formula applicable (Alho, 1992; Mausner & Kramer, 1985; Murray & Lopez, 1994). B. Diseases of the Heart Heart disease is one of the most important public health concerns in the United States. Medical care utilization and health care service utilization are high among people with heart disease. Annual physician visits and hospitalization rates are higher for those with heart disease than for those without heart disease. Hospitalization rates are particularly high for people with Ischemic Heart Disease (IHD), which is also called Coronary Heart Disease (CHD). Over three-quarters of the population with CHD have been hospitalized for heart disease (National Academy on an Aging Society [NAAS], 2000). People of age 70 and over with heart disease are more likely to use prescription drugs and other services such as social workers, adult day care, rehabilitation, transportation, or Meals on Wheels, than are those without heart disease (NAAS, 2000). Heart disease limits physical activity and work conditions among individuals who experience it. In the U.S. labor force, IHD is a leading cause of premature, permanent disability. Among 5,201 community-dwelling persons aged 65 and over, 16% of men and 13% of women claimed some difficulty in physical tasks due to heart disease (Ettinger et al., 1994). Labor force participation is lower in the population with heart disease as compared to the population without heart disease. Experiencing heart disease appears to be a factor pushing people toward early retirement and reducing satisfaction with retirement (NAAS, 2000). 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It is also pointed out that there is a sex difference in physical functional ability due to heart disease. A follow-up study of the 1976-1977 cohort that experienced hospitalization due to angina or other acute coronary diseases, researchers found that female survivors with angina showed a significantly higher risk for diminished functional capacity when compared to men as well as to women with other heart disease (Nickel & Chirikos, 1990). Heart disease has been the leading cause of death in the United States since 1921 (Anonymous, 1999b). IHD is the most common type of heart disease. It is the most important cause of death in both middle aged and older persons. CHD is also a major cause of morbidity (Marmot, 1996). Heart disease accounts for one third of all deaths, more than 40% of deaths of those over 65 and close to half of all deaths of those over age 85 (American Heart Association [AHA], 1999; Cohen, Van Nostrand, & Furner, 1993). The number of deaths from IHD was approximately 476,000 in the United States in 1999 (AHA, 1999; National Center for Health Statistics [NCHS], 1999a). The prevalence of IHD is estimated at 12 million cases in the United States (AHA, 1999). Ischemic heart disease arises from the condition called atherosclerosis. When vascular blood supply to the heart muscle is limited, the heart muscle receives oxygen insufficient to its demand. Eventually, the decrease in the blood flow in the coronary vessels leads to chest pain (called 'angina pectoris'). Sudden occlusion of blood flow to the coronary vessels causes acute myocardial infarction (Ml), or a heart attack, which results in the death of heart muscle where the oxygen supply was stopped. If blood flow is chronically diminished with heart muscle damage, this can lead to congestive heart failure (AHA, 1999). 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A large number of men have heart disease. According to the National Health and Nutrition Examination Survey III (NHANES III) of 1988-1994, 5.9 million men have IHD, and 2.3 million men have congestive heart failure (CHF), separately. While IHD refers to a problem of the coronary artery in the heart, CHF refers to a dysfunction of the heart pumping that results from the blood staying or backing up in the veins and lungs. In the category of IHD, 4.4 million men have Ml, and 2.3 million men have angina pectoris (AHA, 1999). Although it is commonly recognized that the prevalence of heart disease is more frequent among men than among women (AHA, 1999), the actual number of women with a disease of the heart is larger than the number of men for some types of heart disease. This is due to the difference in the population size for men and women. Women have a greater number of angina pectoris than men in one of the categories of IHD (AHA, 1999). According to the NHANES III for 1988-1994, 6.2 million women have IHD, and 2.4 million women have CHF. Among those within IHD, 2.8 million women have Ml, and 3.9 million women have angina pectoris (AHA, 1999). The prevalence of heart disease increases with age and shows a significant sex difference, however, this sex difference disappears after age 70 (Agner, 1985; AHA, 1999; Timiras, 1994). With an examination of IHD prevalence at ages 70 and 80, one study shows a significantly higher prevalence rate of IHD was found at age 70 for men than women. Women reached the prevalence level of men at age 80. Although the prevalence rate of CHD is 15% for men and 9% for women of those aged 70, the rates for both men and women become 20% at age 80 (Agner, 1985). 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C. Mortality Trends Many studies examine heart disease mortality from all diseases of the heart combined, ischemic heart disease (IHD), or myocardial infarction (Ml). In this study, death resulting from any type of heart problems is called overall heart disease mortality. Ischemic heart disease includes Ml and angina pectoris, and Ml mostly contributes to heart disease mortality. For instance, mortality rates from all disease of the heart and IHD are 271.6 and 174.2 per 100,000 U. S. population, respectively (NCHS, 1999b). Overall mortality due to diseases of the heart has declined since the 1950s (Anonymous, 1999b). Overall heart disease mortality for both sexes declined 27% between 1980 and 1991 (NCHS, 1999a). Mortality rates from IHD have declined since the mid-1960s (Anonymous, 1999b; Gillum, 1994; Thom & Maurer, 1988). According to the MONICA project of the World Health Organization, the United States and Australia had the largest IHD mortality decline (38%) between 1972 and 1984. This was greater than that in countries of Europe, the Asian-Pacific region, and America (Epstein, 1988). The approximate annual rate of decline in IHD was 3.0% between 1963 and 1985 in the United States (Thom & Maurer, 1988). Another study in the U.S. found an even faster annual decline (3.5%) in IHD mortality during the 1980s (Gillum, 1994). Other studies based on specific geographic areas of the United States also indicate a decline in mortality due to diseases of the heart. Continuing decline in IHD mortality from 1968 to 1988 among both men and women has been found in the Minnesota Heart Survey (Burke, Sprafka, Folsom, Luepker, Norsted, & Blackburn, 1989; Demirovic, Blackburn, McGovern, Sprafka, & Doliszny 1993). In Olmsted 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. County, Minnesota, researchers found a decline parallel to the national mortality decline between 1979 and 1994 (Roger, Jacobsen, Weston, Bailey, Kottke, & Frye, 1999). Mortality due to IHD declined 3.6% annually among people aged 35 to 74 from 1987 to 1994 in four communities of the United States (Rosamond, Chambless, Folsom, Cooper, Conwill, Clegg et al., 1998). A difference in the trend of overall heart disease mortality by sex has been observed in previous studies. Heart disease mortality for women peaked in the 1930s and in the 1960s among men. The trend in heart disease mortality among men was a steady increase until the 1950s, it then leveled off for about two decades, and then started coming down after the late 1960s. The pattern for women was different. Heart disease mortality rates for women rose until the 1930s; then gradually declined until about 1950; and declined more steeply after that (Feinleib, 1995). The trend in heart disease mortality decline differs by sex in different time periods. In a 20-year follow-up of the Framingham Heart Study, mortality due to IHD declined significantly for men (44%) and women (61%) aged 50-59 between the 1950 and the 1970 cohorts (Sytkowski, D'Agostino, Belanger, & Kannel, 1996). Overall decline of mortality due to Ml was 28% for men and 31% for women between 1987 and 1994 in four communities (Rosamond et al., 1998). From the late 1970s through the end of the 1980s, overall heart disease mortality rates among men have been decreasing faster than among women in the United States (Feinleib, 1995). The decline in overall heart disease mortality was faster for men (39%) than for women (33%) between 1979 and 1993 (NCHSa, 1996). In the Rochester Epidemiology Program Project, it was reported that the overall heart disease death rate declines 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were 4.2% per year for men and 2.5% per year for women between 1979 and 1994 (Roger etal., 1999). An earlier study with a shorter period from 1965 to 1974 suggested that the decline in IHD mortality was less for women than for men in Alameda County (Kaplan, Cohn, Cohen, & Guralnik, 1988). Data from the Minnesota Heart Survey suggests that this trend has continued. During the period 1979 through 1988, a slowing of the rate of decline was observed among women, whereas it accelerated among men in Minnesota. Ischemic heart disease mortality rates among men fell proportionately more each year in the period 1970-1988 (4.8%) than in the period of 1970-1978 (3.9%). On the contrary, the average annual percentage decline in IHD death rate in women was lower in the period 1979-1988 (3.6%) than in 1970-1978 (4.8%) in the Minnesota Heart Survey (Demirovic et al., 1993). Over the long period of 1970 through 1993, men and women experienced similar rates of decline in mortality from heart disease (NCHS, 1996a). Mortality from IHD fell by approximately 25% for both men and women between 1985 and 1990 in the Twin Cities area (McGovern, Pankow, Shahar, Doliszny, Folsom, Blackburn et al., 1996). D. Survival Trends A number of researchers examined survival for individuals once they have been diagnosed with heart disease. Survival over the short term is indicated by in- hospital mortality. Longer survival may be indicated by living for a certain number of years after the diagnosis. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In-hospital mortality has improved among both men and women for many years. However, women have had poorer mortality rates from myocardial infarction in hospitals than men for many years. Women experienced poorer in-hospital survival than men in every year except for the last examination year in a multi hospital population-based study in Worcester, Massachusetts between 1975 and 1988 (Goldberg, Gorak, Yarzebski, Hosmer, Dalen, Gore et al., 1993). There is a decline in the mortality rate from Ml in hospitals for men and women aged 30-74 between 1970 and 1985 in the Minnesota Heart Survey. Age-adjusted in-hospital mortality rates due to Ml declined from 20.9% to 11.4% (55% drop) for men and from 26.5% to 16.5% (62% drop) for women (Burke et al., 1989). In the Minnesota Heart Survey, on the other hand, the decline in IHD mortality rates in the hospital became faster among men, increasing from 2.4% annually in 1970-1980 to 8.3% annually in 1980-1988. The increase in the in-hospital mortality decline was not as fast for women during the same period (3.7% vs. 4.1%) (Demirovic et al., 1993). It seems that both sexes have similar in-hospital mortality rates in recent years in some studies. In-hospital mortality from IHD for men and women aged 35-74 fell at almost the same annual rate of 5% among people aged 35 to 74 in four communities between 1987 and 1994 (Rosamond et al., 1998). Researchers usually use a 28-day mortality rate as a criterion of short-term survival. Mortality rate within 28 days after the event of Ml was significantly lower in 1980 than in 1970 in both men and women, but no change occurred from 1980 to 1985 in the Minnesota Heart Survey. Mortality rates within 28 days among men who experienced Ml declined from 18% in 1970 though 12% in 1980 to 13% in 1985. Women had even a larger decline in mortality than men; from 27% in 1970 through 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22% in 1980 to 18% in 1985 (McGovern, Folsom, Sprafka, Burke, Doliszny, Demirovic et al., 1992). In a later study from the Minnesota Heart Survey, researchers found that mortality within 28 days after the onset of Ml was 25% for men and 15% for women from 1985 to 1990. Statistical significance was not found for women (McGovern et al., 1996). In a more recent study, the mortality rate within 28 days for those who were hospitalized due to Ml showed a continuous improvement in survival. In the period from 1987 through 1994, the annual rate of mortality decline within 28 days was 4% for men and 10% for women (Rosamond et al., 1998). Studies of three to four-year survival after Ml revealed more improvement for men than women. As compared to the 1970 cohort, the 1980 cohort that experienced Ml had a four-year survival rate that was 33~35% higher for men and 27~28% higher for women in the Minnesota Heart Survey (Gomez-Marin, Folsom, Kottke, Wu, Jacobs, Gillum et al., 1987; McGovern et al., 1992). However, there was no significant change in survival between the 1980 cohort and the 1985 cohort (McGovern et al., 1992). Survival among patients hospitalized for Ml has improved substantially in recent years. Comparing the 1990 and the 1985 cohorts, the death rates within three years after hospitalization for Ml dropped from 27 to 21 (24%) for men and from 33 to 28 (16%) for women aged 30-74 (McGovern et al., 1996). Looking at long-term survival, women with heart disease experienced better survival than men with heart disease, however, no improvement over time was seen for either sex. With a follow-up of 10 years for people who developed CHF from 1948 to 1988 in the Framingham Study, researchers pointed out that men experienced a significantly poorer long-term survival than women with CHF (Ho, 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Anderson, Kannel, Grossman, & Levy, 1993). This difference by sex is also confirmed in another study. Between 1975 and 1988, men experienced poorer long term survival than women among those aged 25 and over who were discharged with an Ml diagnosis in Worcester, Massachusetts (Goldberg et al., 1993). No differences in 9-year survival after Ml were found between the 1971 and 1980 cohorts of persons 65 and over of either sex in the Kaiser Foundation Health Plan (Haan, Selby, Rice, Quesenberry, Schofield, Liu et al., 1996). With age adjustment, no significant improvement was noted in long-term survival of up to 14 years for either sex of those aged 25 and over with Ml between 1975 and 1988 in Worcester, Massachusetts (Goldberg et al., 1993). E. Incidence Trends The trend in heart disease incidence differs by sex in many studies. In an early study of the 30 years from 1952 to 1982, the age-adjusted incidence of IHD for men decreased 11% in the Rochester Epidemiology Program Project, a population- based study in Rochester, Minnesota (Elveback, Connolly, & Melton, 1986). However, an examination of the incidence trend with a long period of follow-up in the Framingham Heart Study had a different result. Researchers compared three same age cohorts, aged 50-59, for the years 1950,1960, and 1970 with a 20-year follow- up and found no significant reduction in incidence rates of IHD among the 1950, 1960, and 1970 male cohorts (Sytkowski et al., 1996). The observed decline in IHD incidence was due mainly to changes in the incidence of Ml within the category of IHD. For instance, the incidence rates of Ml declined by 20% in men from 1952 to 1982 in the Rochester Epidemiology Program 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Project. In this study, the greatest changes were found in men aged 50-69. There was a 32% decline in Ml incidence within this population between 1965-1969 and 1979-1982 (Elveback et al., 1986). Other studies have reported declines for men between 1970 and 1980. A computer simulation estimated the annual decline in the IHD incidence rate to be 2.0% between 1970 and 1980 for men aged 35-84 (Hunink, Goldman, Tosteson, Mittleman, Goldman, Williams et al., 1997). In the same period, hospitalization rates for Ml in persons 30 to 74 years old declined 8% among men in the Minnesota Heart Survey (Gillum, Folsom, Luepker, Jacobs, Kottke, Gomez-Marin et al., 1983). Other studies report different heart disease incidence decline rates for men. Men experienced large declines in Ml incidence at ages 55 and over. Between 1975 and 1988, the age-specific incidence rates of Ml declined by 26% in men aged 25-54 years, 15% in those 55-64 years, 25% in those 65-74 years, 33% in those 75-84 years of age, and 54% in those 85 years of age and over. Although the age-adjusted incidence rates of Ml declined by 26% for men in the period of 1975 through 1988, age-adjusted incidence rates of Ml increased for men in the shorter 1975 to 1981 period (Goldberg et al., 1993). A later study also reported that the rates of hospitalization for Ml declined by 5% to 10% between 1985 and 1990 (McGovern et al., 1996). Some studies show no changes in heart disease incidence. Using a different diagnostic approach, researchers found no significant changes occurred in the definite Ml attack rate from 1970 to 1985 in men. After examining the trend in Ml using discharge diagnoses and accounting for hospital coding practices for the years 1970,1980, and 1985 in the Minnesota Heart Survey, researchers found that there 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was no significant age-adjusted change for the definite Ml rates in men. In this study, Ml incidence rates for each 10-year age-interval (i.e., 30-39, 40-49, 50-59, 60-69, and 70-74 due to 30-74 age range) also indicated no statistically significant changes in the period of 1970, 1980 through 1985 (Burke et al., 1989). Another study also found no significant change in Ml incidence for elderly men from the 1970s through the 1980s. Examining the 9-year Ml incidence rate among those aged 65 and over in 1971 and 1980, researchers found that there was no cohort difference in incidence among men in the Kaiser Foundation Health Plan (Haan et al., 1996). A more recent study from four communities revealed no significant trend in hospitalization due to Ml. Using surveillance of hospital admissions for Ml among residents aged 35 through 74 between 1987 and 1994, researchers found that the age-adjusted incidence of hospitalization for Ml changed little over the 8-year period. For men, the hospitalization rate due to Ml was 4.1 per 1,000 in both 1987 and 1994. The age- adjusted rate of hospitalization for Ml increased 0.1% annually from 1987 to 1994 for men, but they concluded that there was no evidence of a decline in the incidence of hospitalization for a first myocardial infarction among men (Rosamond et al., 1998). The trend in heart disease incidence for women also differs across studies. In an early study of the 30 years from 1952 to 1982, the age-adjusted incidence of IHD increased 9% in women in the Rochester Epidemiology Program Project in Minnesota (Elveback et al., 1986). However, some studies reported declines for women between 1970 and 1980. A computer simulation estimated the annual decline in the IHD incidence rate to be 2.0% for women aged 35-84 between 1970 and 1980 (Hunink et al., 1997). One study examined the trend in incidence over a long period of follow-up for three cohorts aged 50-59. Researchers in the 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Framingham Heart Study compared those aged 50-59 in 1950, 1960, and 1970 with a 20-year follow-up. They found a significant reduction in IHD incidence rates between the 1950 and 1970 cohorts of women. The 20-year incidence of each diagnostic category of CHD declined among the female cohorts except for Ml. Although the 20-year incidence rate of Ml for women fell between the 1950 and 1960 cohorts, it rose 37% between the 1960 and 1970 cohorts (Sytkowski et al., 1996). These changes in trend were due mainly to changes in the incidence of Ml within the category of IHD. In the Rochester Study, which included all residents, the incidence rates of IHD increased by 28% in women aged 50-69 from 1965 though 1969 to 1979 through 1982. The greatest changes were found in the Ml incidence rates. Women aged 50-69 experienced a 44% increase of Ml incidence during the period (Elveback et al., 1986). Age-adjusted incidence rates of Ml increased for women in the period 1975 through 1981 (Goldberg et al., 1993). Between 1970 and 1980, hospitalization rates for acute Ml in women 30 to 74 years old declined 26% in the Minnesota Heart Survey (Gillum et al., 1983). The age-adjusted incidence rates of Ml declined by 22% in women in the period 1975 through 1988. Between 1975 and 1988, the age- specific incidence rates of Ml in women declined by 1% among those 25-54 years of age, 29% for those 55-64 years, 15% in those 65-74 years, 24% in those 75-84 years, and 55% in those 85 years old and over (Goldberg et al., 1993). A later study also reported that the rates of hospitalization for Ml declined by 5 to 10 percent between 1985 and 1990 (McGovern et al., 1996). Additional studies show no changes in heart disease incidence for women. Examining the trend of Ml through discharge diagnoses and after accounting for 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hospital coding practices in the years of 1970, 1980, and 1985 in the Minnesota Heart Survey, researchers found that there were no significant age-adjusted changes in the definite Ml rates in women. In this study, incidence rates of Ml for 10-year age groups did not change statistically from 1970 to 1980 or 1985 (Burke et al., 1989). Another study found no significant change in Ml incidence from the 1970s through the 1980s among the elderly. Through the examination of 9-year incidence rates for those age 65 and over in 1971 and 1980, researchers found no cohort difference in Ml incidence among women in the Kaiser Foundation Health Plan (Haan et al., 1996). A more recent study from four communities revealed no significant trend change in hospitalization due to Ml. Using surveillance of hospital admissions for Ml among residents aged 35 through 74 between 1987 and 1994, researchers found that the age-adjusted incidence of hospitalization for Ml changed little over the 8-year period. The rate of hospitalization due to Ml among women was 1.9 per 1,000 persons in 1987 and 1.8 per 1,000 in 1994. The age-adjusted rate of hospitalization for Ml from 1987 to 1994 declined 0.2% annually for women, but they concluded that there was no evidence of a decline in the incidence of hospitalization for a initial Ml among women (Rosamond et al., 1998). F. Prevalence Trends Heart disease prevalence is the number of persons with heart disease in the population. Examining the heart disease prevalence is a useful way for determining the needs of the population for medical facilities and manpower related to the disease. One study based on the National Health Interview Survey (NHIS) indicates that the nationally representative prevalence rate of overall heart disease increased about 50% for those aged 55-64 between 1972 and 1989 (Feinleib, 1995). In a later 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. study, researchers indicated that the prevalence rate of CHF among people who are aged 45 and over more than doubled from the period of 1976-1980 to the period of 1988-1991 (measured from Figure 3 in Hall, Ferrario, Moore, Hall, Flack, Cooper et al., 1997). The following table from the NHIS indicates IHD prevalence for age groups 55-64, 65-74, and 75 and older for three periods of 1979-1981, 1982-1984, and 1985-1987, The IHD prevalence for all age groups increased between 1979 and 1987. However, there are differential changes between periods. Most of the increase for those less than 75 occurred between the first and the second periods. The increase for the oldest age group was more continuous over the whole time (Schick & Schick, 1994, Table 2.1). There are different patterns of change by age in IHD prevalence among the three periods of 1979-1981, 1982-1984, and 1985-1987 for men aged 55 and over. In this table, three age groups of 55-64, 65-74, and 75 and older were used. All men who are aged 55 and over have increases in IHD prevalence from the period of 1979-1981 to the period of 1982-1984. Although those 65 and over show a continuous increase in IHD prevalence between 1982-1984, there is a decrease in IHD prevalence for age 55-64 men during the period. Women aged 55 and over also had an increase in IHD prevalence between 1979-1981 and 1982-1984. Although women aged 75 and over had a continuous increase in IHD prevalence between 1982-1984 and 1985-1987, women aged 55-64 and 65-74 experienced a decrease in prevalence of IHD between the periods. 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.1 Average Annual Number of Ischemic Heart Disease (IHD) Per 1,000 Persons Aged 55 Years and Over, by Sex and Age Sex and age 1979-81 1982-84 1985-87 Both sexes 55-64 58.8 75.7 70.1 65-74 105.0 115.7 114.1 75 and over 101.3 117.1 133.6 Men 55-64 77.6 107.5 98.9 65-74 133.9 149.9 150.7 75 and over 106.6 128.9 156.3 Women 55-64 42.1 47.9 44.7 65-74 82.8 89.4 85.3 75 and over 98.2 110.4 120.4 (Cited and modified from Schick & Schick, 1994, p. 118) Other researchers estimated that the annual decline in IHD prevalence was 0.6% for those 35-84 of both sexes between 1980 and 1990 (Hunink et al., 1997). However, in this study, the trend in IHD prevalence differs by age. The prevalence of IHD among the elderly increased in the period between 1980 and 1990. Moreover, it was pointed out that the absolute number of IHD cases increased for ail ages among men and women aged 35-84 over the same period. A study examined the age-adjusted prevalence of three categories of heart disease - Ml, angina pectoris, and CHF - for men aged 65 and over between 1971 and 1980 in the Kaiser Foundation Health Plan. The prevalence of Ml declined slightly from 15.0% in 1971 to 14.4% in 1980 for men. The prevalence of CHF also declined slightly from 11.6% in 1971 to 10.8% in 1980 for men. However, these changes are not statistically significant. The prevalence of angina pectoris increased significantly from 10.8% in 1971 to 13.1% in 1980 (Haan et al., 1996). 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For women, outcomes from the study of the Kaiser Foundation Health Plan showed age-adjusted prevalence of three categories of heart disease - Ml, angina pectoris, and CHF - for women aged 65 and over between 1971 and 1980. The prevalence of Ml slightly declined from 8.7% in 1971 to 7.7% in 1980 for women. The prevalence of CHF slightly increased from 8.8% in 1971 to 9.6% in 1980 for women. On the other hand, angina pectoris for women increased significantly from 9.8% in 1971 to 11.8% in 1980. None of the other changes for three heart disease categories for women are statistically significant (Haan et al., 1996). G. Trends in Risk Factors, Diagnoses, and Treatments Many researchers have pointed out that changes in primary and secondary risk factors at least partly explain the decline in IHD incidence and mortality and improved survival among men and women in the United States. Favorable trends in risk factors include declines in serum cholesterol level, smoking, level of hypertension, salt intake, and the increased use of anti-hypertensives, anti- lipidemics, and aspirin (Burke et al., 1989; Cutler & Sheiner, 1998; Haan et al., 1996; Hunink et al., 1997; McGovern et al., 1996; Sytkowski, Kannel, & D'Agostino, 1990). On the other side, there is an unfavorable trend in body mass index (BMI), which has increased in recent years (Burke et al., 1989; McGovern et al., 1996). Improved medical treatments and the introduction of new treatments for heart disease have contributed to improved heart disease survival. In recent decades, new therapeutic procedures such as thrombolytic therapy and angioplasty have been implemented and widely used. Furthermore, the use of the therapies such as angiography, coronary artery bypass surgery, and heparin use have increased 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Burke et al., 1989; Cutler & Sheiner, 1998; Doliszny, Luepker, Burke, Pryor, & Blackburn, 1994; Gillum et al., 1983; Haan et al., 1996; McGovern et al., 1996; Rosamond et al., 1998). Improvements in physical or occupational therapy have also affected heart disease survival (Cutler & Sheiner, 1998). In addition to more effective treatments, diagnostic practice also has improved in recent decades. For instance, testing for abnormal enzyme levels, creatinine phosphokinase (CPK) and creatinine kinase in cardiac muscle bands (CK-MB) is used more frequently to diagnose Ml (Burke, Edlavitch, & Crow, 1989; Haan et al., 1996). For the detection of angina pectoris, the use of treadmill monitoring has increased (Haan et al., 1996). H. Comprehensive Views of Mortality, Survival, Incidence, and Prevalence In sum, mortality from diseases of the heart has been declining for the last three decades for both men and women. In recent years, the mortality decline in IHD has been faster than ever. Men have experienced more rapid IHD mortality decline than women in the past, but recent studies show that currently the rate of decline in IHD mortality is larger for women than for men. In terms of survival, men had higher in-hospital and one-month survival rates than women. Survival rates for men and women have increased in recent years. Women have shown better improvement than men in recent years. Moderate length or four-year survival has improved for both sexes, and men showed more improvement than women. Men survived less than women in the long term, and long-term survival trends have not changed significantly. 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The trend in heart disease incidence is less clear than the trend in heart disease mortality for men and women. Although there was a pattern of decline in IHD incidence for men between the 1950s and the 1980s, some increase of IHD incidence was observed for women in the same period. One study states that there is no decline in incidence for men in recent years. It seems that there has been a period when women had larger declines in heart disease incidence than men, but recent studies show faster declines in heart disease incidence for men. However, these research findings are not all comparable because of issues of sampling, diagnosis, and examination methodology. The prevalence trend of heart disease is difficult to summarize for men and for women. The trend differs by the category of heart disease. For instance, the time trend of angina pectoris is different from those for Ml and CHF. The prevalence trend also differs by age and the period of examination for men and for women. Changes in mortality due to diseases of the heart are explained by changes in incidence and survival. After a review in the IHD-related literature, researchers conclude that roughly 1/3 of the reduction in IHD mortality is attributable to a reduced incidence of IHD, and survival after a IHD incident accounts for the remaining 2/3 of the improvement in IHD mortality (Cutler & Sheiner, 1998). Heart disease mortality decline is due to favorable changes in risk factors, health behaviors, and medical treatments. Reduction of IHD incidence can be attributed to primary prevention resulting from favorable trends of risk factors. Improvement in survival is attributed to secondary prevention including treatment and health behaviors after an event of IHD. Since the late 1960s, 1/2 to 3/4 of the health improvement is the result of primary and secondary prevention, rather than advances in high-tech intensive medicine (Cutler & 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sheiner, 1998). Another study concludes that 25% of the decline in IHD incidence in the 1980s was attributable to primary prevention and more than 70% to the effects of reductions in risk factors or improvements in the treatment of patients with IHD (Hunink et al., 1997). The role of acute treatment and secondary prevention in improved survival varies over time. Older literature examining changes in IHD mortality in the late 1960s and 1970s found that secondary prevention was more important than treatment, but more recent studies from the 1980s place more weight on intensive treatments (Goldman et al., 1984; Hunink et al., 1997). There have been studies that produce combined observations of mortality, survival, incidence, and prevalence of heart disease. In terms of IHD, a strong increase in prevalence has been observed along with a survival gain while there was a small decline in incidence in the Framingham Heart Study (Sytkowski et al., 1990). Different mortality rates in age groups result in different changes in prevalence with age. The prevalence of IHD is higher at older ages because IHD mortality decline was proportionately greater among people aged 30-74 from 1970 to 1985 (Burke et al., 1989). The changes by age in heart disease prevalence may be due to different changes by age in the pattern of incidence. For instance, Ml incidence was delayed to a later age among those aged 35-74 between 1978 and 1985 in the rural Pee Dee area of South Carolina (Keil, Gazes, Litaker, Saunders, Weinrich, Baroody et al., 1989). Another study demonstrated changes in Ml prevalence profile by age for the Dutch population. By finding different outcomes in incidence and mortality among younger, middle-aged, and older adults, researchers implied that prevalence rates would decrease among the young and middle-aged but increase among the elderly. Using simulation models, it was also found that decreasing incidence along with the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. increase in survival results in lower prevalence rates in the young and increasing prevalence rates in older age groups (Bonneux, Barendregt, Meeter, Bonsel, & Van derMaas, 1994). Characteristics of heart disease differ somewhat for men and women. Women on average were 10 years older than men at an initial manifestation of IHD and 20 years older at the event of an Ml (Wenger, 1992). Because women have Ml at older ages, they are more likely than men are to die from them within a few weeks (AHA, 1999). As a result, 42% of women who have an Ml die within a year as compared with 24% of men (AHA, 1999). Thus, IHD is a serious problem for women as well as men. I. Research Questions Past studies indicate trends in mortality, survival, incidence, prevalence, and risk factors for heart diseases. However, these patterns vary by heart disease category, sex, age group, and time. A comprehensive examination of trends in all aspects of diseases of the heart has not performed. It is necessary to have a consistent examination of diseases of the heart. From examinations of the time trends in heart disease using the same age range and length of time, we will clarify how mortality and morbidity from heart disease are related at the population level. Then, the pattern of disability among those with heart disease will be clarified. On the other hand, we do not know how current trends in heart disease affect an individual life's expectations. The trajectories of an individual life with and without heart disease through the later years of life and death are not well-documented. By calculating transition rates into heart disease and to death, years of life with and 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without heart disease and prevalence rates for diseases of heart in the middle-aged and older populations are produced. To clarify these issues, the following research questions are addressed in this study. (1) Trends ■ What is the trend in mortality from heart disease in recent years? ■ Does it differ from the long-term trend? ■ Does it differ by type of disease, age, or sex? ■ What is the trend in heart disease prevalence in recent years? ■ Does it differ from the long-term trend? ■ Does it differ by type of disease, age, or sex? ■ What is the trend in length of time since onset of heart disease among those who have heart disease? ■ Does it differ by age or sex? ■ What is the trend in likelihood of death among those with and without heart disease? ■ Does it differ by age or sex? ■ What is the trend in disability status among persons with heart disease? ■ Does it differ by types of limitation, age, or sex? (2) Multistate model ■ What is length of life with and without heart disease? ■ Does it differ by age or sex? ■ How do morbidity and mortality transitions affect the years of life and implied prevalence of heart disease? ■ Does it differ by age or sex? 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER III MATERIALS AND METHODS A. Data Sources, Samples, and Measures In this study, four large data sets are used for a comprehensive examination of the trends in heart disease in the United Status. Trend analyses use data sets from the Vital Statistics of the United States and the National Health Interview Survey (NHIS). Multistate life table estimates of years of life with and without heart disease and implied prevalence rates of heart disease are used in analyses of transition rates from analyses of the Health and Retirement Study (HRS), and the Assets and Health Dynamics of the Oldest Old (AHEAD). Descriptions of these data sources, samples, and measures are provided in this chapter. 1. Vital Statistics of the United States - Mortality In this study, mortality data are derived from the Vital Statistics in the United States. In the United States, state laws require death certificates to be completed for all deaths, and Federal law mandates national collection and publication of deaths and other vital statistics data (Anonymous, 2000b). Mortality data from the Vital Statistics of the United States are therefore available to the public. Coding of medical information on the death certificate follows World Health Organization (WHO) rules specified in the International Classification of Diseases (ICD). The Vital Statistics data of the United States provide mortality data with a parallel classification of causes of death from the International Classification of 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Diseases (ICD). Death rates from two heart disease categories are used. One is overall heart disease which is coded as "Diseases of Heart (390-398, 402, 404-429)", and another is heart attack which is categorized as "Acute Myocardial Infarction (410)" in the Vital Statistics (Table 3.1). Table 3.1 The Criteria and Code Number of Heart Disease in the International Classification of Diseases: 9th Revision (ICD-9) in the Vital and Health Statistics of the Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Center for Health Statistics Diseases of Heart (390-398,402,404-429) Rheumatic fever and rheumatic heart disease (390-398) Hypertensive heart disease (402) Hypertensive heart and renal disease (404) Ischemic heart disease (410-414) Acute myocardial infarction (410) Other acute and subacute forms of ischemic heart disease (411) Angina pectoris (413) Old myocardial infarction and other forms of chronic ischemic heart disease (412,414) Other diseases of endocardium (424) All other forms of heart disease (415-412,425-429) In this study, death rates from 1968 to 1996 (a total of 29 years) are collected for men and women 50 years of age and over. The death rates are reported based on 100,000 population. During this period, the ICD coding changed from the eighth revision to the ninth revision. However, there is no effect in the heart disease categories used in this study. Death rates for age 50 and over are given for 5-year 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. age intervals for men and women. These eight age categories are 50-54, 55-59,60- 64, 65-69, 70-74, 75-79, 80-84, and 85 and over. For trend examinations, change is examined over 1968 to 1996,1982 to 1996, and 1987 to 1996. These examined periods are for the longest time trend with a consistent heart disease classification, for an adjusted period with other time trend studies, and for the last decade. 2. National Health Interview Survey (NHIS) The National Health Interview Survey (NHIS) provides information on the health of the civilian noninstitutionalized population who are 18 years of age and over in the United States. This is an annual cross-sectional household interview survey, and sampling and interviewing are continuous throughout each year since 1957. NHIS data are collected annually from approximately 43,000 households that include approximately 106,000 persons. The annual response rate of the eligible household sample in NHIS is more than 90 percent. Under the operation of this data collection program by the National Center for Health Statistics (NCHS), the Department of Health and Human Services (DHHS) uses collected data for the purpose of monitoring trends in disease, illness, disability, and achievement of national health objectives. The Survey includes a variety of health issues such as acute conditions, injuries, restriction in activity, prevalence of chronic conditions, limitation of activity due to chronic conditions, respondent-assessed health status, and the use of medical services. Questions about heart disease are included in one of the six sets of specific chronic disease questionnaires that are given to one of six subsamples of the entire sample (Anonymous, 2000a). The National Health Interview Survey (NHIS) provides information on heart disease prevalence for the noninstitutionalized population in the United States. Since 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. heart disease questions are included as one of six sets of chronic conditions, they are asked of only one-sixth of the whole sample each year. Respondents are asked to indicate the presence or absence of a variety of the categories of heart disease at the interview. The NHIS diagnostic categories are closely classified according to the standard International Classification of Disease (ICD) groupings (Table 3.2). Table 3.2 The Criteria and Code Number for Diseases of Heart in the National Health Interview Survey (NHIS) - Diagnostic Record (B) 2-digit 3-digit Recode Recode Category NHIS Diagnostic Code Inclusions 25 26 27 28 250 251 259 260 269 271 280 281 289 Rheumatic fever and rheumatic heart disease Acute rheumatic fever Chronic rheumatic heart disease Residual Hypertensive disease Hypertensive heart disease Residual Ischemic heart disease Ischemic heart disease Disease of pulmonary circulation and other Forms of heart disease Pulmonary embolism Cardiac dysrhythmias Residual Selected variables for the study Diseases of the heart Ischemic heart disease (IHD) 390, 392-398, 399A 390, 392 393-398 399-A 401-405 402, 404 401,403, 405 413.414 413.414 415-417, 420-428, 429.0-5, 8, 9 415.1 427 415.0, 416,417, 420-426, 428, 429.0-5. 8, 9 (390-398,402,404-429) (413, 414) 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The National Health Interview Survey (NHIS) provides heart disease prevalence information for the U.S. noninstitutionalized population for a variety of types of heart disease. In this study, two kinds of heart disease information are included. Heart disease variables are categorized as overall heart disease (502-507) and Ischemic heart disease (IHD) (502). Men and women age 50 and over from the NHIS data are included in this study. The study sample consists of 29,717 men (average age is 63.8 with a 9.7 standard deviation) and of 37,244 women (average age is 65.3 with a 10.3 standard deviation). The distribution of the study sample by year is shown in Table 3.3. Since the NHIS sample is divided into six subgroups for different chronic disease lists, a weight variable that contains a sample weight and an annual weight is provided in the NHIS data for making the sample representative of annual event. Table 3.3 Eligible Sample Size for Heart Disease Study from the National Health Interview Survey (NHIS) Among Those Aged 50 and Over, by Sex, by Year Men___________ Women Total 29,717 37,244 Year 1982 2,042 2,432 1983 2,048 2,548 1984 2,018 2,564 1985 1,711 2,144 1986 1,180 1,492 1987 2,294 2,971 1988 2,334 2,923 1989 2,109 2,685 1990 2,222 2,806 1991 2,282 2,902 1992 2,249 2,833 1993 2,062 2,518 1994 2,246 2,790 1995 1,804 2,279 1996 1.116 1.357 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Timing of heart disease onset is also reported. In NHIS, the timing of the onset of heart disease is originally divided into five categories (Table 3.4). These categories are simplified into two categories with either one of two different time criteria: "onset within one year" and "onset within five years." Table 3.4 Heart Disease Related Variables in the National Health Interview Survey (NHIS) Onset variable Onset 1. During the past 2 weeks 2. Over 2 weeks to 3 months 3. Over 3 months to 12 months 4. Over 1 year to 5 years 5. Over 5 years Activity related variables Activity Limitation Status 1. Unable to perform major activity 2. Limited in kind/amount major activity 3. Limited in other activities 4. Not limited Activity Lim itation Status M easured b y “Activity to W ork” (limited up to age 69) 1. Unable to work 2. Limited in kind/amount of work 3. Limited in other activities 4. Not limited Heart disease information in the NHIS can be also linked to information on activity limitation. Two activity limitation variables are extracted for determining disability among those with heart disease from the data set. These are “activity Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. limitation status” and “activity limitation status measured by ability to work.” Each variable contains four categories (Table 3.4). For analysis, these categories are simplified into two categories with either one of two different criteria: unable to perform major activity (or work) or not and no limitation in performing activity (or work) or not. These focus on severe limitation and being free of limitation. In addition to annual health survey data, the National Center for Health Statistics provides the National Health Interview Survey (NHIS), Multiple Cause of Death Data File for survey years 1986-1994. This file links information from NHIS respondents to the National Death Index (NDI). Linkage of these two data sets provides a longitudinal component and allows for ascertainment of vital status of past respondents. Beginning with survey year 1986, the NHIS Multiple Cause of Death Data File for survey years 1986-1994 contains vital status through the end of 1995 with the criteria of ICD-9 cause of death and the NCHS cause of death codes (Anonymous, 2000c). Death from heart disease are coded 320 though 410. Trend analyses for the NHIS data are based on the years since 1982 and 1987. For the death trend analyses using the NHIS Multiple Cause of Death Data File for survey years 1986-1994, two variables indicating death during a number of years from the year of interview are created. One is for mortality before the end of the next year after the interview year (1.5 averaged years), and another is for mortality within the end of the following year (2.5 averaged years). For use of the 2.5- year duration, data from 1994 are eliminated. 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3. Health and Retirement Study (HRS) and Assets and Health Dynamics of the Oldest Old (AHEAD) The Health and Retirement Study (HRS) is a nationally representative study conducted by the University of Michigan with funding from the National Institute on Aging (NIA). This study is a panel study with a longitudinal design. The sample in HRS is representative of the noninstitutionalized US population aged 51-61 at baseline. These baseline respondents are the birth cohorts of 1931-1941 and their spouses and consist of over 7,600 households with 12,654 individuals. The Health and Retirement Study has several major objectives such as the explanation of the antecedents and consequences of retirement, examination of the usage and accumulation of wealth, monitoring of health, income, and wealth and their interrelations. Follow-ups are conducted every second year with proxy interviews for deceased individuals who participated in a previous interview. The first data collection was conducted from April, 1992 through March, 1993, and the second data collection began in April, 1994 and concluded in December, 1994. The third collection began in May, 1996 and concluded in February, 1997. The Health and Retirement Study oversamples Hispanics, Blacks, and Florida residents. Weighting is used to adjust the sample to the noninstitutionalized US population (Anonymous, 1999c). Public release data sets are available through the website of the University of Michigan (http://www.umich.edu/~hrswww/). The Assets and Health Dynamics of the Oldest Old (AHEAD) is a national panel study also funded by the National Institute on Aging (NIA). This study is an auxiliary survey of The Health and Retirement Study. The AHEAD Study features a longitudinal design which calls for re-interviews of the sample every two years. The study also includes respondents and his/her spouse and concentrates on special 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concerns of the family and household dynamics. The first data collection was conducted from October, 1993 through July, 1994, and the second collection began in November, 1995 and concluded in June, 1996. At baseline, this study sample represents the non-institutionalized elderly who are aged 70 and over who were born in 1923 or before, as well as their spouses. When possible, respondents report their own health status. When this is not possible, interviews are done by proxy. At follow- up, proxies also respond for those who are institutionalized or deceased. The AHEAD Study oversamples Hispanics, Blacks, and Florida residents. Weighting is used to adjust the sample to the noninstitutionalized US population (Anonymous, 1999a). Public release data sets are available through the website of University of Michigan (http://www.umich.edu/~hrswww/). The AHEAD Study is now currently joined with the HRS in a single data collection effort known as Health, Retirement and Aging (HRA'98). This single data set includes the fourth wave of HRS and the third wave of AHEAD. Data from Wave I and Wave II for the Health and Retirement Study (HRS) and the Assets and Health Dynamics of the Oldest Old (AHEAD) are merged for the present study. Using the HRS/AHEAD Tracker File, known deaths are identified in the sample. Participants missing from the second interview are eliminated from the study. Those who were age 50-97 at the first wave are selected for this study. Because heart disease is the focus of this study, individuals who did not answer heart disease questions are also omitted from this study. The final sample size in this study is 8,050 for men (average age is 64.4 with a 10.6 standard deviation) and 10,110 for women (average age is 65.7 with an 11.6 standard deviation). These numbers include individuals who were deceased between Wave I and Wave II. Detailed sample selection is described in Table 3.5. 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.5 Heart Disease Study Sample for the Merged Data from the Health and Retirement Study (HRS) and the Assets and Health Dynamics of the Oldest Old (AHEAD) - Numbers of Sample in Each Step towards the Final Sample Selection Number H R S A H E A D H R S /A H E A D Step 1 Participation in Wave I (Original) Total (Deaths) 12,652 (241) 8,222 (836) 20,874 (1,077) Step 2 Omit missing individuals (no identification of dead/alive at Wave II) Total (Deaths) 11,555 (241) 7,788 (836) 19,343 (1,077) Step 3 Omit age < 50,97 < age Total (Deaths) 10,419 (225) 7,749 (823) 18,168 (1,048) Step 4 Omit invalid cases for heart disease for Heart Disease Total (Deaths) 10,418 (225) 7,742 (821) 18,160 (1,046) Detailed criteria for each step Each stage follows the previous step Step 1 All Wave II participants Step 2 Wave I participants but omit missing people at Wave II Missing people: who cannot be identified alive/missing Step 3 Keep only selected age group Step 4 Omit invalid cases for Heart Disease Invalid: who did not answer the question at Wave I who answered 'No Problem' at Wave 1 and Missing at Wave II (no alive/dead specification) Detailed manipulation for invalid cases: Omit this category -> Keep this category -> Keep this category -> Wavel Wave II Death status No Missing, Do not know, Alive or maybe Refused dead Yes Missing, Do not know, Alive Refused No/Yes Any Dead 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In the merged data from HRS and AHEAD, two heart disease categories are selected for study. One category is overall heart disease, and another category is heart attack. Overall heart disease includes heart problems such as heart attack, coronary heart disease, angina, congestive heart failure, or other heart problems. Participants were asked whether they had been told about any of these heart problems by a doctor at the Wave I interview. Heart attack involves myocardial infarction. Participants were asked whether they had this heart problem by a doctor before or in the last five years at the Wave I interview. At the Wave II interview, participants were asked whether they had been told that they had heart disease by a doctor (overall heart disease or a heart attack) since the date of the Wave I interview. These questions are listed in Table 3.6. Table 3.6 Heart Disease Questions in the Health and Retirement Study (HRS) and the Assets and Health Dynamics of the Oldest Old (AHEAD), Wave I and Wave II Wave I Overall Heart Disease Has a doctor ever told you that you had a heart attack, coronary heart disease, angina, congestive heart failure, or other heart problems? (HRS) Has a doctor ever told you that you had a heart attack, coronary heart disease, angina, congestive heart failure, or other heart problems? (AHEAD) Heart Attack (Myocardial Infarction: Ml) Did you have a heart attack or myocardial infarction? (HRS) Have you had a heart attack (include myocardial infarction) in the past five years? (AHEAD) Wave II Overall Heart Disease Since (Wave I date) has your doctor told you that you had a heart attack, coronary heart disease, angina, congestive heart failure, or other heart problems? (HRS) Since your interview in (Wave I date) has a doctor told you that you have had a heart attack, have coronary heart disease, angina, congestive heart failure, or other heart problems? (AHEAD) Heart Attack (Myocardial Infarction: Ml) Did you have a heart attack or myocardial infarction since (Wave I date)? (HRS) Have you had a heart attack or myocardial infarction since (Wave I date)? (AHEAD) 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Persons in the sample can be categorized as "No heart disease," "Overall heart disease without heart attack," and "Heart attack" for Wave I. Another category "Death" between the Waves is added to those heart disease categories at Wave II for analyses. For descriptive purposes, age is defined using a five-year interval for prevalence and incidence. This creates ten age categories: 50-54, 55-59, 60-64, 65- 69, 70-74, 75-79, 80-84, 85-89, 90-94, and 95 and over. Age is treated as an interval variable for transition analyses. Duration between two waves is set as two years. Both data sets of HRS and AHEAD oversample Mexican-American Hispanics, African-Americans, and households in the State of Florida. Weighting factors are already created and provided to adjust for the differences in age, geographic factors, and race group sampling rates (Anonymous, 1995a, 1995b). Since each weighting factor is data set-specific for HRS and AHEAD, the weight for HRS is used for the age group of 51-61, and the weight for AHEAD is used for those aged 70 and over. Among these, 59 individuals were originally assigned zero in their weight value. Individuals of other ages are weighted as assigning a value of one (Table 3.7). 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.7 Weight Variables for the Health and Retirement Study (HRS) and the Assets and Health Dynamics of the Oldest Old (AHEAD) Data Sets V15 (Original for HRS): Weight value for person level analysis WTRNORM (Original for AHEAD): Centered respondent weight M_WT_R (new variable for HRS/AHEAD merged data (SAS program for weight value for the HRS_AHEAD merged data) if 51<=age<=61 then m_wt_r=newwgt1; else if 70<=age then m_wt_r=wtrnorm; else m_wt_r=1; (1) With Original Sample Variable N Mean Std Dev Minimum Maximum V15 12,652 1,864.31 1,349.90 NEWWGT1 12,652 0.77656 0.56228 WTRNORM 8,222 0.90574 0.46019 0 12,312.84 0 5.12877 0 2.85681 (2) With Selected Sample (51<=age<=61 for HRS, 70<=age for AHEAD) Variable N Mean Std Dev Minimum Maximum V15 9,902 2,350.81 1,074.37 NEWWGT1 9,902 0.97920 0.44751 WTRNORM 7,512 0.98930 0.38552 0 12,312.84 0 5.12877 0 2.85682 (3) New Weight Variable for HRS_AHEAD merged data Variable N Mean Std Dev Minimum Maximum M_WT_R 20,644 0.99726 0.37317 0 5.12877 Among 20,644 individuals, 59 individuals (37 age 51. 9 age 61, and 13 age 70) were originally assigned zero in their weight value. Age V15 WTRNORM Frequency 5 1 0 61 0 70 0 37 9 13 Total 59 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. B. Analytical Methods A variety of analytical methods are used in this study. The methods for trend analyses include ordinary least squares (OLS) regression models, log-linear regression models, and logistic regression models. For the creation of multistate life tables, hazard models and multistate models are used. 1. Analytical methods for trends in heart disease with population (non individual) data Trends in heart disease can be presented in a variety of ways. Data obtained from the Vital Statistics of the United States, for instance, are age-specific rates and not for individuals. These are, therefore, limited in the types of statistical analyses that can be performed. One of the substantial approaches to examine heart disease trends is graphical. This approach gives an overall view of the heart disease trend and can be used for study preliminary to statistical analyses. This approach is used to provide an overview of the trend in heart disease. One of the statistical approaches in examining mortality from heart disease is a regression model, which enables us to identify the statistical significance of changes in rates of heart disease. A linear regression model can be used to analyze death rates that are treated as continuous dependent variables. For example, the annual change in morality rates was calculated using linear regression to show the absolute changes for the time period of 1968-1988 (Armstrong, Wing, & Tyroler, 1995; Sytkowski et al., 1996). 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In this study, this method is used for each age group. The regression equation based on ordinary least square (OLS) is expressed as: Y = a + bX, (Eq. 3.1) where ‘a’ is an intercept, 'b' is a coefficient representing number of years since the beginning of the period. For instance, parameter estimates are calculated from regression analyses of the mortality rates from heart problems for the age group 50- 54 on the number of years since 1982 for both men and women. Another way to examine the heart disease mortality change is through a calculation of the annual percent change with 95% confidence intervals (CIs), using a general linear models (GLM) procedure. Annual percentage change between two years is used in many studies (Clayton & Hills, 1993, p. 197-198; Kleinman, 1986; Lee & Carter, 1992; Pollard, Yusuf, & Pollard, 1974, p. 21), including those examining heart disease (Armstrong et al., 1995; Demirovic et al., 1993; Rosamond et al„ 1998). Average annual percent change in heart disease morality rates will be calculated using this procedure. In this case, all mortality rates between two years are included in the model. This model involves a natural logarithmic (In) transformation of all mortality rates between the two years of interest. For each age group, the percentage change of heart disease rates is expressed as: Ln(Y) = a + bX, (Eq. 3.2) where ‘a’ is an intercept, ‘b’ is a coefficient of the duration in years from a certain year. 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2. Analytical methods for trends in heart disease with individual data With individual data, we can use logistic regression models to examine trends in heart disease. Mortality, prevalence, incidence, and other heart-related trends can be examined using these models. The logistic regression models are appropriate when a dichotomized variable is the dependent variable. By treating mortality, prevalence, incidence, and other heart-related data as dichotomized dependent variables, logistic regression models can be used to reveal time trends in these dependent variables by including an indicator of time since the beginning of the period or year as an independent variable. In all models, a variable representing chronological age is also included. The logistic regression model is expressed as: LogY = a + b,Xi + b2 X2 , (Eq. 3.3) where X,’ and 'X2 ' represent age and the number of years from the beginning of the period. For instance, the prevalence of heart disease is indicated by determining whether a person has it or not, and this can be examined by logistic regression models. Odds ratios (ORs) are calculated from these logistic regression analyses indicating the effect of dependent variables. To examine interaction effects, some models may add an interaction term for age and the duration in years. This allows us to test whether the trend differs by age. Regression equations with the interaction term are expressed as: LogY = a + ^X , + b2 X2 + bsX^z (Eq. 3.4) where ,X1 X2 ’ represents an interaction term of age and the time in years from the base year. The interaction in this study consists of these two variables with an 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interval scale. The outcomes can be interpreted differently from variables with other scales (Silverstein, Chen, & Heller, 1996). 3. Event history analysis Event history analysis is used to examine a longitudinal record of the timing of specified events among a group of persons. In essence, the risk (or hazard) of an event is expressed, in the form of a regression equation, as a function of a specified set of explanatory variables. The regression coefficients are then estimated to maximize the likelihood of observing what has actually been observed. In a discrete time hazard approach, status differences between the beginning and the end of intervals are used to determine transitions in heart-related health status. Thus the data impose the assumption that only one transition occurs within each two-year interval; such an assumption very likely results in underestimating transitions. Within each observation interval, events (i.e., a death or a health status transition, and censoring) are assumed to occur in the middle of the interval, and exposure is adjusted to produce central rates. Thus the transition rates are assumed to be constant within intervals, although rates may vary across intervals (Allison, 1995). A(x) = lim P x+n / n, (Eq. 3.5) n-*0 where P(x, n) represents the probability that a transition from one state (e.g., no heart disease) to another state (e.g., heart disease) occurs in the interval x to x+n, given that the person is in a specific state at exact age x (Crimmins, Hayward, & Saito, 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1994; Saito et al., 1999). In this study, n equals 2. These rates are estimated on the basis of a model which assumes that the process depends only on age. The general form of the model in this study is expressed as; M -(x ) = exp(p0 + p,x) (Eq. 3.6a) or, equivalently, ln(x) = p0 + pix. (Eq. 3.6b) In estimating these models, age is treated not as a continuous variable with a single age but as a 2-year interval by combining each two years together ranging from 50 to 97 years. This treatment is appropriate for the feature of a two-year interval between data collections in this study. For example, age 50 and 51 are combined and set as age 50, age 52 and 53 as age 52, and so on. Models of this form are analogous to piecewise exponential models used in discrete-time hazards analyses (Allison, 1995). Survival analysis deals with dichotomous variables such as dead or alive. Unlike logistic regression analysis, survival analysis incorporates methods for dealing with censoring and time-dependent covariates. This study uses the survival analysis SAS program “LIFEREG" because it estimates regression models with censored, continuous-time data. Censoring can be of three basic types. The first one is right censoring which occurs if a heart problem is observed after the interview. The second one is left censoring that occurs when a heart problem already occurred before the interview. The third one is interval censoring that occurs if a heart problem occurred sometime between the interviews (here Wave I and Wave II), but we don't know when it exactly happened. LIFEREG accommodates left censoring and interval 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. censoring (Allison, 1995). The hazard function is measured as a rate from survival analysis to estimate the risk that an event will occur at a certain time. In this study, two events are measured: onset of heart disease and death. Using the outcomes (rates) of the hazard model, age-specific transition schedules for the onset of heart disease (from no heart problem to heart problem) and death (from no heart problem to death and from heart problem to death) are estimated. Several researchers illustrate flow diagrams that include states and possible transitions across states (Crimmins et al., 1994; Saito et al., 1999; Schoen & Land, 1979). In this study, six hazard models are created and estimated based on a state space for heart disease as defined in Figure 3.1. After running hazard analyses, the parameter estimates from each model are used to obtain schedules of predicted age- specific transition rates, Mx (Crimmins et al., 1994). Figure 3.1 Status of Disease in the Heart Disease Model Death No Heart Disease Heart Disease (exci. Heart Attack or Myocardial Infarction: Ml) Heart Attack or Myocardial Infarction (Ml) 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4. Multistate Life Table Multistate models are known as increment-decrement models. They recognize more than one state and allow a population to move from one state to another state. Although the basic life table is a simple type of hazard model as it analyzes mortality events whose risk varies with the explanatory variable age, multistate models can include additional explanatory variables as they add states (Crimmins et al, 1994; Saito et a., 1999; Hayward, Crimmins, & Wray; 1994). The muitistate life table is estimated by transforming the transition rates into a matrix of transition probabilities from which all other life table functions are derived in the multistate model. Multistate models can be used in the analysis of many issues, including different states of heart disease. Multistate estimates of health expectancy were introduced in the 1980s and developed in the 1990s. Recently, there has been great interest in the estimation of health expectancy among both policy makers and members of the research community (Saito et al., 1999). Multistate estimates of health expectancy are derived from a set of transitions between multiple states of life to death. Because these measures incorporate indicators of both mortality and morbidity, they are highly appropriate as summary measures of the effects of changing health status and mortality schedules in populations. Multistate estimates of heart-related life expectancy are derived from a set of transitions to and from states of life with heart disease, without heart disease, and death. They describe state-specific incidence, implied prevalence, and mortality from the states of having and not having heart disease. These transition estimates are derived from longitudinal studies. Heart-related life expectancy computed by this 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. measure is an estimate of the average length of heart disease-related life for a cohort that lives all of its remaining life exposed to current conditions of mortality and morbidity. This measurement provides an indicator of the health of individual life cycles as well as population health expected if current conditions continue. Multistate methods can be used to produce two types of life tables. One is a life table for the population as a whole, and another is a life table for population subgroups. These are called population-based life table and status-based life tables, respectively. In this study, both population-based and status-based multistate life tables are used. Population-based multistate life tables describe potential life cycle events for the whole population. Population-based multistate life tables indicate the life expectancy and the health state of the entire life table population as it is distributed across health states at all ages of the life table. Population-based tables also allow us to compute the implied age-specific prevalence of health states in the life table population. The prevalence rates of health states are a consequence of flows both into and out of the various heart-related health states and of the individuals remaining in each state. To estimate population-based multistate life tables, the radix population is set at 100,000 for the initial distribution according to the observed prevalence in each state for the U.S. population aged 50 (Crimmins et al., 1994; Crimmins, Hayward, & Saito, 1996). In comparison, status-based multistate life tables can be used to measure the prospective life cycles of those who reach specified ages in different heart-related health states. Status-based life tables indicate the heart disease health status and life expectancy of a group that enters the life table in one particular heart disease 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. health state rather than that of the entire population (Saito et al., 1999). To estimate status-based multistate life tables, the 100,000 initial population is in one state. In this study, the 100,000 of people are set at each of these states: without heart disease, with heart disease (without heart attack), and with heart attack. This method has been used by some researchers (Rogers, Rogers, & Branch, 1989,1990). Those who are interested in information on the health of a population as a whole should pay attention to outcomes from the population-based multistate life table. On the contrary, those who are interested in a particular subgroup or heart disease-related state, can satisfy their interests with the status-based multistate life table. a. Estimated years of life with and without heart disease To construct life tables for the heart disease model, the matrix of transition rates, Mx, is used to calculate a transition probability matrix, Px (Crimmins et al., 1994; Saito et al., 1999; Rogers & Ledent, 1976). The relationship between the matrix of transition rates and the transition probability matrix is as follows; Px = (I + 1/2 n • Mx )'1 (I -1/2 n Mx ). (Eq. 3.7) For instance, the transition probability matrix, Px, indicates the probability that a person in one state at age x will move into another state at age x+n. Survivorship values can be calculated from the computed transition probability matrix (Hayward & Grady, 1990; Rogers, 1975; Schoen & Woodrow, 1980). The survivorship formula is as follows: 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ix+n - Ix ' P x - (Eq- 3.8) Ix + n represents a survivorship matrix which represents the number of persons whose movement is from one state at age x to another state at age x+n. I* stands for a survivorship matrix similar to !*♦„. The element, lx , however, represents the number of persons in two different states at exact age x. Its diagonal element at another state represents the number of survivors in initial state at age x. After the calculation of the survivorship values, the expected magnitude of flows between states for all ages is computed. The computation includes the expected number of transfers from one state to another state above a certain age or between ages. Calculation of transfers between states for each age, Dx , is based on the matrix of transition rates, Mx , and a matrix of person-years lived, Lx (Crimmins et al., 1994; Saito et al., 1999). This calculation is as follows: Dx = Lx M (Eq. 3.9) where assuming linearity in the gross flow function: Lx = 1/2 • n • (lx + U ). (Eq. 3.10) This allows the calculation of state-specific life expectancy estimates. Life expectancy in one particular state at age x is thereby calculated as: ex = Tx/lx, (Eq. 3.11) 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where T* represents the number of person-years lived in a particular state beyond age x by the age cohort that survived to age x. ex stands for the life expectancy in one state at age x (Crimmins et al., 1994; Saito et al., 1999). Life expectancy here is the average number of years expected in that specified state. Total life expectancy is the sum of life expectancy in each of the heart-related states. b. Estimated prevalence rate In addition to heart-related life expectancy, the multistate method provides information on the implied prevalence of each health state. The prevalence is measured through the processes of health change leading to the estimated years in specified states. The prevalence rate consists of the stock of persons in the particular state of interest: inflows from some states, and outflows toward other states (Schoen, 1988). In the case of heart disease, the prevalence rates in the heart disease state is a consequence of flows both into and out of the state of heart disease. One of the uses of the multistate life table is to clarify the demographic forces that underlie the prevalence measures of health status in the population. The connections between the prevalence rates and the incidence rates can be simply illustrated by adopting a life table model. These connections consist of three states: without heart disease (o), with heart disease (h), and death (§). The following formula explains the heart disease prevalence rate in terms of life-table notation (Crimmins et al., 1994; Schoen & Woodrow, 1980; Saito et al., 1999). 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. \ x PREV(h)x = -------— h L x + ° L x (Eq. 3.12) In this expression, h L x refers to the person-years lived in the state of heart disease during the specified time period from x to x+n. ° L X refers to the person-years lived in the state of without heart disease (Crimmins et al., 1994; Saito et al., 1999). The person-years lived in some state is expressed as: The following two equations denote the person-year values and ultimately the heart disease prevalence rates. The first equation explains that the number of persons in the state of no heart disease at age x+n equals the number of persons in the same state at age x (i.e., no heart disease) minus the number of movers from the state of no heart disease to the state of heart disease, minus those persons in the state of no heart disease at age x who died during the interval. The second equation explains that the number of persons in the state of heart disease at age x+n equals the number of persons in the same state at age x (i.e., heart disease) plus the number of movers from the state of no heart disease to the state of heart disease, minus those persons in the state of Lx = 1/2 • n • (lx + U ). (Eq. 3.13) (Eq. 3.14) (Eq. 3.15) 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. heart disease at age x who died during the interval. These two equations contribute to the numerator and denominator of the prevalence rate (Crimmins et al., 1994; Saito etal., 1999). 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER IV RESULTS - TRENDS IN HEART DISEASE This chapter examines trends in heart disease for men and for women. These trends include mortality, prevalence, onset, survival, and limitations among those with heart disease. These trends are examined using multiple sources of data. A. Mortality Trends Trends in mortality from overall heart disease and acute myocardial infarction (AMI) are examined for men and women. First, to examine heart disease mortality trends, mortality rates of heart diseases are illustrated separately for different age groups of men and women in graphs. For more detailed trend examinations, average annual changes and average annual percent changes in mortality rates are statistically analyzed for different age groups of men and women for three periods. These three periods are: (1) from 1968 to 1996 (29 years), (2) from 1982 to 1996 (15 years), and (3) from 1987 to 1996 (10 years). 1. Trends in overall heart disease mortality rates Men Death rates from overall heart disease in 5-year age groups for men for 1968 through 1996 are shown in Table 4.1. These trends are graphed in Figure 4.1. The death rates from overall heart disease are much higher for older age groups than younger groups. From the figure, it is clear that at least at the older ages, death rates have been decreasing relatively steadily over time. 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T ab le 4 .1 D e a th R ates fo r Diseases o f Heart fo r Men, b y 5-Year A g e Group, Vital Statistics 1968-1996 IC D coding: 390-398, 402,404-429 I I ) 1 - in 0 ) m oo eg O 00 rv 00 to 03 eg m h* 0 - 03 o C O o o C O 03 8 0 ) cd a i T “ eg cd eg o i in s i th * o O) 00 03 d (0 eg o in o 03 O o 03 in 0 ) eg q N . co O) to o 10 oo m o 00 eg (O eg o o> 12 8 eg h* 8 00 00 o ’ m’ in y ’ in’ Y Y V V Y Y V y " V cn C O C O c o ’ < o " c o " c o ’ q s . < < q c p * i — c o c o i n o' S O ) n m « v Y(0CMmiA (Oi-coc»)r*ooffl^-Y;^»“ flpr-iortcopscoNcqrtflpo«»-o)a) Kegdddr^cdd^r^0: ddcb^;cba>dincde«-o:cddego><df'*^ O ) o o o O ) 0 ) q o n i n i n i n ^ ^ n o o o 9 $ ® r s . < o i n c o e g c n o , o_ c o ’ o ’ o ’ o ’ o ’ o ’ o* o * o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o* e g * e g ’ e g * e g ’ e g ’ e g * e g * e g ’ e g * e g * e g * e g ’ cocMNOtDOT-ifiYegos a)«fgfs.ino^tf)egh»oo> oocooof^r^r-comin^^co o^ooaococoegrs. ^oiricpqjqjr-cgcjioijfljjj o«o^ego(0«-Nifl O N N N N f f O O O C O 03 C O 00 o m v — to o o m 0* 03 r - o to C O o in C O o 5 q s q eg fs. q 00 o q £ 00 C O , e g ’ e g ’ eg ’ e g ’ T “ r * r— T “ 03 O) q 0» O 03 O * 00 00 0* Y cd s s Y C O o co eg 3 eg oo 0» 3 uS eg 00 m ©_ 0" T “ 0 * 0" 0 * C O 03 q C O * co r * C O C O C O eg 03 8 C O r*. 00 m 00 03 C O C O oo 00 00 rv in m f"* cd C O r*. 03 C O S ** o 0) ci d m m o co co o> eg eg (jj < Q < p l 0 3 0 0 C O e g ( O i A l O ( O N Y O ( O K ) M N r i A ( M f l p ( 0 i A ( O O l ( ) N ( p S N f l q q q p S cd^^rotdihNl^r^^r^tpoirsJcD^^oitQoi^egWoi^coQdcd oosioio^MAM^OiaMOfeg^ooioiflnf-OfiOcoiQiou) C 0 « N N N N N ( 0 < 0 ( 0 ( 0 m i n m i n i f l i f l i f l Y Y Y Y Y Y r t O n O O 0)0)ff)MOOiOOiflrYClOriflYNNnNO)(p(OOiflO\00 d d d ^ d g jd ^ ^ d ^ d ^ N in ^ in e g d n io id ^ in in 88 3 55 0 0 *nr-od>NMO»ni,)T-»-0)co(OinvnNnNi'|[M t<t'r^T'f»^nnnrtnnnnnNNNPiN(M iNNNNN S . (O • C 0 Z 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Data sources: GMWK291 - Death Rates fo r 7 2 Selected Causes b y 5-Year A g e Groups, Race, a n d Sex: United States, 1979-97 a t http://www.cdc.gov/nchs/datawh/statab/unpubd/mortabs/gmwfc291 htm. Vital Statistics o f th e United States: Vol II - Mortality, Part A. Published annually b y U.S. DHHS. Various issues, 1968-1978. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.1 Death Rates for Diseases of the Heart for Men, Vital Statistics 1968-1996 10,000 9,000 a 3 8 a > -o G > IS 8.000 7,000 a i . « c 6,000 c o -5 5,000 3 a. o a o o o © o 4,000 I 3,000 8. j§ a 2, 000 o c ■ ■ > < — -X + • .- > * 3v- ■ A ~ A — A — A — A - ~ * X J > M t~ 1,000 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 •-5 0 -5 4 | A - 55-59 i -K—60-64 j • —65-69 j A—70-74 j -4 — 75-79 80-84 85+ Year ui Outcomes from regression models show that overall heart death rates for each 5-year age group have decreased for three decades. The overall heart disease mortality rates have declined in each period (1968-1996, 1982-1996, and 1987-1996). Parameter estimates for all age groups show statistically significant annual declines in overall heart disease mortality rates per 100,000 population in each age category for men in each period (Table 4.2). Although death rates for each age group decline over the three decades, the size of the mortality rate decline differs in each period. Mortality rate declines for the 29-year period of 1968-1996 are smaller for those aged less than 70 years old, but are larger for those aged 70 and over than those declines over the period of the last 15 years (Figure 4.2). The average annual percentage changes with 95% confidence intervals for heart disease mortality rates for different age groups of men are shown in Table 4.3. These percentage changes are shown for the three periods that parallel the regression models shown in the previous table. The average annual percentage changes indicate larger declines for younger age groups than older age groups over the entire period. The average annual percentage declines are 3.2% for those aged 50-54 and 0.9% for those aged 85 and over since 1968. When comparing the average annual percentage changes for 1968-1996 with those for 1987-1996, the annual percentage declines are larger for men over age 65 (Figure 4.3). Between 1987 and 1996, the average annual percentage changes in overall heart disease mortality rates appear to have increased at older ages. In the last 10 years, the average annual mortality decline rate in percentage is 2.4% for those aged 50-54 and 1.3% (the lowest) for those aged 85 and over. Those 70-74 have the highest decline rate of 3.3%. 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T ab le 4 .2 Parameter Estimates o f t h e Annual Change i n D e a th f r o m Overall Heart Problems, b y Sex, b y 5-Year A g e Group, Vital Statistics 1968-1996 a : S C O O ' 0 ) o * mC O C O C O CO < & G O G O G O C O O O G O 0 ) G O O )C f )O ) 0 ) G O i O )o > ©O )O ) 0 ) O )G O o o Oo' o o o O ? o b o o‘o' o b o < r • • « • . • • • • • • • « • • • « « • • ! « « e • • « • • « « * G O C M C O o i n 0 )G O h - CT) N -o C O i n O )o 0 ) o mi no ; C O E C M o mC O O )C O i nC O v - o ^ « ? 0:C O o ;c d C M O ' * i nc do' c d i i i C M i T «? G O » p - t 1 i i C M 1 T r * - i r- • • • • • • « • • • t • • « • « • « • « • • « • « « « « « • « C M mi no O ' C O C O o C O M ’ C M C O h - o C M i n C O C O O ) o G O G O t - 0 )C O o C M 0 ) C O C O C O N » o C O i ni r i c d i n t ) w r - i i n r * a d c d o1 o' O ) o O ) c o O ' i n <b o O ) v 0 ) C O O ) C O CD C O i n G O m G O G O t! T — C O i n O S CM C M CO » — T - C O h - o o o o o o o o i ni ni ni nm m i n i n T * T ” r~ T ~ T— T“ r - S S GO 0 ) O ' © h * O ) GO O ) O ) O ) o o o o ' o o ' o ' o ' o ' • « • • « • e • • • • ■ * • i i n *-* m c o r- o - i n o m 1 ^ . o 0 ) CM O ) O ' 04^ GO i i n i O ) CM i © O ' i t c d I * * * i • • • • • • • • • • • • • • • • CO GO CO o m c o in CO 0 ) C O h - c o c o i n o ) r- r ^ c d o' b o' o; m c o T “ c o c m o CM o- GO T . m . cm. cm_ cm” v go" 0 )0 )00 )0 ) 0 )0 )0 ) C M C M C M C M C N C N C M C N co 2 3 0 5 < 0 > w 0 ) > o O ' O ) O ' 0 ) o * O ) O ' *o c i n i n < 9 CO hj. CO ( Q o i n o i n o i n o i n i n m CO CO r- G O G O m m m o i n o m U)(OIDNN O O $ 5 ? ® i C T Q O < $ a ) > o O ) o' O ) O ' O ) o* X J C i n i n i < 9 < 9 N » t GO i ( Q o m o m o i n o i n i n m C OC O r*^ GO GO e ( o o c n c o o j i n r - o OOOr-COCOOOiO ci oiod ^ t oi r rfOlfiSOJO) i i i i i i i GO s CO 0 ) i n 0 ) CO 0 ) m O ) TT O ) CO GO o' o ' o ' o ' o ' o' o' o e • • • • • o m CO GO GO o GO o CO CO CO CO CM • CM O ' Is* o o o o o o o o CM GO a o r^ . GO r^ - CO 0 ) o > O ) 0 ) O ) 0 ) O ) 0 ) o ' o o ’ b o o ' o o ' • e • ■ • • • • • • • • • • i n GO o CM O ) GO o CO CO T“ m GO CM 0 ) in CM i n i n’ b b 1 CM CO m fs . O ) o i i i i i 1 i • • # • • # • • • • • • • • • CM CO O ) 0 ) CO GO CO o GO GO GO O O GO o CM o ' GO CO ^ CO m CM 0 ) N » CM CO in GO CO o O N ^ c m c o o * o o GO 0 ) 0 ) O i 0 ) GO r - - CM O ) 0 ) 0 ) O ) 0 ) 0 ) 0 ) GO o ' o ' o ' O o ' o ’ o o • • « # • • • • • • • • « • • m CO CO O ) GO o * o O ) r * . GO CM r ^ h * O ' O ) o ' c d c d b CM CO m GO r ^ i n m m i n i n m i n i n 7 fr» r*T *T -r*^ -r- ^ • CO CO 0 0 CM CO CM i c m co co m O ) • < 0 i n CM CO 0 0 0 0 i r N ( O r " 0 ) l f l C O • oo c m _ co _ ao ©_ r- < - v - cm- ^ i n b 0 ) 0 )0 ) 0 )0 ) 0 ) 00 ) C M C M C M C M C M C M C M C M 8 O V a Q 3 O So " g 5 < O < D > O * o ^ ^ O ) ^ C> O ' c mincfupr^r^ao < o bidainbinbm m m cocor^r^coco So * g i O v o t o x r o v c i n m c^> up r - cp <u o i n o m o i n o m m incocor^h»oooo Q So • & o o < $ O ' 0)O 'O ) O 'a > m i in « ? « ? r^ - • o ino ino in i n in C O C O < o > o •o o o ’ v a 8 o V a 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Time: Years sin ce beginning o f period Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.2 Average Annual Change for Overall Heart Disease Mortality Rates, by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 Men 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Women 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ 0 -10 c - o -20 1 a. -30 & s -40 o o' o -50 k g -60 < a V a t c gm -70 cm JZ a -80 T 5 e -90 k < -100 -110 c o Q . O a o o o © o g (0 o o > e < 9 a c < □ 1968-1996 01982-1996 ■ 1987-1996] [□1968-1996 01982-1996 B1987-1996] a > o 8 Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.3 Average Annual Percentage Changes (and 95% Confidence Intervals) in Overall Heart Disease Mortality, by Sex, by 5-Year Age, Vital Statistics 1968-1996 _______ Men_________ _______ Women 1987-1996 1987-1996 Aae GrouD (95%Cls) Aae GrouD <95%Cls) 50-54 -2.4 ( -2.5 . -2.4 ) 50-54 -2.3 ( -2.4 , -2.3 ) 55-59 -3.1 ( -3.2 . -3.1 ) 55-59 -2.6 ( -2.7 . -2.6 ) 60-64 -2.9 ( -3.0 , -2.8 ) 60-64 -2.5 ( -2.5 , -2.4 ) 65-69 -3.0 ( -3.1 . -3.0 ) 65-69 -3.1 ( -3.1 . -3.0 ) 70-74 -3.3 ( -3.3 . -3.2 ) 70-74 -2.9 ( -3.0 , -2.9 ) 75-79 -3.1 ( -3.2 . -3.1 ) 75-79 -3.0 ( -3.0 , -2.9 ) 80-84 -2.4 ( -2.5 , -2.4 ) 80-84 -2.5 ( -2.5 . -2.4 ) 85 and over -1.3 ( -1.4 . -1.3 ) 85 and over -1.3 ( -14 . -1.3 ) 1982-1996 1982-1996 Aae GrouD l95%Cls) Aae Grouo f95%Cls) 50-54 -3.3 ( -3.4 . -3.2 ) 50-54 -2.6 ( -2.6 , -2.5 ) 55-59 -3.5 ( -3.5 . -3.4 ) 55-59 -2.5 ( -2.6 , -2.5 ) 60-64 -3.2 ( -3.3 , -3.1 ) 60-64 -2.6 ( -2.7 . -2.5 ) 65-69 -3.2 ( -3.3 . -3.2 ) 65-69 -2.9 ( -3.0 , -2.9 ) 70-74 -3.3 ( -3.4 . -3.2 ) 70-74 -3.0 ( -3.1 . -3.0 ) 75-79 -3.1 ( -3.1 . -3.0 ) 75-79 -2.9 ( -3.0 . -2.9 ) 80-84 -2.4 ( -2.5 , -2.4 ) 80-84 -2.6 ( -2.6 , -2.5 ) 85 and over -1.4 ( -1.5 . -1.4 ) 85 and over -1.1 ( -1.1 . -11 ) 1968-1996 1968-1996 Aae GrouD (95%Cls) Aae Grouo /95%Cls) 50-54 -3.2 ( -3.3 , -3.1 ) 50-54 -2.4 ( -2.5 . -2.3 ) 55-59 -3.2 ( -3.3 , -3.1 ) 55-59 -2.4 ( -2.5 . -2.3 ) 60-64 -3.0 ( -3.1 , -2.9 ) 60-64 -2.4 ( -2.5 , -2.4 ) 65-69 -2.8 ( -2.9 . -2.8 ) 65-69 -2.6 ( -2.7 . -2.5 ) 70-74 -2.7 ( -2.8 . -2.7 ) 70-74 -2.8 ( -2.9 , -2.7 ) 75-79 -2.5 ( -2.6 , -2.4 ) 75-79 -2.8 ( -2.8 . -2.7 ) 80-84 -1.8 ( -1.8 , -1.7 ) 80-84 -22 ( -2.3 , -2.2 ) 85 and over -0.9 < -1.0 , -0.9 ) 85 and over -1.1 ( -11 . -10 ) Average percentage change based on log-linear regresson Figure 4 .3 Comparison o f t h e Average Annual Percentage Change fo r Overall Heart Disease Mortality Rates, b y Sex, b y 5-Year A g e Group, Vital Statistics 1968-1996 saBueqo aBeiueaied lenuuy aafiueqo {uaoiad |enuuy Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ! □ 1968-1996 B 1982-1996 ■ 1987-1996 □ 1968-1996 B1982-1996 ■ 1987-1996 Women The overall heart disease mortality rates per 100,000 population for women in 5-year age groups from age 50 and over for each year since 1968 are shown in Table 4.4 and Figure 4.4. The death rates from overall heart disease per 100,000 population increase with age, but it is clear that these death rates have decreased since 1968. Parameter estimates of regression analyses for 5-year age groups indicate that declines in overall heart disease mortality rates for women are statistically significant for each of the three time periods shown from 1968 to 1996 (Table 4.2). There are relatively small variations in average annual changes among these three periods indicating that change was fairly similar over the entire period. The average annual change in mortality rates in both periods of 1982-1996 and 1987-1996 show somewhat smaller declines in almost all age groups relative to declines in the period of 1968-1996. Only the average annual changes in mortality rates for those aged 85 and over are larger in magnitude in the recent time period (Figure 4.2). The average annual percentage changes in overall heart disease mortality rates for 5-year age groups of women are shown with 95% confidence intervals in Table 4.3. Most percentages for those less than 85 indicate rates of decline around 2.4% to 3.0% in each of the three periods: 1968-1996,1982-1996, and 1987-1996. The average percentage declines are larger for the age groups of 65-69, 70-74, and 75-79 than the decline in other age groups. The average annual percentage declines in overall heart disease mortality rates for women are larger for all age groups 55 and over in the period of 1987-1996 than those in the period of 1968-1996 (Figure 4.3). 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tab le 4 . 4 D e a th R ates fo r Diseases o f Heart fo r Women, b y 5-Year A g e Group, Vital Statistics 1968-1996 IC D coding: 390-398, 402,404-429 (A 0 ) * 5 (X o o o o > r* ■ 1 0 a 3 O O ) * a a > 2 & £ c o 5 3 a o a o 8 o “ o i * 0 C M C O o C O oo ps O i C O 0 0 0 ps 0 ps. T - 0 T» O O ) O) 0 o 0 px. 0 o 8 C M 3 O i to ps 10 0 8 cd 5 . p . co 0 _ o 0 0 0 £ 0 0 0 cd Ps 0 P - 0 0 o 0 0 od Ps o 0 3 C M O i o . i Psl 0 O s 0> cd 0 o px. 0 o C M 0 0 ps. ^ r C M 8 3 C M 5 C M 0 C M px. 0 C M 0 o a i a i ps" 00* p ." p .* p .' 0 * 0 * 0 * 0 * 0 ’ Is** n T 0 * psT 0 * psT 0* 0* 0 * 0* 0* 0* 0* 0* 0* 0* 0* ps K C O C O C M 0 C O N . 0 ) C O 0 0 0 O 0 fs . 0 o 0 0 * 0 O C M 0 ) o C O C M C M cd co C M 0 O i 10 C O o 00 C M O) C M 0 ps i £ i 0 ? 00 C D 0 S C M 0 Ps C M C M S 8 C M cd a i 0 o o px. 0 ) px. 0 0 0 8 cd o |Xs s 0 0 C M 0 o 0 0 C M O V 0 0 cd 0 * V V C O * V C O * co’ C O * C O * C O * C O * 0 * 0 * 0 * 0 * 0 * 0 * 0 * C M * C M * C M * C M * C M * C M * C M * C M * C M * C M * C M * C O C O o O O O ) C O o C M 0 00 0 0 0 C M o ps. 0 0 O) 0 0 0 T — O 0 O Ps 0 0 C M 10 b C O n 10 ps b o 10 10 cd C M cb 00 C M 0 ) 0 0 0 o s o cd 0 0 cd 0 d ps. pv V fx . p . C O 0 fs_ o Ps. cd 0 0 o s cd 0 0 _ 0 * C M 0 0 o 0 0 8 0 _ 0 0 _ b 0 C M 3 C M co C M c m ' C M * C M * C M * C M * C M * C M * C M * cm" C M * C M * * - *** C O ps. 10 O i 10 0 0 C O 0 C O 0 0 ps ps P^ o P^ 0 0 0 C M 0 0 O ) 0 O O i cd C O 10 a i O i s o 10 i f C O o n m a i s. 0 01 C M _ o C M C M cvi ps 0 C M s r» 0 Ps O . S 0 0 o 0 C M o _ o C M o _ a i 0 0 ) o px. 0 P^ 3 a i O i C M 0 0 0 0 3 0 b Px. px. b 0 Ps o 0 Px. b C M ps C M C M O px. T *» T“ T “ o o o 00 0 o 0 0 O C O o ps. C M C M 0 px. 0 C M 0 O i V ps. Px. 0 px. 0 ps 0 o o cd 1 0 0 0 cd o o o C O o o ps 8 Ps C M C M Ps 8 0 p - 3 C M C M 0 0 8 0 0 0 0 C M 0 b o 0 cd 0 0 0 ps. 0 ps 0 0 8 0 o 3 0 0 0 0 0 O 0 * *s 0 0 ? 0 0 p x ^ 0 cd 0 px. O i 0 O ) ps O ) C M 0 o a> 0 O ps 0 0 0 C O O T** Ps. C M C M C M 0 O 0 0 0 ps ps b 5 b b C M 8 0 0 ps C M 0 C M 0 0 0 0 o C M C M 0 0 0 0 V C M O) Ps Ps 0 0 0 s 0 C M C M O 0 Ps 0 0 0 0 0 ■ M * 4 ^ 0 0 0 0 0 0 0 0 0 0 0 0 C M C M C M C M C M C M C M C M C M C M q ps o> Ps o> O ) 0 C M ps r - 0 0 O i 0 0 0 C M 01 C M 0 0 C M 0 C M 0 b ps! 0 0 0 0 Ps C M b 8 C M cd 0 s r * T“ C M b cd C M 0 P^ C M 0 8 ps 0 0 C M O O i O i O) 0 0 0 ps ps ps 0 0 0 0 C M C M C M P M C M C M C M C M C M C M T— T “ *“ 0 0 C M o 0 0 0 0 O i 0 0 o> q 0 0 O i 0 0 0 px. 0 0 ps 0 Ps o c d C O o C M0 c d c d c d b 0 8 t— 0 b b 0 0 0 C M0 o c d 0 0 ps 0 r» 0 0 0 C M r * o o T * o o O o 0 O ) 0 0 0 0 Ps 0 ps ps ps ps oaio<-Nn^mifiNooaiOT-Nn«iniOMoaio>-Nn«iAio lOVNNNSNSSNSKQOOOflOflDttBDOOtOtt0AO)O)AO)O)w OIOlO)Od)4t0)O)0)OIO>O)OO)OIO)OIO)OtOO01OIO)0IOkOI0)O) e o E o 5 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Data sources: GMWK291 - Death Rates fo r 7 2 Selected Causes b y 5-Year A g e Groups. Race, a n d Sex. United States, 1979-97 a t http://www.cdc.gov/nchs/datawh/statab/unpubd/mortabs/gmwk291 htm Vital Statistics o f th e United States: Vol II - Moitallity, Part A. Published annually b y U.S. DHHS. Various issues, 1968-1978. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.4 Death Rates for Diseases of the Heart for Women, Vital Statistics 1968-1996 ■ ■ a 10,000 9,000 8,000 7,000 U & « l c 6,000 c o 5,000 Q . O e o o o a. & « a N c e. 4,000 3,000 2,000 1,000 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 - m - 50-54 | - A - 55-59 X - 60-64 -*-6 5 -6 9 | -e-70-74 | — t— 75-79 ! 80-84 | — 85+ Year o> tn For 1987-1996, the smallest decline is 1.3% for that of aged 85 and over, and the highest rate of decline 3.1% occurs among the age 65-69. 2. Trends in acute myocardial infarction (AMI) mortality rates Men Annual death rates for acute myocardial infarction (AMI) per 100,000 population in 5-year age groups for men aged 50 and over by year between 1968 and 1996 are shown in Table 4.5 and Figure 4.5. The AMI death rates are higher at older ages and decline over time. Parameter estimates from regression analyses indicate that AMI mortality rates for all age groups of men have declined significantly in all three periods: 1968- 1996, 1982-1996, and 1987-1996 (Table 4.6). In each time period, the size of the annual declines in AMI mortality rates per 100,000 population increases with age until age 80-84, then it becomes smaller. The changes in AMI mortality rates during the more recent periods are lower than those in the entire period of 1968-1996 for all age groups. Declines in the period of 1987-1996 are generally lower than those in the 1982-1996 period. Only one age group (85 and over) shows a slightly higher rate of decline in the most recent period than for the period of 1982-1996 (Figure 4.6). In other words, the average annual rate of decline in AMI mortality has decreased in recent years in general. The average annual percentage declines in AMI morality rates for men are smaller at older ages. The average annual percent declines are 5.7% for those aged 50-54 and 2.0% for those aged 85 and over for men (Table 4.7). However, the relationship between the size of decline and age becomes less clear in more recent years. As compared with the period of 1968-1996, the decline in AMI for age groups 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T ab le 4 .5 D e a th R ates fo r A c u te Myocardial Infarction (AMI) fo r Men, b y 5-Year A g e Group, Vital Statistics, 1968-1996 CL 3 O ■ o d ) o 0) a < /) c o 3 a o a o o o 0) a c o u 0 O (A £ C O Q L S T 5 * - ' c o c o < r c o a o i O ) C D 0 ) C N * - * i n C N O flO i n o ( 0 i n Q Ot r i g n C O C O C N r - a d c o c o8 s C N 8 c o C N r * - o O 8 o o s a 8 C N C N C N C N C N C N C N C N C N C N C N T - C N C N O) O ) coto<n<oo>0)^-r^coo»-:OCN®cNcor^*r^ro)<o^<o<oc)r*.*-co<o M(P5rtrtNO®fW©M(*)<OM5{0*-^U2WOI3N»-00©® n W ^ C N C O N f - O O O ) ® CO CO S N t f © (O U1 T l ' . n W N N ^ ^ O O cm ’ CN CN CN CN CN CN CN CN e- * r-‘ f - ' - r - ~ ^ 1 - ' T- ~ r e-‘ OJinMCCOWSOOWO^WWCMONn^^Ol^fOONOIACOCO ......................................... ^doi^rco^flocd^^r'ioicscN ‘* fs -2bQw^o6(OOcoco 000)0)(OCONNN(OCD CNCDCO^^•^•CN<OCNQO)<OC)CNp^-OG), 8 cn?<CQcqincB<oa>rra><£ — g>tp«-tf)i ®®o>a)cor^(o<oininco{ocoN?'iCN*-< oocoio^-^co^wp^tncoinoioino^-o^co^ipin^pvin*-*- c o w w c M i r i t O N f t N r t o i t p ^ n o d a i N w W r ^ t o i i A i r i i o o j i r i i A - ■ — — ■ * “ l o i n T i *-3*-w«-(orto)(D(nbg«(0 aicocp^r-totowinioin^^^ 0 ) ^ T (0 < n ^ rc o r -* -T fc o * - a o (n rs -tD » -c n c n c o ir3 u o o o c N o a 3 r*.c o a o w O f-»-S f-fN N f(D *-rtf-Q O N3 0 fC1V(flgNfW{0 (» ) in»-o<05®5co50)i75cN4»-r^in*-to^g)<e3C40)®mincnv-*- *-*-oooo>®cocor'-r'«»r'»h*<0io<oi75in^^^^iO(Ococococo(O ^^ ,o<oir)GO(O<o^ts)flO'rin(no>cooo^-ts>cNOco«-®a><otft<0(o c n i n c o k c n < 0 b c o f ^ b b b c n b c n b 5 ® i n c n c o o c n « 2 g c o c o u7iAin^^r: ?<r^«(n(nnncMWwcNNNr-r-i-i-^t-f-i-t-»- Q O O )» -® lA O IA (D fO ffl 5eo^-*-a)cor^inco -■ c o < 0 c o c o C O M / J N O C M C M C N C M C M C N N0)O(n(OMU)ONf-^t-<O(*)«-^MlAOOOOC4(A(M(O«-(DOO -r ^ -i g j ^ ^ ^ g ^ ^ ffl C O O N W 5 5 CO SN(OI07PiM OOO)OIOicOCOCOCO ( 0 N O O 9 ) O r N C O n I ft CD I** 60 0> © * - CN C O V l f l C O h»h»h»K6O flO coao0O ooaoaoeoco&&o>a>0>©& 0)0)^0)0)90)0)00)90)0)0)0)0)0)0)0)0)0) • a s © O ^ M c o * 1 0 “i IN N S N N N 10 0 9 0 ) 0 0 _ ^ ----------------- m jv e 4) s 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. D ata sources: GMWK291 - Death Rates fo r 7 2 Selected Causes b y 5-Year A g e Groups, Race, a n d Sex: U nited States, 1979-97 a t http://www.cdc.gov/nchs/datawh/statab/unpubd/mortabs/gmwk291 htm. Vital Statistics o f th e U nited States: Vol II - Mortallity, Part A Published annually b y U S DHHS. Various issues, 1968-1978. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.5 Death Rates for Acute Myocardical Infarction (AMI) for Men, Vital Statistics 1968-1996 3,500 3,000 a a S | 2,500 iC t .£ 2,000 e o % 9 C L & o o o 1,500 o~ o 1,000 500 * A A A 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 .•■••••50-54 -A- 55-59 > • 60-64 —■—65-69 70-74 —♦ — 75-79 80-84 85+ Year o> oo Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.6 Parameter Estimetes of the Annual Change in Death from Acute Myocardial Infarction (AMI), by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 Men 1987-1996 Aae Group N Intercept Time R-sauare 50-54 10 112.74 *** -3.80 0.89 55-59 10 191.25 — -7.07 — 0.98 60-64 10 300.45 ” • -10 67 — 0.96 65-69 10 454.85 ••• -17 23 — 0.98 70-74 10 691.09 *“ -28.28 ••• 0.97 75-79 10 1,006.11 ••• -39.73 — 0.98 80-84 10 1,470.01 -48 55 — 0.98 85 and over 10 2,126.48 **• -36 83 •“ 0.89 1982-1996 Aae Group N Intercept Time R-sauare %0-S4 " " 16 159.00 -6.43 — 0.92 55-59 15 261.81 ■ “ -10.48 “ • 0.95 60-64 15 407.31 -15 82 — 0.95 65-69 15 610.07 •“ -23.92 •“ 0.96 70-74 15 913.44 -36.07 — 0.97 75-79 15 1276.70 — -46.59 — 0.99 80-84 15 1738.26 — -50 93 — 0.99 85 and over 15 2301.08 •** -36 13 — 0.94 1968-1996 Aae Group N Intercept Time R-sauare 50-54 29 323.86 “ • -9.94 ” * 0.97 55-59 29 523.96 — -15 89 0.97 60-64 29 796.48 *** -23.62 *•* 0.97 65-69 29 1,118.66 **• -32 05 — 0.98 70-74 29 1,565.27 — -43 02 — 0.99 75-79 29 2,045.93 ••• -52 02 — 0.99 80-84 29 2,452.64 — -51 13 — 0.99 85 and over 29 3,063.48 — -48.83 — 0.92 *p <0.05, **p <0.01, "■p <0.001 Women 1987-1996 Aae Group N Intercept Time R-sauare 50-54 10 32.5? — -1.10 ••• 0.91 55-59 10 63.53 “ • -2.15 *** 0.97 60-64 10 116.90 “ * -3.97 *•* 0.96 65-69 10 200.95 — -7 86 ••• 0.98 70-74 10 341.96 ••• -13.14 •** 0.99 75-79 10 560.07 *** -20.82 **• 0.99 80-84 10 915 27 •" -29.00 “ • 0.99 85 and over 10 1,587.69 **• -21.10 — 0.78 1982-1996 Aae Grouo N Intercept Time R -sauare 50-54 15 41.52 — -1.44 •** 0.95 55-59 15 80.43 *** -2.74 •“ 0.97 60-64 15 148.25 — -5.07 **• 0.97 65-69 15 254.15 — -9.18 **• 0.99 70-74 15 432.17 — -15.48 *•* 0.99 75-79 15 688.10 — -23.09 0.99 80-84 15 1056.17 -28.61 **• 1.00 85 and over 15 1641.52 — -16.27 — 0.84 1968-1996 Aae Group N Intercept Time R-sauare 50-54 29 71.06 "• -1.89 *•* 0.97 55-59 29 135.66 — -3 56 *** 0.97 60-64 29 248.82 — -6.48 — 0.98 65-69 29 423.63 -11.24 •" 0.96 70-74 29 714.49 — -18.74 — 0.98 75-79 29 1,104.75 ••• -27.53 0.98 80-84 29 1,540.20 •“ -32.82 “ • 0.98 85 and over 29 2,256 22 — -35.56 0.78 Time: Years since beginning of period Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.6 Average Annual Change for Acute Myocardial Infarction (AMI) Mortality Rates, by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 Men Women 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ -60 [□1968-1996 01982-1996 ■1987-1996 [□1968-1996 Q 1982-1996 ■1987-1996] - J O Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.7 Average Annual Percentage Changes (and 95% Confidence Intervals) in Acute Myocardial Infarction (AMI) Mortality by Sex, by 5-Year Age Group, Vital Statistics 1968-1996 Men Women 1987-1996 Aae Grouo (95%Cls) 1987-1996 Aae Grouo (95%Cls) 50-54 -3.9 ( -3.9 , -3.8 ) 50-54 -3.9 ( -4.0 . -3.8 ) 55-59 -4.4 ( -4.5 , -4.3 ) 55-59 -4.0 ( -4.1 , -4.0 ) 60-64 -4.2 ( -4.3 , -4.1 ) 60-64 -4.0 ( -4.0 . -3.9 ) 65-69 -4.6 ( -4.6 , -4.5 ) 65-69 -4.7 ( -4.8 . -4.6 ) 70-74 -5.0 ( -5.1 , -4.9 ) 70-74 -4.6 ( -4.7 . -4.6 ) 75-79 -4.8 ( -4.9 , -4.7 ) 75-79 -4.5 ( -4.6 , -4.4 ) 80-84 -3.9 ( -4.0 . -3.8 ) 80-84 -3.7 ( -3.8 , -3.6 ) 65 and over -1.9 ( -1.9 . -1.9 ) 85 and over -1.4 ( -15 . -1.4 ) 1982-1996 1982-1996 Aae Grouo (95%Cls) Aae GrouD (95%Cls) 50-54 -5.5 ( -5.6 . -5.4 ) 50-54 -4.5 ( -4.6 , -4.4 ) 55-59 -5.5 ( -5.6 . -5.4 ) 55-59 -4.5 ( -4.6 , -4.4 ) 60-64 -5.3 ( -5.4 . -5.1 ) 60-64 -4.5 ( -4.6 , -4.4 ) 65-69 -5.3 ( -5.5 , -5.2 ) 65-69 -4.9 ( -5.0 . -4.8 ) 70-74 -5.5 ( -5.6 , -5.3 ) 70-74 -4.8 ( -4.9 , -4.7 ) 75-79 -4.9 ( -5.1 . -4.8 ) 75-79 -4.4 ( -4.5 , -4.3 ) 80-84 -3.7 ( -3.8 , -3.6 ) 80-84 -3.4 ( -3.5 , -3.3 ) 85 and over -1.8 ( -1.8 . -1.7 ) 85 and over -1.1 ( -11 . -11 ) 1968-1996 1968-1996 Aae GrouD (95%Cls) Aae GrouD (95%Cls) 50-54 -5.7 ( -5.8 . -5.5 ) 50-54 -4.4 ( -4.5 , -4.2 ) 55-59 -5.5 ( -5.7 , -5.4 ) 55-59 -4.3 ( -4.4 , -4.1 ) 60-64 -5.3 ( -5.5 , -5.2 ) 60-64 -4.2 ( -4.4 . -4.1 ) 65-69 -5.0 ( -5 2 . -4.9 ) 65-69 -4.3 ( -4.5 , -4.2 ) 70-74 -4.7 ( -4.9 , -4.6 ) 70-74 -4.3 ( -4.4 , -4.2 ) 75-79 -4.2 ( -4.3 , -4.0 ) 75-79 -4.0 ( -4.1 , -3.9 ) 80-84 -3.0 ( -3.1 , -3.0 ) 80-84 -3.1 ( -3.2 , -3.0 ) 85 and over -2.0 ( -2 1 . -2.0 ) 85 and over -1.9 ( -2.0 , -1.8 ) Average percentage change based on log-linear regresson. from 65 through 84 have larger increases in more recent years in the period 1982- 1996 (Figure 4.7). The decline pattern differs in the short period. In the period 1987- 1996, the largest average annual percent decline is 5.0% for those aged 50-54 while the smallest average annual percent decline is 1.9% for those aged 85 and over. Women The age-adjusted death rates for AMI per 100,000 population in 5-year age groups for women aged 50 and over by year from 1968 to 1996 are given in Table 4.8 and Figure 4.8. The AMI death rates for women increase with age in all years, but death rates have declined since 1968. Parameter estimates from regression analyses indicate that annual changes in AMI mortality rates for women in all age groups have declined in each of the three periods: 1968-1996, 1982-1996, and 1987-1996. These declines in AMI death rates are all statistically significant at p<0.001 (Table 4.6). The annual declines in AMI mortality rates per 100,000 population are higher at older ages in the period of 1968- 1996. In the periods of 1982-1996 and 1987-1996, the annual decline increases with age up to age 80-84, at age 85 and over the decline is less. The average annual changes in AMI mortality rates in the more recent periods of 1982-1996 and 1987- 1996 are smaller than those in the period of 1968-1996 in all age groups (Figure 4.6). Between 1968 and 1996, average annual percentage changes for AMI mortality rates for women are around 4.3% for age groups less than 80. They are smaller at the higher ages, 3.1% for age 80-84 and 1.9% for age 85 and over (Table 4.7). When comparing the percentage declines in AMI mortality rates in the overall period of 1968-1996 to those for the most recent years 1987-1996, rates for age groups 65 through 84 are higher in recent years (Figure 4.7). In the period of 1987- 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.7 Comparison of the Average Annual Percentage Change for Acute Myocardial Infarction (AMI) Mortality Rates, by Sex, by 5-Year Age Group, Vital Statistics 1982-1996 Men Women 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ 0.0 - 1.0 M flt S -2.0 o e o a. c < -3.0 -4.0 -5.0 - 6.0 M a a > c < ■ ■ C o a t at 3 c o e S . 0.0 - 1.0 - 2.0 -3.0 r -4.0 -5.0 - 6.0 ID1968-1996 131982-1996 »1987-1996 □ 1968-1996 □ 1982-1996 ■ 1987-1996 "j C O T ab le 4 .8 D e a th R ates fo r A c u te Myocardial Infarction (AMI) fo r Women, b y 5-Year A g e Group, Vital Statistics 1968-1996 IC D codina: 410 + IA C O in o i t S n t t / i j inMflSSNNQ O 0 > m O o C M lh O ) C M 01 C O m »-in<ors- 0 ) in f s- ■ o 0) c o 0) a (A C o a o a . o o o o* o 0 ) a ( A Q > (9 cr O i (A * « ? O HI 2 8 * . 8 o o C O C M o c oC O C O < 0O O C O C O o i * C O q o c o C O o oo oC M $ r - * . C M 8 C O C O C O . C O < 5 C O . t h C O r * - . o i 8 . s a C O c o 5 C O O i i n C M i n i n s i C M i n5 8 i n C O m i n o i n C M I s - $ o 5 i n s. * " < ? i n c oO i C M r * « -i n C M o C M C O o o C O C O o o o i C M . C M C O c d C M O i S o . 1 C M o . § § 8 O i 0 ) C M s o i C O 0 0 C M o o§ d r - I s * . C O p - s r - r - c o C O C O C O r “ C M C O O ) C O r > »o i i n O i 0 O i n o mq C M mo C O § C M i n o o i n C O C O 8 r - C O C O h . o o o C O 5 5 C O C M p * . C O 18 5 5 m C O i n t m C M m C M O i n r * - C M — 3 3 5 r * ^ 0 ) C O o i P - c o COP^O^CMCMCMP^lO^-lOCO CMf^ddP^di^cocjNlipd WNfliotonflictQSt NN(os(5iouiinuii/)T^ — d oi coo»-otncoiA(o*-(n(Oco*“ ddr^dNcM^iocrioigjd^ “ ( o frtfQ O M fltn n C O C O C O C O W C M C M C M C M C M C M ^■Tr^ooifn^r^oico^-cncMOiOico^rnincMr^aooiCNcn © r t s d « « cd o ’ o ir i j o W u i o V io n n ®NSioi7)^(n?i«“ OC7)cBNii)i7i^Tr^(n CMCNCMCMCNCMCMCMeMCMT-*-*-*-*-*-*-*-*- GicnOiocMt/iCMinco (b^r*«-0tp5NO m F ^ ^ ^ c o co co co co 0)f-(Oco(O^cot0)0)0iCNN0^iAO(sinroino)oiA«-a)O«-co Wdcodd^ddindirid^io^gidcodd^cQcodco^oids IOTTt(n(MOO9)QN(0(OlAl/)tt(O^NfOO9iO)a)COScO qK)iflONO)*-Nrtf0^n(*)con»Nt«N»«pOi-^rt^ (O^tcbcn'^oi^NcviddtncNaiNrtJdcn^NuiN^doiuii v(onNNCM^ooo)0)fli®cocoNSso{0(OU)uiininiA^77 fvr.|s.^fs.fs(OONO(C»-^NOOCOa3^Nr-0(OCMNOi«-0)0) in w d « ^ t «5 ® w q n « j j n w « n v i * • o s co id i f u i r i n rs *r^h*<oioioinioinin^Mr,T^‘ ^^rcncocococococM CM CNCM CM CM (M flO0)OrNCO^U)(OSflOO)O^NrO^iO<ON0OOiO^(M(n^lO(O 8 ( O K h . ^ N K h » N N h S Q O C O f l O O O f l O G O 0 C O f l O C O O i O ) O ) 0 ) O ) O i O i 0 ) 0 ) 0 ) O) O > O > O iO > O > O > O > O > 0 > O ) O > O iO ) O i O) 0 ) 0 ) 0 ) 0 ) 0 ) 0 ) 0 9 ) 0 ) e 0) E o 5 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. D ata sources: GMWK291 - Death Rates fo r 7 2 Selected Causes b y 5-Year A g e Groups, Race, a n d Sex: U nited States, 1979-97 a t http://www.cdc.gov/nchs/datawh/statab/unpubd/mortabs/gmwk291 htm Vital Statistics o f th e U nited States. Vol II - Mortality, Part A Published annually b y U.S. DHHS. Various issues, 1968-1978 Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.8 Death Rates for Acute Myocardical Infarction (AMI) for Women, Vital Statistics 1968-1996 3,500 3,000 Q . O ) 2,500 I M C 2,000 c 0 1 Q . & s O © o 1,500 £ 1.000 M 0 1 s a c 500 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 -50-54 A 55-59 60-64 -*-6 5 -6 9 70-74 — I— 75-79 — 80-84 85+ Year -> i cn 1996, the largest average annual percentage decline is 4.7% for those aged 65-69 while the smallest annual percentage decline is 1.4% for those aged 85 and over. In sum, heart disease mortality represented by overall heart disease or Ml has declined continuously in the last three decade for both sexes. Men basically show faster annual percent decline in heart disease mortality rates than women except for the overall heart disease mortality among older age groups. B. Prevalence Trends Trends in reported prevalence rates from overall heart disease and ischemic heart disease (IHD) are examined in this section. First, prevalence rates from overall heart disease and IHD for men and women aged 50 and over in 5-year age intervals are calculated for each NHIS interview year from 1982 to 1996. Trends in these prevalence rates are graphed. To smooth out the annual variability in prevalence rates, three-year average prevalence rates are also calculated. Third, these prevalence trends for different age groups in the periods of 1982-1996 and 1987- 1996 are analyzed. Fourth, statistical analyses of the relationships between prevalence and age, time or year, and their interaction effects are examined. These analyses are conducted separately for men and for women. 1. Overall heart disease prevalence trends Men The prevalence or percent of the groups with overall heart disease for men aged 50 and over by 5-year age intervals from 1982 to 1996 is shown in Table 4.9. The prevalence of overall heart disease generally increases with age. For some years, however, this is not true at age 70 and over. The average prevalence from 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.9 Annual Prevalence of Overall Heart Disease, by Sex, by 5-Year Age Group, NHIS 1982-1996 (Percent with disease) Men Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Average 1982-1996 1987-1996 Age Group 50 and over 17.1 18.3 19.1 18.3 17.7 18.5 19.3 17.5 18.9 18.6 19.9 19.6 19.5 20.1 19.5 18.7 18.9 50-54 9.2 10.5 10.1 8.3 12.1 8.8 7 0 6.7 7.7 11.2 9.2 9.8 9 0 9.2 8.6 9 2 8.8 55-59 13.0 14.8 11.3 17.0 13.0 14.0 144 13.4 14.8 16.2 15.7 11.7 115 14 1 12.3 13.8 13.8 60-64 19.6 16.2 21.5 14.1 14 0 17.1 18 4 19.2 19.4 16.4 18.1 17.5 20 9 21.4 17.0 18.0 18.5 65-69 21.0 22.1 23.5 23.5 21.1 222 25.5 19.3 19.9 18.9 23.5 24.5 19 5 20 9 17.1 21 5 21 2 70-74 21.5 29.2 29.9 22.2 26.4 27.1 24.7 23.9 21.3 25.8 264 27.5 26 5 24 7 23.3 25.4 25.2 75-79 23.9 26.2 25.8 22 5 23 8 20.1 28.0 26.6 30.4 27.4 31.4 29.9 33.2 31 2 26.4 27.4 28.7 80-84 19.3 23.1 27.9 31.0 29.9 29.6 30.8 32 3 36.2 28.1 26.4 35.1 34 5 36 7 40.1 31.0 32.9 85+ 38.9 27.5 31.8 47 8 19.7 34.1 31.6 27.6 32.7 24.6 35.0 32.3 36.3 34.3 38.9 33.5 33.0 Women Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Average 1982-1996 1987-1996 Age Group 50 and over 15.1 17.4 16.4 16.4 15.6 15.6 16 2 15.1 15.7 16.9 16.9 15.7 16.3 16.1 15.3 16.0 16.0 50-54 6.7 8.6 7.1 6 9 10 3 8.7 7 3 7.3 8.1 10.5 7.5 7 8 6 3 9 2 6.9 7.9 7.9 55-59 10.1 12.7 8.8 12.2 11.2 11.1 12.5 9.5 10.6 9.4 10.6 8.5 9.9 9.9 13.4 10.7 10.5 60-64 13.9 15.0 16.2 14 0 9 9 9.7 126 12.6 11.8 12.4 14.7 14.2 13 3 126 9.8 12 8 12.3 65-69 18.0 20.1 20.2 15 8 16 8 19 1 17.2 14 1 19.0 16.8 18.0 17 2 17 9 15.3 19.1 17.7 17.4 70-74 14 5 21.4 19.7 23 8 17.0 19.0 21 7 18 3 18.6 19 9 18.4 166 19 0 20.5 176 19.1 19.0 75-79 23.0 27.0 30.4 24 1 26 9 23 5 276 24 2 22 7 274 28.3 23.1 28 8 22 8 19.1 25.2 24.7 80-84 27.8 26.6 28.0 24 4 27 9 264 24 8 28 7 24.2 31.6 26 7 28 3 30.6 27 8 27.4 27.5 27.8 85+ 36.4 29.1 20.3 38 9 29.0 29.0 23 0 294 26.7 29.0 346 37 8 22 5 27 3 26.0 29.0 28.4 ■ n i -v l 1987 through 1996 is 18.9% for all men aged 50 and over, ranging from 8.8% for age 50-54 to 33.0% for age 85 and over. The trends in prevalence for overall heart disease for different age groups are graphed in Figure 4.9. While younger age groups show slight changes in prevalence rates, older age groups show large variations of prevalence rates over the 15 years from 1982 to 1996. This variability may be due to the sample size of these groups in the NHIS. Three-year average prevalence was calculated to smooth out the prevalence and present a clearer picture of change. Prevalence is graphed at the middle year of the three-year range in Table 4.10 and Figure 4.10. As an alternative, the overall heart disease prevalence for men aged 50 and over with a 10-year age interval is also provided in Table 4.11 and Figure 4.10. It looks like there was not much change in heart disease at the younger ages but an increase at least in recent years at the older ages. To examine the trends in overall heart disease prevalence for different age groups, regression analyses are used. Table 4.12 shows parameter estimates from the regression analyses of the prevalence rates of heart problems for 5-year age cohorts on the number of years since 1982 and since 1987. The trend analyses for 15 years between 1982 and 1996 shows that there are increasing levels of overall heart disease prevalence for those aged 75-79 and 80-84 among men. There is no statistical significance in the prevalence trends for other age groups. The trend analysis for 10 years from 1987 to 1996 shows that there are no statistically significant trends in overall heart disease prevalence rates for any age groups aged 50 and over. The average annual percentage change in overall heart disease prevalence for men is an increase by 0.5% between 1982 and 1996 (Table 4.13). It has slowed 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.9 Annual Prevalence of Heart Disease for Men, NHIS 1982-1996 50 45 40 35 30 o > 25 20 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 - i - 50-54 55-59 - A - 60-64 65-69 70-74 75-79 —i— 80-84 —— 85+ Year co Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.10 Prevalence of Overall Heart Disease, by Sex, by 5-Year Age Group, NHIS 1982-1996 (Average of a 3-Year Range) (Percent with disease) Men Range 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Average of Years 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1982-1996 Age Group 50-54 9.9 9.7 10.2 9.8 9 3 7.5 7.1 8.6 9.4 10.1 94 9.3 8.9 9.2 55-59 13.0 14.4 13.8 14.7 13 8 13.9 14.2 14.8 15.6 14.5 13 0 12.4 12.6 13.8 60-64 19.2 17.3 16.5 15.1 16 5 18 2 19.0 18.3 17.9 17.3 18.9 19.9 19.8 18.0 65-69 22.2 23.0 22.6 22.2 22 9 22 3 21.5 19.4 20 8 22.4 22.5 21.7 19.2 21.5 70-74 26 9 26.9 26.0 25.3 26.1 25 3 23.3 23.7 24 6 26.6 26.8 26.3 24.8 25.4 75-79 254 24.9 24.1 22.3 24.4 254 28.4 28.1 29.7 29.5 31.5 31.4 30.1 27.4 80-84 23.4 27.3 29.8 30.2 30.1 30 8 33.3 32.5 30.3 29.8 31 8 35.4 36.7 31.0 85+ 32.4 36.9 36.0 37 1 29 7 31 0 30.8 28.4 31 0 30.9 34 5 34.3 36.5 33.5 Women Range of Years 1982 1984 1983 1985 1984 1986 1985 1987 1986 1988 1987 1989 1988 1990 1989 1991 1990 1992 1991 1993 1992 1994 1993 1995 1994 1996 Average 1982-1996 Age Group 50-54 7.5 7.5 8.1 8.6 8.7 7.7 7.5 8.7 8.7 8.6 7.2 7.8 7.5 7.9 55-59 10.5 11.2 10.7 11.5 116 111 10.9 9.8 10.2 9.5 9 7 9.4 11.0 10.7 60-64 15.0 15.1 13.4 11 1 10 7 116 12.3 12.3 12.9 13.8 14.1 13 4 119 12.8 65-69 19.5 18.7 17.6 17.3 17.7 16 9 16.8 16 7 17.9 17.3 17.7 16.8 17.5 17.7 70-74 18.5 21.6 20.2 20 0 19 3 196 19.5 19.0 19.0 18.3 18.0 18.8 19.1 19.1 75-79 26.8 27.1 27.1 24 8 26.0 25 1 24.8 24.7 25 9 26.2 26.7 24 7 236 25.2 80-64 27.5 26 3 26.7 26.2 26.4 26 7 26.0 28.3 27.6 28.8 28.5 28.9 28.6 27 5 85+ 28.9 28.7 28.7 31.5 27 1 27.2 26.4 28 2 29.8 33.6 31 2 28.8 25.2 29.0 F ig u re 4 .1 0 Prevalence o f Heart Disease fo r Men, N H IS 1982-1996 (3-Year Average) | 1 j ^ O ) O ) ^ O ) ^ ' U j) up (O N N C O + O L O o i n o in o m in ifi(0 (ossooao O o C O in C N o v— o aBeiuaaiad 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 9 8 3 1 9 8 4 1 9 8 5 1 9 8 6 19 8 7 19 8 8 1 9 8 9 19 9 0 1991 1 9 9 2 1 9 9 3 1 9 9 4 1995 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.11 Annual Prevalence of Overall Heart Disease, by Sex, by 10-Year Age Group, NHIS 1982-1996 (Percent with disease) Men Average Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1982-1996 1987-1996 Age group 50 and over 17.1 183 19.1 18.3 17.7 18.5 19.3 17.5 18.9 186 19.9 19.6 19.5 20.1 19.5 18.7 18.9 50-59 11.1 126 10 7 12.8 12.6 11.5 10.7 10.1 11.2 13 7 12.3 10.7 10.1 11.3 10.3 11.4 11.2 60-69 20.3 19.0 22 4 18.2 17.2 19.5 21.7 19.2 19.7 176 20.8 21.0 20.2 21.1 17.0 19.7 19.8 70-79 22.5 27.9 28 1 22.3 25.3 24.8 26.2 25.0 24.9 26 4 28.3 28.4 29.3 27.5 24.7 26.2 26.6 80 and over 24.7 24.2 29 4 36.5 27.6 30.8 31.0 30.9 35.2 27.1 28.9 34.2 35.0 36 0 39.6 31.8 32.9 Women Average Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1982-1996 1987-1996 Age group 50 and over 15.1 17.4 16.4 16.4 15.6 15.6 16.2 15.1 15.7 16.9 16.9 15.7 16.3 16.1 15.3 16.0 16.0 50-59 8.5 10.7 7.9 9.6 10.7 9.9 9.8 8.4 9.3 10.0 9.0 8.1 8.1 9.5 9.9 9.3 9.2 60-69 15.8 17.5 18 0 14.8 13.3 14.0 14 8 13.3 15 1 14 6 16.4 15.8 15.7 13.9 14.4 15.2 14.8 70-79 18.1 24.0 24 2 24.0 21.5 21.0 24 4 20.8 20.5 23 0 22 4 19.3 23 0 21.6 18.3 21.7 21.4 80 and over 30.7 27 5 25 6 28.1 28.3 27 3 24.2 29.0 25.2 30 7 29.0 31.4 27 8 27.6 26.9 28.0 28.0 C D ro T ab le 4 .1 2 Parameter Estimates o f t h e Annual Change i n Overall Heart Disease Prevalence, b y Sex, b y 5-Year A g e Group, N H IS 1982-1996 0) h - C O T — © 05 c oin E O O o o CN CN o F o • o' o o d i o i o o r ^ - v- o i n o © T - e g o e g O 0 5 c o© o o o o o o o T - O o ' o o ' o ' d o ' o o ' « a « « « a a a * a « a a « a a a a « a a a a a in C O r** i n CO © CO v— CN in o C O © o C O C O oo o ' CN F 05 CD d G O t — CN CN CN OOOOOOOO o CO CO r - r - T“ CN © o T — 05 ino 05 05 CN o o o o o o OO o ' o ' o ’o’o o ' o d 0^(0 05^0^0 O O r O O C N r (N J 0000000 0 « a a a a a a a a a a a a a a a a a a a a a a a CN © © 05 © 05 CO © 05 05 I s- CN C O © © F o ' cd G O 05 © © o T “ T — CN CN CO in tn in to in in u ) ia < 0 0 05 CO U 5 O U 5 O U 5 CO © X 3 c CO (Q O lO CO CO 05 IT) in M 0 ) © CN © © 0 5© o E CN © O © © h- F o' o' i o ' o ' o * o o d © CN r — CN 05 © © r - © o CN I s- © T * o o © © CN o d d d d o o O a a a a a a a a a a a a a a a a a a a a a a a a © CN © O * r O 05 © © I s- CN © T * © F < r F CN © © 05 05 T - CN CN CN CN CN Q 3 C O o 2 ° I o s s < O ) OOOOOOOO 0 ) > o _ -a O i * T O i C © © up © r*^ co c o 6©o©o©o© U)lO(O(0NNC0fl0 © CN 05 © 05 © © © CN CN © o © CN o OOO o © o OOo d d o ' oo 0 ) © © © o © © © r- E o o CN o © 05 F o • o’ t o o 1 o i d o 'o ^r-©05©05t“ © mcMT-ooscNO^ d*rh-c\i©cd^rcN t- t- C N C M C M C N C O ©©©©©©©© Z © © i n © T- ^ ^ V C D O l g. I ol S5< . V O ) ^ ^ O ) dl© © © u p r- ~ o in © © o in ©©©©*- r- > o ■ o c 0 0 (Q o m c o c o o o o ' v a o o v a in o o v a 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Time: Y ears s in c e th e beginning o f th e period Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.13 Average Annual Percentage Changes (with 95% Confidence Intervals) in Overall Heart Disease Prevalence, by Sex, by 5-Year Age Group, NHIS 1982-1996 1996 Men (95%Cls) 1987-1996 Women (95%Cls) Age group 50 and over 0.2 ( 0.2 0.3 ) Age group 50 and over -0.2 ( -0.2 , -0.2 50-54 2.2 ( 2.1 23 ) 50-54 -0.9 < -10 , -0.9 55-59 -1.7 ( -1.8 -1.6 ) 55-59 -0.2 ( -0.3 , -0.1 60-64 0.7 ( 0.7 0.8 ) 60-64 0 6 ( 0.6 , 0.7 65-69 -1.7 ( -1.8 -1.7 ) 65-69 0.1 ( o.o , 0.1 70-74 0.0 ( -0.1 0.0 ) 70-74 -0.8 ( -0.8 , -0.7 75-79 2.7 ( 2.6 2.8 ) 75-79 -1.4 ( -1.4 . -1.3 80-84 2 5 ( 2.4 2.6 ) 80-84 1.1 ( 10 , 1.1 85 and over 2.1 ( 2.0 2 2 ) 85 and over 0.1 ( oo , 0.2 1982-1996 1982-1996 (95%Cls) (95%Cls) Age group 50 and over 0.5 ( 0.5 0.5 ) Age group 50 and over -0.1 ( -0.1 . 0.0 ) 50-54 -0.5 ( -0.6 -0.4 ) 50-54 0.0 ( -0 1 , 0 1 ) 55-59 -0.4 ( -0.5 -0.4 ) 55-59 -0.4 ( -0.5 , -0.3 ) 60-64 0.8 ( 0.7 0 8 ) 60-64 -1.0 ( -11 . -0.9 ) 65-69 -1.0 ( -11 -0 9 ) 65-69 -0 5 ( -0 6 , -0 5 ) 70-74 -0.1 ( -0.2 -0.1 ) 70-74 -0 1 ( -0 2 , 0.0 ) 75-79 2.0 ( 2 0 2 1 ) 75-79 -0 9 ( -0 9 , -0.8 ) 80-84 3.3 ( 3.2 3.4 ) 80-84 0 5 ( 0.5 , 0.5 ) 85 and over 0.5 ( 0.4 07 ) 85 and over -0.6 ( -0.7 , -0.5 ) Average percentage change based on log-linear regresson. down to 0.2% in the last 10 years from 1987 to 1996. Looking at individual 5-year age groups, men have different levels of average annual change in overall heart disease prevalence by age. It seems that younger age groups, those less than 75, have declines in prevalence except for those aged 55-59 from 1982 to 1996. On the contrary, older age groups show an increase of overall heart disease prevalence (Table 4.13 and Figure 4.11). In the examination of the broader 10-year age intervals, declines in overall heart disease prevalence for those under age 70 and increases after age 70 are found. Between 1987 and 1996, the average annual decline in prevalence for 10-year interval age groups under age 70 is around 0.5%. In the same period, the average annual percentage increases for men aged 70-79 are 0.9% and 2.4% for those aged 80 and over (Table 4.14 and Figure 4.12). Next, time trends are examined for the prevalence of overall heart disease for men aged 50 and over using individual level data. Table 4.15 shows the odds ratios (ORs) from logistic regression analyses of the prevalence of overall heart problems. For the sample men aged 50 and over, the presence of overall heart disease was regressed on age and years since 1982 and since 1987 (i.e., 15 years and 10 years duration, respectively). In addition to main effects of age and time, an interaction term of age and time is included in an additional regression model. In the results, the main effects of age and time are statistically significant in all models for men. The prevalence rates of overall heart disease for men increase in recent years. As age increases, the odds ratio (OR) of having any kind of heart disease slightly increases in both periods. In addition to main effects, an interaction effect between age and time is significant. This indicates that the time trend differs by age. An additional effect on the increase on prevalence rate as age increases is observed in recent 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.11 Comparison of the Average Annual Percentage Change for Overall Heart Disease Prevalence, by Sex, by 5-Year Age Group, NHIS 1982-1996 Men W om en 4.0 □ 1982-1996 ■1987-1996] ? 2.0 - 2.0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Age 4.0 3.0 S 2.0 O r C o 0 1 8 « a. "5 3 e c < 1.0 0.0 -10 - 2.0 □ 1982-1996 ■ 1987-1996 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Age 0 3 o> Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.14 Average Annual Percentage Changes (with 95% Confidence Intervals) in Overall Heart Disease Prevalence Rates, by Sex, by 10-Year Age Group, NHIS 1982-1996 Men Women 1987-1996 1987-1996 Age group 50 and over 50-59 60-69 70-79 80 and over (95%Cls) Age group 0.2 ( 0.2 0 3 ) 50 and over -0.2 ( -0.2 -0.2 ) -0.5 ( -0 5 -0 4 ) 50-59 -0.5 ( -06 -0.5 ) -0 5 ( -0.5 -0 4 ) 60-69 0 5 ( 0.5 0.6 ) 0.9 < 0.9 10 ) 70-79 -1.1 ( -1.2 -1.1 ) 24 ( 2.3 2 5 ) 80 and over 0.7 ( 0.7 0 8 ) (95%Cls) 1982-1996 1982-1996 Age group 50 and over 50-59 60-69 70-79 80 and over (95%Cls) Age group 0.5 ( 0.5 0 5 ) 50 and over -0.1 ( -0 1 0 0 ) -06 ( -0 7 -0 6 ) 50-59 -0.3 ( -0 3 -02 ) -0.1 ( -0.1 0 0 ) 60-69 -06 ( -0 7 -0.6 ) 0.8 ( 0.7 0.8 ) 70-79 -0.6 ( -0.6 -0.5 ) 2.3 ( 2.2 2 3 ) 80 and over 0.1 ( 0.1 0 1 ) (95%Cls) Average percentage change based on log-linear regresson 0 0 Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.12 Comparison of the Average Annual Percentage Change for Overall Heart Disease Prevalence, by Sex, by 10-Year Age Group, NHIS 1982-1996 Men Women « a t O f O ) C O + + e § 0 1 Q . « 3 C c < □ 1982-1996 ■ 1987-1996 -0 5 - 1.0 o » O ) C O e o » o O f C L T i a -0.5 □ 1982-1996 ■ 1987-1996 j 50-59 60-69 70-79 80 and over 50-59 60-69 70-79 80 and over Age Age Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.15 Odds Ratios from the Logistic Regression Analyses of the Prevalence of Overall Heart Disease, by Sex, NHIS 1982-1996 Men Women 1987-1996 N Age Time Age*Time df -2 Log L 20,718 1.049 *** 1.003 *** 2 1,642.636 0.0001 20,718 1.045 *** 0 946 *** 1 001 * * * 3 1,647,126 00001 1987-1996 N Age Time Age*Time df -2 Log L 26,064 1.047 *** 0.998 *** 2 1,795,619 00001 26,064 1.046 *** 0.994 *** 1.000 * * * 3 1,795,646 0.0001 1982-1996 1982-1996 N Age Time Age*Time df -2 Log L 29,717 1 047 1.003 2 2,155,803 0.0001 29,717 1 040 *** 0 944 *** 1.001/ * * 3 2,172,050 0.0001 _ N Age Time Age*Time df -2 Log L 37,244 1.047 *** 0.995 *** 2 2,645,787 0.0001 37,244 1.047 *** 0.998 *** 1.000/ * * 3 2,645,787 0.0001 *p<0 05, **p<0.01, ***p<0 001 Time: Years since the beginning of the period Age'Time: Interaction of Age and Time o o u > years among men. Figure 4.13 shows a trend of parameter estimates at specific age in two periods. A negatively larger parameter estimate indicates a lower odds ratio of overall heart disease prevalence. The trend patterns are same for the periods of 1982-1996 and 1987-1996. The main and interaction effects are relatively small. Women Annual prevalence of overall heart disease for women aged 50 and over by 5- year age intervals from 1982 to 1996 are shown in Table 4.9. The average prevalence for women aged 50 and over is 16.0%. The prevalence of overall heart disease for women generally increases with age, but again there are some reverses at the higher ages. Average prevalence increases from 7.9% for those aged 50-54 to 28.4% for those aged 85 and over in the period of 1987-1996. To see the trends, the prevalence rates of overall heart disease for different age groups are graphed in Figure 4.14. To smooth out the variation, averaged three-year prevalence rates are also shown in Table 4.10 and Figure 4.15. In addition, the overall heart disease prevalence rates for women with 10-year intervals are also shown in Table 4.11. Regression analyses for the time trends for 5-year age groups in two periods of 1982-1996 and 1987-1996 indicate no statistical significance of the time trend during those periods (Table 4.12). This means that there is no trend in overall heart disease prevalence for women in any of these age groups. The average annual percentage change in overall heart disease prevalence rates for women was a decrease of 0.1% between 1982 and 1996 and of 0.2% in the most recent 10 years. Change in individual age groups varies from small increases to small decreases (Table 4.13 and Figure 4.11). Age groups underage 80 generally have some decreases in overall heart disease prevalence, but there is some recent 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.13 Parameter Estimates for Overall Heart Disease Prevalence with the interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 1987-1996 oo 3 -1 .0 | - 2 0 5 - 3 . 0 3 E -4 .0 s S-5 0 -6.0 MEN Year 1 9 8 7 1 9 8 B 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 -Age 50 - Age 60 - Age 70 -Age 80 . WOMEN Year 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 00 | - 1.0 m 1 -2 0 ?-,» 1-.. c 2 - 5 0 - 6 0 -Age 50 - — Age 60 Age 70 — Age 80 1982-1996 1 9 8 2 -60 -■ 1 9 8 4 1 9 8 6 MEN Year 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 I L * 5 0 -Age 50 — Age 60 - Age 70 - Age 80 1 9 8 2 1 9 8 4 1 9 8 6 1 9 8 8 WOMEN Year 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 0 0 -1 0 -2 0 - 3 0 - 5 0 • - -6 0 - -Age 50 -Age 60 Age 70 - Age 80 <o aBeiuaojad Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.15 Prevalence of Heart Disease for Women, NHIS 1982-1996 (3-Year Average) 40 30 —i- 25 0 1 » a o t a — -II .... : : 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Mid-Year 50-54 55-59 -A-~ 60-64 ■ ■ ■ ■ y C - - 65-69 70-74 75-79 ~ + - 80-84 - 85+ increase in older age groups. The age pattern for the prevalence trend is not as clear for 10-year age groups as 5-year age groups of women (Figure 4.12). Between 1982 and 1996, women under age 80 experienced some decrease in overall heart disease prevalence (Table 4.14). Between 1987 and 1996, women aged 50-59 and 70-79 have shown some decrease while those aged 60-69 and 80 and over have shown some increase in the same period. The largest annual percent decrease is 1.1% for those aged 70-79 and the largest annual increase is 0.7% for those aged 80 and over. Odds ratios (ORs) from logistic regression analyses of the prevalence of overall heart problems for women are shown in Table 4.15. Overall heart disease increases significantly with age among women. The odds of having any kind of heart problem has slightly decreased since 1982 and since 1987. In addition to the main effect of time, an interaction effect between age and time is statistically significant. This indicates that the time trend has a differential effect by age although the odds ratios is 1.000. The interactions of age and time is thus a very small effect (Figure 4.13). 2. Ischemic heart disease prevalence trends Men Prevalence rates for ischemic heart disease (IHD) for men in 5-year age groups for each year from 1982 to 1996 are shown in Table 4.16 and Figure 4.16. The average prevalence rate of IHD for men of aged 50 and over in the period of 1987-1996 is 11.6%. The prevalence rates of IHD for men increase with age until age 75-79 and then drop moderately. The average prevalence rates are 5.3% for age 50-54, 17.5% for age 75-79, and 16.8% for those aged 85 and over in 1987- 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.16 Annual Prevalence of Ischemic Heart Disease (IHD), by Sex, by 5-Year Age Group, NHIS 1982-1996 (Percent with disease) Men Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Average 1982-1996 1987-1996 Age group 50 and over 10.2 10.8 12.3 11.4 11.6 11.7 11.7 11.5 12.1 119 12 5 10.7 11.3 11.7 10.8 115 11.6 50-54 5.0 6.7 6.0 62 8.6 4.5 4.1 4.7 5.2 7.9 4.9 5.5 5.3 5.1 6.0 5.7 5.3 55-59 8.9 10.0 7.6 12.9 9.1 9.5 10.3 8.5 8.4 112 10 0 6.2 5.4 8.7 7.3 9.0 8.6 60-64 13.4 9.3 14.9 7.7 8.8 11.7 12.8 15.1 13 8 9.8 13.8 10.0 13.7 12 9 9.9 11.8 12.3 65-69 11 9 12.8 15.9 13.0 14 5 14.3 15.6 11.0 13 7 12 3 14 8 14 9 12.3 12.6 12.3 13.5 13.4 70-74 13.7 17 3 21.1 16.1 15.5 17.0 12.4 14.7 13.0 14.0 192 14.9 17.6 14.7 14.2 15.7 15.3 75-79 11.0 13.9 12.7 11.5 16.4 14.7 17.7 17.5 20.5 17.9 19 1 14.6 20.9 17.0 15.5 16.2 17.5 80-84 7.9 13.7 19.5 17.8 15.5 16.5 15.1 21.4 20.0 18.6 12.6 16.4 13.9 18.5 21.0 16.5 17.1 85+ 176 5.5 11.8 20.8 19.7 20.1 10.5 12.8 20.3 17.7 11.4 18.0 8.4 23.7 23.2 16.3 16.8 Women Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Average 1982-1996 1987-1996 Age group 50 and over 5.7 7.6 7.0 7.2 6.7 6.6 6.9 6.0 6 5 6.7 7.3 6.0 72 6.6 6.8 6.7 6.7 50-54 3.5 3.7 1.7 2.2 3.6 2 6 2.4 16 2 4 4 0 2.4 2.1 1.7 3.3 3.0 2.7 2.6 55-59 2.5 53 2.7 4 7 3.9 2.6 5.1 2 8 32 2 9 4.1 2.6 4.0 3 7 3.2 3.6 3.4 60-64 46 6 5 7.3 7.7 3.8 4.0 5.7 4.7 5 2 4 7 5.9 4.1 5.8 4.7 3.6 5.2 4.8 65-69 92 10.4 9.6 6.8 8.5 9.4 5.9 54 7 9 7 7 6.7 6.8 7.0 6.8 8.8 7.8 7.3 70-74 4.2 9.4 10.1 11.1 62 9.2 11 5 82 10 3 8.9 6.4 7.4 9.3 8.8 10.1 8.8 9.0 75-79 10 7 10.7 12.1 9.5 14.1 9.8 14.2 10.0 10 0 10.9 15 7 9 5 15.5 11.2 9 8 11.6 11.6 80-84 10.7 10.2 13.3 134 13.1 13.5 8.9 15.2 10.9 113 14 3 11.6 12 6 92 12.4 12.1 12.0 85+ 76 13 2 7.3 106 10 9 137 7 3 14 2 7 9 10 8 18 8 18.5 11.5 113 14.7 120 12.8 < o w Fig u re 4 .1 6 Annual Prevalence o f Ischemic Heart Disease (IHD) fo r Men, N H IS 1982-1996 : rr 0 5 ^ O ) ^ o if) in (ip s n eo + o w o w o i o o m l Ol f i ( D( DNN®® O o in m CM aSejuaojad 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 9 8 2 1 9 8 3 1 9 8 4 1 9 8 5 1 9 8 6 19 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1991 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1996 1996. Prevalence of IHD for men in 5-year age groups vary from year to year more than the overall heart disease prevalence. To smooth the variation in IHD prevalence, average prevalence rates for three years were calculated. These are shown at the mid-year in Table 4.17 and Figure 4.17. In addition, the IHD prevalence rates for men aged 50 and over for 10-year intervals are also shown in Table 4.18. In the regression analyses of IHD prevalence trends for different age groups for 1982 through 1996, the group aged 75-79 shows a statistically significant increase in prevalence rates (Table 4.19). There are no other significant time trends in this period for men. No trend in the prevalence of IHD for any age group is found between 1987 and 1996. The annual change in IHD prevalence for men was average percentage increase of 0.2% between 1982 and 1996. The annual change in IHD prevalence was a decline of 0.5% from 1987 to 1996. Age groups less than 70 show a decline in IHD prevalence except those aged 50-54 while older age groups show an increase prevalence of 2.0% to 2.5% (Table 4.20 and Figure 4.18). Prevalence of IHD for men aged 55-59 decreases from 1987 to 1996 at an annual rate of 4.3%. Increases in IHD prevalence for the age groups 75 and over are seen between 1982 and 1996. Looking at annual changes in prevalence for 10-year age groups for men, younger age groups have been experiencing improvement over the period while older age groups have experienced deterioration. Between 1987 and 1996, the annual percentage change in IHD prevalence was a decline of 1.8% for those aged 50-59 to an increase of 1.0% for those aged 80 and over (Table 4.21 and Figure 4.19) Outcomes from the regression analyses with age and time trend for the IHD prevalence for men in the two periods of 1982-1996 and 1987-1996 indicates that Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.17 Prevalence of Ischemic Heart Disease (IHD), by Sex, by 5-Year Age Group, NHIS 1982-1996 (Average of 3-Year Range) (Percent with disease) Men Range of Years 1982 1984 1983 1985 1984 1986 1985 1987 1986 1988 1987 1989 1988 1990 1989 1991 1990 1992 1991 1993 1992 1994 1993 1995 1994 1996 Average 1982-1996 Age group 50-54 5.9 6.3 6.9 6.4 5.7 4.4 4.7 6.0 6.0 6.1 5.2 5 3 5.5 5.7 55-59 8.9 10 2 9.9 10 5 9.6 9.4 9 0 9.4 9 9 9.1 7.2 6.8 7.1 9 0 60-64 12.7 10.8 10.5 9.4 11.1 13.1 13.9 12.8 12 4 11.1 12.5 12 2 12.2 11.8 65-69 13.5 13.9 14 5 14 0 14.8 13 6 13.5 12.4 13.6 14.0 14.0 13.3 12.4 13.5 70-74 17.4 18.1 17.4 16 2 15.0 14.8 13 3 13.9 15 5 16.1 17.2 15.7 15.5 15.7 75-79 12.6 12.8 13.7 14 3 16.4 16.8 18.5 18.6 19 1 17.2 18.2 17 5 17.7 16.2 80-84 13.7 16.8 17.5 16.7 15.7 17.4 18.7 19.9 16 9 15.6 14.2 16.3 17.4 16.5 85+ 11.4 13.7 17.1 20 3 16.3 14.4 14.9 17.1 16.2 15.5 12.6 17 0 18.9 16.3 W om en Range of Years 1982 1984 1983 1985 1984 1986 1985 1987 1986 1988 1987 1989 1988 1990 1989 1991 1990 1992 1991 1993 1992 1994 1993 1995 1994 1996 Average 1982-1996 Age group 50-54 3.0 2.6 2.5 2.8 2.9 2.2 2 1 2.7 2.9 2.8 2.1 2.4 2.7 2.7 55-59 3.5 4.2 3.8 3 8 3.8 3 5 3.7 2.9 3.4 3.2 3.6 3.5 3.7 3.6 60-64 6.1 7.2 6 3 5.1 4 5 4 7 52 4.9 5 3 4.9 5.3 4.9 4.7 5.2 65-69 9.8 9.0 8 3 8.3 7.9 6 9 64 7.0 7.4 7 1 6.8 6.9 7.5 7.8 70-74 8.0 10.2 9.2 8 9 9.1 9.6 10.0 9.1 8 5 7.6 7.7 8 5 9.3 8.8 75-79 11.2 10.7 11.9 11.1 12.6 114 11.4 10.3 12 0 12.0 I3.5 12.0 12.1 11.6 80-84 11.4 12.4 13 3 13.3 11.8 12.6 11.8 12.5 12 3 12.5 12.9 11.1 11.4 12.1 85+ 9.4 10.5 9.6 119 10.8 118 9.6 10.6 12 1 15.9 16.1 13.6 12.6 12.0 < 0 o o Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.17 Prevalence of Ischemic Heart Disease (IHD) for Men, NHIS 1982-1996 (3-Year Average) 0 > a < o c V a V C L 25 20 15 10 5 0 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 —• — 50-54 __JH— 55-59 - A - 60-64 65-69 — x t— 70-74 75-79 —I — 80-84 ---------- 85+ Mid-Year co co Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.18 Annual Prevalence of Ischemic Heart Disease (IHD), by Sex, by 10-Year Age Group, NHIS 1982-1996 (Percent with disease) Men Average Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1982-1996 1987-1996 Age group 50 and over 10.2 10.8 12.3 11.4 11.6 11.7 11.7 11.5 12.1 11.9 12.5 10 7 11.3 11.7 10.8 11.5 11.6 50-59 6 9 8 3 6.8 9.6 8.8 7.1 7.2 6.6 6 8 9.5 7.3 5.8 5.4 6.7 6.6 7.3 6.9 60-69 12 7 11.0 15.4 10.0 11.3 12.9 14.1 13 1 13.8 10.9 14.3 12.5 13.0 12.8 11.0 12.6 12.8 70-79 12.6 15.9 17.4 14.3 15.9 16.3 14.8 15.9 16.0 15.6 19.2 14.8 19.0 15.7 14.8 15.9 16.2 80 and over 106 11.7 16.6 18.8 164 17.4 13.9 18 8 20.1 18 4 12.3 16 9 12.5 20.0 21.9 16.5 17.0 Women Average Year 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1982-1996 1987-1996 Age group 50 and over 5.7 7.6 7.0 7.2 6.7 6.6 6.9 6.0 6.5 6.7 7.3 6 0 7.2 6.6 6.8 6.7 6.7 50-59 3.0 4 5 22 3 5 3 8 2.6 3.7 22 2 8 3.4 3.2 2 3 2.9 3.5 3.1 3.1 3.0 60-69 6 8 8.4 8.3 7.3 6.1 6.4 5.8 5.0 6 5 6.3 6.3 5.5 6.4 5.7 62 6 5 6.0 70-79 7.0 10.0 11.0 104 9.8 9.5 12.7 89 10 1 9.7 10.2 8 3 11.8 9.9 10 0 10.0 10.1 80 and over 9.6 11.3 11.4 12.7 12.4 13.5 8.4 14 9 9.7 11.2 15.7 13 8 12.2 9.9 13.2 12.1 12.3 o o Table 4 .1 9 Parameter Estimates o f t h e Annual Change i n Ischemic Heart Disease (IHD) Prevalence, b y Sex, b y 5-Year A g e Group, N H IS 1982-1996 e © E o 0) E m o r* O ^ 3 * O C O o o r- C D O o T — 0 ) C O F o ' d o ' 1 o ' o i o o i d WrOinSlflOO tOOJO VOCNO) OOOOOOOO o o o ' o o ' o ’ o ' o' a a a a a « a « : a a a a a a a a t « a « « C OG OC N N- G O C NC O C O C O O T — N - C O C NC O i n ’ l < d C NT- T * T - OOOOOOOO C O 3 C . a o o t f l I X C O C O C O(0 C Oa- O ) o o *- o o O C O p p p o ‘ o * o o o ' C O G O N 1 r- (0 C O T — O C N C No O O C N o ' o ' o ' o ' O a « « i a a a a a a a a a a o C O Is - C O o> C O r - r* r* CO ob £ CN d inininiototninm Q 3 < 0 O o > 5® i a fc< 0 ) > o N* 0 ) *r a > O) N 1 o C m i m i < 9 < ? h - i r- • G O i (9 o m o m o i n O i n i n i n C Oco I s - r- G OG O a 3 <0 O 9 ) i O s < O ) 0 ) > o *D ^ O ^ 0> ^ ® ^ c t o u p (^ 3 CO N - r« j. G O (0 d i n o t n d m d i n ioincocofs -rs -aoco e a> co C O0 ) C Oo C Nr-. C N E C O T “ o o i n i- o o ' o ' 1 o * d o ' o o G Oa > i n O o i n C N C OO ) G Oo C No o G O C No v — O o o o o ’ o o ' o ' o ' o o o OOOOOOOO ( 0 o|o ( A I X nmco^rcoincocN ^OOfNCOOinCD V - o O o ^ r- O 0 0 0 0 0000 ^O^lOOCOCNCO OrO Qr 6 0 0 0 0 0 0 0 gO C N lO O O T T O C O tO " 6 o ' T - co 6 c o ^ co lotnminminmin < 0 O !;< ^ O ) ^ o ) o ) u ?«?t?<¥r T rr o i n o i n o i n IOU)(O(0SN 0 ) > o * o ^ c 0 0 ( 0 o in a o c o < 0 o V - © • oj N< ( 9 > O T 3 ^ 0> ^ O J ^ 0> ^ C i n i n u p up r - > r — c p ( Q d in o in o m d m U)U)(D(OSNOOQO o o o ' V Q . o o’ V a m o o ' v a 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Time: Y ears s in c e th e beginning o f th e period Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.20 Average Annual Percentage Changes (with 95% Confidence Intervals) in Ischemic Heart Disease (IHD) Prevalence, by Sex, by 5-Year Age Group, NHIS 1982-1996 1987-1996 Men (95%Cls) 1987-1996 Women (95%Cls) Age group 50 and over -0.5 ( -0.6 -0.5 ) Age group 50 and over -0.2 ( -0.2 , -0.2 50-54 26 ( 2.5 2.7 ) 50-54 1.8 ( 17 , 2.0 55-59 -4 1 ( -4 3 -4 0 ) 55-59 0.7 ( 0.6 , 0.9 60-64 -1.6 ( -17 -1.5 ) 60-64 -0.9 ( -10 , -0.8 65-69 -11 ( -12 -11 ) 65-69 0.7 ( 06 , 0.8 70-74 0.7 ( 0.6 0 8 ) 70-74 -1.1 ( -12 , -1.0 75-79 0.1 ( 0.0 0.1 ) 75-79 0.5 { 0.3 , 0.6 80-84 0.3 ( 0.2 04 ) 80-84 -0.7 ( -0.8 , -0.5 85 and over 2.5 ( 2.3 2.7 ) 85 and over 3.5 ( 3.3 , 3.7 1982-1996 1982-1996 O (O Average percentage change based on log-linear regresson. (95%Cls) (95%Cls) iV ge group 50 and over 0.2 ( 0.2 0.3 ) Age group 50 and over -0.1 ( -0.2 -0.1 50-54 -0.7 ( -0.8 -0.6 ) 50-54 -0.8 ( -1.0 -0.7 55-59 -2 3 ( -2 4 -2 2 ) 55-59 -0.4 ( -0.5 -0.2 60-64 0.5 ( 0.4 06 ) 60-64 -2.0 ( -21 -19 65-69 -0 4 ( -0.4 -0.3 ) 65-69 -19 ( -2.0 -1.8 70-74 -0.6 ( -0.7 -0 5 ) 70-74 12 ( 11 1.4 75-79 2 9 ( 28 3 0 ) 75-79 0.4 ( 0.3 0.5 80-84 2 4 ( 2.2 2 5 ) 80-84 -0.1 ( -0 2 0.0 85 and over 26 ( 2.4 28 ) 85 and over 3.4 ( 3.2 3.5 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.18 Comparison of the Average Annual Percentage Change for Ischemic Heart Disease Prevalence, by Sex, by 5-Year Age Group, NHIS 1982-1996 Men Women [□1982-1996 ■ 1987-1996 -5.0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Age 4.0 □ 1982-1996 ■ 1987-1996 -5.0 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+ Age o co Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.21 Average Annual Percentage Changes (with 95% Confidence Intervals) for Ischemic Heart Disease (IHD) Prevalence, by Sex, by 5-Year Age Group, NHIS 1982-1996 Men Women 1987-1996 1987-1996 Age group 50 and over 50-59 60-69 70-79 80 and over (95%Cls) Age group -0.5 ( -0.6 -0.5 ) 50 and over -0.2 ( -0.2 -0.2 ) -1.8 ( -1.9 -1.7 ) 50-59 1.2 ( 1.1 1.3 ) -1.3 ( -1.4 -1.3 ) 60-69 0.2 ( 0.2 0.3 ) 0.3 ( 02 0 3 ) 70-79 -0.3 ( -0.4 -0.2 ) 1.0 ( 0.9 1.1 ) 80 and over 0.8 ( 0.7 0 9 ) (95%Cls) 1982-1996 1982-1996 (95%Cls) (95%Cls) Age group Age group 50 and over 0.2 ( 0.2 0.3 ) 50 and over -0.1 CM o 1 's™ *' -0.1 ) 50-59 -1.7 ( -1.8 -1.7 ) 50-59 -0.6 ( -0.8 -0.5 ) 60-69 0.1 ( 0.0 0.2 ) 60-69 -1.9 ( -1.9 -1.8 ) 70-79 0.8 ( 0.7 o 0 0 70-79 0.6 ( 0.6 0.7 ) 80 and over 2 2 ( 2.1 2.3 ) 80 and over 0.9 ( 0 8 1.0 ) Average percentage change based on log-linear regresson. o ■ A . Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.19 Comparison of the Average Annual Percentage Change for Ischemic Heart Disease Prevalence, by Sex, by 10-Year Age Group, NHIS 1982-1996 Men Women □1982-1996 ■ 1987-1996 □ 1982-1996 ■1987-1996] 50-59 60-69 70-79 80 and over 50-59 60-69 70-79 80 and over Age Age o cn both main effects and an interaction effect of age and the time trend are statistically significant, but that the effect is small (Table 4.22). The chance of having ischemic heart disease increases with age in all the equations. The odds of having IHD slightly decreases in both periods of 1982-1996 and 1987-1996. The decline of the IHD prevalence rate is larger in 1982-1996 than that in 1987-1996. Despite a decline indicated by the main effect, the time trend by age interaction indicates an additional effect that contributes to increasing IHD prevalence as age increases. However, the interaction is very small. Figure 4.20 shows a trend of parameter estimates for IHD prevalence at specific age in two periods. A negatively larger parameter estimate indicates a lower odds ratio of IHD prevalence. Since the interaction effects are relatively small, very little time changes by different ages are seen for the IHD prevalence trends. Women The prevalence of ischemic heart disease (IHD) for women aged 50 and over for 5-year age intervals from 1982 through 1996 is illustrated in Figures 4.21. The average prevalence of IHD is 6.7% for both periods of 1982-1996 and 1987-1996. The IHD prevalence increase with age. The prevalence rates are 2.6% for age 50-54 and 12.8% for aged 85 and over in 1987-1996 (Table 4.16). Similar to the trends for men, IHD prevalence varies from year to year. The three-year average prevalence is shown in Table 4.17 and Figure 4.22. In addition, IHD prevalence figures for women in 10-year age groups are presented in Table 4.18. Parameter estimates of the effect of time trends indicate that there are no statistically significant changes in any 5-year age groups for the 1982-1996 period or the 1987-1996 period (Table 4.19). 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.22 Odds Ratios from the Logistic Regression Analyses of the Prevalence of Ischemic Heart Disease (IHD) by Sex, NHIS 1982-1996 Men Women 1987-1996 N Age Time Age*Time df -2 Log L 20,718 1.039 *** 0.991 *** 2 708,386 0.0001 20,718 1.035 *** 0.944 *** 1.001 » » * 3 710,550 0.0001 1987-1996 N Age Time Age*Time df -2 Log L 26,064 1 051 *** 1.00 1 *** 2 1,048,324 0.0001 26,064 1.051 *** 0.995 *** 1.000 *** 3 1,048,353 0.0001 1982-1996 N Age Time Age*Time df -2 Log L 1982-1996 29,717 29,717 37.244 1.049 *** 0 995 1.036 1.028 0.999 0.935 1 .0 0 1 Age*Time 2 1,384,926 878,597 892,097 0 0 0 0 1 0.0001 0 0 0 0 1 37,244 1.042 *** 0.939 *** 1 0 0 1 *** 3 1,392,864 0.0001 */><0.05, **p<0.01, ***p<0.001 o - v j Time: Years since the beginning of the period Age*Time: Interaction of Age and Time Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.20 Parameter Estimates for Ischemic Heart Disease Prevalence with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 1987-1996 M E N W O M E N Y u r 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 r e -z 2 -3 I 5 m “-6 - 7 0 0 0 — 0 \ Z - 0 r — t ° k 0 i— 0 '--- -Age 50 — Age 60 — Age 70 — — Age 80 Year 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 00 S ' 10 |-20 fl’30 1-0 f - 5 0 2 - 6 0 - 7 0 -Age 50 • Age 60 — •— Age 70 — Age 80 1982-1996 men women Yaar Year 1 9 8 2 1 9 8 4 1 9 8 6 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 1 9 8 2 1 9 8 4 1 9 8 6 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 0 .0 ■ > 00 ..................... •1 0 ■ -1 0 ,------------------------------------------------------------------ 5 ; S o 0 0 jj-2 0 S - 3 0 U J 5 -4 0 • -Age 50 — Age 60 — Age 7 0 Age 80 — Age 50 — Age 60 — Age 70 — Age 80; Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.21 Annual Prevalence of Ischemic Heart Disease (IHD) for Women, NHIS 1982-1996 25 20 15 10 x ’j 5 A x--- 0 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 50-54 55-59 ~A~ 60-64 65-69 70-74 75-79 —i— 80-84 ------- 85+ Year Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.22 Prevalence of Ischemic Heart Disease (IHD) for Women, NHIS 1982-1996 (3-Year Average) 25 « 09 a e « a « a 20 15 1 0 A A 5 H fr * 0 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 50-54 . . . . . . . 55-59 A 60-64 -ii— 65-69 70-74 75-79 —I — 80-84 ------- 85+ Mid-Year o Average annual percentage change in IHD prevalence for women decreased by 0.1% between 1982 and 1996 and by 0.2% in the last 10 years. For the age group 85 and over, the annual increase was around 3.5% (Table 4.20 and Figure 4.18). In general, women over 50 experienced decreases in the prevalence of IHD for 10-year age groups. Over the entire period 1982-1996, younger women improved and women over 60 deteriorated in prevalence. In the period of 1987-1996, women aged 50 through 69 and those over 80 experienced an increase in prevalence (Table 4.21 and Figure 4.19). Next, age and time trends in the prevalence of IHD for women aged 50 and over are examined using individual data. The main effects of age and the time trend are statistically significant in all models for women (Table 4.22). As age increases, the chance of having IHD increases. Time trends are different in the two periods for women. Although IHD prevalence increased over the whole period (1982-1996), it decreased in recent years (1987-1996). The interaction of age and time trend indicates that there is a very small effect to change IHD prevalence rates in recent years as age increases (Figure 4.20). C. Other Trends • Onset, Activity/Work Limitations, and Survival Trends One facet in evaluating trends is whether there is a trend in the time since onset of a disease. If the length of time people have heart disease changes, this is important in evaluating time trends. Disability information for those with heart disease also provides us information of how sick people are with heart disease. The trend in disability informs us of how heart disease affects people’s daily lives. A time trend in survival among those with heart disease implies an increase or a decrease in the 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. length of time people live after experiencing heart disease. It also is important in indicating a potential source of influence in the time trend of other heart disease trends such as prevalence and mortality. In this section, trends in time since onset, activity/work limitations, and survival from all causes of death and heart disease for those with and without heart disease are examined for people aged 50 and over. These trends are based on data from the National Health Interview Survey (NHIS). These trends are examined basically for 1982-1996 and 1987-1996. The trends are examined separately tor men and for women. 1. Time since the onset of heart disease Men Among men aged 50 and over who had heart disease, the percentage who experienced onset of heart disease within the last year before the survey and within the last five years between 1982 and 1996 are shown in Table 4.23. On average, 9.4% of men experienced heart disease onset within one year of the interview date between 1987 and 1996. In the same period, 37.3% of men experienced onset of their heart problem within five years before the interview. Trends in the length of time from the date of a heart disease onset up to the interview date from 1982 to 1996 are illustrated in Figure 4.23. The figure indicates few trends among men. The outcomes from logistic regression analyses of the likelihood of having onset of heart disease within a specified time of the interview date for men aged 50 and over from individual data is shown in Table 4.24. Age and time indicating the time since 1982 to 1987 are included as main effects in the initial model and the interaction effect of age and time is added in a second model. These models are 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.23 Percentage of Interviewees who Experienced Heart Disease Onset within 1 and 5 Years before Interview, by Sex, NHIS 1982-1996 Men Women Year Within 1 year Within 5 years Year Within 1 year Within 5 years 1982 10.3 % 36.1 % 1982 11.5 % 37.0 % 1983 11.7 % 40.1 % 1983 10.9 % 36.8 % 1984 9.4 % 38.9 % 1984 9.0 % 40.5 % 1985 8.8 % 41.9 % 1985 8.8 % 39.9 % 1986 8.0 % 44.2 % 1986 9.7 % 41.3 % 1987 9.5 % 40.6 % 1987 9.9 % 34 .5 % 1988 8.4 % 41.8 % 1988 9.6 % 36.7 % 1989 8.5 % 35.5 % 1989 12.2 % 42.4 % 1990 11.0 % 39.3 % 1990 12.0 % 39.2 % 1991 10.4 % 39.6 % 1991 12.0 % 38.6 % 1992 9.4 % 34.3 % 1992 13.7 % 37.4 % 1993 9.5 % 34.1 % 1993 10.0 % 33.6 % 1994 9.5 % 37.6 % 1994 11.0 % 37.0 % 1995 8.3 % 35.1 % 1995 9.0 % 33.4 % 1996 9.6 % 36.2 % 1996 16.2 % 42.5 % Average Average 1987-1996 9.4 % 37.3 % 1987-1996 11.5 % 37.5 % 1982-1996 9.5 % 38.3 % 1982-1996 11.0 % 38.0 % C O Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.23 Percentage of Interviewees who Experienced Heart Disease Onset within 1 and 5 Years before Interview, by Sex, NHIS 1982-1996 50 c 0 S 0 Q . 45 30 f Men - Onset within 1 year Men - Onset within 5 years Women - Onset within 1 year Women - Onset within 5 years 25 20 15 10 5 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Year Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.24 Odds Ratios of Experiencing the Onset of Heart Disease within 1 and 5 Years before Interview, by Sex, NHIS 1982-1996 Men Women 1987-1996 Onset within one year Onset within five years 1987-1996 0.980 0.986 0.994 0.986 Age Time Age*Time 1.119 0.976 0.889 1.000 0.998 1 .00 1 Age Time 14.058 15.649 48.720 51,678 -2 Log L -2L o g L 0.0001 0 0 0 0 1 0.0001 0.0001 Onset within one year 4,196 4,196 1 .0 0 1 1.025 2 4,586 0.0001 0.991 0.892 1.002 3 7,867 0 0001 Onset within five years 4,196 4,196 0.997 1 .0 0 1 2 2,095 0 .0 0 0 1 1.0 0 1 1.069 0.999 3 3,769 0 .0 0 0 1 1982-1996 Onset within one year Onset within five years 1982-1996 1.002 1.020 0.987 0.993 0.988 0.988 Age Time Age*Time 0.997 1.052 0.986 0.986 Time 0.999 1.000 Age Time -2L o gL 18.145 58,167 58,167 -2 Log L 10,674 16,878 0 0 0 0 1 0 .0 0 0 1 0 .0 0 0 1 0 0 0 0 1 Onset within one year 6,001 6,001 0.0001 1.002 1.014 1.000 3 10,695 0.0001 Onset within five years 6,001 6,001 1.000 0.996 2 1,107 0.0001 1.005 1.048 0.999 3 4,653 00001 *p<0.05, **p<0.01, *" p <0.001 Time: Years since the beginning of the period Age*Time: Interaction of Age and Time fitted for 1982-1986 and 1987-1996. All effects except for the main effects of time without the interaction term in 1987-1996 are statistically significant. In all models, the main effects of age and time independently related to the onset of heart disease for men. With higher age, individuals are less likely to have experienced heart disease onset either within the previous one year or within five years. This means that older men have had longer experience with heart disease than younger men. For men, the chance of experiencing heart disease within one year decreased between 1982 and 1987. Interaction models in both periods indicate that time makes an additional negative contribution to heart disease onset within one year as age increases. Figure 4.24 shows a trend of parameter estimates with an interaction effect for the onset of heart disease within one year before interview at specific age in two periods. Younger age groups are expected to experience the onset of heart disease within one year more than those for older age groups in recent years in 1982- 1996 and 1987-1996 due to the interaction effect. In terms of onset within five years, main effects indicate that the chance of experiencing heart disease within five years decreases over years and also decreases with increased age in the periods 1982-1996 and 1987-1996. The interaction model for the period 1987-1996 indicates that time has a positive contribution to heart disease onset within five years with higher age for men. In recent years of the period, the onset of heart disease within five years is expected to be similar in any age among men. Figure 4.25 shows a trend of parameter estimates with an interaction effect for the onset of heart disease within five year before interview at specific age in two periods. 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.24 Parameter Estimates for Experiencing the Onset of Heart Disease within 1 Year Before Interview with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 1987-1996 MEN WOMEN Year 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 0 0 £ - 0 5 -1 0 -1 5 S -2 0 - 2 5 - -Age 50 -Age 60 - Age 70 - Age 80 Y u i 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 00 5 - 0 5 -1 0 iS - 2 .0 - 2 5 Age 80 j Age 50 Age 60 Age 70 1982-1996 1 9 8 2 MEN Yaar 1 9 8 4 1 9 8 6 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 00 - 0 5 • -2 0 I - 2 5 -Age 50 — Age 60 —*— Age 7 0 Age 80 1 9 8 2 oo | - 0 5 m E S -1 0 u i f -1 5 E S £ -2 0 - 2 5 1 9 8 4 1 9 B 6 WOMEN Yaar 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 -Age 50 -Age 60 — Age70 — Age 80, Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.25 Parameter Estimates for Experiencing the Onset of Heart Disease within 5 Years Before Interview with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 1987-1996 m en w om en 0 0 Women Among women aged 50 and over who have a heart disease, the percentage of those who experienced the onset of the disease within one year and within five years between 1982 and 1996 is shown for each interview year in Table 4.23. On average, the percentage of women who experienced onset of heart disease within one year is 11.5% in 1982-1996, and within five years is 37.5%. Trends in the time of a heart disease onset from 1982 to 1996 are shown in Figure 4.23. Trends and levels of the onset for women and men are very similar. The results from the logistic regressions of women at the individual level show statistically significant effects for the main effects of age and time and for the interaction term of age and time in both periods of 1982-1996 and 1987-1996 (Table 4.24). The main effect models indicate that the likelihood of women experiencing heart disease within one year increases with older age. This means that older women are more likely to experience recent onset of heart disease than are younger women. The main effect of time indicates that women who have heart disease are more likely to have gotten it within the previous year. However, the interaction model provides a somewhat different view. The Interaction models in the period 1987-1996 indicate that time makes an additional positive contribution to the situation of having heart disease within one year as age increases among women. There is a crossover around year 1991 for the onset of heart disease within one year among different age groups. Before 1991, younger age groups experienced more frequent onset of heart disease within one year than older age groups, but after 1991 this situation was reversed (Figure 4.24). 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In terms of experiencing heart disease within five years, the signs of the main effects of time and age differ in the periods of 1982-1996 and 1987-1996. In both periods, interaction models indicate that there is an additional effect of time that negatively contributes to experiencing heart disease with five years as age increases. As shown in Figure 4.25, there is a crossover between years 1988 and 1989 for the onset of heart disease within five years among different age groups. Up to year 1988, older age groups experienced the onset of heart disease within one year more than younger age groups, but this situation was reversed after 1989. 2. Limitation of activity/work for those who have heart disease Knowing the presence of disability among those with heart disease is important in evaluating the effect of heart disease on daily life. Disability also serves as an indication of the severity of heart disease. The NHIS data set includes two disability variables that can be linked to heart disease. These are general activity limitation and work limitation. Examinations of limitations are focused on two statuses: no limitation and inability to do their major activity or work. Information on activity limitation evaluation is for those age 50 and over for the years from 1982 to 1996. Information about work limitation among those with heart disease is not available in 1982, so the analysis is for those aged 50-69 from 1983 to 1996. Men The NHIS dataset provides four different disability categories for activity limitation and work limitation among those with heart disease. The percentages of those who are unable to do their major activities and those who are not limited among those with heart disease for each year of the study period are shown in Table 4.25. The trend of activity limitations is also illustrated in a bar graph in Figure 4.26. Men 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.25 Percentages in Specified Disability States for Those Aged 50 and Over with Heart Disease, by Sex, NHIS 1982-1996 Activity Limitations (for aged 50 and over), 1982-1996 Not Limited (includes unknowns) Unable to Perfom Major Activity Year Men Women Year Men Women 1982 27.2 30.3 1982 35.6 30.3 1983 37.8 35.7 1983 33.2 24.4 1984 32.1 38.7 1984 36.2 21.7 1985 32.9 35.6 1985 33.8 24.8 1986 43.6 40.0 1986 30.2 23.4 1987 38.2 37.8 1987 30.5 22.0 1988 40.7 40.7 1988 31.2 22.7 1989 42.3 38.7 1989 26.1 21.2 1990 39.9 39.4 1990 29.3 20.6 1991 46.9 38.4 1991 28.0 22.1 1992 38.5 35.2 1992 28.6 17.5 1993 40.0 39.1 1993 26.5 17.6 1994 42.8 36.3 1994 25.4 18.6 1995 45.1 38.8 1995 26.9 19.3 1996 43.0 39.0 1996 25.3 17.3 Average Average 1987-1996 41.8% 38.3% 1987-1996 27.7% 19.9% 1982-1996 39.6% 37.6% 1982-1996 29.6% 21 5% Work Limitations (for aged 50-69), 1983-1996 Not Limited (includes unknowns) Unable to Work Year Men Women Year Men Women 1983 38.0 42.8 1983 28.3 21.0 1984 33.4 42.5 1984 34.7 20.7 1985 40.9 42.3 1985 30.8 23.0 1986 51.1 53.9 1986 25.8 18.3 1987 43.5 48.1 1987 31.6 19.5 1988 43.6 51.5 1988 31.2 17.2 1989 44.4 50.4 1989 29.0 17.2 1990 43.6 59.0 1990 29.1 16.7 1991 54.2 49.5 1991 24.6 15.7 1992 44.2 51.0 1992 27.8 14.3 1993 44.0 51.2 1993 26.2 15.4 1994 46.1 50.2 1994 26.9 8.4 1995 56.0 52.4 1995 21.7 15.7 1996 52.1 44.3 1996 22.7 15.0 Average Average 1987-1996 47.2% 50.7% 1987-1996 27.1% 15.5% 1983-1996 45.3% 49.1% 1983-1996 27.9% 17.1% 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.26 Activity Limitation Among Men with Heart Disease Aged S O and Over, NHIS 1982-1996 No Activity Limitation vs. Some to More Activity Limitation 1 0 0 ipBi Hi ! i 90 - i l - f i S W A ! M I? 2 0 * - - - - - - - - - - - - - - - 1 0 - - - - - - - - - - - - - - - - 0 ' ^ I * * ~ * I ** ' l | l I i I 1,1 ■ >|L I l | l I | L — I , L .1,1 1,1- 1,1 I I 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Year j □ Not limited P Some limitation to Unable to perform major activity No to Some Activity Limitation vs. Unable for Major Activity 1982 1983 1984 1985 1986 1987 1988 1989 1990 19 9 1 1992 1993 1994 1995 1996 Year □ Not limited to Limited some B Unable to perform major activity 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. aged 50 and over who have heart disease have better functioning or less disability in recent years. On average, 41.8% of men aged 50 and over and with heart disease reported no activity limitation during 1987-1996 while 39.6% of those men reported no limitation during the whole 1982-1996 period. On the other hand, 27.7% of men reported that they were unable to perform their major activity in the period 1987-1996 as compared with 29.6% of those in the period of 1982-1996. The trend in work limitations for those men with heart disease is shown in Table 4.25. Two different limitation status criteria are used separately: no work limitation and inability to work. The trend in work limitation among those with heart disease is illustrated in Figure 4.27. On average, 45.3% of these men reported that they were not limited in their work with heart disease in the period of 1983-1996. In the period of 1987-1996, a higher average rate of 47.2% of men reported no work limitation among them with heart disease. On the other hand, the status of inability in work among them with heart disease has decreased in recent years. Among those with heart disease, the average percentage of men who reported that they were not able to work was 27.9% in 1983-1996 and 27.1% in 1987-1996. To examine the trend in activity limitation, logistic regression analyses of individual data are performed with age and time trend as the independent variables. An interaction term of age and time trend is also included. Results of these examinations are shown in Table 4.26. Results show that the time trend is one of decreasing disability. Main effect models indicate that the status of no activity limitation has increased in recent years. The condition of no activity limitation becomes less prevalent with age in both periods of 1982-1996 and 1987-1996. Interaction models show that time makes a negative contribution to the status of no 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.27 Work Limitation Among Men with Heart Disease Aged 50-69, NHIS 1983-1996 No Work Limitation vs. Some or More Work Limitation s a w 8) 60 f 3 5 50 I < 0 30 2 0 1 0 i ! M M a 11 m 1 1 1 m s w w m IS I:-# : • x- kS- l l s « w il: l|i; 1 i M S S W s & W If s i S S iS : ass i l l - 1 . 4 hs' i | | | i i m U £ iJ S S t B S 1 1 1 m §!|;L. ill fill m • $ > : # | | | 11 S S S S 1983 1984 1985 1986 1987 1988 1989 1990 19 91 1992 1993 1994 1995 1996 Year □ Not limited QSome limitation to Unable to work No to Some Work Limitation vs. Unable to Work 1 0 0 90 « S P s £ o . 1 80 - 70 60 50 40 4 - 30 1 2 0 | 1 0 I 1 I I I I I I I I I I I l i l t I I I I I I I I I I I 1 . 1 I I l l . l _ l _ l — ! ■ ! I 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Year j □ Not limited to Limited in other activity B Unable to work j 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.26 Odds Ratios of Activity Limitation Status for Individuals with Heart Disease Aged 50 and Over, by Sex, NHIS 1982-1996 Men 1987-1996 N Age Time Age*Time df -2 L o g L No Activity Limitation 0.0001 0.0001 Unable for Major Activity 3,914 0.995 0.997 Age*T ime 11,060 23,152 •2 Log L 3,914 0.987 0.975 2 37,225 0.0001 3,914 0.979 0.866 *** 1.002 *** 3 41,298 0.0001 1982-1996_ N Age Time Age*Time 0.996 1.006 0.999 Age*Time 73,992 83,231 0.0001 0.0001 5,534 0.989 0.967 2 92,730 0.0001 5,534 0.987 *** 0.952 *** 1.000 *** 3 92,999 0.0001 Women 1987-1996 No Activity Limitation Unable for Major Activity Age Time Age*Time 0.997 0.983 Age*Time 1.000 60,582 60,622 -2 L o g L -2 L o g L 0.0001 0.0001 4,196 1.009 *** 0.965 *** 2 28,011 0.0001 4,196 1.000 0.836 *** 1.002 3 33,319 0.0001 1982-1996 6,001 Age Time Age*Time 0.988 Age*Time 1.000 82,859 84,234 -2 L o g L -2 L o g L 0.0001 0.0001 6,001 1.009 0.961 2 34,955 0.0001 6,001 1.005 0.920 1.001 3 36,738 0.0001 •p <0.05, "p <0.01, ***p <0.001 Time: Years since the beginning of the period Age'Time (Interaction of age and time) References: No activity limitation: Unable to perform major activity, limited in kind/amount major activity, and limited in other activities Unable for major activity: Limited in kind/amount major activity, limited in other activities, and not limited 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. activity limitation as age increases in the periods of 1982-1996 and 1987-1996. Figure 4.28 shows a trend of parameter estimates for the status of no activity limitation among those with heart disease with an interaction effect at specific age in two periods. There is a crossover in the effect of time for the status of no activity limitation among different age groups in the periods of 1982-1996 (between 1986 and 1987) and 1987-1996 (between 1989 and 1990). After the year of crossover, younger age groups report no activity limitation status more than older age groups do. In terms of the status of inability to perform a major activity, main effect models indicate that this status decreases in recent years in both periods. They also indicate that the status of being unable to perform a major activity is generally higher among younger age groups in both periods. Interaction models reveal that time additionally makes a positive contribution to this condition as age increases in these periods. Although the main effects and the interaction effects are statistically significant in all models, the effect size was very small especially in 1982-1996. Figure 4.29 shows a trend of parameter estimates for the condition of being unable to perform a major activity among those with heart disease with an interaction effect at specific age in two periods. In the examination period 1987-1996, the condition of being unable to perform a major activity becomes similar among different age groups in men. In terms of work limitation, both the main effects of age and time and the interaction effect are statistically significant in all models (Table 4.27). Work limitation among those with heart disease is ameliorated in recent years. Odds ratios from main effect models indicate that men aged 50-69 have more chance of being no work limitation status in later years. The main effect of the time trend indicates that the condition of no work limitation increases in recent years. Interaction effect models 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.28 Parameter Estimates for the Status of No Activity Limitation Among Those with Heart Disease with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 1987-1996 MEN WOMEN Year 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 0 5 0 .0 -1 .0 Age 80 Age 60 Age 70 Age 50 Year 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 1 0 | 0 5 1 " 00 a S - 0 5 a - 1 0 -Age 50 —1 •— A g e60 — Age 70 -Age 80 1982-1996 MEN Year 1 9 8 2 1 9 B 4 1 9 8 6 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 1 .0 0 . 5 0 .0 S - 0 5 1 0 Age 60 — Age 70 —— Age 80 Age 50 WOMEN Year 1 9 8 2 1 9 8 4 1 9 8 6 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 - 1 0 - Age 50 —•— Age 60 —*— Age 70 —-— Age 80; Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.29 Parameter Estimates for the Status of Unable to Perform A Major Activity Among Those with Heart Dise with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 1987-1996 MEN WOMEN Year 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 0 5 i-20 0 . - 2 5 - 3 0 -Age 50 -Age 60 — Age 70 -A ge 80 Year 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 0 5 g - 0 5 1-10 S -1 5 1-20 £ - 2 5 Age 60 Age 70 Age 80 Age 50 I 1982-1996 MEN WOMEN 1 9 8 2 1 9 8 4 1 9 8 6 Year 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 0 5 a - 0 5 1 5 I -2 0 - 3 0 -Age 50 -Age 60 — Age 70 -Age 80 1 9 8 2 Year 1 9 8 4 1 9 8 6 1 9 8 8 1 9 9 0 1 9 9 2 1 9 9 4 1 9 9 6 0 5 00 5 20 2 5 3 0 -Age 50 — Age 60 — Age70 Age 80 ro CD Table 4.27 Odds Ratios of Work Limitation Status for Individuals with Heart Disease Aged 50-69, by Sex, NHIS 1983-1996 Men No Work Limitation Unable to Work 1987-1996___________________________ N 2 ,231 2,231 N 2,231 2,231 Age 0.942 *** 0.944 *** Age 1.039 *** 1.046 *** Time 1.036 *** 1.066 *** Time 0.954 *** 1.047 *** Age*Time 1.000 *** Age*Time 0.998 *** d f 2 3 df 2 3 -2 L o g L 148,869 148,932 -2 L o g L 63,371 63,883 P 0.0001 0.0001 P 0.0001 0.0001 13-1996 N 3,024 3,024 N 3,024 3,024 Age 0.947 *** 0.954 *** Age 1.029 1.013 *** Time 1.043 *** 1.118 *** Time 0.969 *** 0.834 *** Age'Time 0.999 *** Age'Time 1.002 d f 2 3 df 2 3 -2 L o g L 201,011 202,092 -2 L o g L 56,886 60,950 P 0.0001 0.0001 P 0.0001 0.0001 Women 1987-1996 No Work Limitation Unable to Work 0.987 0.808 17,348 20,619 0.0001 0.0001 2,070 1.025 *** 0.955 *** 2 21,722 0.0001 2,070 1.022 * ’ * 0.919 *** 1 .001 *** 3 21.782 0.0001 1983-1996_ N Age Time Age*Time 1.018 1.004 Age*Time 62,658 76.897 0.0001 0.0001 2,843 1.039 *** 0.952 *** 2 76,708 0.0001 2,843 1.060 1.168 0.997 3 81,696 0.0001 *p<0.05, ” p<0.01, " ‘p <0.001 Tim e: Y ears since th e b eg in n in g of the p e rio d Age'Tim e (In teractio n of age an d tim e) References: N o work lim itatio n : Unable to w ork, lim ited in kind/am ount of w ork, and lim ited in other activities Unable to w ork: L im ited in kind/am ount of work, lim ited in other activities, and not lim ited 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reveal different contributions to no work limitation status. Figure 4.30 shows a trend of parameter estimates for the status of no work limitation among those with heart disease with an interaction effect at specific age in two periods. In recent years, younger men show a slightly different change in the parameter estimates as compared with older men. In terms of the status of being unable to work, main effect models indicate that its status decreases among increased age groups in both the 1983-1996 period and the 1987-1996 period. Figure 4.31 shows a trend of parameter estimates for the status of unable to work among those with heart disease with an interaction effect at specific age in two periods. Interaction effect models show different time trends in the status of being unable to work by age. While the interaction effect is positive in the period of 1983-1996, it is negative in recent years. Women The percentages of women aged 50 and over with heart disease by activity limitation statuses in the periods of 1982-1996 and 1987-1996 are shown in Table 4.25. The trend in disability status is also shown in Figure 4.32. While 37.6% of women aged 50 and over reported that they had no activity limitation in 1982-1996, the average percentage increased to 38.3% in 1987-1997. In 1982-1996, the average percentage of women with heart disease reporting that they were unable to perform their major activity is 21.5%. A smaller percentage of 19.9% is in the same category in 1987-1996. Trends in work limitation for women aged 50-69 with heart disease are shown in Table 4.25 and Figure 4.33. On average, 50.7% of women with heart disease reported no limitation in work ability in 1987-1996, while 49.1% of women reported the same status in 1983-1996. At the other end of the continuum, 15.5% of women 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.30 Parameter Estimates for the Status of No Work Limitation Among Those with Heart Disease with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 1987-1996 MEN WOMEN 4 0 £ § 3 0 i f»1.0 S o.o ; 4 0 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 Year • — Age 50 — Age 60 —• Age 70 —— Age 80 3 0 m u i h . I E S a a . 2 0 00 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 Year -Age 50 • Age 60 — Age70 -A ge 80 1982-1996 MEN 4 0 S n 3 .0 E 1 5 E S m a . oo ; 1 9 B 3 1 9 8 4 1 9 8 5 1 9 B 6 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 Year — Age 50 —— Age 60 —— Age 70 — Age 80 4 0 WOMEN 3 0 20 10 00 1 9 8 3 1 9 8 4 1 9 8 5 1 9 8 6 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 Year -Age 50 -Age 60 — < » — Age 70 — Age 80 GO Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.31 Parameter Estimates for the Status of Unable to Work Among Those with Heart Oisease with the Interaction of Age and Time at Specific Age, by Sex, NHIS 1982-1996 1987-1996 MEN WOMEN Year 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 -Age 50 Age 60 - Age 70 —— Age 80 Year 1 9 8 7 1 9 B B 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 00 -Age 50 -Age 60 -Age 70 - Age 80 1982-1996 MEN WOMEN CO ro Year 1 9 8 3 1 9 8 4 1 9 8 5 1 9 8 6 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 00 Year 1 9 8 3 1 9 8 4 1 9 8 5 1 9 8 6 1 9 8 7 1 9 8 8 1 9 8 9 1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 0 0 2 -2 0 i i -4 0 •50 -Age 50 -Age 60 -Age 70 - Age 80 3- 2 0 - 3 0 • - 4 0 •5 0 -Age 50 -Age 60 Age 70 - Age 80 Figure 4.32 Activity Limitation Among Women with Heart Disease Aged S O and Over, NHIS 1982-1996 No Activity Limitation vs. Some to More Activity Limitation 100 90 80 70 8) 60 s § 50 s £ 40 30 2 0 1 0 0 1982 1983 1984 1985 1986 1987 1988 1989 1990 1 9 9 1 1992 1993 1994 1995 1996 Year □ N o t lim ite d □ Som e lim ita tio n to U n ab le to p erfo rm m ajor a c tiv ity I No to Some Activity Limitation vs. Unable for Major Activity 100 90 80 70 - n - R - B - R 1 60 1 50 if £ 40 30 I 20 - 10 I 1982 1983 1984 1985 1986 1987 1988 1989 1990 19 9 1 1992 1993 1994 1995 1996 Year □ N ot lim ite d to L im ited som e B U nable to p e rfo rm m ajo r a c tiv ity Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.33 Work Limitation Among Women with Heart Disease Aged 50-69, NHIS 1983-1996 No Work Limitation vs. Some or More Work Limitation w a s s is s 3 0 - - - - - - - - - - - - - - 20 r _ — _ — _ — — — 1 0 - - — - i o J — U — U — U — U — U — LI— U — LJ— U — U — U — LI— LI— 1983 1984 1965 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Year | □ N o t lim ite d O S om e lim ita tio n to U nable to w ork | N o to Some Work Limitation vs. Unable to Work 100 90 80 70 1 60 c 1 1 50 u s a 40 30 20 10 0 I I I I I I I I I I I J I I I l _ l l - l _ l - l „ U U I a i ■ 1983 1984 1985 1986 1987 1988 1989 1990 19 9 1 1992 1993 1994 1995 1996 Year D N ot lim ite d to L im ite d in o th er a ctivity B U nable to w o rk 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with heart disease reported that they were unable to work in 1987-1996 and 17% reported this in 1983-1996. Results from logistic regression models indicated that the status of no activity limitation decreases with age even when the interaction term is included (Table 4.26). The state of unable to work decreases in recent years among women with heart disease. The main effect models for no activity limitation among women aged 50 and over indicate different time trend patterns between the two periods of 1982-1996 and 1987-1996. Between 1982-1996, the condition of no activity limitation increased. However, this state decreased in the years 1987-1996. The main effect of age reveals that the condition of no activity limitation decreases with age in both periods. The interaction of time and age is seen, but its effect is very small. Figure 4.28 shows a trend of parameter estimates for the status of no activity limitation among those with heart disease with an interaction effect at specific age in two periods. In terms of the condition of being unable to perform major activity, the main effect of the time trend indicates that the condition decreases in recent years in the periods of 1982-1996 and 1987-1996. The main effect of age indicates that being unable to perform major activity is higher among older women. Interaction models for both periods indicate that time has an additional positive contribution to this condition as age increases among women. Figure 4.28 shows different trends of parameter estimates for the status of unable to perform a major activity among those with heart disease with an interaction effect at specific age in two periods. In terms of work limitation for women aged 50-69, the time trend shows different results in the two periods (Table 4.27). While the long period from 1982 to 1996 has an increase in the condition of no work limitation, the short period from 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1987 to 1996 exhibits a decrease in the condition among women. Interactions of the time trend and age are significant in both periods. They indicate that time has an additional positive contribution as age increases. Figure 4.30 shows a trend of parameter estimates for the condition of no work limitation among those with heart disease with an interaction effect at specific age in two periods. Main effect models for the condition of unable to work indicate the recent decreases in time trends in both periods of 1982-1996 and 1987-1996. Significant interactions of time trend and age explain that time has an additional negative contribution at the older ages in the long term (1983-1996) but a positive contribution in the shorter term (1987-1996) as age increases. These interaction effects indicate different trends in severe work limitation among different age groups in women (Figure 4.31). 3. Survival trends Heart disease survival trends are examined using the data from the National Health Interview Survey (NHIS) with the matched Multiple Cause of Death Data File for the survey years 1986-1994. These data provide individual mortality information for people interviewed in the survey. By examining the trend in the likelihood of death for those with and without heart disease within a certain period of time after the interview year, we can evaluate trends for those with and without heart disease. In this study, all deaths and deaths due to heart disease are examined. Deaths occurring within 1.5-year and 2.5-year intervals after the interview are examined. Men Overall mortality information by different heart disease status at the time of interview for men aged 50 and over is shown annually for the 1986 to 1994 period in Table 4.28 and Figure 4.34. The average mortality rate for those with heart disease 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.28 Percentages of People Deceased by Time since Interview, by Sex, NHIS 1986-1994 Men Time since interview Death within 1.5 years Death within 2.5 years Without With Without With Interview year Heart Disease Heart Disease Heart Disease Heart Disease 1986 2.1 6.1 4.1 11.7 1987 2.8 11.2 5.6 16.2 1988 2.7 7.7 4.4 13.9 1989 2.2 7.8 4.2 11.3 1990 3.0 8.5 5.1 16.6 1991 2.6 4.7 5.4 9.0 1992 3.3 9.7 6.2 15.0 1993 3.3 7.8 5.9 12.5 1994 2.9 8.3 Average 2.8% 8.0% 4.5% 11.8% Women Time since interview Death within 1.5 years Death within 2.5 years Without With Without With Interview year Heart Disease Heart Disease Heart Disease Heart Disease 1986 1.4 6.8 3.1 10.5 1987 2.0 4.9 3.9 9.6 1988 1.8 5.7 3.1 7.7 1989 1.9 4.4 3.3 8.4 1990 1.8 5.8 3.2 8.4 1991 1.8 6.7 3.5 10.2 1992 2.4 4.7 3.9 10.0 1993 1.9 5.0 3.5 8.0 1994 2.1 5.3 Average 1.9% 5.5% 3.0% 8.1% 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.34 Percentage Deceased among Men Aged 50 and Over by Time since NHIS Interview, by interview Year Death within 1.5 years Death within 2.5 years n 18 S S I ■ I I I 1986 1987 1968 1989 1990 1991 1992 1993 1994 Interview Year I dWithout Heart Disease ■With Heart Disease 1986 1987 1988 1989 1990 1991 1992 1993 Interview Year □Without Heart Disease BWith Heart Disease j CO C O is higher than the rate for those without heart disease at the time of interview. The average percentage dying within 1.5 years of interview is 2.8% for men who reported no heart disease and 8.0% for those who reported heart disease. The average death rate within 2.5 years is 4.5% for men without heart disease and 11.8% for those with heart disease. Odds ratios from the logistic regression analyses for men indicate that the death within both 1.5 years and 2.5 years is associated with age (Table 4.29). There is no time trend in the likelihood of death within 1.5 years; however, death rates within 2.5 years decrease with time. The patterns of change in 2.5-year death likelihood are found among those with and without heart disease. Deaths due to heart disease increase with age. This pattern is the same men with and without heart disease. Deaths due to heart disease occurring within 2.5 years decrease among those with heart disease in recent years. Women Percentages of death by time since interview for women are shown in Table 4.30 and Figure 4.35. The average mortality rate is higher for those with heart disease than for those without heart disease at the time of interview. The average death rate within 1.5 years for women without heart disease is 1.9% and 5.5% for those with heart disease. The average death rate within 2.5 years is 3.0% for those without heart disease and 8.1% for women with heart disease. Odds ratios from logistic regression analyses for women indicate that the rates of both 1.5-year deaths and 2.5-year deaths due to heart disease are higher with advancing age. As for men, there is a decrease in 2.5-year death rates (Table 4.30). Statistically significant changes are seen for 2.5-year deaths due to heart disease and 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.29 Odds Ratios for Deaths During Specific Time After Interview for Men, by Heart Disease Conditions, by Period, NHIS 1986-1994 Within 1.5 years Within 2.5 years Death in General (Among All) N 18.978 18.890 Age 1.08 *** 1.08 *** Duration 1.02 0.94 *** df 2 2 -2 Loo L 475.65 705.24 P 0.0001 0.0001 (Among those N 15.407 13.604 Without Heart Disease) Age 1.09 *** 1.09 *** Duration 1.04 0.96 ** df 2 2 -2 Log L 332.42 471.77 P 0.0001 0.0001 (Among those N 3.571 3.128 With Heart Disease) Age 1.06 *** 1.06 ” * Duration 0.99 0.90 *** df 2 2 -2 Loo L 75.48 141.39 P 0.0001 0.0001 Death due to Heart Disease (Among All) N 18.978 18.890 Age 1.09 *** 1.09 *** Duration 1.01 0.93 *** df 2 2 -2 Loa L 189.64 297.41 P 0.0001 0.0001 (Among those N 15.407 13.604 Without Heart Disease) Age 1.10 — 1.10 — Duration 1.04 0.96 df 2 2 -2 Loa L 105.93 191.64 P 0.0001 0.0001 (Among those N 3.571 3.128 With Heart Disease) Age 1.05 *** 1.05 *** Duration 0.97 0.90 *** df 2 2 -2 Loa L 36.76 54.91 P 0.0001 0.0001 *p<0.05, "p<0.01, *"p<0.001 1.5-year (am ong 1986-1994 interview ees) 2.5-year (am ong 1986-1993 interview ees) D uration: Years since 1986 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.30 Odds Ratios for Deaths During Specific Time After Interview for Women, by Heart Disease Conditions, by Period, NHIS 1986-1994 Within 1.5 years Within 2.5 years Death in General (Among All) N 23.920 23.850 Age 1.08 *** 1.07 *** Duration 1.01 0.92 *** df 2 2 -2 Loa L 351.55 533.52 P 0.0001 0.0001 (Among those N 20.067 17.727 Without Heart Disease) Age 1.08 **’ 1.08 *** Duration 1.03 0.93 *** df 2 2 -2 Loa L 240.89 362.85 P r 0.0001 0.0001 (Among those N 3.853 3.403 With Heart Disease) Age 1.05 *** 1.05 *** Duration 0.97 0.90 *** df 2 2 -2 Loa L 53.32 90.91 P 0.0001 0.0001 Death due to Heart Disease (Among All) N 23.920 23.850 Age 1.09 *** 1.09 *** D uration 0.95 0.88 *** df 2 2 -2 Loa L 162.32 289.42 P 0.0001 0.0001 (Among those N 20.067 17.727 Without Heart Disease) Age 1.10 **’* 1.10 — Duration 0.94 0.88 *** df 2 2 -2 Loa L 116.64 200.93 P 0.0001 0.0001 (Among those N 3.853 3.403 With Heart Disease) Age 1.04 *** 1.05 *** D uration 0.96 0.88 *** df 2 2 -2 Loa L 18.89 45.61 P 0.0001 0.0001 *p<0.05, **p<0.01, ***p<0.001 1.5-year (am ong 1986-1994 interview ees) 2.5-year (am ong 1986-1993 interview ees) D uration: Years since 1986 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 4.35 Percentage Deceased among Women Aged 50 and Over by Time since NHIS Interview, by Interview Year Death within 1.5 years Death within 2.5 years 18 16 14 1 2 1 0 & m C L 8 * l l l l l I l H 1986 1987 1988 1989 1990 1991 1992 1993 1994 Interview Year I DWithout Heart Disease ■With Hreart Disease 18 16 14 12 1986 1987 1988 1989 1990 1991 1992 1993 Interview Year □Without Heart Disease BWith Heart Disease -u ro all deaths among those with and without heart disease at the interview. There is no trend in the 1.5-year death rates due to all conditions or heart disease among women with and without heart disease. D. Summary In the time trend analyses, both overall heart disease and myocardial infarction (Ml) mortality rates have declined continuously for three decades. The decline rates are faster in the most recent decade than those in the previous three decades for older men and women. Men show faster annual percentage declines in overall heart disease and Ml than women, except for the Ml mortality change for older age groups (aged 70 and over) over the last three decades. While the overall heart disease prevalence rate has increased in recent years, the ischemic heart disease (IHD) prevalence rate has decreased among men in each period examined. Among women, the pattern is different. While the overall heart disease prevalence rate has decreased in recent years, the IHD prevalence rate has increased in the most recent decade after decreasing in the previous three decades. Age and time trend interactions are seen in every model. In terms of the time trend between the onset of heart disease and the interview date, men and women have different trends. Men experienced declines in the likelihood of having heart disease onset within a year and within five years in the most recent decade. On the other hand, women show increases in experiencing heart disease onset within a year and within five years in the most recent decade but not for the long examination period (interview years 1982-1996). The interaction between age and the trend is statistically significant for both men and women. 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In terms of disability, both men and women with heart disease tend to show decreased level of disability and work limitation in recent years. More men with heart disease are free from activity limitation and work limitation in recent years. Additionally, severe activity and work limitations have decreased in recent years. For women, the results are almost the same except for the ten-year examination period (1987-1996). In this period, the status of being free from activity and work limitations has decreased for women with heart disease. The time trend in disability also has a significant interaction effect with age for both sexes. Death rates from all causes for those with heart disease have decreased in recent years for men and for women. Death rates due to heart disease have also decreased in recent years for both sexes, but women have experienced a larger decline than men. 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER V RESULTS - HAZARD MODELS, TRANSITION SCHEDULES, AND MULTISTATE LIFE TABLES Multistate life tables, created using the outcomes from hazard models, produce two important outcomes: (1) expected years of life with and without heart disease and (2) estimated implied heart disease prevalence rates. Expected years of life with and without heart disease provides information on how men and women experience heart disease and how long on average they live with and without heart disease over their expected life cycle. Estimated implied prevalence rates inform us how the population will experience heart disease if these rates continue. Population- based and status-based estimation techniques are used to show these two outcomes. This chapter presents the results from transition analyses and a multistate life table based on the merged dataset from the HRS and AHEAD samples. It starts with a description of the transitions in heart disease statuses between Wave I and Wave II. First, the observed number of events and the percent making state transitions from Wave I to Wave II for men and women are described. Second, outcomes from hazard analyses of the transitions between states of heart disease and death for each sex are reported. Third, two main outcomes from the multistate life table created using the transition rates are investigated. One outcome is the expected years of life with and without heart disease. The other outcome is the estimated implied prevalence rates of disease of the heart. Each outcome is derived using population-based and status-based life tables. 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A. Transition in heart disease status The observed heart disease status at baseline (Wave I) and transitions between Wave I and Wave II for men and for women are described in numbers and percentages in this section. The observed heart disease prevalence at Wave I and incidence of heart disease between Wave I and Wave II with age are also shown. Men Numbers of events and percentages making transitions across the different heart disease states from Wave I to Wave II in the merged dataset of HRS and AHEAD are shown in Table 5.1. For men, 6,107 (76.8%) reported no heart disease experience, 1,111 (14.0%) reported heart disease without a heart attack (or myocardial infarction: Ml), and 737 (9.3%) had a heart attack before the Wave I interview. From the Wave I to Wave II, most men stay in the same heart disease categories (80 to 90%). Among men who reported no heart disease at the first interview, 205 men (3.4%) experienced heart disease without a heart attack and 104 men (1.7%) had a heart attack between the waves (Table 5.1). Among men who reported a heart problem without a heart attack at the interview, 55 individuals (5.0%) experienced a heart attack between Wave I and Wave II. Between the waves, 470 (5.9%) men died. Among men with heart disease but without a heart attack, 11.8% were deceased by Wave II. The second highest death percentage is those in the heart attack category (9.5%). The observed prevalence rate of heart disease including heart attacks for men aged 50 and over increases from 12.0% for those aged 50-54 to 40.4% for those aged 80-84, then the rate decreases in the more advanced ages (Table 5.2 and 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 5.1 Observed Numbers and Percentages for Heart Disease Transition, HRS/AHEAD Wave I - Wave II, by Sex Men Numbers s iiililllifll No Heart Heart Heart Death Disease Disease Attack No Heart Disease Heart Disease Heart Attack 6,107 1,111 737 5,529 205 104 269 924 55 131 666 70 7,955 5,529 1,130 826 470 100.0% 100.0% 100.0% 100.0% Women Numbers W ave) l i i i i i i i l i s No Heart Disease Heart Disease Heart Attack Death No Heart Disease 8,096 7,406 292 85 313 Heart Disease 1,498 — 1,297 54 147 Heart Attack 427 - - 371 57 10,021 7,406 1,589 509 517 100.0% 100.0% 100.0% 100.0% Heart disease does not include a heart attack. Heart Attack: Heart attack or myocardial infarction (Ml) Some numbers shown in percentage may not be summed up to 100 due to rounding. l i i i i i i i i ll Percentages No Heart Disease Heart Disease Heart Attack Death No Heart Disease 80.8% 91.5% 3.6% 1.0% 3.9% Heart Disease 14.9% — 86.6% 3.6% 9.8% Heart Attack 4.3% - - 86.8% 13.3% 4 00.0% 73.9% 15.9% 5.1% 5.2% Percentages j v» m m m m a t d m m m M m m m m m m m No Heart Heart Heart Death Disease Disease Attack No Heart Disease Heart Disease Heart Attack 76.8% 14.0% 9.3% 90.5% 3.4% 1.7% 4.4% 83.2% 5.0% 11.8% 90.4% 9.5% 100.0% 69.5% 14.2% 10.4% 5.9% Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 5.2 Observed Prevalence of Heart Disease at Wave I, by Sex, by 5-Year Age Group, HRS/AHEAD Men Age Group Overall Heart Disease Heart Attack 5 0 -5 4 12.0 % 6.7 % 5 5 -5 9 16.4 9.3 6 0 -6 4 20.7 11.5 6 5 -6 9 26.1 14.7 7 0 -7 4 34.1 7.3 7 5 -7 9 36.5 9.7 8 0 -8 4 40.4 11.2 8 5 -8 9 30.7 7.2 9 0 -9 4 34.1 11.3 Overall heart disease includes a heart attack Women Age Group Overall Heart Disease Heart Attack 5 0 -5 4 8.3 % 2.0 % 5 5 -5 9 11.7 3.4 6 0 -6 4 14.1 4.0 6 5 -6 9 19.8 4.4 7 0 -7 4 24.3 5.3 7 5 -7 9 28.7 4.7 8 0 -8 4 33.1 7.3 8 5 -8 9 36.2 6.5 9 0 -9 4 37.3 7.4 o o Figure 5.1). In terms of heart attack, the prevalence rate for men varies across ages (Table 5.2 and Figure 5.1). In terms of heart disease incidence, the two-year incidence rate of heart disease without a heart attack increases with age from 2.8% for those aged 50-54 to 8.4% for those aged 75-79, then the rate decreases in advanced ages for men (Table 5.3 and Figure 5.2). The two-year incidence of a heart attack is less related to age. It increases from 1.2% for those 50-54 to 3.8% for those aged 70-74, then the rate decreases after age 75 (Table 5.3). Women Observed numbers of transitions and the percentages making the different transition status from Wave I to Wave II for women are shown in Table 5.1. Among 10,021 women at Wave I, 8,096 (80.8%) reported no heart disease, 1,498 (14.9%) reported heart disease but without a heart attack, and 427 (4.3%) reported that they had had a heart attack. From Wave I to Wave II, most women stay in the same heart disease state (87 to 92%). Among women who reported no heart disease at Wave I, 292 (3.6%) experienced heart disease without a heart attack and 85 (1.0%) had a heart attack between the waves (Table 5.1). Among women who reported a heart problem but no heart attack at the first interview, 54 (3.6%) experienced a heart attack between the two waves. Between the waves, 517 women (5.2%) died. Those from heart attack to death were most frequent (13.3%). The second highest death rate was for those with heart disease but without a heart attack (9.8%). The prevalence of overall heart disease including Ml increases with age from 8.3% for those aged 50-54 to 37.3% for those aged 90-94 among women (Table 5.2 and Figure 5.1). Heart attack prevalence also increases with age among women (Figure 5.1). The rate gradually increases from 2.0% for those aged 50-54 to 7.4% 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.1 Observed Prevalence of Heart Disease at Wave I, by Sex, HRS/AHEAD 45 40 30 A * c 0 ) 2 a ' O ------------ -O' a 50 - 54 55 - 59 60 - 64 65 - 69 70 - 74 75 - 79 80 - 84 85 - 89 90 - 94 —A—Heart Disease (Men) ♦ Heart Disease (Women) -■A-Heart Attack (Men) • O "Heart Attack (Women) Age (5-year interval) cn O Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 5.3 Observed Incidence of Non-Fatal Heart Disease between Wave I and Wave II, by Sex, by 5-Year Age Group, HRS/AHEAD Men Age Group Overall Heart Oisease Heart Attack 5 0 -5 4 2.8 % 1.2 % 5 5 -5 9 4.1 1.8 6 0 -6 4 4.7 2.3 6 5 -6 9 6.8 3.2 7 0 -7 4 7.9 3.8 7 5 -7 9 8.4 2.1 8 0 -8 4 7.0 3.3 8 5 -8 9 7.1 1.8 9 0 -9 4 6.8 1.2 Women Age Group Overall Heart Oisease Heart Attack 5 0 -5 4 3.1 % 0.7 % 5 5 -5 9 2.1 0.4 6 0 -6 4 4.4 1.3 6 5 -6 9 5.0 1.8 7 0 -7 4 5.1 1.7 7 5 -7 9 8.1 2.7 8 0 -8 4 8.2 2.7 85-89 10.5 2.8 90-94 10.7 3.4 Overall heart disease does not include a heart attack ui Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.2 Observed Incidence of Non-Fatal Heart Disease between Wave I and Wave II, by Sex, HRS/AHEAD o i CL 12 10 8 6 4 2 0 50 - 54 55 - 59 60 - 64 65 - 69 70 - 74 75 - 79 80 - 84 85 - 89 90 - 94 -A —Heart Disease (Men) - • —Heart Disease (Women) ■ ■ & ■ ■ ■ Heart Attack (Men) -o ~ Heart Attack (Women) Age (5-year interval) cn to for those aged 90-94 (Table 5.2). Heart disease incidence for women increases with age from 3.1% for those aged 50-54 to 10.7% for those aged 90-94. The two-year incidence of a heart attack for women gradually increases with age: from 0.7% for those 50-54 to 3.4% for those aged 90-94 (Table 5.3 and Figure 5.2). In sum, the observed two-year transitions into and out of heart disease status differ between men and women. Men have a higher rate of having heart attacks than women at the baseline. Men with heart disease but not with a heart attack have the highest rate of death after two years. On the other hand, women with a heart attack show the highest death rate in the same period as compared with men. B. Hazard models and annual transition rates To examine transitions in heart disease status for men and for women, a variety of transitions of heart disease status are examined using hazard analyses. The analyses include six different heart disease transitions for each sex. These status changes are (A) from "No heart disease” to "Heart disease without a heart attack,” (B) from “No heart disease" to “Heart attack," (C) from “Heart disease without a heart attack" to “Heart attack,” (D) from “No heart disease" to “Death,” (E) from “Heart disease without a heart attack" to “Death,” and (F) from “Heart attack" to “Death." Men The equations for the six hazard models for men are summarized in Table 5.4. Except for the transition from “Heart disease” to “Heart attack,” age is significantly related to the other heart disease transitions. From the hazard models, the annual transition rates for the six heart disease transitions are calculated for men 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 5.4 Summary of Equations of Hazard Analyses for Heart Disease Transitions, by Sex, by Transitions, HRS/AHEAD Men Transition of Heart Disease Status N #of events Intercept Age (A) 'No Heart Disease' to 'Heart Disease' 6,071 211 6.936 *** -0.044 *** (B) 'No Heart Disease' to 'Heart Attack' 6,071 107 5.886 *** -0.018 * (C) 'Heart Disease' to 'Heart Attack' 1,115 56 4.163 *** -0.008 (D) 'No Heart Disease' to 'Death' 6,071 291 10.115 *** -0.093 *** (E) 'Heart Disease' to 'Death' 1,115 142 9.282 *** -0.088 *** (F) 'Heart Attack' to 'Death' 739 77 7.313 *** -0.063 *** Women Transition of Heart Disease Status N # of events Intercept Age (A) 'No Heart Disease' to 'Heart Disease' 8,087 293 6.617 *** -0.039 (B) 'No Heart Disease’ to 'Heart Attack' 8,087 92 8.756 *** -0.052 *** (C) 'Heart Disease’ to 'Heart Attack' 1,473 54 5.396 *** -0.020 (D) 'No Heart Disease' to 'Death' 8,087 310 10.244 *** -0.089 *•* (E) 'Heart Disease' to 'Death' 1,473 147 8.401 *** -0.071 *«« (F) 'Heart Attack' to 'Death' 454 61 6.558 -0.053 **• *p <0.05, **p<0.01, ~*p <0.001 'Heart Disease' does not include heart attack or myocardial infarction. 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. aged 50-97 with a two-year age interval. The results are shown in Table 5.5 and graphed in Figure 5.3. The annual transition rates to death are higher than those to heart disease. The transition rates differ by transitions. Women The summary of the results of the equations of the six hazard models for women is shown in Table 5.4. Except for the transition from “Heart disease” to “Heart attack," age is significantly related to all transitions. The annual transition rates implied by the equations for the six heart disease transitions are calculated for women aged 50-97 in two-year age intervals (Table 5.5). The rates are shown in Figure 5.3. The annual transition rates to death are higher than those to heart disease. The transition rates differ by transitions. Calculated annual transition rates for heart disease and death for men and for women are somewhat different. Although the likelihood of dying is higher than the likelihood of changing to any heart disease status for both sexes, annual transition schedules are somewhat different for men and for women. In some transition rates for heart disease and death, there is a crossover in annual transition rates between men and women. C. Multistate life tables By using the annual transition schedules from the six heart disease transitions, two kinds of multistate life tables are created: the population-based life table and the status-based life table. The population-based life table represents the entire population in each age category. This table takes into account the initial distribution of the population by heart disease status. In other words, it describes the 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.3 Annual Transition Rates for -State Heart Disease, by Transition, by Sex, HRS/AHEAD (A) 'No Heart Disease* to 'Heart Disease' (B) 'No Heart Disease' to 'Heart Attack' (C) 'Heart Disease' to 'Heart Attack 0 .1 0 ----------------------------------------------- 0.08 • ------------------------------------------ 0.10 0.08 0.06 £ Men 0.04 Women 0.02 0.00 70 50 60 80 90 Age Women 0.06 — Men _ _ 0.02 Women 0.00 50 60 70 80 Age 90 (D) 'No Heart Disease' to 'Death' 0.50 0.40 0.30 £ m Men D C 0.20 0.10 Women 0.00 70 50 60 80 90 Age (E) 'Heart Disease* to 'Death* 0.50 r ------------------------------------------ 0.40 Men s 0.30 n “ 0.20 0.10 Women 0.00 50 70 80 90 60 (F) 'Heart Attack' to 'Death' 0 .5 0 -------------------------------------- 0.40 £ o c Men 0.20 0.10 0.00 80 90 50 60 70 Age Age cn o > 'Heart Disease* does not include heart attack or myocardial infarction 7 T ab le 5 . 5 Annual Transition R ates fo r Multi-State Heart Disease fo r Those A g e d 50-97, b y Sex, HRS/AHEAD J C 0 1 I X ( f t IV 0 ) ° o I t: S S 9 “ o z u T 9 (A 5 tr IS 01 IS St .2 n ° I IV 4 > O 0 ) « (V 0 ) to o tr S o ffi v > C O 9 C O 5 « IV 4 > X O u J tr <v X o to 9 to S 5 t 3 x CM i n Q GO o 5 CO CO O CM i n f* - 5 GO CM CM CM CM CM CO CO CO £ O o o o © o o o o o d o o d o o GO o CM CO 0 ) 0 ) m m GO © CO c o 0 ) c o GO CM CM CM CM CO CO o o o o o O o o o o o o o o o o h- C M e g c m in t o cm h - o 0 0 0 0 0 S m t o c p o o o 8 0 ) O ) O W N o o o o o S (fi O 0) IO (O o ) (O (0 N S J W (O S o o o o O C O G O 0) N . C O C M O C M ^ S O ) 00 C M G O f - * I A A C O S O § Qt-r>^i-i-MCMCMCOn O O O O O O O O O O O 0 0 0 0 0 0 0 0 0 0 0 0 Ni-OOlAlAt-CO^^SCO 2 5 m i n o o o o O) 8 o o d O NO)OcOiAflOMIOrN^<oSinbSCO^(OCOOt-i-CS| OQ^rfPCNNCOCOVlOCpNOlOCMIOflOrCOrNt O O Q O O O O O Q O Q O O O O * - ' - y- y-CMCMCOCO'T d d d d d d d d d d d d d d d d d o O O IA O) N h * in CO * C O y— 0 > o 8 C OC ON . 0) Y— CO i n C O Is * 0 ) o CO GO CM < 0 GO o Q O © O o V y* CM CM £ CO o © o © O o o o O O o o o o o o o o o d o o o o o o o o s 3 o o o ID C M ^ m i n o CM h * CO CM r * - 0 ) N . GO 0 ) 8 s CO i n CO GO c o T » Y“ CM CM o o o o o o m o CO GO 8 8 i n r* * CM CM m s . CM m O i Y“ CM CM CM o d o o d o CM CO o i n CO o m o CO i n o CM £ CO Y “ T - CM o o o o o o CM n h - o o c o CO GO o Y— CM GO c o Y— h * 5 CM CM CO CO 4 o o o d d o o CM i n CO m r s GO CO T“ CO 0 ) o r ^ 0 ) CO m 0 ) o o Y - Y— o o o d o o C M ^ C M ^ O S 00 ? C fl (D N 00 O Q O O O O t- O C D co in CO N CO f l) N C O CD GO C M N C M O ) N N O O O O O O O O O O O O O O O O O ^ * < 0 ^ - 1 ^ * c m c m c o c o O O O O d o d o $ N in C O f O ) C O N C O 0 ) O r - 1 - C M C O O O o o o o d o o o o o o o o o o o d o d d d d d d d d C O Y“ § o C O o C M 8 £ C M h * m O Y“ Y- Y» C M C M C O o o o o o o 8 h - GO 0 ) o C O h . GO o C M C M C M C M C M C O O O O O O o o o o o o o o s GO CM i n 0 ) CO h * m o GO CO h * CM <P 8 o m o m CO CO * m m CO CO N» h* h - GO CO 0 ) $ O CM CM CO CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM CO CO CO £ CO CO CO o O o o O O o O o O O O o o o o o o o o o o o o d o d d o o d o d o o o o o d o d o o o o d o o . N C M 05 N C O r- C M C O m C O G O O C O t - C O ( O D C M ( O O J ? D t Q ^ t C M r Q r CMCMCMCMCOrtCOjJlAfifiNCOOOr .. .. _ _ __ _ § B § S S § S o o S o o S S S o o o o o o o o o d d d d d d d d d d d d d d d o d d d d d d d d O C M I f ) 0 0 r r s . r * » I s * . k ~ KQCO^Q^QOCMCO^-inOtnOtnOCDCM 8 . w w w = a » 2 2 ® O O O r r ( M ( M n c 0 ^ ^ l f l l A ( 0 q o o o o q q q o * “ *“ ^ ^ y * y * y * ^ ^ y “ ^ ^ ^ ^ o o o o o o o o o o o o o o o o o o o o o o o d d d d d d d d d d d d d d d o d d d d d d d d oo c o o o m . t t ifl ® N 0 5 0 C M ' T ( O C O O IflC O r-O O r C M Ifi 050t-CMCO^in(0 § ^^^y-^^t-*^t-CMCMCMCMCMCO o o o o o o o o o o o o o o o d d d d d d d d d d d d d d d o is . G O o C O O O C O 8 C M o> 0) co C M rm C O o C M m C O G O C M C O C M C O C M G O C M Y“ C O 8 C O o o O o o o O O O o o o o O o o d o d o o o o o o d o o o o in o) C O m r * . 0 ) C O m 0 ) C O in m m C O C O C O C O C O N . t** f s h* oo G O ? r r C M N f 5 G 0 cocbocMinocoh* c o « e o ? t t i f l i o o o o o o o o o o o o o o o* o o co cm co cp 00 CO G O CO 0 ) S i n m co co o o o o © d o d o a 5^ 2 05 9 < in N 0 5 r CO i f l S G O G O 00 05 05 05 05 OCM^rCOCOQCM*<paoaCM*COCOOCM3C0OOQCM^Cp i n m i A U ) U ) ( D ( 0 ( D(0 C 0 N N S N N f l 0 0 0 « 00Q0 0 ) a ) O 0) 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 'Heart Disease' d o e s n o t in c lu d e heart attack o r myocardial infarction. whole population. On the other hand, the status-based life table is based on categorizing the population into-three groups based on their heart condition at a given age: heart-disease-free, having some Kind of heart disease but not having a heart attack, and having a heart attack. In this study, two outcomes are examined from each multistate life table. They are (a) expected years of life with and without heart diseases and (b) the estimated implied heart disease prevalence rates. Each outcome is specified for men and for women. 1. Expected years of life with and without heart diseases Expected years of life with and without heart disease provides an estimate of the an average person’s experience with heart disease over his/her life cycle using population-based life tables. Status-based estimates provide estimates for three different groups in three different tables: persons free of heart disease, persons who have heart disease but not a heart attack, and persons who have had a heart attack. Status-based outcomes are graphed for four ages: 50-51,60-61, 70-71, and 80-81 assuming that 100% of the group begins life in the specified state at each of the ages. Men Expected years of life with and without heart disease for men in the population-based analyses are shown in Table 5.6. Total life expectancy and years of life with and without heart disease are also shown in Figure 5.4. At age 50-51, men have 27.0 years of expected life remains. These years consist of 17.8 years of life without heart disease, 4.4 years of life after heart disease but without a heart attack, and 4.8 years of life after a heart attack. At age 70-71, with 12.1 years of expected life, men have more years with some kind of heart disease than expected years without heart disease. They are expected to have 2.8 years of life with heart 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T ab le 5 . 6 Expectations o f Y ears o f L if e w ith a n d without Heart Disease a t Certain A g e fo r A g e d 5 0 a n d Over, Population-Based, b y Sex C O C M O O O 0 0 C O C O C M C M C M C M Q O C M c o M ; o c o s ( b o o C O C O C M C M C M o o in C O r * ^ C M N M C M C M C N f i - C O i O N * O i * “ C O lO h ' » 0 ‘ ~ S N 00 00 00 CO C O I I I I I » I C O oo © C M ^ CO 00 CO CO N N N O ) C M o o N in M ^ O ) s m M ; M p C O C O C O C O C O C M C M C M C M C M C M 0 0 m ; cm o oo cp CO CO CO C M C M C M 00 00 O C O C O n cc co in t (OCOOCONr-^m CMCMCMCMCMT-^^- N O) T - CO m N n > oo oo oo oo (6 O O 6 C M 4 C O 0 0 Q O O O 0 0 r CO in N 01 i n m i n m in i i t i i Q C M ^ CO 00 i n in in >n in co 6 i o i o o i • _i i i » oo o cm 5 cp O O 01 Ol Ol O ) O l o Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.4 Expected Age of Onset of Heart Disease and Heart Attack, by Age and Sex, Population-Based Men 90 85 80 75 0 > o > < ■ o a 70 o & 1 1 1 65 60 55 50 50-51 7.1 12.1 HBa s s s s 18.9- 27.0 ------------ — ■ _ ------------------ 10.9 17.8 2.9 6.0 ■After heart attack □With heart disease □No heart disease : IBase age adjustment 60-61 70-71 80-81 Age Women a 70 1 BAfter heart attack 50-51 □With heart disease □No heart disease I IBase age adjustment 60-61 70-71 80-81 Age o > O disease and 3.3 years after having a heart attack. At any given age, the years of life after having a heart attack are longer than the years of life with heart disease but no heart attack for men. Status-based expected years of life with and without heart disease for men who are free of heart disease in the initial status are shown in Table 5.7. The years of life without heart disease are always longer than the years of life with some kind of heart disease for men ages over 50. The expected years of life with and without heart disease at specific ages among men who do not have heart disease at the specified age are graphed in Figure 5.5. As shown in Table 5.7, the years of life with heart disease are longer than years of life after having a heart attack. The expected years of life by heart disease state among those who have already experienced a heart problem but not a heart attack at the specified age are graphed in Figure 5.6. Compared to men who never had a heart problem at the specified age, those who have already experienced some kind of heart disease but not a heart attack are expected to have shorter years of life expectancy, longer years of life with heart disease, and longer years of life after having a heart attack (Tables 5.7 and 5.8). After age 74-75, the estimated life expectancy of men who have already experienced a heart attack becomes longer than that of those who have already experienced some kind of heart disease but not a heart attack. Compared to men who never had a heart problem at the specified age, those who have already experienced a heart attack in the initial status live more years with a heart attack but live fewer years in total (Tables 5.7 and 5.9). 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T ab le 5 . 7 Expectations o f Y ears o f L ife b y Heart Disease Status a t Certan A g e fo r A g e d 5 0 a n d Over, Status-Based (Initial Status: Without Heart Disease), b y Sex, HRS/AHEAD C O to CO oo co o d d C M C O c o c o o CO C M C M C M C M ^ C O 0 0 0 0 K C O C O v- Ci h- co ^ CO M N C M C M I D N O ) r - C O 0 0 0 0 c& 5 C O 00 O C M 0 0 0 0 O ) O ) C O C O 00 I C N 00 v- c o in to m in o > 0 0 m o o o o C M C M O N * ID C O C M C M C M C M C M C M C O 2 * O O CO 00 C M C O C O C M C M r^co^cM^-OJoocoin CMCMCMCMCMf^^^ < o iflso )< -rtin N P*- 0 ) C O CO (O N N i T I i C O O O O C M in i o o a o 00 CO 0 ) O ) oo oo Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.5 Expected Age of Onset of Heart Disease and Heart Attack for Those who are Free of Heart Disease at Specific Age, by Sex, Status-Based Men ■After heart attack □With heart disease □Without heart disease '.Base age adjustment 50-51 60-61 70-71 Age 80-81 55 50 50-51 Women 27.0 79 9 ■After heart attack □With heart disease □Without heart disease IBase age adjustment 60-61 70-71 Age 80-81 o > U T ab le 5 . 8 Expectations o f Y ears o f L ife b y Heart Disease Status a t Certain A g e fo r A g e d 5 0 a n d Over, o < U l X < 0) 0 £ X X o (0 > A « ■ > C O C M o d d IO C M C O C O C O Q O d C O O ) m m s V ^ C O C M C M C M C M C M C M C M C M o in Q O Q O i C O O ) C O C M C O e <A n o > (A O t m e X (A 3 3 ( 0 T S e ( A c o 1 0 I (0 2 ( 0 O ) C O C O < 0 o C O if) in in Ifl N O J r C O SN|s.«flO 4 C O C O d C M h - oo oo Ifi M 3 ) r n i f l N 00 C O C O O O) O 9) i i • ■ i i i S C O 0 0 O M 5 ( D G O C O O ) O ) O u ) C O O C M C M Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T able 5 . 9 Expectations o f Y ears o f L ife b y Heart Disease Status a t Certain A g e fo r A g e d 5 0 a n d Over, Status-Based (Initial Status: W ith Heart Attack), b y Sex, HRS/AHEAD 1 1 * 5 id C O c o C O CM C M C M C O lO 00 C O C O 6 n t G O G O C O O ) 0 0 c o co O ) C M C O NCOQOCOOOCO0)O)O)O) C O 6 C M r- o o c o O ) o C M C O 0 0 0 0 0 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Women The population-based expected years of life with and without heart disease for women aged 50 and over are shown in Table 5.6. Expected years of life with and without heart disease at the specified age are shown in Figure 5.4. At age 50-51, women have 30.5 years of expected total life that includes 21.8 years of life free of heart disease, 6.2 years of life with some kind of heart disease but no heart attack, and 2.5 years after having a heart attack. For women up to age 76-77, the years of life without heart disease are longer than the years of life with some kind of heart disease. After age 76-77, women are expected to spend more years of life with heart disease than without heart disease. Unlike men, women are expected to experience fewer years after having a heart attack than years with other forms of heart disease. Status-based expected years of life with and without heart disease for women who are free of heart disease in the initial status are shown in Table 5.7. At every age, women who never experienced a heart problem are expected to spend more years of life without heart disease than years of life with some kind of heart disease. Years of life with heart disease but not a heart attack are longer than years of life after having a heart attack at every age among women (Table 5.7). The expected years of life at the specific age among women who had not experienced a heart problem before that age are shown in Figure 5.5. Women who experienced some kind of heart disease but not a heart attack are estimated to have shorter total remaining years of life but longer years of life with some kind of heart disease than those of women without any heart disease. Longer life after having a heart attack is expected among women with heart disease but not a heart attack (Table 5.8). Status-based expected years of life at the specified age among women who have a 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. heart problem but not a heart attack are shown in Figure 5.6. Women who already have experienced a heart attack have shorter expected total years than other two groups (Tables 5.7, 5.8, and 5.9). The expected years of life for women who have a heart attack by a certain age are shown in Figure 5.7. 2. Estimated implied heart disease prevalence rates Expected implied heart disease prevalence rates include population-based and status-based estimates. The population-based estimated implied heart disease prevalence rates include all statuses of heart disease in the initial status. The status- based estimated implied heart disease prevalence rates can be separated for three different entries: persons free of heart disease, persons who have heart disease but not a heart attack, and persons after having a heart attack. However, the status of persons after having a heart attack is eliminated because there is no other outcome category. The prevalence shown is for the midpoint for the age interval. Men In the population-based multistate table for men, the prevalence rate of heart disease without having a heart attack at age 50-51 is 8.1%, and the rate with having a heart attack is 6.9% in the same age (Table 5.10). The overall heart disease prevalence rate increases and passes 50% at age 76-77 for men. The prevalence rate of heart disease without having heart attack increases up to age 86-87 (26.6%), then it decreases after that age for men. The prevalence rate of heart attack for men increases with age is faster than the prevalence rate of heart disease without heart attack (Figure 5.8). At age 60-61, the heart attack prevalence rate reaches the same level as heart disease without heart attack (14.2%). The heart attack prevalence rate at age 96-97 is 56.7% for men. 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.6 Expected Age of Onset of Heart Attack for Those with Heart Disease at Specific Age, by Sex, Status-Based Men 2 70 ■After heart attack ■With heart disease I .'Base age adjustment 50-51 60-61 70-71 80-81 Women 50-51 Age 60-61 ■After heart attack ■With heart disease I IBase age adjustment 80-81 O ) o o Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.7 Expected Years of Life after Heart Attack who have had Heart Attack at Specific Age, by Sex, Status-Based Men Women S 70 After heart attack Base age adjustment 50-51 60-61 70-71 80-81 After heart attack .'Base age adjustment 50-51 60-61 70-71 80-81 o > c o T a b le 5 .1 0 Im plied Population-Based Age-Specific Prevalence Rate, b y Sex « E I O ( 0 if ( Q 0 ) X 0 ) i/) C O 0 ) C C / 1 I 5 ^ r ( 0 0 ) X < D c / i o * 2 £ b > C O d ) X O l c o C Oo 0 0 C Oi n i n i n C O0 0 o C M i n 0 0 C O0 0 i n C ON * I s - C MC Orr i n C Or * . o o O l o C O i n C Oo o O l C Oi n o o o o o o o o o o o o r — T — C MC MC M C MC O o d o o o o o o d o d o o o o d o d d o d o o o c o T “ r - C O o C O O l C O C M 0 1 in * — r ^ - C MN . C M C O o C O C O C O C O h * O l o C M N * i n r- G O o C M C O in G O O *— C O N - C D h » o o O l o o o r - T * r - T “ T — C M C M C M C M C M C M C O C O C O co C O C O C O C O O ' o o o o o o o d o d o o o o o o o o o d o o o o O l N . o £ in 01 C O C O O l T - N* C O oo o T - C M C M 01 m o oo C O C M 01 h - O ' T~ 01 C O C O r “ 00 in C N o h * M ’ T “ GO r - GO O l GO oo GO GO h * f * - h - C O C O C O C O in in in in N * M ’ N ' C O C O C O C M o o d o o d o o o d o d o d o o o o o o o o o o 0 3 § * O ) O < < 5 in C O in in m r^ in O l in 5 C O C O in C O h - C O O l c o r^ C O r * - m r * » h * . N - 0 1 r ^ - V * 0 0 C O o o in G O 1 " - o o 0 1 G O O l C O O l in O l O l o 1 0 CM m 4 m C O m CO m o C O CM C O 4 c o C O C O d c o d r- CM r^ N * r* - c o n . 0 0 r * * - o o o CN 0 0 4 0 0 C O G O d G O d O l CM O l O l C O O l in o m c m o _ C M O n m c o C O C O C O C O O O C O © C O m m o o co C N C N o o o o C O C O C O n- c n ci C N C N T - o o d c o c o c o c o c o i n C N C M C N d o o 0 0 co n - m o © C O o o o o C O o o C N 0 0 C O d C O i n QO o i n i n o o O l O Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.8 Population-Based Age-Specific Prevalence Rate (Initial Status: Without Heart Disease), by Sex Men Women 100% 100% 2 50% g 50% BWith Heart Attack □With Heart Disease □Without Heart Disease □With Heart Attack □With Heart Disease □Without Heart Disease 50- 54- 58- 62- SI 55 59 63 66- 70- 74- 78- 82- 67 71 75 79 83 86- 90- 94- 87 91 95 50- 54- 58- 62- 66- 70- 74- 78- 82- SI 55 59 63 67 71 75 79 83 86- 90- 94- 87 91 95 Age Age 453971991499519955 In the status-based estimation derived for men with an initial status of free of heart disease, a similar pattern of heart disease prevalence rate changes is seen but at different ages (Figure 5.9). At age 50-51, 0.9% of men are estimated to have heart disease without having a heart attack, and 0.7% of men are estimated to have heart attack. The heart disease prevalence rate exceeds 50% at age 82-83 (Table 5.11). This means that more than a half of men without heart disease after age 82-83 are estimated to experience some kind of heart disease. The estimated implied prevalence rate of heart disease without having a heart attack peaks at 27.0% at age 88-89 among those who never experienced a heart problem. The crossover of prevalence of heart disease and the prevalence of heart attack is seen somewhere between age 80-81 and age 82-83. The prevalence of heart attack continuously increases, and it reaches 50.1% at age 96*97 among those with any heart disease as the initial status (Table 5.12 and Figure 5.10). Women In the population-based multistate life table for women, the prevalence rate of heart disease without a heart attack at age 50-51 is 7.6%, and the heart attack prevalence rate is 1.4% in the same age group (Table 5.10). The overall heart disease prevalence rate increases with age, reaches 50% at age 82-83, then it continues to increase in advanced ages for women (Figure 5.8). The prevalence rate of heart disease without a heart attack continues to increase with age for women. The prevalence rates of heart disease without a heart attack are 15.6% at age 60-61, 23.9% at age 70-71, and 31.7% at age 80-81. The prevalence rate of heart attack also increases with age but is lower than the rate of heart disease without having a 172 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.9 Status-Based Age-Specific Prevalence Rate (Initial Status: Without Heart Disease), by Sex Men W om en 100% 90% 80% 70% 60% 50% 40% 30% 20% a > a. 100% OWithout Heart Oisease □With Heart Disease □ W th Heart Attack « i 0 0 1 a . 10% 0% □Without Heart Disease □With Heart Disease □With Heart Attack 50- 54- 58- 62- 66- 70- 74- 78- 82- 86- 90- 94- SI 55 59 63 67 71 75 79 83 87 91 95 Age 50- 54- 58- 62- 66- 70- 74- 78- 82- 86- 90- 94- SI 55 59 63 67 71 75 79 83 87 91 95 Age G > Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 5.11 Implied Status-Based Age-Specific Prevalence Rate (Initial Status: Without Heart Disease), by Sex Men Women Age Group Without Heart Disease With Heart Disease With Heart Attack Age Group Without Heart Disease With Heart Disease With Heart Attack 50-51 0.984 0.009 0.007 50-51 0.989 0.009 0.002 52-53 0.953 0.026 0.021 52-53 0.965 0.028 0.007 54-55 0.921 0.043 0.036 54-55 0.941 0.047 0.012 56-57 0.889 0.061 0.051 56-57 0.916 0.06S 0.018 58-59 0.856 0.078 0.066 58-59 0.890 0.086 0.025 60-61 0.823 0.095 0.082 60-61 0.863 0.105 0 032 62-63 0.790 0.113 0.098 62-63 0.835 0 125 0.040 64-65 0.757 0.130 0.114 64-65 0.807 0.145 0.049 66-67 0.724 0.146 0.129 66-67 0.778 0.164 0.058 68-69 0.692 0.163 0.146 68-69 0.749 0.184 0.068 70-71 0.660 0.179 0.162 70-71 0.719 0.203 0.078 72-73 0.628 0.194 0.178 72-73 0.689 0.222 0.089 74-75 0.597 0.209 0.194 74-75 0.659 0.241 0.100 76-77 0.567 0 223 0.210 76-77 0.629 0.259 0.112 78-79 0.538 0.236 0.227 78-79 0.599 0.277 0.124 80-81 0.509 0.247 0.244 80-81 0.569 0.294 0.137 82-83 0.481 0.256 0.262 82-83 0.538 0.311 0.151 84-85 0.454 0.264 0.282 84-85 0.508 0.327 0.165 86-87 0.427 0.269 0 305 86-87 0.477 0.343 0.181 88-89 0.400 0.270 0.330 88-89 0.445 0 358 0.197 90-91 0.373 0.267 0.361 90-91 0.413 0.372 0.215 92-93 0.344 0.259 0.398 92-93 0.379 0.386 0.235 94-95 0.313 0244 0.443 94-95 0.345 0 398 0257 96-97 0.277 0.221 0.501 96-97 0.308 0 409 0.283 -u T ab le 5 .1 2 Im plied Status-Based Age-Specific Prevalence R a te (Initial Status: W ith Heart Disease), b y Sex C M in ® C M v — co m co o o o o d o d o < d C D co c o C O m c m o c o o o o c o c o ^ 00 C O V - C O C M C M o e oo m o o < b o o o o C M 0 0 m o C D C O in d o o in m i n m I s * r* h * r - co m o o o o o o d cm 4 0 0 o o 0 0 00 0 0 00 C M 0 0 C M C M e i C O C O O) in 0 0 o o o o C M C M 0 0 o o C O C M o o f C O i n N O r C O i n N O i - l O l O N h»r*.h»r^r*»QOoooocooo0>o>o>0> J l l l l J I t l l l l l l OCM^COOOOCM^COOOQCM^Cp rs.p^r^[^f^oooocoooooo><Da)3> O ) o 0 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 5.10 Status-Based Age-Specific Prevalence Rate (Initial Status: With Heart Disease), by Sex Men Women 100% e 50% ■BBiBiniaBi With Heart Attack □With Heart Disease 50- 54- 58- 62- 66- 70- 74- 78- 82- 86- 90- 94- SI 55 59 63 67 71 75 79 83 87 91 95 Age 100% 90% 80% 70% 60% § 50% o C L 40% 30% 20% 10% 0% With Heart Attack □With Heart Disease 50- 54- 58- 62- 66- 70- 74- 78- 82- 86- 90- 94- SI 55 59 63 67 71 75 79 83 87 91 95 Age -> i o > 8770 5 heart attack. The prevalence rates of a heart attack for women are 4.8% at age 60- 61, 9.5% at age 70-71, and 15.5% at age 80-81. From the status-based estimates of implied heart disease prevalence among women with no heart disease with age as the initial status, there is an increase pattern of heart disease prevalence (Figure 5.9). At age 50-51, 0.9% of women are estimated to have a heart disease except for a heart attack, and 0.2% of women are estimated to have a heart attack. The estimated implied overall heart disease prevalence rate passes 50% at age 86-87 (Table 5.11). This means that more than a half of women without heart disease after age 86-87 are estimated to experience some kind of heart disease. The estimated implied prevalence rate of heart disease with the initial entry of persons with heart disease but not a heart attack for women is shown in Table 5.12 and Figure 5.10. The implied prevalence rate of heart disease increases with age. The rates are 10.5% at age 60-61, 20.3% at age 70-71, and 29.4% at age 80-81. The estimated implied prevalence rate of heart attack for women also continues increasing with age. The heart attack prevalence rates are 3.2% at age 60-61, 7.8% at age 70-71, and 13.7% at age 80-81. Among women with heart disease as the initial status, the prevalence rate of heart attack increases continuously with age. The prevalence rates are 1.2% at age 50-51,12.6% at age 60-61, 22.8% at age 70-71, and 31.6% at age 80-81. D. Summary The observed two-year transitions into and out of heart disease status differ for men and women. Men have a higher rate of having heart attacks than women. For each sex, men with heart disease but who have not had a heart attack have the 177 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. highest rate of death after two years, while women with a heart attack show the highest death rate in the same period. Calculated annual transition rates for heart disease and death are somewhat different for men than for women. The likelihood of dying is higher than the likelihood of changing to any heart disease status for both men and women. However, men and women differ in transition schedules. In some transitions, there is a crossover in annual transition rates between men and women. These different annual transition schedules result in differences in estimated years of life with and without heart disease and implied prevalence rates of heart disease for each sex. Men and women live different numbers of years of life with and without heart disease. Women relatively live fewer years with having had a heart attack than men. Men who experienced a heart attack at a young age live longer than women in the same state. The implied estimated prevalence rate of heart disease increases with age for both men and women. While the prevalence rate of heart attacks dominates men with heart disease with advanced age, heart attack prevalence is always lower than heart disease without having had a heart attack at each age among women. 178 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER VI SUMMARY AND CONCLUSIONS In this chapter, the results for men and women are compared, findings from this study are reviewed and compared to results from past studies, and limitations of this study are explored. After discussing these issues, suggestions for future research are proposed. In this study, it was found that heart disease mortality has declined continuously over the last three decades for men and for women but that patterns of decline are different by sex. This decline over time characterizes both the overall heart disease mortality rate and the myocardial infarction (Ml) mortality rate. Average annual percentage declines for overall heart disease mortality rates are small for men aged less than 65 but are larger for those aged 65 and over in the last decade. For women, the average annual percentage changes for overall heart disease mortality rate decline are relatively small for those aged 50-54, but are large for those aged 55 and over in the most recent decade. In the most recent decade, the average annual percentage declines for overall heart disease mortality vary by age for men and women. In terms of Ml mortality, both men and women show continuous decline over the last three decades. Up to age 70, larger mortality declines occur among men in the short term as compared to the long term. Men aged 70 and over show higher average annual percentage declines in mortality than do younger men. The pattern of the average annual percentage mortality change with age is less clear for men 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. than it is for women in the examination of 5-year age groups aged 50 and over in the most recent decade. Relatively large declines in AMI mortality rates are seen in some age groups among men and women. In the comparison of the declines in AMI mortality rates between men and women, the decline rates are mostly larger for men than they are for women in the two periods of 1982-1996 and 1987-1996. Previous studies indicated decline in heart disease mortality over the last three decades. These results are observed among men and women in this study. However, this study also finds that the speed of decline in mortality vanes by age. Older age groups show faster declines in mortality rates than younger age groups in this study. Mortality declines have slowed in recent years among younger age groups. In contrast, the mortality declines among the elderly are faster in recent years. These patterns are seen in all heart disease mortality and Ml mortality among both sexes. Annual mortality declines in the most recent decade are faster for men than for women for both overall heart disease and AMI. However, over the entire three decades, women aged 70 and over show faster overall heart disease mortality declines than do men aged 70 and over. The time trend in heart disease prevalence differs by heart disease category and sex. Overall heart disease prevalence has increased, but the IHD prevalence has decreased in recent years for men. The prevalence of heart disease other than IHD has increased in recent years more than the size of decline of the IHD prevalence among men. For women, despite a recent decrease in overall heart disease prevalence over the past three decades, the IHD prevalence rate has increased. Among women, although the IHD prevalence rate increased over time, heart disease other than the IHD decreased more than the magnitude of the IHD 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. prevalence change. Therefore, overall heart disease prevalence has declined in recent years among women. Despite the time trend patterns from individual based analyses, results from an age group examination show inconclusive time trend patterns for men and for women. There are different time trends since heart disease onset among men and women. This trend seems to change with IHD prevalence rather than with overall heart disease prevalence. Furthermore, onset profiles are somewhat different according to sex. Men have a decline in the likelihood of onset less than one year before the interview in recent years, while women have an increase in the likelihood of having heart disease less than one year before the interview in recent years. Men also show a recent decline in the likelihood of heart disease onset of less than five years before the interview while women show a decline over the long period, but an increase over the short period in the likelihood of having heart disease in the year before the interview in recent years. These trends may imply that men are less likely to have a recent onset of heart disease in recent years, while women are more likely to have had heart disease in recent years. In general, disability trends among those with heart disease have improved in recent years for both men and women. Men show an increase over time in the prevalence of having no activity limitation and a decrease in being unable to perform their major activity. Men also have a higher likelihood of having no work limitation in recent years, while the status of being unable to work has decreased. Women show almost the same disability trends. Therefore, it appears that people who have heart disease suffer less disability in recent years, and that men and women with heart disease have less problem working in recent years. 181 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The likelihood of death from all causes and death due to heart disease within 1.5 years since the interview date has not changed in recent years for men or for women. However, the likelihood of death from all causes within 2.5 years after the interview date among men and women with a heart problem has declined in recent years. Additionally, the likelihood of death due to heart disease within 2.5 years after the interview date among men and women has declined in recent years among both those with and without heart disease at the interview. Men and women seem to experience transitions to heart disease and to death somewhat differently. Although all transition rates increase with age, and transitions to any heart disease statuses are much lower than the transitions to death from any statuses for both sexes, each transition rate differs by sex. Transition rates are generally higher for men than those for women, especially at older ages, except for the transition rates from the status of having no heart disease to the status of having a heart attack. Crossover in the transition rates are seen except for the transitions from the status of having heart disease to the status of having a heart attack and from the status of having no heart disease to the status of death. Transitions toward the status of having a heart attack accelerate faster for women than for men. In the general population aged 50 and over, the expected years of life with heart disease increases with age. However, the profiles of years of life with and without heart disease for men and for women differ because of their different transition schedules. At age 50, men are expected to have 34% of those remaining life with some kind of heart disease. Men are expected to live more years with heart disease than without heart disease after age 70. On the other hand, women are expected to have approximately 30% of their remaining lives with some kind of heart 182 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. disease at age 50. Women are expected to live more years with heart disease than without heart disease after age 78. Men are expected to have more years of life after a heart attack than with any other heart disease. However, women are expected to live fewer years after having a heart attack than the years of life with any other heart disease. Although expected years of life with heart disease without a heart attack are shorter for men than for women, expected years of life with a heart attack are longer for men than for women at any age. Among those who experienced no heart disease at specified ages in the status-based life table, years of life with heart disease decrease as age increases for men and for women. Years of life with a heart attack are longer for men than for those for women of the same age, while years of life with a heart attack are slightly shorter than the years of life with other heart disease for men as compared with the case for women. Among those who experienced a heart problem other than a heart attack by age 50, men are expected to live 6.5 years with a heart attack (27% of remaining years), and women are expected to have 4.5 years after a heart attack (17% of remaining years). Years of life after a heart attack for those who experienced a heart problem but not a heart attack decrease as the age at which the heart problem is experienced. Among those who experienced a heart attack, men are expected to live longer than are women until a certain age. Men who have a heart attack by age 50 are expected to live 21.8 years while women at age 50 are expected to live 20.8 years. Men and women who already experienced a heart attack at age 64 are expected to live 15.6 years and 15.3 years, respectively. Women who have had a heart attack by age 70 or 80 are expected to live longer than men. 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The estimated implied prevalence rate of heart disease increases with age for men and for women. It is implied from the multistate life table that 50% of women aged 78-79 and 50% of women aged 82-83 have some kind of heart disease. For men, the heart attack prevalence rate increases with age. Although the prevalence rate of heart disease without a heart attack for men increases with age up to age 84 (26.5%), it decreases after that age. For women, the implied prevalence rate of heart disease without a heart attack and the prevalence rate of heart attack both continuously increase with age. The overall heart disease prevalence rate is higher for men than it is for women. However, the prevalence rate of heart disease without a heart attack is higher for women aged 54 and over than it is for men. Among individuals who have never experienced a heart problem at a given age, expected overall heart disease prevalence increases with age for men and for women. The expected overall heart disease prevalence rate for those who never experienced a heart problem is higher for men than it is for women at any age. The prevalence rate of heart disease without a heart attack among people who have never experienced a heart problem increases with age up to age 88 (27.0%) and then decreases for men while women expect the increase of prevalence rate of heart disease without having a heart attack to be continuous. Both men and women are expected to see an increase in the heart attack prevalence rates with age, with higher rates for men than for women. For those who already have a heart problem besides a heart attack, both men and women are expected to have increasing heart attack prevalence with age. Although half of those men aged 80 and over are expected to have a heart attack, the prevalence of having a heart attack never becomes so high among women. 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Those who are interested in information on the health of a population as a whole should use population-based multistate methods. However, those policy makers who are interested in a particular group by heart disease status for policy relevant to medical institutions, health care personnel, medication, and others may want to look at the results of the status-based method. However, it should be noted that years of life with and without heart disease, as computed in this study, are indicators of the average length of life for a corresponding cohort whose members live their remaining years of life exposed to current conditions of mortality and morbidity. These estimated indicators are based on current rates for individual life cycles and population. Therefore, monitoring of future population trends such as risk factors, incidence, and change in mortality from heart disease is needed to see if such model implications are likely to be borne out in the future. This study’s strength involves the use of a sample representing the population aged 50 and over in the United States. By using a national sample, study outcomes provide information on a nationally based sample. However, there are some limitations and possible design problems in this study. First, mortality information on the time trend obtained from the Vital Statistics cannot be broken down to an individual level, rather this information is limited to the data of abridged 5-year age groups. This data source limits further analysis of all people aged 50 and over as a whole population in this study. Second, the trend in duration of heart disease used in this study is based on the period of time between the date of the interview and 1.5 or 2.5 years after the interview. Since the data set covers the interview period between 1986 and 1994, using a longer time period shortens the period used in the trend study. Thus, there is a trade-off in the choice between the time period of heart 185 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. disease and the examination period. Third, due to the limitation of obtaining data on the incidence trend for people aged 50 and over, as well as the difficulty in using heart disease duration data, complete examination of the relationship among prevalence, incidence, and duration of heart disease can not be conducted. Fourth, the sample at Wave II includes survivors from new heart problems and deceased individuals after the Wave I interview. Therefore, the drop-outs from the original Wave I sample are not included, and some of them may have died. In addition, those who died due to heart problems cannot be identified in this study at this time. Fifth, all trends except for the mortality rates from heart disease, are based on the noninstitutionalized U.S. population. This could result in an underestimation of the trend of heart disease for the whole U.S. population aged 50 and over. There is a potential for change in the diagnosis of heart disease over time which could affect outcomes in the study. The change is potentially caused by several factors. The influence of physicians' behaviors is probably greater than that of other social systems such as governmental health policies and organizational priorities (McKinlay, 1996). For example, there are no differences in the measurement of heart disease between the ICD-8 and the ICD-9 for the heart disease diagnostic criteria in this study. However, health care policy by the government affects physicians' diagnostic behaviors. For example, by examining patients who were discharged from several hospitals in communities of Rhode Island and Massachusetts between 1980 and 1988, one study indicates that the prospective payment system (PPS) that was enacted in 1983 with the use of diagnosis-related groups (DRGs) has influenced the assignment of hospital discharge codes in a way that would increase reimbursement to hospitals. In this study, researchers found that 186 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. there was an increase in the use of the diagnostic codes for acute (410 and 411 in the ICD code) rather than chronic (412 through 414 in the ICD code) forms of IHD (Assaf, Lapane, McKenney, Carleton, 1993). As a result of this diagnostic change, health care providers are paid more for a new acute episode of IHD than for an old IHD episode. It also indicated that organizational priorities contribute to the social pattern of heart disease diagnosis outcomes. For instance, health maintenance organizations (HMOs) may seek better efficiency through cost-containing practices that affect behaviors to diagnose heart disease (McKinlay, 1996). As a result, it is possible for them to persuade doctors to diagnose heart disease in the most cost effective way. There are some unknown issues about sex differences in the trend of heart disease and expected years of life with heart disease and without heart disease. The differences may be due to sex-specific risk factors (Corrao, Becker, Ockene, & Hamilton, 1990; Nettleman and Klein, 1998). Although many studies conclude the differences by sex are due to unequal treatment or diagnoses of heart disease for men and those for women, these are controversial issues (Chiriboga, Yarzebski, Goldberg, Chen, Gurwitz, Gore, Alpert, & Dalen, 1993; Kannel, Cupples, & Gagnon, 1990; Krumholz, Douglas, Lauer, & Pasternak, 1992; Malacrida, Genoni, Maggioni, Spataro, Parish, Palmer, Collins, & Moccetti, 1998; Maynard, Litwin, Martin, & Weaver, 1992; Roger, Farkouh, Weston, Reeder, Jacobsen, Zinsmeister, Yawn, Kopecky, & Gabriel, 2000; Shaw, Miller, Romeis, Kargl, Younis, & Chaitman, 1994). Furthermore, different trends in heart disease by sex are confounded by the issue of measurement. This issue includes changes in the diagnosis and treatment of heart disease along with the development of medical technology. For instance, 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. different criteria of diagnosis result in different outcomes (Madans, Reuben, Rothwell, & Eberhardt, 1995). The measurement issue further involves a limitation of heart disease detection. Many studies point out that there are a certain number of unrecognized elderly individuals for the detection of heart disease because of their atypical symptoms (Aronow, 1987; Kannel, Cupples, & Gagnon, 1990; Healy, 1991; Maheshwari, Laird-Fick, Cannon, & DeHart, 2000; Nadelmann, Frishman, Ooi, Tepper, Greenberg, Guzik, Lazar, Heiman, & Aronson, 1990; Sheifer, Gersh, Yanez, Ades, Burke, & Manolio, 2000). In this study, the time trends in mortality and prevalence have declined over recent years. These trends are based on age-specific rates and leave the issue of absolute numbers of people who experience these problems undiscussed. For example, increases in the prevalence of heart disease in the future are expected because the aging of the population will increase CHD prevalence and annual incidence, mortality, and costs by about 40-50% by the year 2010, assuming that there have been no changes in risk factors or in the efficacy of treatments after 1980. By the year 2010, it is projected that the annual number of persons experiencing their first CHD event will increase by about 38% in the population aged 35-84, with an estimated 10% decline in CHD incidence rates. The prevalence of CHD in the US population is projected to increase 50% in absolute terms, from 6.0 million in 1980 to 9.0 million in 2010. Since the projected prevalence decline is larger than the decline of incidence, they imply an increase in years of life among people with CHD. They also project a future increase in the absolute annual number of CHD deaths despite the decline of the CHD mortality rate (Weinstein, Coxson, Williams, Pass, Stason, & Goldman, 1987). These patterns indicate increases in the absolute number of heart 188 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. problems despite declines in their rates, this will lead to an increase in demand for medical institutions, treatments, and health care personnel. Current heart disease trends may continue in the future. It should be noted that even if mortality, incidence, and prevalence rates decline in the future, the absolute number of heart disease cases may increase due to population aging in the United States. Population aging will continue as baby boomers who were born between 1946 and 1964 grow old. People who practice good health behaviors and have lower heart risk factors may not have heart problems while they are young but may develop heart disease later in the future. Disability cases may also increase. It was reported that CHD accounts for 19% of disability allowances by the Social Security Administration in the United States (AHA, 1999). If the number of older persons increases in the future even if heart disease is less disabling, the number of disability cases may increase. In addition to the issues addressed above, further studies should link the heart disease trend to the social burdens linked to heart disease. This could produce effective indicators that could be linked to the future development of medical technology, diagnosis, and treatments. It will also be useful in the future to include risk factors and health behaviors as well as race and education in estimating the burden of heart disease. Links to death reports and medical records should also be made to determine the health experience and cause of death of persons in longitudinal samples so they can be linked to the multistate life table in future study. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES Agner, E. (1985). Epidemiology of coronary heart disease in the elderly patient. In E. L. Coodley (Ed.). Geriatric heart disease (pp. 114-125). Littleton, MA: PSG. Alho, J. M. (1992). On prevalence, incidence, and duration in general stable populations. Biometrics. 48(21. 587-592. Allison, P. D. (1995). Survival analysis using the SAS system: A practical guide. Cary, NC: SAS Institute Inc. American Heart Association. (1999). 2000 Heart and Stroke Statistical Update. Dallas, TX: American Heart Association. Anonymous. (1995a). Codebook: Assets and Health Dynamics Among the Oldest Old (AHEAD). 1993 (Wave 1). Retrieved February 2, 2000 from the World Wide Web: http://www.umich.edu/~hrswww/center/codebks Anonymous. (1995b). Codebook: Health and Retirement Study (HRS). 1992 (Wave 1). Retrieved February 2, 2000 from the World Wide Web: http://www.umich.edu/~hrswww/center/codebks Anonymous. (1999a). AHEAD Overview. Retrieved February 2, 2000 from the World Wide Web: http://www.umich.edu/~hrswww/overview/ahdover.html Anonymous. (1999b). Decline in deaths from heart disease and Stroke- United States, 1900-1999. MMWR - Morbidity & Mortality Weekly Report. 48(30). 649-656. Anonymous. (1999c). HRS Overview. Retrieved February 2, 2000 from the World Wide Web: http://www.umich.edu/~hrswww/overview/hrsover.html Anonymous. (2000a). Description. Division of Health Interview Statistics. National Center for Health Statistics. Centers for Disease Control and Prevention. Retrieved February 2, 2000 from the World Wide Web: http://www.cdc.gov/nchs/about/major/nhis/hisdesc.htm Anonymous. (2000b). Mortality Data from the National Vital Statistics System. Division of Health Interview Statistics. National Center for Health Statistics. Centers for Disease Control and Prevention. Retrieved February 2, 2000 from the World Wide Web: http://www.cdc.gov/nchs/nvss.htm 190 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Anonymous. (2000c). National Health Interview Survey Multiple Cause of Death Public-Use Data Files. Division of Health Interview Statistics. National Center for Health Statistics. Centers for Disease Control and Prevention. Retrieved February 2, 2000 from the World Wide Web: http://www.cdc.gov/nchs/products/catalogs/subject/nhis/nhismcd.htm Armstrong, D. L., Wing, S. B., & Tyroler, H. A. (1995). United States mortality from ill-defined causes, 1968-1988: Potential effects on heart disease mortality trends. International Journal of Epidemiology. 24(3). 522-527. Aronow, W. S. (1987). Prevalence of presenting symptoms of recognized acute myocardial infarction and of unrecognized healed myocardial infarction in elderly patients. American Journal of Cardiology. 60(14). 1182. Assaf, A. R., Lapane, K. L., McKenney, J. L., & Carleton, R. A. (1993). Possible influence of the prospective payment system on the assignment of discharge diagnoses for coronary heart disease. New England Journal of Medicine. 329(13). 931-935. Bonneux, L., Barendregt, J. J., Meeter, K., Bonsel, G. J., & Van der Maas, P. J. (1994). Estimating clinical morbidity due to ischemic heart disease and congestive heart failure: The future rise of heart failure. American Journal of Public Health. 84(1). 20-28. Burke, G. L, Edlavitch, S. A., & Crow, R. S. (1989). The effects of diagnostic criteria on trends in coronary heart disease morbidity: The Minnesota Heart Survey. Journal of Clinical Epidemiology. 42(1). 17-24. Burke, G. L. & Manolio, T. A. (1993). Epidemiology of established major cardiovascular disease risk factors. In A. S. Kapoor and B. N. Singh (Eds.) Prognosis and risk assessment in cardiovascular disease (pp. 61-72). New York: Churchill Livingstone. Burke, G. L., Sprafka, J. M., Folsom, A. R., Luepker, R. V., Norsted, S. W., & Blackburn, H. (1989). Trends in CHD Mortality, Morbidity and Risk Factor Levels from 1960 to 1986: The Minnesota Heart Survey. International Journal of Epidemiology 18. S73-S81. Chiriboga, D. E., Yarzebski, J., Goldberg, R. J., Chen, Z., Gurwitz, J., Gore, J. M., Alpert, J. S., Dalen, J. E. (1993). A community-wide perspective of gender differences and temporal trends in the use of diagnostic and revascularization procedures for acute myocardial infarction. American Journal of Cardiology. 71(4). 268-273. Clayton, D., & Hills, M. (1993). Statistical models in epidemiology. Oxford, NY: Oxford University Press. 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cohen, R. A., Van Nostrand, J. F., & Fumer, S. E. (Eds.) (1993). Chartbook on health data on older Americans: United States. 1992. Hyattsville, MD: National Center for Health Statistics. Corrao, J. M., Becker, R. C., Ockene, I. S., Hamilton, G. A. (1990). Coronary heart disease risk factors in women. Cardiology. 77(SupdI. 2), 8-24. Crimmins, E. M., Hayward, M. D., & Saito, Y. (1994). Changing mortality and morbidity rates and the health status and life expectancy of the older population. Demography. 31. 159-175. Crimmins, E. M., Hayward, M. D., & Saito, Y. (1996). Differentials in active life expectancy in the older population of the United States. Journal of Gerontology. Social Sciences. 51 B(3). S111-S120. Cutler, D. M., & Sheiner, L. (1998). Demographics and Medical Care Spending: Standard and Non-Standard Effects. NBER Working Paper No. 6866. Cambridge, MA: National Bureau of Economic Research. Demirovic, J., Blackburn, H., McGovern, P., Sprafka, J. M., & Doliszny, K. (1993). Sex differences in coronary heart disease mortality trends: The Minnesota Heart Survey, 1970-1988. Epidemiology 4(1). 79-82. Doliszny, K. M., Luepker, R. V., Burke, G. L., Pryor, D. B., & Blackburn, H. (1994). Estimated contribution of coronary artery bypass graft surgery to the decline in coronary heart disease mortality: The Minnesota Heart Survey. Journal of the American College of Cardiology 24(1). 95-103. Elveback, L. R., Connolly, D. C., & Melton, L. J., III. (1986). Coronary heart disease in residents of Rochester, Minnesota. VII. Incidence, 1950 through 1982. Mavo Clinic Proceedings. 61(11), 896-900. Epstein, F. H. (1988). International trends in mortality and morbidity from ischemic heart disease. Annals of Clinical Research. 20(1-2). 21-25. Ettinger, W. H. Jr., Fried, L. P., Harris, T., Shemanski, L., & Schulz R. R. J. (1994). Self-reported causes of physical disability in older people: The Cardiovascular Health Study. CHS Collaborative Research Group. Journal of the American Geriatrics Society. 42(10). 1035-1044. Feinleib, M. (1995). Trends in heart disease in the United States. American Journal Medical Sciences 310(SuppI. 1), S8-S14. Freeman, J., & Hutchison, G. B. (1980). Prevalence, incidence and duration. American Journal of Epidemiology. 112(5). 707-723. 192 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Gillum, R. F. (1994). Trends in acute myocardial infarction and coronary heart disease death in the United States. Journal of the American College of Cardiology. 23(6). 1273-1277. Gillum, R. F., Folsom, A. R., & Blackburn, H. (1984). Decline in coronary heart disease mortality. Old questions and new facts. American Journal of Medicine. 76(6), 1055-1065. Gillum, R. F., Folsom, A., Luepker, R. V., Jacobs, D. R. Jr., Kottke, T. E., Gomez-Marin, O., Prineas, R. J., Taylor, H. L., & Blackburn, H. (1983). Sudden death and acute myocardial infarction in a metropolitan area, 1970-1980: The Minnesota Heart Survey. New England Journal of Medicine. 309(22). 1353-1358. Goldberg, R. J., Gorak, E. J., Yarzebski, J., Hosmer, D. W., Jr., Dalen, P., Gore, J. M., Alpert, J. S., & Dalen, J. E. (1993). A communitywide perspective of sex differences and temporal trends in the incidence and survival rates after acute myocardial infarction and out-of-hospital deaths caused by coronary heart disease. Circulation. 87(6). 1947-1953. Gomez-Marin, O., Folsom, A. R., Kottke, T. E., Wu, S. C., Jacobs, D. R., Jr., Gillum, R. F., Edlavitch, S. A., & Blackburn, H. (1987). Improvement in long-term survival among patients hospitalized with acute myocardial infarction, 1970 to 1980. The Minnesota Heart Survey. New England Journal of Medicine 316(22). 1353-1359. Haan, M. N., Selby, J. V., Rice, D. P., Quesenberry, C. P., Jr., Schofield, K. A., Liu, J., & Fireman, B. H. (1996). Trends in cardiovascular disease incidence and survival in the elderly. Annals of Epidemiology 6. 348-356. Hall, W. D., Ferrario, C. M., Moore, M. A., Hall, J. E., Flack, J. M., Cooper, W., Simmons, J. D., Egan, B. M., Lackland, D. T., Perry, M., Jr., & Roccella, E. J. (1997). Hypertension-related morbidity and mortality in the southeastern United States. American Journal of the Medical Sciences. 313(4). 195-209. Hayward M. D., Crimmins E. M., & Wray L. A. (1994). Relationship between retirement life cycle changes and older men's labor force participation rates. Journals of Gerontology. 49(5). S219-S230. Hayward, M. D., & Grady, W. R. (1990). Work and retirement among a cohort of older men in the United States. Demography. 27(31 337-356. Hennekens, C. H., & Buring, J. E. (Eds.) (1987). Epidemiology in medicine. Boston: Little Brown. Ho, K. K., Anderson, K. M., Kannel, W. B., Grossman, W., & Levy, D. (1993). Survival after the onset of congestive heart failure in Framingham Heart Study subjects. Circulation. 88(1). 107-115. 193 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hunink, M. G., Goldman, L., Tosteson, A. N., Mittleman, M. A., Goldman, P. A., Williams, L. W., Tsevat, J., & Weinstein, M. C. (1997). The recent decline in mortality from coronary heart disease, 1980-1990: The effect of secular trends in risk factors and treatment. Journal of American Medical Association. 277. 535-542. Kane, R. L., Ouslander, J.G., & Abrass, I. B. (1999). Essentials of clinical geriatrics (4th ed.). New York: McGraw-Hill, Health Professions Division. Kannel, W. B., Cupples, L. A., & Gagnon, D. R. (1990). Incidence, precursors and prognosis of unrecognized myocardial infarction. Advances in Cardiology. 37, 202-14. Kaplan, G. A., Cohn, B. A., Cohen, R. D., & Guralnik, J. (1988). The decline in ischemic heart disease mortality: Prospective evidence from the Alameda County Study. American Journal of Epidemiology. 127(6). 1131-1142. Keiding, N. (1991). Age-specific incidence and prevalence: A statistical perspective. Journal of Roval Statistical Society, Part A, 154(3). 371-412. Keil, J. E., Gazes, P. C., Litaker, M. S., Saunders, D. E., Jr., Weinrich, M. C., Baroody, N. B., Jr., Lackland, D. T., & Hudson, M. B. (1989). Changing patterns of acute myocardial infarction: Decline in period prevalence and delay in onset. American Heart Journal. 117(5). 1022-1029. Kleinman, J. C. (1986). State trends in infant mortality, 1968-83. American Journal of Public Health. 76(6). 681-688. Krumholz, H. M., Douglas, P. S., Lauer, M. S., & Pasternak, R. C. (1992). Selection of patients for coronary angiography and coronary revascularization early after myocardial infarction: Is there evidence for a gender bias? Annals of Internal Medicine. 116(10). 785-790. Lee, R. D., & Carter, L. (1992). Modeling and forecasting US Mortality. Journal of the American Statistical Association. 87(419). 659-675. Madans, J. H., Reuben, C. A., Rothwell, S. T., & Eberhardt, M. S. (1995). Differences in morbiditymeasures and risk factor identification using multiple data sources: The case of coronary heart disease. Statistics in Medicine. 14(5-7). 643- 653. Maheshwari, A., Laird-Fick, H. S., Cannon, L. A., & DeHart, D. J. (2000). Acute Ml. Age-related presentations and treatment options. Geriatrics. 55(2). 32-34, 37-40. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Malacrida, R., Genoni, M., Maggioni, A. P., Spataro, V., Parish, S., Palmer, A., Collins, R., & Moccetti, T. (1998). A comparison of the early outcome of acute myocardial infarction in women and men. The Third International Study of Infarct Survival Collaborative Group. New England Journal of Medicine. 338(1). 8-14. Marmot, M. (1996). Ischaemic heart disease. In S. Ebrahim & A. Kalache (Eds.), Epidemiology in old age (pp. 253-261). London: BMJ Pub. Group in collaboration with the World Health Organization. Mausner, J. S., & Kramer, S. (Eds.) (1985). Epidemiology: An introductory text. (2nded.). Philadelphia: Saunders. Maynard, C., Litwin, P. E., Martin, J. S., & Weaver, W. D. (1992). Gender differences in the treatment and outcome of acute myocardial infarction. Results from the Myocardial Infarction Triage and Intervention Registry. Archives of Internal Medicine. 152(5). 972-976. McGovern, P. G., Folsom, A. R., Sprafka, J. M., Burke, G. L., Doliszny, K. M., Demirovic, J., Naylor, J. D., & Blackburn, H. (1992). Trends in Survival of Hospitalized Myocardial Infarction Patients between 1970 and 1985. Circulation 85. 172-179. McGovern, P. G., Pankow, J. S., Shahar, E., Doliszny, K. M., Folsom, A. R., Blackburn, H., & Luepker, R. V. (1996). Recent trends in acute coronary heart disease. New England Journal of Medicine 334. 884-890. McKinlay, J. B. (1996). Some contributions from the social system to gender inequalities in heart disease. Journal of Health & Social Behavior. 37(1). 1-26. Murray, C. J., & Lopez, A. D. (1994). Quantifying disability: Data, methods and results. Bulletin of the World Health Organization. 72(3). 481-494. Nadelmann, J., Frishman, W. H., Ooi, W. L., Tepper, D., Greenberg, S., Guzik, H., Lazar, E. J., Heiman, M., & Aronson, M. (1990). Prevalence, incidence and prognosis of recognized and unrecognized myocardial infarction in persons aged 75 years or older: The Bronx Aging Study. American Journal of Cardiology. 66(5). 533-537. National Academy on an Aging Society (2000). Heart disease: A disabling vet preventable condition. No 3. Washington, DC: National Academy on Aging Society. National Center for Health Statistics. (1996). Health. United States. 1995. Hyattsville, MD: Public Health Service. National Center for Health Statistics. (1999a). Healthy People 2000 Review. 1998-99. Hyattsville, MD: Public Health Service. 195 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. National Center for Health Statistics. (1999b). Death Rates for 72 Selected Causes bv 5-Year Age Groups. Race, and Sex: United States. 1979-97. http://www.cdc.gov/nchs/data/gm291_1.pdf Nettleman, M. D., & Klein, W. S. (1998). Comparison of the early outcome of acute myocardial infarction in women and men. New England Journal of Medicine. 338(21). 1543-1544. Nickel, J. T„ & Chirikos, T. N. (1990). Functional disability of elderly patients with long-term coronary heart disease: A sex-stratified analysis. Journal of Gerontology. Social Sciences. 45B(2). S60-S68. Pollard, A. H., Yusuf, F., & Pollard, G.N. (1974). Demographic techniques (3rd ed.). Rushcutters Bay, N.S.W. Australia: Pergamon Press. Roger, V. L., Farkouh, M. E., Weston, S. A., Reeder, G. S., Jacobsen, S. J., Zinsmeister, A. R. Yawn, B. P., Kopecky, S. L., Gabriel, S. E. (2000). Sex differences in evaluation and outcome of unstable angina. Journal of American Medical Association. 283(5). 646-652. Roger, V. L„ Jacobsen, S. J., Weston, S. A., Bailey, K. R., Kottke, T. E„ & Frye, R. L. (1999). Trends in heart disease deaths in Olmsted County, Minnesota, 1979-1994. Mavo Clinic Proceedings. 74(7). 651-657. Rogers, A. (1975). Introduction to multi-regional mathematical demography. New York: Wiley. Rogers, A., Rogers, R., & Branch, L. G. (1989). A multistate analysis of active life expectancy. Public Health Reports. 104. 222-226. Rogers, A., Rogers, R., & Branch, L. G. (1990). Longer life but worse health? measurement and dynamics. The Gerontologist. 30. 640-649. Rogers, A., & Ledent, J. (1976). Increment-decrement life tables: A comment. Demography. 13. 287-290. Rosamond, W. D., Chambless, L. E., Folsom, A. R., Cooper, L. S., Conwill, D. E., Clegg, L., Wang, C. H., & Heiss, G. (1998). Trends in the incidence of myocardial infarction and in mortality due to coronary heart disease, 1987 to 1994. New England Journal of Medicine. 339(13). 861-867. Saito, Y., E. M. Crimmins, & Hayward, M. D. (1999). Health expectancy: An overview. (Nihon University Research Paper Series, No. 67). Tokyo: Nihon University. Schick, F. L., & Schick, R. (Eds.). (1994). Statistical handbook on aging Americans. 1994 Edition. Phoenix, AZ: Oryx Press. 196 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Schoen, R. (1988). Modeling multigroup populations. New York: Plenum Press. Schoen, R., & Land, K. C. (1979). A general algorithm for estimating a Markov-generated increment-decrement life table with applications to marital-status patterns. Journal of the American Statistical Association. 74(368). 761-776. Schoen, R., & Woodrow, K. (1980). Labor Force Status Life Tables for the United States. Demography. 17(3). 297-322. Shaw, L. J., Miller, D. D., Romeis, J. C., Kargl, D., Younis, L. T., & Chaitman, B. R. (1994). Gender differences in the noninvasive evaluation and management of patients with suspected coronary artery disease. Annals of Internal Medicine. 120(7). 559-566. Sheifer, S. E., Gersh, B. J., Yanez, N. D., III., Ades, P. A., Burke, G. L., & Manolio, T. A. (2000). Prevalence, predisposing factors, and prognosis of clinically unrecognized myocardial infarction in the elderly. Journal of the American College of Cardiology. 35(1). 119-26. Sytkowski, P. A., D'Agostino, R. B., Belanger, A., & Kannel, W. B. (1996). Sex and time trends in cardiovascular disease incidence and mortality: The Framingham Heart Study, 1950-1989. American Journal of Epidemiology. 143(4). 338-350. Silverstein, M., Chen, X., & Heller, K. (1996). Too much of a good thing?: Intergenerational social support and the psychological well-being of older parents. Journal of Marriage and the Family. 58. 970-982. Sytkowski, P. A., Kannel, W. B., & D'Agostino, R. B. (1990). Changes in risk factors and the decline in mortality from cardiovascular disease. The Framingham Heart Study. New England Journal of Medicine, 322(23). 1635-1641. Thom, T. J. & Maurer, J. (1988). Time trends for coronary heart disease mortality and morbidity. In M. W. Higgins & R. V. Luepker (Eds.) Trends in coronary heart disease mortality: The influence of medical care (pp. 7-15). New York: Oxford University Press. Timiras, P. S. (1994). Cardiovascular alterations with age: Atherosclerosis, coronary heart disease, hypertension. In P. S. Timiras (Ed.), Physiological basis of aoino and geriatrics (2nd ed., pp. 199-214). Boca Raton: CRC Press. Weinstein, M. C., Coxson, P. G„ Williams, L. W., Pass, T. M„ Stason, W. B., & Goldman, L. (1987). Forecasting coronary heart disease incidence, mortality, and cost: The Coronary Heart Disease Policy Model. American Journal of Public Health. 77(11), 1417-1426. 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wenger, N. K. (1992). Coronary heart disease in women: A 'new' problem. Hospital Practice. 27(11). 59-62, 64, 67, 71, 72, 74. 198 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX A Appendix A is a collection of supplemental tables and figures that are comprised of the data that underlie tables, figures, and text found in the Chapter IV in this dissertation. In addition to the examinations of linear relations between mortality and time trend with age, curvilinear relations between heart disease mortality rates and age factors, age, age-squared, and age-cubed terms are examined sequentially in regression models for both overall heart disease and AMI for men and women for the period of 1968-1996. Models M 1 and W1 only use a single age term. Models M2 and W2 add an age-squared term to Models M 1 and W1, and age-cubed terms are additionally included in Models M3 and W3. Original models (Models M1 and W1) show statistically significant relationship between age and the death rate of heart problem for men and women. Added terms show the significant relationship with the death rate of heart problem for men and women while other terms lose their significant relationships. To examine whether there is a better fitting model for heart disease prevalence, age-squared, age-cubed, and age-logarithm terms were included in the logistic regression models. These models were examined for both overall heart disease prevalence and ischemic heart disease (IHD) prevalence for the period of 1982-1996. This was done separately for men and women. All age terms were statistically significant at p<0.001. In addition to determining the odds ratios for the 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. different age terms, the probabilities of having heart disease at a particular age and year were calculated by using the parameter estimates from the logistic regression models including linear, quadratic, cubic, and logarithm equations. These are illustrated for comparisons. In sum, the age-cubic models seem to provide better estimations to the actual prevalence rates for both overall heart disease and IHD for men and women, but the effect of the time trend is essentially the same in all models. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. A.1 Parameter Estimates from Regression Analyses for Death Rates from Overall Heart Disease for Aged 50 and Over, by Sex, Vital Statistics 1982-1996 (per 100,000 population) N Intercept Age5 Age5-squared Age5-cubed R-square Men Model M1 8 -859.09 904.98 •• 0.783 Model M2 8 659.55 -613.66 216.95 “ 0.963 Model M3 8 156.94 702.72 -28567 47.87 * 0.994 Women Model W1 8 -1018.42 746 99 * 0.677 Model W2 8 562.79 -834.22 225.89 “ 0.925 Model W3 8 -60.79 798.96 -397.69 59.39 * 0.985 •p<0 05, "pcO.OI, —p <0.001 Age5:5-year in terval of age started at aged 50-54 A.2 Parameter Estimates from Regression Analyses for Death Rates from Overall Heart Disease for Men Aged 50 and Over, by Year, Vital Statistics 1982-1996 (per 100,000 population) Year Model N Intercept Age5 Age5-squared Age5-cubed R-square 1982 M 1 8 -833.5 994.0 " 0.813 M 2 8 705.5 -545.0 219.9 ” 0.972 M 3 8 235.1 687.0 -250.6 44.8 * 0.995 1983 M 1 8 -870.6 1,007.1 ** 0.807 M 2 8 715.7 -579.1 226.6 ** 0.971 M 3 8 226.4 702.4 -262.7 46.6 ’ 0.995 1984 M 1 8 -867.3 986.7 “ 0.805 M2 8 699.8 -580.4 223.9 " 0.971 M 3 8 216.4 685.5 -259.5 46.0 " 0.996 1985 M 1 8 -886.3 987.1 " 0.801 M 2 8 700.0 -599.2 226.6 ** 0.970 M 3 8 209.9 684.3 -263.4 46.7 *• 0.996 1986 M 1 8 -860.2 954.2 ” 0.801 M 2 8 670.2 -576.2 218.6 •• 0.969 M 3 8 193.3 672.8 -258.3 45.4 ' 0.995 1987 M 1 8 -864.2 937.6 ” 0.795 M 2 8 664.9 -591.5 218.4 •* 0.967 M 3 8 180.3 677.6 -266.1 46.1 ’ 0.995 1988 M 1 8 -902.9 946.1 " 0.787 M 2 8 669.5 -626.3 224.6 " 0.964 M 3 8 154.7 722.0 -290.2 49.0 * 0.994 1989 M 1 8 -851.3 889.3 ** 0.785 M 2 8 638.6 -600.6 212.8 ” 0.964 M 3 8 149.9 679.4 -275.9 46.5 " 0.995 1990 M 1 8 -859.7 876.3 ** 0.779 M 2 8 629.9 -613.4 212.8 ** 0.962 M 3 8 134.7 683.7 -282.4 47.2 ** 0.994 1 9 9 1 M 1 8 -843.1 853.0 " 0.773 M 2 8 623.8 -614.0 209.6 ** 0.959 M 3 8 121.4 702.0 -292.9 47.9 • 0.993 1992 M 1 8 -837.8 836.1 ” 0.765 M 2 8 627.2 -628.9 209.3 •* 0.956 M 3 8 115.2 712.0 •302.7 48.8 * 0.993 1993 M 1 8 -873.1 854.3 " 0.759 M2 8 645.6 -664.4 216.9 ** 0.955 M 3 8 109.9 738.4 -318.7 51.0 ** 0.993 1994 M 1 8 -848.3 825.0 " 0.752 M2 8 637.7 -661.0 212.3 ** 0.952 M 3 8 103.5 738.0 -321.9 50.9 * 0.992 1995 M 1 8 -843.9 813.9 " 0.750 M 2 8 632.5 -662.4 210.9 ** 0.951 M 3 8 101.7 727.7 -319.9 50.6 * 0.992 1996 M 1 8 -843.9 813.9 " 0.750 M2 8 632.5 •662.4 210.9 ” 0.951 M 3 8 101.7 727.7 -319.9 50.6 * 0.992 *p<0.05. **p<0.01. ***p <0.001 Aqe5: 5-year interval of age started at aged 50-54 202 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A.3 Parameter Estimates from Regression Analyses for Death Rates from Overall Heart Disease for Women Aged 50 and Over, by Year, Vital Statistics 1982-1996 (per 100,000 population) Year Model N Intercept Age5 Age5-squared Age5-cubed R-square 1982 W 1 8 -1,051.8 806.8 *• 0.717 W2 8 545.3 -790.3 228.2 •* 0.946 W3 8 -22.9 697.9 -340.1 54.1 * 0.991 1983 W 1 8 -1,075.6 817.7 " 0.708 W2 8 570.2 -828.1 235.1 " 0.942 W3 8 -28.4 739.8 -363.5 57.0 * 0.990 1984 W 1 8 -1,056.2 800.7 " 0.705 W2 8 560.0 -815.5 230.9 •* 0.940 W3 8 -35.5 744.1 -364.6 56.7 * 0.990 1985 W 1 8 -1,069.6 801.8 * 0.695 W2 8 577.4 -845.2 235.3 ** 0.934 W3 8 •46.0 787.4 -388.1 59.4 * 0.988 1986 W 1 8 -1,064.7 792.5 • 0.690 W2 8 576.2 -848.3 234.4 •* 0.931 W3 a -53.7 801.3 -395.5 60.0 * 0.987 1987 W 1 8 -1,052.3 779.8 * 0.683 W2 8 580.0 -852.5 233.2 " 0.927 W3 8 -58.4 819.5 -405.2 60.8 * 0.986 1988 W 1 8 -1,067.8 783.0 * 0.677 W2 8 591.0 -875.7 237.0 " 0.924 W3 8 -65.1 842.6 -419.1 62.5 * 0.985 1989 W 1 8 -1,019.3 742.9 * 0.674 W2 8 562.0 -838.3 225.9 " 0.924 W3 8 -63.6 800.0 -399.7 59.6 ’ 0.985 1990 W 1 8 -996.5 722.6 * 0.668 W2 8 555.6 -829.5 221.7 * 0.919 W3 8 -69.6 808.0 -403.5 59.5 * 0.983 19 9 1 W 1 8 -980.7 707.8 * 0.664 W2 8 550.8 -823.7 218.8 * 0.918 W3 8 -71.2 805.3 -403.2 59.2 * 0.983 1992 W 1 8 -964.1 693.5 * 0.658 W2 8 549.5 -820.1 216.2 * 0.914 W3 8 -73.3 811.0 -406.5 59.3 * 0.983 1993 W 1 8 •991.9 708.9 * 0.656 W2 8 562.8 -845.8 222.1 * 0.913 W3 8 -79.7 837.1 -420.5 61.2 * 0.982 1994 W 1 8 -969.2 689.2 * 0.652 W2 8 553.5 -833.5 217.5 * 0.911 W3 8 -79.8 825.2 -415.8 60.3 ’ 0.982 1995 W 1 8 -970.7 687.1 * 0.646 W2 8 560.2 -843.7 218.7 * 0.907 W3 8 -85.9 848.3 -427.4 61.5 * 0.980 1996 W 1 8 -946.1 670.4 * 0.647 W2 8 547.6 -823.2 213.4 ' 0.909 W3 8 -78.7 816.9 -412.9 59.6 * 0.981 *p<0.05. **p<0.01, ***p <0.001 AgeS: 5-year interval of age started at aged 50-54 203 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. A.4 Parameter Estimates from Regression Analyses for Death Rates from Ischemic Heart Disease (IHD) for Aged 50 and Over, by Sex, Vital Statistics 1982-1996 (per 100,000 population) N Intercept Age5 Age5-squared Age5-cubed R-square Men Model M1 8 -143.18 258.16 ••• 0.888 Model M2 8 167.09 • -52.10 44.32 *•* 0 993 Model M3 8 105.05 " 110.38 * -17.72 5.91 0.999 Women Model W1 8 -205.60 188 35 " 0.803 Model W2 8 100.45 -117.69 43.72 ** 0.976 Model W3 8 15.54 104.69 -41.19 8.09 “ 0.997 •p<0.05, “ p <0.01. — p <0.001 Age5.5-year in te rva l of age started at aged 50-54 A.5 Parameter Estimates from Regression Analyses for Death Rates from Acute Myocardial Infarction (AMI) for Men Aged 50 and Over, by Year, Vital Statistics 1982-1996 (per 100,000 population) Year Model N Intercept Aae5 Aqe5-squared Age5-cubed R-square 1982 M 1 8 -68.0 298.0 •** 0.941 M 2 8 190.3 *•* 39.7 • 36.9 "• 0.999 M 3 8 167.4 *•* 99.5 ** 14.0 2.2 * 1.000 1983 M 1 8 -86.8 295.6 "• 0.935 M2 8 183.3 "* 25.4 38.6 *•* 0.999 M 3 8 160.7 *" 84.8 ** 15.9 * 2.2 ■ 1.000 1984 M 1 8 -101.0 289.4 *•* 0.930 M 2 8 175.3 •** 1 3 .1 39.5 •” 0.999 M 3 8 15 4 .1 ” • 68.8 ** 18.2 ** 2.0 •* 1.000 1985 M 1 8 -121.5 288.6 0.919 M 2 8 174.8 ” -7.7 42.3 "• 0.998 M 3 8 1 3 9 .1 •" 85.8 ” 6.6 3.4 "• 1.000 1986 M 1 8 -119.3 269.1 “ • 0.919 M2 8 158.4 ” -8.7 39.7 *•* 0.999 M 3 8 128.8 •” 68.9 "• 1 0 .1 * 2.8 •** 1.000 1987 M 1 8 -134.5 264.1 — 0.905 M 2 8 160.5 ” -30.9 42.1 *•* 0.997 M 3 8 116.7 **• 83.9 •* -1.7 4.2 ** 1.000 1988 M 1 8 -155.2 264.0 "• 0.894 M 2 8 15 6 .1 * -47.4 44.5 *•* 0.995 M 3 8 102.6 92.8 ** -9.0 5 .1 " 1.000 1989 M 1 8 -170.5 265.1 **• 0.877 M2 8 165.7 * -71.1 48.0 *•* 0.992 M 3 8 94.7 ” 114.8 “ -23.0 * 6.8 •** 1.000 1990 M 1 8 -176.2 258.7 "• 0.867 M2 8 165.4 -82.9 48.8 “ • 0.990 M 3 8 90.2 " 114.0 * -26.4 * 7.2 “ * 1.000 19 91 M 1 8 -174.1 248.2 •** 0.856 M 2 8 164.4 -90.4 48.4 •** 0.986 M 3 8 80.4 * 129.8 * -35.7 • C D O • • 0.999 1992 M 1 8 -178.2 241.9 “ 0.841 M 2 8 16 8 .1 -104.4 49.5 " 0.982 M 3 8 74.2 141.5 * -44.4 * 8.9 " 0.998 1993 M 1 8 -171.6 234.0 ** 0.843 M 2 8 162.2 -99.7 47.7 " 0.982 M 3 8 7 1 .1 138.8 * -43.4 * 8.7 " 0.999 1994 M 1 8 -164.3 222.9 ** 0.831 M 2 8 162.6 -104.0 46.7 •* 0.977 M 3 8 65.8 149.6 * -50.1 * 9.2 " 0.997 1995 M 1 8 -163.2 216.4 “ 0.827 M 2 8 159.5 -106.3 46.1 “ 0.977 M 3 8 65.0 141.3 * -48.5 * 9.0 ** 0.998 1996 M 1 8 -163.2 216.4 — 0.827 M2 8 159.5 -106.3 46.1 ** 0.977 M 3 8 65.0 141.3 * -48.5 * 9.0 ** 0.998 *p <0.05, "p <0.01, *"p <0.001 Age5:5-year in te rv a l of age started a t aged 50-54 205 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A.6 Parameter Estimates from Regression Analyses for Death Rates from Acute Myocardial Infarction (AMI) for Women Aged S O and Over, by Year, Vital Statistics 1982-1996 (per 100,000 population) Year Model N Intercept Aqe5 Aqe5-squared Aqe5-cubed R-square 1982 W 1 8 •196.9 211.6 ” • 0.864 W2 8 89.0 -74.3 40.8 "* 0.993 W3 8 36.3 63.6 -11.8 5.0 ” 0.999 1983 W 1 8 -198.1 209.0 •** 0.858 W2 8 90.7 -79.8 41.3 "• 0.991 W3 8 34.2 68.3 -15.3 5.4 “ 0.999 1984 W 1 8 -197.2 204.3 **• 0.855 W2 8 88 .1 -81.0 40.8 •“ 0.991 W3 a 32.4 65.1 * -15.0 5.3 ** 1.000 1985 W 1 8 -201.7 202.5 ** 0.842 W2 8 93.0 -92.2 42.1 •** 0.987 W3 8 27.5 79 .1 -23.3 6.2 ** 0.999 1986 W 1 8 -202.1 196.8 " 0.835 W2 8 91.2 -96.5 41.9 •** 0.986 W3 8 23.2 81.4 -26.0 6.5 ** 0.999 1987 W 1 8 -201.7 192.3 ” 0.824 W2 8 94.4 -103.8 42.3 •** 0.984 W3 8 22.7 84.0 -29.4 6.8 " 0.999 1988 W 1 8 -208.3 191.5 ** 0.809 W2 8 98.7 -115.5 43.9 *• 0.979 W3 8 16.4 100.0 -38.4 0 0 0.998 1989 W 1 8 -222.3 195.3 " 0.789 W2 8 106.9 -134.0 47.0 " 0.973 W3 8 11.8 115.3 -48.1 9 .1 ** 0.997 1990 W 1 8 -217.7 188.8 ” 0.783 W2 8 105.3 -134.2 46.1 ** 0.969 W3 8 8.7 118.9 -50.5 9.2 ” 0.996 1 9 9 1 W 1 8 -216.9 185.2 " 0.768 W2 8 111.0 -142.7 46.8 ** 0.965 W3 8 8.4 126.2 -55.8 9.8 ” 0.995 1992 W 1 8 -210.5 178.1 " 0.764 W2 8 108.4 -140.8 45.5 “ 0.963 W3 8 6.8 125.2 -56.0 • 9.7 “ 0.995 1993 W 1 8 -209.9 174.9 ” 0.754 W2 a 108.6 -143.6 45.5 ** 0.958 W3 8 2.6 134.2 -60.5 1 0 .1 •• 0.994 1994 W 1 8 -204.0 169.5 " 0.748 W2 8 108.5 -143.0 44.6 ** 0.956 W3 8 2.3 1 35 .1 -61.5 1 0 .1 •• 0.993 1995 W 1 8 -201.1 165.2 " 0.739 W2 8 107.7 -143.6 44.1 ** 0.950 W3 8 -1 .1 141.5 -64.7 10.4 * 0.991 1996 W 1 8 -195.7 160.4 ” 0.740 W2 8 105.2 -140.5 43.0 *• 0.953 W3 8 1.0 132.4 -61.2 9.9 •* 0.993 *p <0.05, **p<0.01, —p <0.001 A ge5: S - y e a r in te rv a l of age started at aged 50-54 206 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. A. 7 Odds Ratios from Logistic Regression Analyses for Overall Heart Disease Prevalence Rates, by Sex, NHIS 1982-1996 N Age Time Age-squared Age-cubed Age-logarithm df -2 Log L Model M1 29,717 1.047 1.003 2 2,155,803 0.0001 Men Model M2 Model M3 29,717 1.227 — 1.004 — 0.999 •** 3 2,272,044 0.0001 29,717 1.700 1.004 0.994 1.000 4 2,335,315 0.0001 Model M4 29,717 1.003 22.363 2 2,241,959 0 0001 Model M1 37,244 1.047 0 995 2 2,645,787 0.0001 Women Model M2 Model M3 37,244 1.133 — 0.995 — 0.999 *•* 3 2,724,177 0.0001 37,244 0.891 0.995 1.003 1.000 4 2,706,882 0.0001 Model M4 37,244 0.995 22 958 •" 2 2,687,177 0.0001 *p<0.05, “ pcO.01, ***p<0.001 Time: Years since 1982 ro o ~ -j A.8 Estimated and Actual Prevalence Rates of Overall Heart Disease at Certain Ages in Selected Years, by Sex, NHIS (Tables: Part I) Aqe Year -1 9 8 w/ aqe Men 2 w/ aqeA 2 w/ aqeA 3 Actual w/ aae Women w/ aqeA 2 w/ aqeA 3 Actual 50 10.1% 7.9% 7.4% 7.8% 8.4% 7.2% 7.7% 8.3% 55 12.4% 11.5% 11.5% 9.8% 10.3% 9.8% 9.8% 11.6% 60 15.1% 15.6% 15.9% 16.9% 12.6% 12.7% 12.4% 12.2% 65 18.2% 19.8% 20.1% 17.8% 15.3% 16.0% 15.7% 15.4% 70 21.9% 23.8% 23.7% 24.5% 18.5% 19.5% 19.4% 18.6% 75 26.0% 27.1% 26.6% 27.5% 22.2% 23.0% 23.3% 24.5% 80 30.6% 29.5% 28.8% 29.3% 26.4% 26.4% 26.9% 29.1% 85 35.7% 30.8% 30.8% 30.0% 31.1% 29.5% 29.8% 25.5% 90 41.1% 30.8% 32.6% 35.8% 36.1% 32.2% 31.4% 27.6% Year »1990 Aqe w/ aqe w/ aqeA 2 w/aaeA 3 Actual w/ aqe w/ aqeA 2 w/aqeA 3 Actual 50 10.3% 8.1% 7.6% 7.8% 8.0% 7.0% 7.4% 8.3% 55 12.6% 11.8% 11.8% 9.8% 9.9% 9.4% 9.4% 11.6% 60 15.4% 15.9% 16.3% 16.9% 12.1% 12.3% 12.0% 12.2% 65 18.6% 20.3% 20.6% 17.8% 14.8% 15.4% 15.2% 15.4% 70 22.3% 24.3% 24.2% 24.5% 17.9% 18.8% 18.8% 18.6% 75 26.5% 27.7% 27.2% 27.5% 21.5% 22.3% 22.6% 24.5% 80 31.2% 30.1% 29.5% 29.3% 25.6% 25.6% 26.1% 29.1% 85 36.3% 31.4% 31.4% 30.0% 30.2% 28.7% 28.9% 25.5% 90 41.7% 31.4% 33.3% 35.8% 35.2% 31.3% 30.5% 27.6% Year»1996 Aqe w/ aqe w/aqeA 2 w/aqeA 3 Actual w/ aqe w/ aaeA 2 w/ aqeA 3 Actual 50 10.5% 8.3% 7.8% 7.8% 7.8% 6.8% 7.2% 8.3% 55 12.9% 12.0% 12.0% 9.8% 9.6% 9.1% 9.1% 11.6% 60 15.6% 16.2% 16.6% 16.9% 11.8% 11.9% 11.7% 12.2% 65 18.9% 20.6% 20.9% 17.8% 14.4% 15.0% 14.8% 15.4% 70 22.6% 24.7% 24.6% 24.5% 17.4% 18.4% 18.3% 18.6% 75 26.9% 28.1% 27.6% 27.5% 21.0% 21.8% 22.0% 24.5% 80 31.6% 30.6% 29.9% 29.3% 25.0% 25.0% 25.5% 29.1% 85 36.7% 31.8% 31.9% 30.0% 29.5% 28.1% 28.3% 25.5% 90 42.2% 31.8% 33.8% 35.8% 34.4% 30.7% 29.9% 27.6% w/ age: Linear equation model w/ ageA 2: Quadratic equation model w/ ageA 3: Cubic equation model Actual: Average of the prevalence from 1982 to 1996 (closest) 208 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A.9 Estimated and Actual Prevalence Rates of Overall Heart Disease at Certain Ages in Selected Years, by Sex, NHIS (Tables: Part II) Aae Year » 1982 Aae-loq Men w/ aqeA 3 Actual Aae-loa Women w/ ageA 3 Actual 50 9.2% 7.4% 7.8% 7.6% 7.7% 8.3% 55 12.0% 11.5% 9.8% 10.0% 9.8% 11.6% 60 15.2% 15.9% 16.9% 12.7% 12.4% 12.2% 65 18.7% 20.1% 17.8% 15.7% 15.7% 15.4% 70 22.4% 23.7% 24.5% 19.1% 19.4% 18.6% 75 26.4% 26.6% 27.5% 22.6% 23.3% 24.5% 80 30.5% 28.8% 29.3% 26.4% 26.9% 29.1% 85 34.6% 30.8% 30.0% 30.2% 29.8% 25.5% 90 38.7% 32.6% 35.8% 34.1% 31.4% 27.6% Year * 1990 Aoe Aae-loa w/ aqeA 3 Actual Aae-loa w/aqeA 3 Actual 50 9.5% 7.6% 7.8% 7.3% 7.4% 8.3% 55 12.3% 11.8% 9.8% 9.6% 9.4% 11.6% 60 15.5% 16.3% 16.9% 12.2% 12.0% 12.2% 65 19.1% 20.6% 17.8% 15.2% 15.2% 15.4% 70 22.9% 24.2% 24.5% 18.4% 18.8% 18.6% 75 26.9% 27.2% 27.5% 21.9% 22.6% 24.5% 80 31.0% 29.5% 29.3% 25.6% 26.1% 29.1% 85 35.2% 31.4% 30.0% 29.3% 28.9% 25.5% 90 39.3% 33.3% 35.8% 33.2% 30.5% 27.6% Year * 1996 Aae Aae-loa w/ aqeA 3 Actual Aae-loa w/ aaeA 3 Actual 50 9.6% 7.8% 7.8% 7.1% 7.2% 8.3% 55 12.5% 12.0% 9.8% 9.3% 9.1% 11.6% 60 15.8% 16.6% 16.9% 11.9% 11.7% 12.2% 65 19.4% 20.9% 17.8% 14.8% 14.8% 15.4% 70 23.3% 24.6% 24.5% 18.0% 18.3% 18.6% 75 27.3% 27.6% 27.5% 21.4% 22.0% 24.5% 80 31.4% 29.9% 29.3% 25.0% 25.5% 29.1% 85 35.6% 31.9% 30.0% 28.7% 28.3% 25.5% 90 39.8% 33.8% 35.8% 32.5% 29.9% 27.6% log: age logarithm w/ ageA 3: Cubic equation model Actual: Average of the prevalence from 1982 to 1996 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A. 10 Estimated and Actual Prevalence Rates of Overall Heart Disease at Certain Ages in Selected Years, by Sex, NHIS (Figures: Part i) Year =1982 w / a g e : L in e a r e q u atio n m o d el w / ageA 2 : Q u adratic eq u atio n m o d e l w/ ageA 3 : C ubic e q u a tio n m o d e l A c tu a l: A verage of the prevalen ce fro m 1982 to 1996 (clo sest) M en 40% 30% 2 2 0 % 10% & — w/ageA 2 ■ # — w / ageA 3 — • - A ctu al 0% 50 55 60 65 70 75 80 85 90 Age W omen 40% 30% u 20% 10% 0 % 50 55 60 65 70 75 80 85 90 Age Years 1990 c V u h . 0 1 a. M en 40% 30% 20% 10% 0 % 50 55 60 65 70 75 80 85 90 Age 0% 4- W om en u 2 0 % & w/ ageA 2 * — w/ageA 3 - - - A ctu al 50 55 60 65 70 75 80 85 90 Age Year =1996 M en W om en 0 1 a . 40% 30% 20% 1 0 % — A— w / ageA 2 ■ H K — w/ageA 3 - — - - A c tu al 0 % 50 55 60 65 70 75 80 85 90 40% 30% 10% ■*— w/ageA 3 — - - A ctu al 0 % 50 55 60 65 70 75 80 85 90 Age Age 210 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A.11 Estimated and Actual Prevalence Rates of Overall Heart Disease at Certain Ages in Selected Years, by Sex, NHIS (Figures: Part II) lo g : age lo g a rith m w / ag e A 3 : C u bic e q u a tio n m o d e l A c tu a l: A verage o f the prevalen ce fro m 1982 to 1996 Year=1982 M en W omen 3 20% • • A g e -lo g w/ageA 3 A c tu a l 50 55 60 65 70 75 80 85 90 Age 40% 30% • 0 s a 20% 0 1 a 10% 0 % 'J L — s fc - — X- -Age-log - w / ageA 3 j . - A ctu al 50 55 60 65 70 75 80 85 90 Age Year=1990 M en Women e s * 4 1 a . 0 % -•£ A g e -lo g — X— w / ageA 3 - — - - A c tu a l 50 55 60 65 70 75 80 85 90 Age 40% 30% 1 8 20% 4 1 & 1 0 % t^par i o % i— — ■ & A g e-lo g • * — w / ageA 3 j • • — • - A c tu a l 50 55 60 65 70 75 80 85 90 Age Year = 1996 M en Women 40% 30% Si 20% -Age-log ! 10% - - A ctu al 0 % 50 55 60 65 70 75 80 85 90 u 20% A - A g e-lo g X— w / ageA 3 -A ctual Age 50 55 60 65 70 75 80 85 90 Age Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. A. 12 Odds Ratios from Logistic Regression Analyses for Ischemic Heart Disease (IHD) Prevalence Rates, by Sex, NHIS 1982-1996 N Age Time Age-squared Age-cubed Age-logarithm df -2 Log L Model M1 29,717 1.036 ■ * 0.999 *' 2 878,597 0.0001 Men Model M2 Model M3 29,717 1.312 1.000 0.998 3 1,053,355 0.0001 29,717 1.522 1.000 0.996 1 000 4 1,124,651 0.0001 Model M4 29,717 0.999 11.578 •“ 2 949,049 0 0001 Model M1 37,244 1.049 0.995 2 1,384,926 0.0001 Women Model M2 Model M3 37,244 1.262 '*• 0.996 ••• 0.999 “ • 3 1,460,523 0.0001 37,244 0.809 0.995 1.005 1.000 4 1,529,319 Model M4 0.0001 37,244 0.995 28.476 1,444,575 0.0001 *p<0.05, **p<0.01, ***p <0.001 Time: Years since 1982 A.13 Estimated and Actual Prevalence Rates of Ischemic Heart Disease (IHD) at Certain Ages in Selected Years, by Sex, NHIS (Tables: Part I) Aqe Year = 198 w/ aqe Men 2 w/ aqeA 2 w/aqeA 3 Actual w/ aqe Women w/ ageA 2 w/ aqeA 3 Actual 50 7.1% 4.9% 4.8% 4.7% 3.1% 2.1% 2.4% 2.5% 55 8.4% 7.5% 7.5% 6.1% 3.9% 3.3% 3.4% 3.0% 60 9.8% 10.3% 10.4% 10.5% 4.9% 4.9% 4.8% 4.8% 65 11.5% 13.1% 13.2% 11.3% 6.2% 6.8% 6.6% 6.4% 70 13.4% 15.4% 15.4% 15.7% 7.7% 8.8% 8.7% 8.4% 75 15.6% 16.7% 16.5% 15.8% 9.6% 10.6% 10.9% 10.9% 80 18.1% 16.9% 16.7% 17.5% 11.8% 12.0% 12.5% 11.1% 85 20.9% 15.8% 15.9% 16.4% 14.5% 12.9% 13.1% 12.1% 90 23.9% 13.8% 14.3% 18.9% 17.7% 13.0% 12.3% 11.4% Year ■ 1990 Aae w/ aqe w/ aqeA 2 w/ aqeA 3 Actual w/ aqe w/ aqeA 2 w/ aqeA 3 Actual 50 7.1% 4.9% 4.7% 4.7% 3.0% 2.0% 2.3% 2.5% 55 8.3% 7.4% 7.4% 6.1% 3.8% 3.2% 3.2% 3.0% 60 9.8% 10.3% 10.4% 10.5% 4.8% 4.8% 4.6% 4.8% 65 11.4% 13.1% 13.2% 113% 6.0% 6.6% 6.4% 6.4% 70 13.3% 15.3% 15.3% 15.7% 7.4% 8.5% 8.4% 8.4% 75 15.5% 16.7% 16.5% 15.8% 9.2% 10.3% 10.5% 10.9% 80 18.0% 16.8% 16.6% 17.5% 11.4% 11.7% 12.1% 11.1% 85 20.7% 15.8% 15.8% 16.4% 14.1% 12.5% 12.7% 12.1% 90 23.8% 13.7% 14.3% 18.9% 17.2% 12.6% 11.9% 11.4% Year »1996 Aqe w/ aqe w/ aqeA 2 w/ aqeA 3 Actual w/ aqe w/ aqeA 2 w/ aqeA 3 Actual 50 7.0% 4.9% 4.7% 4.7% 2.9% 2.0% 2.2% 2.5% 55 8.3% 7.4% 7.4% 6.1% 3.7% 3.1% 3.2% 3.0% 60 9.7% 10.3% 10.4% 10.5% 4.6% 4.6% 4.5% 4.8% 65 11.4% 13.1% 13.2% 11.3% 5.8% 6.4% 6.2% 6.4% 70 13.3% 15.3% 15.3% 15.7% 7.2% 8.3% 8.2% 8.4% 75 15.4% 16.6% 16.5% 15.8% 9.0% 10.0% 10.2% 10.9% 80 17.9% 16.8% 16.6% 17.5% 11.2% 11.4% 11.8% 11.1% 85 20.6% 15.7% 15.8% 16.4% 13.7% 12.2% 12.4% 12.1% 90 23.7% 13.7% 14.3% 18.9% 16.8% 12.4% 11.6% 11.4% w/ age: Linear equation model w/ ageA 2: Quadratic equation model w/ ageA 3: Cubic equation model Actual: Average of the prevalence from 1982 to 1996 (closest) 213 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A. 14 Estimated and Actual Prevalence Rates of Ischemic Heart Disease (IHD) at Certain Ages in Selected Years, by Sex, NHIS (Tables: Part II) Aqe Year = * 1982 Aqe-loq Men w/ aqeA 3 Actual Aqe-loq Women w/aqeA 3 Actual 50 6.6% 4.8% 4.7% 2.7% 2.4% 2.5% 55 8.1% 7.5% 6.1% 3.7% 3.4% 3.0% 60 9.9% 10.4% 10.5% 4.9% 4.8% 4.8% 65 11.8% 13.2% 11.3% 6.3% 6.6% 6.4% 70 13.8% 15.4% 15.7% 8.0% 8.7% 8.4% 75 15.9% 16.5% 15.8% 9.9% 10.9% 10.9% 80 18.2% 16.7% 17.5% 11.9% 12.5% 11.1% 85 20.5% 15.9% 16.4% 14.3% 13.1% 12.1% 90 22.8% 14.3% 18.9% 16.8% 12.3% 11.4% Y ear* 199C Age Aqe-loq w/ aqeA 3 Actual Aqe-loq w/ aqeA 3 Actual 50 6.5% 4.7% 4.7% 2.6% 2.3% 2.5% 55 8.1% 7.4% 6.1% 3.6% 3.2% 3.0% 60 9.8% 10.4% 10.5% 4.8% 4.6% 4.8% 65 11.7% 13.2% 11.3% 6.1% 6.4% 6.4% 70 13.7% 15.3% 15.7% 7.7% 8.4% 8.4% 75 15.8% 16.5% 15.8% 9.5% 10.5% 10.9% 80 18.0% 16.6% 17.5% 11.6% 12.1% 11.1% 85 20.3% 15.8% 16.4% 13.8% 12.7% 12.1% 90 22.7% 14.3% 18.9% 16.2% 11.9% 11.4% Year * 1996 Age Aqe-loq w/ aqeA 3 Actual Aqe-loq w/aqeA 3 Actual 50 6.5% 4.7% 4.7% 2.6% 2.2% 2.5% 55 8.0% 7.4% 6.1% 3.5% 3.2% 3.0% 60 9.8% 10.4% 10.5% 4.6% 4.5% 4.8% 65 11.6% 13.2% 11.3% 6.0% 6.2% 6.4% 70 13.6% 15.3% 15.7% 7.5% 8.2% 8.4% 75 15.7% 16.5% 15.8% 9.3% 10.2% 10.9% 80 17.9% 16.6% 17.5% 11.3% 11.8% 11.1% 85 20.2% 15.8% 16.4% 13.5% 12.4% 12.1% 90 22.6% 14.3% 18.9% 15.9% 11.6% 11.4% log: age logarithm w/ ageA 3: Cubic equation model Actual: Average of the prevalence from 1982 to 1996 214 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A.15 Estimated and Actual Prevalence Rates of Ischemic Heart Disease at Certain Ages in Selected Years, by Sex, NHIS (Figures: Part I) w/ a g e : L in e a r eq u atio n m o d e l w / ag eA 2 : Q u adratic equ atio n m o d el w / ag eA 3 : C ubic eq u atio n m o d el A c tu a l: A verage of the p revalence fro m 1982 to 1996 (clo sest) Year=1982 W omen s 6 a M en 20% 15% 10% 5% 0% 50 55 60 65 70 75 80 85 90 Age 20% 15% 10% 5% 50 55 60 65 70 75 80 85 90 Age Years 1990 U 10% | — A— w/ ageA 2 — * — w/ageA 3 ! • • — • -A ctual 50 55 60 65 70 75 80 85 90 Age 20% 15% 10% 5% 0% — -A -— w / ageA 2 - - — - -A ctual 50 55 60 65 70 75 80 85 90 Age Year = 1996 Men 20% 1 5 % 5% •A — w/ ageA 2 ■*— w/ ageA 3 — - -Actual 0% 50 55 60 65 70 75 80 85 90 Women 20% 15% « 10% 5% — A— w/ ageA 2 — * — w/ageA 3 • - — - -Actual 0% 50 55 60 65 70 75 80 85 90 Age Age 215 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A. 16 Estimated and Actual Prevalence Rates of Ischemic Heart Disease at Certain Ages in Selected Years, by Sex, NHIS (Figures: Part II) lo g : age lo g a rith m w / ageA 3 : C ubic e q u atio n m o d e l A c tu a l: A verag e of the prevalen ce fro m 1982 to 1996 Year =1982 Men 20% 15% 8 10% ■ & — A g e-lo g j -X— w/ageA 3 — • - A c tu al 5% 0% 50 55 60 65 70 75 80 85 90 Age Women -A - A ge-log — * — w/ageA 3 • — - - A ctu al 50 55 60 65 70 75 80 85 90 Age Year = 1990 15% u 10% * A g e-lo g * — w/ageA 3 ~ - - A ctu al '-r i i~ r - v 50 55 60 65 70 75 80 85 90 Age Women 20% 15% o 10% • s f c A g e-lo g -*— w/ageA 3 — - - A ctu al 5% 50 55 60 65 70 75 80 85 90 Age Year =1996 20% 15% §10% 0 1 & 5% 0% M en y - ’ jip-- ! ..> i ’ — &— A g e-lo g — * — w / ageA 3 • • — • -A ctual 50 55 60 65 70 75 80 85 90 Age Women 20% 15% S10% — &— A ge-log — — w/ ageA 3 - - — - - A ctu al 5% 0% 50 55 60 65 70 75 80 85 90 Age Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX B Appendix B presents the summaries of equations of hazard analyses for heart disease multistate models with age terms for men and for women, underlying tables and figures in Chapter V. In addition to the hazard models using age as an independent variable, an age-squared term was added to these six heart disease transition models. Then, an age-cubed term was also included in each model. Among these models, only the transition from “No heart diseases" to "Heart attack" has a statistically significant age- squared term for men. No other models have statistical significance for either the age-squared or age-cubed terms for men. No statistical significance is seen for the models with the age-squared or the age-cubed terms for women. 217 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. B.1 Summary of Equations of Hazard Analyses for Heart Disease Multi-State Models with Age Terms for Men, HRS/AHEAD (A) 'No Heart Disase* to ‘Heart Disease* (B) ’No Heart Disease* to 'Heart Attack* (C) ’Heart Disease' to ’Heart Attack' N 6,071 6,071 6,071 6,071 1,115 1,115 Number Moved 211 211 107 107 56 56 Intercept 12.84056 *** 23.00349 16.03326 ** -3.20882 10.24980 40.96726 Age -0.21684 * -0.64778 -0.32130 * 0.52003 -0.18460 -1.48812 Age-squared 0.00126 0.00752 0.00227 * -0 01037 0.00129 0.02026 Age-cubed -0.00003 0.00006 -0.00009 Log-likelihood -1,009.02 -1,008.86 -590.67 -590.46 -253.48 -253.25 (D) 'No Heart Disease' to 'Death' (E) 'Heart Disease' to 'Death' (F) 'Heart Attack' to 'Death' N 6,071 6,071 1,115 1,115 739 739 Number Moved 291 291 142 142 77 77 Intercept 12.01238 *** 16.53550 3.52219 -44.60699 -0.68407 -47.47239 Age -0.14586 * -0.33110 0.06700 2.05834 0.16258 2.10488 Age-squared 0.00037 0.00296 -0.00106 -0.02920 -0.00160 -0.02909 Age-cubed -0.00001 0.00013 0.00013 Log-likelihood -1.132.90 -1.132.85 -452.19 -450.95 -263.35 -261.76 *p <0.05, “ p<0.01, " " p <0.001 'Heart Disease' does not include heart attack or myocardial infarction. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. B.2 Summary of Equations of Hazard Analyses for Heart Disease Multi-State Models with Age Terms for Women, HRS/AHEAD (A) 'No Heart Disase* to ’Heart Disease* (B) 'No Heart Disease’ to 'Heart Attack' (C) Heart Disease' to 'Heart Attach' N 8,087 8,087 8,087 8,087 1,473 1,473 Number Moved 293 293 92 92 54 54 Intercept 6.84153 ** -17.04699 12.79910 ** -3.86468 11.20658 -3.98087 Age -0.04573 0.96081 -0.16737 0.53052 -0.18214 0.44752 Age-squared 0.00005 -0.01442 0.00082 -0.00914 0.00114 -0.00777 Age-cubed 0.00007 0.00005 0.00004 Log-likelihood -1,419.93 -1,418.37 -509.59 -509.40 -264.77 -264.69 (D) 'No Heart Disease' to 'Death' (E) 'Heart Disease' to 'Death' (F) 'Heart Attack' to 'Death* N 8,087 8,087 1,473 1,473 454 454 Number Moved 310 310 147 147 61 61 Intercept 9.77507 *** -7.24143 10.81729 ** 15.72620 10.27451 21.26482 Age -0.07661 0.61918 -0.13479 -0.33108 -0.15425 -0.59822 Age-squared -0.00009 -0 00977 0 00042 0.00310 0.00069 0.00681 Age-cubed 0.00004 -0.00001 -0.00003 Log-likelihood -1,36510 -1,364.34 -54312 -543.10 -197.13 -197 09 *p <0.05, “ p <0.01, ***p <0.001 'Heart Disease' does not include heart attack or myocardial infarction.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Filial expectations and social exchange patterns among older Taiwanese parents and their adult children

PDF

Impact of housing arrangements on social support and health status among Chinese American elderly

PDF

Socioeconomic status and the incidence of heart disease among the elderly

PDF

Effects of cultural values and behaviors in sociocultural stress and coping among Korean-American and Korean caregivers

PDF

Factors affecting child survival in Pakistan

PDF

Intergenerational social support and the psychological well-being of older parents in China

PDF

Cross -cultural examination of mental health measures in dementia caregivers: Assessment of the Center for Epidemiologic Studies -Depression Scale and the Zarit Burden Inventory

PDF

Determinants of life satisfaction among the high-functioning elderly: Implications of effects of psychological factors and activity participation

PDF

Dementia caregiving and ethnicity: African American caregivers and the sociocultural stress and coping model

PDF

Gender and the bump: An investigation of the reminiscence effect in the Long Beach Longitudinal Study

PDF

An analysis of nonresponse in a sample of Americans 70 years of age and older in the longitudinal study on aging 1984-1990

PDF

Alcoholism among the elderly: Fact or fallacy

PDF

Evaluation of the Care Advocate Program: Bridging managed care and home community -based services

PDF

Education and intelligence test scores: Predictors of dementia?

PDF

A methodology to identify high -risk patients with diabetes in the California Medicaid populations (Medi -Cal)

PDF

Assessment of prognostic comorbidity in hospital outcomes research: Is there a role for outpatient pharmacy data?

PDF

Controlling for biases from measurement errors in health outcomes research: A structural equation modeling approach

PDF

Effects of a formulary expansion of the use of SSRIs and health care services by depressed patients in the California Medicaid program

PDF

HIV/AIDS knowledge among Hispanics living in Nogales, Arizona

PDF

HIV/AIDS preventive behavior in Botswana: Trends and determinants at the turn of the 21st century

Asset Metadata

Creator Ueda, Hiroshi (author)

Core Title Heart disease among middle -aged and elderly persons in the United States: Trends and a multistate model

School Graduate School

Degree Doctor of Philosophy

Degree Program Gerontology and Public Policy

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Gerontology,health sciences, public health,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Crimmins, Eileen M. (committee chair), Heer, David (committee member), Silverstein, Merril (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-88555

Unique identifier UC11337891

Identifier 3018138.pdf (filename),usctheses-c16-88555 (legacy record id)

Legacy Identifier 3018138.pdf

Dmrecord 88555

Document Type Dissertation

Rights Ueda, Hiroshi

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA