Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Cognitive health and self-rated memory in later life: linkages to race/ethnicity, multimorbidity, and survival
(USC Thesis Other)
Cognitive health and self-rated memory in later life: linkages to race/ethnicity, multimorbidity, and survival
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
COGNITIVE HEALTH AND SELF-RATED
MEMORY IN LATER LIFE:
LINKAGES TO RACE/ETHNICITY,
MULTIMORBIDITY, AND SURVIVAL
by
Yu Jin Jeong
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
GERONTOLOGY
May 2022
ii
Dedication
This dissertation is dedicated to my grandparents, who always love and support me.
Although my grandpa is peacefully asleep now, I know he is proud of me.
To my grandma, you are my forever rockstar.
iii
Acknowledgements
I would like to express my deepest appreciation to my dissertation committee members,
Jessica Ho, Joseph Saenz, and Yuri Jang. First, I’d like to express my gratitude to my advisor,
Jessica Ho, for guiding me on every step during the Ph.D. degree. Without your encouraging
words and advice along the journey, I may not be the person who I am now. Thank you for
always being patient with me, especially when I took more than a semester to calculate age-
standardized death rate. I will never forget the feeling when I calculated the numbers correctly
for the first time after multiple attempts. You helped me learn how to be patient and never give
up throughout ups and downs. Words cannot express how amazing an advisor you are in this
acknowledgement section. Whenever I encounter hardship or challenge in the future, I will think
about how you have taught me. Also, if I meet someone who needs help, I will give them a hand
like how you helped me. The traits I learned from you prepared me for the world after my Ph.D.
degree. You made me a better person not only as a scholar, but also as a human being.
In addition, I would like to extend my gratitude to my dissertation committee members,
Joseph Saenz and Yuri Jang, for guidance throughout my Ph.D. journey. Joseph, I remember
seeing you at the Gerontological Society of America Annual Scientific meeting in Austin (2019)
attending an immigrant health seminar. Although I saw you at the GERO building sometimes, I
didn’t know you well enough to walk up to you and say hi. Now I think back to that time and
regret not saying hi to you. Who would have guessed that you are in my dissertation committee?
I was so happy to hear that you became a faculty member in our department and that I can have
you as a part of my dissertation committee. It has been my pleasure to work with you and thank
you for your insights. Yuri, I am constantly inspired by your leadership. Sometimes, one meeting
with you will end up with multiple research plans/ideas. It was amazing to witness how you plan
iv
your future research ahead of each project. I may not plan be able to plan so much future
research like you, but whenever I encounter beautiful Korean poems about older adults, I may
still email you all of sudden. Thank you for continuing to guide me throughout my research. I’d
also like to thank Elizabeth Zelinski. I remember visiting your office and saying that I might not
be the right person to be here in my first month of the Ph.D. program. I cried my eyes out and
you assured me everyone says that when they start the program and they’ve all made it in the
end. I trusted in your words and you were right—I made it! Thank you so much for encouraging
me and trusting in my ability when I was lost.
Thank you to Dr. Chae-Hee Park, my undergraduate advisor, who helped me build the
foundation of my interest in the aging population and who continues to be a mentor. Also,
special thanks to Dr. Misook Lee and Taewha Jeong, my parents, who supported me during my
academic journey with sympathy, even though we are on the other side of the globe and have a
huge time difference. Look, Mom and Dad, now I am a doctor too!
Most importantly, I would like to extend my sincere thanks to my Anthony, my husband.
I cannot imagine my Ph.D. journey without you and you deserve half of my Ph.D. degree. Thank
you for always proofreading (all) my papers and countless emails and for helping me with
coding too. You are the best proofreader and programmer I could have asked for. My chelito
lindo, I love you. Also, I’d like to thank my families in Korea and the U.S. Thanks to my family
in Korea. Whenever I was not so confident in myself, you cheered for me and gave me
confidence through video calls/phone calls. It is incredible how we were able to maintain a close
family relationship, even though we are so far apart. I know I missed multiple major family
events while I was in the U.S., but Anthony and I will make it up to you once we move back to
Korea. Finally, thank you to my family in San Francisco. We were able to have the most fun
v
time whenever we visited San Francisco. Travelling and dancing with family was my true
pleasure during my degree. It is still an unforgettable moment when we encountered a herd of
bison in Yellowstone National Park. Also, I enjoyed dancing with El Salvador music, Korean
music, Mexican music, and Persian music, which helped me to regain energy to work on my
degree.
vi
Table of Contents
DEDICATION…………………………………….………………………….……………….…..ii
ACKNOWLEDGEMENT……………………….………………………….………………..…. iii
LIST OF TABLES……………………….………………………….……………………….......vii
ABSTRACT…………….………………………….……………………………………….......viii
CHAPTER 1: BACKGROUND…………………….……………………………………………1
CHAPTER 2: DIFFERENCES IN SELF-RATED MEMORY BY RACE/ETHNICITY…….....5
CHAPTER 3: THE RELATIONSHIP BETWEEN MULTIMORBIDITY AND
TYPES OF CHRONIC DISEASE AND SELF-RATED MEMORY…………………………...26
CHAPTER 4: GENDER DIFFERENCES IN SURVIVAL WITH
COGNITIVE IMPAIRMENT AND DEMENTIA………………………………………………48
CHAPTER 5: SUMMARY & DISCUSSION……………………………………………….…..69
REFERENCES…………………………………….………………………….…………………77
APPENDICES…………………………………….………………………….………………….99
vii
List of Tables
TABLE 1-1. DESCRIPTIVE STATISTICS BY RACE/ETHNICITY………………………… 20
TABLE 1-2. ODDS RATIOS FROM LOGISTIC REGRESSION MODELS FOR THE
ASSOCIATION BETWEEN RACE/ETHNICITY AND FAIR/POOR SELF-RATED
MEMORY……………………………………………………………………………………….22
TABLE 1-3. ODDS RATIOS FROM LOGISTIC REGRESSION MODELS FOR THE
ASSOCIATION BETWEEN RACE/ETHNICITY AND FAIR/POOR SELF-RATED
MEMORY, STRATIFYING BY LEP AMONG HISPANICS……………………………….....24
TABLE 2-1. DESCRIPTIVE STATISTICS BY MULTIMORBIDITY………………………...43
TABLE 2-2. ODDS RATIOS FROM LOGISTICS REGRESSION FOR
THE ASSICATION BETWEEN MULTIMORBIDITY AND
TYPES OF CHRONIC DISEASE AND FAIR/POOR SELF-RATED MEMORY…….………46
TABLE 3-1. THE PROBABILITY OF DYING WITHIN 10 YEARS OF BASELINE
BY AGE GROUP AND GENDER………………………………………………….…………..62
TABLE 3-2. CHARACTERISTICS OF THE SAMPLE, HRS 2000…………………………...63
TABLE 3-3. COX PROPORTIONAL HAZARD MODELS BY GENDER…………………...65
TABLE 3-4. COX PROPORTIONAL HAZARD MODELS AMONG RESPONDENTS
WITH CIND AND DEMENTIA………………………………………………………………..67
viii
Abstract
As older adults live longer and the number of older adults is increasing, it is critical to
maintain healthy cognition to live healthy lives. Cognitive health has been an important field of
research in gerontology and the study of aging. Researchers have shown the importance of
cognitive health at older ages, documenting that cognitive health declines with age and is closely
associated with older adults’ quality of life and mortality. Needless to say, maintaining healthy
cognition is important not only for older individuals, but also for their caregivers and public
health planning. Self-rated memory has been highlighted in recent years due to its usefulness as
an indicator of general cognitive health and its ease of administration. Self-rated memory is
commonly utilized in clinical settings to assess cognitive health (Marino et al., 2009) because
perceived memory change can be an early indication of memory decline as well as dementia
onset. Older adults with negative self-rated memory tend to experience various types of negative
health outcomes (e.g., functional limitations, physical disabilities, and psychological distress,
etc.). This dissertation fills a gap in the existing literature by exploring how self-rated memory
varies by race/ethnicity, how multimorbidity is linked to self-rated memory, and how cognitive
impairment impacts the mortality of older adults by gender.
The aim of this dissertation is to further our understanding of the cognitive status of older
adults. The first study explores differences in self-rated memory across non-Hispanic white, non-
Hispanic black, and Hispanic older adults in the U.S., including how limited English proficiency
impacts self-rated memory by race/ethnicity (Chapter 2). This chapter has important implications
for individuals who are responsible for administering cognitive testing. Also, the findings from
this chapter may contribute to timely or early diagnosis of cognitive impairment among older
adults. The second chapter of this dissertation explores whether multimorbidity and specific
ix
types of chronic diseases are associated with poor self-rated memory (Chapter 3). This chapter
explores which of eight individual chronic diseases are linked to reporting poor self-rated
memory. Given that the prevalence of chronic disease and poor self-rated memory increase with
age, this study provides evidence for the link between multimorbidity and self-rated memory
among older adults at earlier stages of cognitive impairment. Finally, the dissertation investigates
how cognitive impairment contributes to the mortality of older adults by gender (Chapter 4), and
it aims to identify possible factors that may contribute to gender differences in mortality. This
dissertation consists of three studies which used secondary data from the National Health and
Aging Trends Study (NHATS) and the Health and Retirement Study (HRS), which are nationally
representative samples of older adults in the United States.
Chapter 1: Background
Our population is becoming grayer as baby boomers become older adults and we
continue to extend our lives. It is projected that the share of the individuals aged 65 and older
will approximately double from 12% in 2015 to 22% in 2050 around the world (World Health
Organization, 2021). Populations are aging faster compared to in the past. Life expectancy is
considered to be an important summary measure of population health. Life expectancy at birth
has considerably increased from 47.3 years in 1900 to 78.7 years in 2018 in the United States
(Arias & Xu, 2020a; Crimmins & Zhang, 2019). Several factors have contributed to these
monumental increases in life expectancy, including reductions in infectious disease mortality at
younger ages and in chronic disease mortality at older ages. Over time, the number of older
adults is projected to double between 2018 and 2060 (Mather et al., 2019). In light of the
increased number of older adults and increases in life expectancy, it is important to maintain
healthy cognition at older ages since cognition is an important factor influencing healthy aging
(Morley et al., 2015).
The number of deaths caused by Alzheimer’s disease has increased and it is 6
th
leading
cause of death in the U.S. in 2019 (Alzheimer’s Association, 2021). It is estimated that 1 in 14
older adults aged 65+ will develop dementia, and that 1 in 6 adults aged 80 and over will develop
dementia. Although recent studies have found a reduction in the prevalence of dementia (Langa
et al., 2017), some projections indicate the absolute number of the older population with
dementia will gradually increase, while deaths caused by other diseases will decline (Prince et
al., 2013). The burden of poor cognitive health is increasing and it is highly associated with loss
of independent living, high costs of healthcare, and burdens on public health. As a result, it
important to know how individuals’ cognitive health will be in the future, and self-rated memory
2
is one measure that can be utilized to understand one’s cognitive health. Individuals’ self-rated
memory has shown to be closely linked to cognitive decline and to play a role as an early
indicator of cognitive decline among older individuals.
In line with this, there are important differences in cognitive status by race/ethnicity.
Non-Hispanic whites generally have better cognitive health compared to non-Hispanic blacks
and Hispanics (Brewster et al., 2014; Zsembik & Peek, 2001). As the proportion of minority
older adults is gradually increasing in the U.S., it is important to expand our knowledge of how
other race/ethnicities perceive their memory status and how cultural and social factors contribute
to these differences. It is well-documented that education level plays an important role in
cognitive health among older adults (Alley et al., 2007; Crimmins & Saito, 2001). Given that
minority older adults tend to have lower education levels and are less likely to have access to
health care, it is critical to investigate how other racial/ethnic groups report their self-rated
memory (Assari, 2018; Holahan & Kim, 2000) as this can contribute to a better understanding of
cognitive health among minority older adults.
Increased longevity and quality of life at later ages are associated with chronic diseases
among older adults. The prevalence and incidence of chronic diseases are also expected to
increase in the future. For individuals aged 65 and older, it is common to have multimorbidity
and it was reported that approximately 77% of older adults are currently diagnosed with at least
two? chronic disease (National Council on Aging, 2018). Having both chronic diseases and
negative self-rated memory worsen older adults’ quality of life, as chronic diseases require
constant care and management (Centers for Disease Control and Prevention, 2021a). Prior
studies have found that having chronic diseases increases the risk of developing cognitive
impairment (Kurella et al., 2005; Strachan et al., 2008). Therefore, accompanying multiple
3
chronic diseases may be risk factors for worsening cognitive status and even dementia (Vassilaki
et al., 2015)(Hill et al., 2016) Given that there is a high prevalence of chronic diseases and
negative self-rated memory among older adults, it is important to investigate the link between
these two health statuses. By understanding how multimorbidity and different types of chronic
diseases are influencing self-rated memory, health professionals will be able to provide
information to family members so that they can prepare for possible cognitive decline in the
future.
Lastly, although life expectancy has increased and people live longer than in the past,
persons with dementia or cognitive impairment live shorter lives compared to older adults
without cognitive impairment. As mentioned earlier, maintaining healthy cognition is important
since it is highly associated with lower levels of disability, independent living, and higher quality
of life. Given that women have a longer life expectancy than men, women are at a higher risk of
developing dementia. It is possible that mortality associated with cognitive impairment and
dementia varies by gender. Expanding our knowledge with regards to cognitive impairment and
dementia mortality would be helpful to older adults and family members since they would be
better able to plan ahead in terms of healthcare utilization, developing support systems, and
allocating financial resources. These possible benefits are not widely recognized and few studies
have investigated how mortality associated with cognitive impairment differs by gender.
Indicating the complexities of cognitive health among older adults, this chapter calls attention to
the need to address how mortality related to cognitive impairment and dementia differs for men
and women and to identify which characteristics may influence these associations.
Based on the aforementioned review, the aim of this dissertation is to examine several
dimensions of cognitive health among older adults in the United States and how cognitive health
4
is influenced by sociodemographic characteristics and psychosocial factors. This dissertation
investigate the following research questions:
1) Are there differences in self-rated memory by race/ethnicity?
2) Does multimorbidity increase the risk of reporting fair/poor self-rated memory?
3) Which chronic diseases are associated with increased risk of reporting fair/poor self-
rated memory?
4) How do cognitive impairment or dementia affect mortality for men and women?
5) Are there gender differences in survival with cognitive impairment without dementia
(CIND) or dementia?
The goals of the dissertation will be achieved by utilizing secondary data from the
National Health and Aging Trends Study (NHATS) and the Health and Retirement Study (HRS).
Chapter 2 uses the NHATS data collected in 2011 (Round 1) to answer how self-rated memory
status differ by race/ethnicity. Chapter 3 also uses the NHATS data from Round 1 to investigate
the association between multimorbidity and self-rated memory, as well as which types of chronic
diseases are closely linked to self-rated memory. In Chapter 4, we use data from the HRS 2000
to investigate differences in mortality associated with cognitive statuses by gender.
5
Chapter 2: Differences in Self-Rated Memory by Race/Ethnicity
Introduction
As the number of adults aged 65 and older grows in the U.S., researchers are increasingly
focusing their attention on the cognition of older adults. Cognitive functioning declines with age
and having healthy cognition is an important factor that contributes to healthy aging (Morley et
al., 2015). Declines in cognition may affect older adults’ abilities to live their lives independently
(Harada et al., 2013). Growing evidence suggests that cognitive health may differ by
race/ethnicity among older adults (Brewster et al., 2014; Zsembik & Peek, 2001). For example,
non-Hispanic whites showed better cognitive functioning compared to Hispanics and non-
Hispanic blacks in a study of older adults (Díaz-Venegas et al., 2016). Factors hypothesized to
contribute to these differences include socioeconomic status, health status, and sociocultural
factors. For example, lower education level has been found to be associated with worse cognition
(Alley et al., 2007), and racial/ethnic minorities tend to have lower education levels compared to
non-Hispanic whites (U.S. Census, 2014).
Individuals’ perceived memory status is frequently used in research and it is often called
self-rated memory (Segel-Karpas & Palgi, 2019). Self-rated memory is also commonly used to
examine cognitive health in clinical settings in addition to objective cognitive status (Marino et
al., 2009). The experience of perceived memory change could be an early sign of memory
decline and dementia onset. Studies found that subjective memory complaints are prevalent in
the population, with prevalence estimates ranging from 25% to 50%, and they tend to increase
with age (Jonker et al., 2000). Older adults with subjective memory complaints may be at higher
risk of developing cognitive impairment or Alzheimer’s disease in the future (Reisberg et al.,
2010). Older adults with subjective memory complaints but without objectively diagnosed
6
cognitive impairment showed high chances of developing dementia compared to older adults
without poor self-rated memory (Mitchell et al., 2014). In line with this, poor self-rated memory
was linked to reduced brain volume and increased amyloid deposition (Barnes et al., 2006;
Jessen et al., 2006). Hence, self-rated memory may be an important indicator of the general
cognitive health of older adults.
Similar to self-rated memory, self-rated health is a self-reported measure of how
individuals perceived their health status. Self-rated health demonstrates the full perspectives of
individuals’ health status and potentially provides even an early sign of disease that is not
diagnosed yet (Idler & Benyamini, 1997). Although self-rated health may provide individuals’
health status, health risk factors, and mortality, it is necessary to take social context into
consideration (Sen, 2002). For example, self-rated health is complex and may be influenced by
individuals’ physical health, socio-cultural, and psychological conditions (McMullen &
Luborsky, 2006; Sen, 2002). Self-rated health has been found to be a strong predictor of
mortality. For example, compared to older adults with higher self-rated health, older adults with
worse self-rated health had two times higher odds of mortality (DeSalvo et al., 2006).
In line with this, a sizable body of literature suggests that self-rated memory may not be
restricted to only cognition. Older adults with poor self-rated memory showed high levels of
functional limitations, disability, as well as psychological distress (Cordier et al., 2019; Ficker et
al., 2014; Hill et al., 2016). For example, older individuals with poor self-rated memory needed
more frequent assistance to perform daily activities compared to individuals without poor self-
rated memory (Cordier et al., 2019). Thus, understanding older adults’ perceived memory status
might provide broader understanding of their health and future trajectory of cognitive
functioning. Based on the existing literature, older adults’ self-rated memory demonstrates
7
critical information for cognitive health, disability, and psychological well-being (Cordier et al.,
2019; Maki et al., 2014; Montejo et al., 2011). Nonetheless, there has been a lack of research on
differences in self-rated memory by race/ethnicity.
Minority older adults, on average, have worse cognitive health and poor self-rated
memory than non-Hispanic whites (Díaz-Venegas et al., 2016; Fornazzari et al., 2009).
Nonetheless, they are less likely to be aware of the issue with poor self-rated memory, although
they complained about their memory status (Fornazzari et al., 2009). Moreover, minority older
adults tend to delay seeking care for cognitive problems and lack access to health care (Cooper et
al., 2010; Mukadam et al., 2013). Particularly, minority older adults with limited English
proficiency tend to underutilize healthcare (e.g., primary care, emergency visits) compared to
those without limited English proficiency (Graham et al., 2008). Contributing factors behind the
lack of access to healthcare among individuals with limited English proficiency includes
difficulty with understanding both doctors and written information at the doctor’s office, as well
as instructions on prescribed medicine (Kim et al., 2011). Therefore, their lack of knowledge
about self-rated memory could lead to delayed diagnosis of cognitive impairment, especially
among individuals with limited English proficiency.
To conclude, few studies have examined differences in poor self-rated memory by
race/ethnicity. Hence, this study uses a nationally representative sample to look at whether there
are differences in the likelihood of reporting poor self-rated memory by race/ethnicity. It then
examines whether these differences can be explained by a set of characteristics of socio-
demographic characteristics (e.g., age, gender, and marital status), health status (e.g., the number
of chronic conditions, depressive symptoms, objective memory status, and functional
limitations), and sociocultural factors (e.g., economic vulnerability, limited English proficiency,
8
and religious services). The research question is the following: “Are there differences in poor
self-rated memory by race/ethnicity and are they explained by socio-demographic characteristics,
health status, and social and cultural factors?” To examine the association of limited English
proficiency and self-rated memory among minority older adults, we divided Hispanics with
limited English proficiency and those without limited English proficiency on self-rated memory.
By understanding the association and identifying contributors to the differences, this project has
important implications for health care providers, especially for those who are responsible for
cognitive testing. In addition, it may also contribute to timely diagnosis of cognitive decline and
an early detection of cognitive impairment among older adults.
Methods
Data
The study used data from the 2011 National Health and Aging Trends Study (NHATS).
The NHATS is a nationally representative study of Medicare beneficiaries aged 65 and older
residing in the U.S. Data were collected through an annual interview and respondents completed
a written informed consent before the interview. The NHATS was established to provide
comprehensive information about older adults’ overall socio-demographic status, health status,
and cognitive functioning. The NHATS included two-hour in-person interviews with self-
respondents or proxy respondents when older adults were unable to complete the survey. A total
of 8,245 respondents (71% response rate) were interviewed in 2011. In this study, respondents
residing in nursing homes were excluded because these respondents were not asked questions
regarding self-rated memory (N = 468). Also, we eliminated proxy interview (N = 755) as the
self-rated memory status question was not asked of proxy respondents. Individuals from other
racial groups, including American Indians, Asians, Native Hawaiians, Pacific Islanders, and
other race, were not included in the analysis due to small sample size (N = 255) and diverse
9
populations. Thus, after excluding respondents who were missing information on variables of
interest, the final sample consisted of 6,583 respondents.
Measures
Self-rated memory status. self-rated memory was assessed using a single-question: “How
would you rate your memory at present?” The response was on a 5-point scale: 1 = excellent, 2 =
very good, 3 = good, 4 = fair, and 5 = poor. We classified the variable into two categories 0 =
excellent/very good/good, and 1 = fair/poor following the prior studies of self-rated memory
(Cutler, 2015).
Race/ethnicity. Race/ethnicity is the main predictor of interest. It is self-reported and the
categories consist of non-Hispanic whites, non-Hispanic blacks, and Hispanics.
Covariates. Covariates include demographic characteristics, socioeconomic status, health
status, and social and cultural factors. The demographic characteristics include age (65-69, 70-
74, 75-79, 80-84, 85+), gender (male and female), and marital status (married/living with a
partner, divorced/separated, widowed, and never married). The two socioeconomic status
measures are education level (less than high school, high school graduate, some college, and
college or more) and income. The NHATS performed imputation of the income variable and
replaced the missing values of the income (Montaquila et al., 2015). Due to skewness and
kurtosis of income, income was log-transformed.
Health status includes the number of chronic conditions, depressive symptoms, objective
memory status, and functional limitations. Respondents were asked to report whether a doctor
had ever diagnosed them with a set of chronic conditions consisting of heart attack, hypertension,
diabetes, and stroke. Hence, the number of chronic diseases ranged from 0 to 4. In addition,
studies have suggested that individuals’ psychological wellbeing may influence their self-rated
10
memory to provide rich information on how psychological factors impact self-rated memory
status. To do so, depressive symptoms were evaluated using the total score of two items from the
Patient Health Questionnaire 2 (PHQ-2): Over the last month, how often have you had little
interest or pleasure in doing things?” and “Over the last month, how often have you felt down,
depressed, or hopeless?” (Kroenke et al., 2003). The responses were: 0 (not at all), 1 (several
days), 2 (more than half the days), 3 (nearly every day). We summed the total scores from the
two items, with the summed score ranging from 0 to 6. Higher scores indicate a higher level of
depressive symptoms. The objective memory status was based on immediate and delayed 10
item word recall tests. For both tests, a list of 10 nouns was read to respondents and they were
asked to recall as many nouns as possible. The measure ranged from 0 to 20 points, with higher
scores indicating better objective memory status.
Functional limitations are measured using activities of daily livings (ADLs) and
instrumental activities of daily livings (IADLs) (Lawton & Brody, 1969). The ADLs are based
on difficulty with eating, bathing, toileting, dressing, getting around the house, getting in and out
of bed, getting in and out of chairs, and going outside without help. The IADLs are based on
difficulty with shopping, housekeeping, managing medications, laundry, using the phone,
cooking, and managing money. The responses are coded as 0 (no difficulty to perform) and 1
(difficult to perform). The total ADLs range from 0 (no functional limitation) to 8 (severe
functional limitations), and the IADLs range from 0 (no functional limitation) to 7 (severe
functional limitations).
Certain social and cultural characteristics may be particularly relevant for racial/ethnic
minorities. This study examines whether economic vulnerability, religious service, and limited
English proficiency influence differences in self-rated memory by race/ethnicity (Lehning et al.,
11
2020; Lukoff et al., 1995; Zhang et al., 2012). Economic vulnerability is evaluated by whether
individuals had Medicaid, which provides coverage to low-income and disabled individuals
(Lehning et al., 2020). Individuals who said they understand and speak English “not well” or
“not at all” were characterized as having limited English proficiency following previous research
(Franco & Choi, 2020). Attending religious services was asked by the following: “In the last
month, did you ever attend religious services?” and the answers were coded into 1 (yes) and 0
(no).
Analytical Strategy
First, we conducted descriptive analyses of the demographic characteristics,
socioeconomic status, health status, and social and cultural factors by race/ethnicity. We
compared non-Hispanic whites with non-Hispanic blacks and Hispanics using t-tests for
continuous variables and chi-squared tests for categorical variables. We used logistic regression
to examine the relationship between race/ethnicity and the odds of having fair/poor self-rated
memory. We step in variables to examine how different categories of covariates influence
differences in self-rated memory by race/ethnicity. Model 1 includes demographic characteristics
(age, gender, and marital status). Model 2 added education level and annual household income to
examine whether socioeconomic factors explain differences in the association between
race/ethnicity and self-rated memory. Model 3 added the measures of the number of chronic
conditions, depressive symptoms, objective memory status, and functional limitations to
determine whether differences in self-rated memory by race/ethnicity were related to health
status. Model 4 added economic vulnerability, religious service, and limited English proficiency
to provide evidence on whether the relationship between race/ethnicity and self-rated memory
were driven by social and cultural factors. Finally, we examined additional models that stratified
12
by limited English proficiency among Hispanics to further unpack the association between
race/ethnicity and fair/poor self-rated memory. All analyses were performed using STATA 14.2
(Stata Corp., College Station, TX) and weighted to account for complex survey design.
Results
Descriptive characteristics of the sample
Table 1-1 shows weighted descriptive characteristics of the sample by race/ethnicity.
Overall, 72.2% of the respondents were non-Hispanic whites, 21.9% were non-Hispanic blacks,
and 5.8% were Hispanics. The percentage of the respondents who were female was fairly similar
across all groups. Non-Hispanic blacks and Hispanics were less likely to be married/living with a
partner than non-Hispanic whites. Compared to non-Hispanic whites and non-Hispanic blacks,
Hispanics had a lower education level as well as lower annual household income. Approximately
57.9% of Hispanics had less than a high school education compared to 37.6% of non-Hispanic
blacks and 15.4% of non-Hispanic whites. The number of chronic conditions and religious
service were higher among non-Hispanic blacks, while depressive symptoms and the level of
functional limitations were higher among Hispanics compared to non-Hispanic whites. Nearly
half (47.8%) of Hispanics reported limited English proficiency, while almost no non-Hispanic
whites (0.4%) and non-Hispanic blacks (0.4%) reported limited English proficiency. The
prevalence of having fair/poor self-rated memory was substantially higher among Hispanics
(38.0%) and non-Hispanic blacks (25.1%) than among non-Hispanic whites (14.5%). This was
consistent with the objective memory status scores, which were higher among non-Hispanic
whites than among non-Hispanic blacks and Hispanics.
Association between race/ethnicity and fair/poor self-rated memory
13
Table 1-2 presents the results from logistic regression models predicting fair/poor self-
rated memory status. The baseline model shows that non-Hispanic blacks and Hispanics had 2.03
(95% confidence interval [CI] = 1.67-2.42) and 3.78 (95% CI = 2.90-4.93) higher odds of
reporting fair/poor self-rated memory compared to non-Hispanic whites, respectively, controlling
for demographic factors (age, gender, and marital status). Non-Hispanic blacks and Hispanics
were significantly more likely to report fair/poor self-rated memory than non-Hispanic whites
even after controlling for socioeconomic characteristics (education level and income) (Model 2)
and health status (number of chronic conditions, depressive symptoms, objective memory status,
and functional limitations) (Model 3). Once social and cultural factors (economic vulnerability,
religious services, and limited English proficiency) were included in the model (Model 4),
Hispanics continued to have significantly higher odds of reporting fair/poor self-rated memory
than non-Hispanic whites, but the size of the odds ratio was substantially reduced (odds ratio
[OR] = 1.56, 95% CI = 1.09-2.21). Also, non-Hispanic blacks were significantly more likely to
report fair/poor self-rated memory than non-Hispanic whites (OR = 1.33, 95% CI = 1.11-1.60).
Respondents with limited English proficiency were more likely to report fair/poor self-rated
memory (OR = 1.81, 95% CI = 1.14-2.88). Being male, having less than a high school education,
having lower income, having higher levels of depressive symptoms, having a lower objective
memory status score, having higher levels of IADLs, and not attending religious services were
associated with significantly greater odds of reporting fair/poor self-rated memory.
Association between race/ethnicity and fair/poor self-rated memory, stratifying by limited
English proficiency
We found that limited English proficiency was a significant predictor of fair/poor self-
rated memory. To further investigate the role limited English proficiency plays in the association
14
between race/ethnicity and fair/poor self-rated memory, we classified Hispanics into two groups:
those who had limited English proficiency and those without limited English proficiency. The
results from logistic regression models predicting self-rated memory by race/ethnicity and
stratifying by limited English proficiency among Hispanics are presented in Table 1-3. The
results for non-Hispanic blacks were very similar to those presented in Table 1-2. When health
status was controlled (Model 3), Hispanics without limited English proficiency were not
significantly more likely to report having fair/poor self-rated memory than non-Hispanic whites
(OR = 1.42, 95% CI = 0.98-2.05). However, Hispanics with limited English proficiency
continued to experience substantially higher odds of reporting fair/poor self-rated memory status
(OR = 3.16, 95% CI = 2.20-4.54), adjusting for demographic characteristics, socioeconomic
status, and health status (Model 3). This association remained significant even after controlling
for social and cultural factors in model 4 (OR = 3.02, 95% CI = 2.05-4.46).
The impact of survey language on the association between race/ethnicity and good/fair/poor self-
rated memory
Researchers have drawn attention to the fact that the language of interview used in
surveys can affect the reporting of self-rated health. For instance, Viruell-Fuentes and colleagues
(2011) found that Spanish-language respondents reported more poor self-rated health due to the
translation of the English word “fair” into “regular” in the Spanish language survey. These
researchers demonstrated that respondents who were interviewed in Spanish tended to report
poorer health status than they would if they were interviewed in English. The NHATS Spanish
translation of the response classification of the self-rated memory was classified into “excelente
(excellent)”, “muy buena (very good)”, “buena (good)”, “regular (fair)”, and “mala (bad)”. We
ran additional logistic regression models where the outcome was reporting good, poor, or fair
15
(versus excellent) self-rated memory to investigate whether the results remain robust to an
alternate classification of self-rated memory. This may provide evidence about whether survey
language is an issue that leads to differences in self-rated memory by race/ethnicity as the “fair”
category is no longer used as the cutoff.
In these models, non-Hispanic blacks and Hispanics remained significantly more likely to
report having good/fair/poor self-rated memory than non-Hispanic whites, controlling for age,
gender, and marital status in the baseline model. These associations remained significant even
after controlling for socioeconomic status and health status (Appendix A). Contrary to when
using fair/poor self-rated memory as the outcome measure (Table 2), Hispanics were not
significantly more likely to predict having good/fair/poor self-rated memory than non-Hispanic
whites, after adjusting for social and cultural factors in Model 4 (OR = 1.29, 95% CI = 0.94-
1.77). However, respondents with limited English proficiency showed 2.13 higher odds of
reporting good/fair/poor self-rated memory (95% CI = 1.29-3.52). In the final model, older age,
being never-married, having less than a high school education, having low annual household
income, having more chronic conditions, having higher levels of depressive symptoms, and
having lower objective memory status were significantly associated with higher odds of
reporting good/fair/poor self-rated memory status.
The impact of survey language on the association between race/ethnicity and
good/fair/poor self-rated memory, stratifying by limited English proficiency. To test whether
limited English proficiency has an impact on good/fair/poor self-rated memory, we performed
logistic regression to examine the relationship between race/ethnicity and good/fair/poor self-
rated memory stratifying by limited English proficiency in Appendix B. Compared to using
fair/poor self-rated memory (Table 1-3), Hispanics without limited English proficiency did not
16
have significantly higher odds of reporting good/fair/poor self-rated memory than non-Hispanic
whites, adjusting for demographic characteristics and socioeconomic status (OR = 1.35, 95% CI
= 0.97-1.89). This did not change when controlling for health status (Model 3) and social and
cultural factors (Model 4). On the other hand, Hispanics with limited English proficiency had
significantly higher odds of reporting good/fair/poor self-rated memory than non-Hispanic white
in all the models, which is consistent when using fair/poor self-rated memory status. For
instance, Hispanics with limited English proficiency had 2.87 higher odds of reporting
good/fair/poor self-rated memory while controlling for demographic characteristics,
socioeconomic status, health status, social and cultural factors (95% CI = 1.87-4.41). Thus, the
results remained consistent when we used an alternate classification of self-rated memory.
Discussion
As older adults are living longer, a number of older adults are experiencing memory
change in later years. At the same time, the proportion of minority older adults is increasing in
the U.S. Given that minority older adults tend to have worse cognitive functioning compared to
non-Hispanic whites, it is possible that the older minority population may rate their memory
status worse than non-Hispanic whites. Individuals’ self-rated information has been commonly
used in research in the past. One example would be self-rated health. Self-rated health is a
perception of individuals’ health status and it is a strong predictor of individual’s morbidity and
mortality (DeSalvo et al., 2006; G. A. Kaplan et al., 1996). Self-rated health has been widely
used due to its beneficial merits (Borrell & Dallo, 2008; DeSalvo et al., 2006; McMullen &
Luborsky, 2006). Similar to self-rated health, self-rated memory may add to our understanding
of cognitive functioning.
17
This study expanded the scope to investigate differences in self-rated memory by
race/ethnicity in a nationally representative sample of older adults. First, we found that the older
minority population had greater odds of reporting fair or poor self-rated memory than non-
Hispanic whites. There were significant differences in the likelihood of reporting poor self-rated
memory by race/ethnicity, adjusting for the full set of covariates (demographic characteristics,
socioeconomic status, health status, and social and cultural factors). This finding is consistent
with previous research, which showed that the prevalence of poor self-rated memory was higher
among the minority (Jang, Choi, et al., 2021). Respondents with low education, more depressive
symptoms, worse objective memory status, and who had IADL limitations were significantly
more likely to report fair/poor self-rated memory. To the best of our knowledge, there were
limited studies on the association between chronic disease and self-rated memory. However, we
were able to locate research on chronic disease and self-rated health. Although previous studies
found a positive association between chronic diseases and poor self-rated health (Molarius &
Janson, 2002), our findings showed that the number of chronic diseases was not a significant
predictor of fair/poor self-rated memory. More research is necessary to investigate the
relationship between chronic diseases and self-rated memory, and what types of chronic diseases
are highly associated with self-rated memory.
Prior studies have suggested that limited English proficiency is an important predictor for
disability and psychological distress, and it also has critical implications for cognitive health
among minorities (DuBard & Gizlice, 2008; Franco & Choi, 2020; Ponce et al., 2006a).
Although few studies have investigated the relationship between limited English proficiency and
self-rated memory, previous literature demonstrated that limited English proficiency was
associated with worse self-rated health among racial/ethnic minorities (Kim et al., 2011). Ours is
18
the first study to examine whether limited English proficiency is contributes to poor self-rated
memory among older adults. Our findings suggest that Hispanics with limited English
proficiency had higher odds of fair/poor self-rated memory than non-Hispanic whites, while
there were no significant differences in the likelihood of reporting fair/poor self-rated memory
between Hispanics without limited English proficiency and non-Hispanic whites. Individuals
with limited English proficiency are less likely to utilize health care and seek medical assistance
for cognitive health issues as well as chronic diseases (Jang & Kim, 2019; Kim et al., 2011).
Thus, individuals with limited English proficiency are more likely to experience worse health
status than individuals without limited English proficiency due to difficulty understanding
written information from their doctors regarding on prevention and management of chronic
diseases (Betancourt et al., 2012; Kim et al., 2011), which may negatively impact cognition. This
may contribute to racial/ethnic disparities in cognitive health among older adults.
The limitations of the current research should be noted. First, the variables used in this
research are based on respondents’ self-reports. However, while we do not have information
about clinical diagnoses, research has established the reliability of self-reported information on
chronic diseases as well as health utilization (Najafi et al., 2019; Short et al., 2009). Second, the
NHATS is representative of the Medicare population aged 65 and older. Hence, the findings may
not be generalized to the older adult population. However, given that 83% of older adults are
Medicare beneficiaries (Kaiser Family Foundation, 2016), the sample in the study represents a
large proportion of older adults residing in the U.S. Among the Medicare beneficiaries, 76.4%
are non-Hispanic whites, 9.4% non-Hispanic blacks, and 9.0% Hispanics (Kaiser Family
Foundation, 2016). Third, despite the use of a nationally representative sample of older adults,
19
the sample size of Hispanics was small in this study (N = 388). Future studies are necessary to
see if the results hold in sample with larger numbers of Hispanic older adult population.
Despite these limitations, this research contributes to the literature on differences in self-
rated memory by race/ethnicity and demonstrates how limited English proficiency influences
these associations among Hispanics. Our findings have important implications and practice for
health professionals, especially for those who are in involved in conducting and interpreting
cognitive evaluation. Health professionals should be mindful of self-rated memory of non-
Hispanic blacks and Hispanics, since they tend to report poor self-rated memory than non-
Hispanic whites, especially for older adults with limited English proficiency. Greater efforts to
promote cognitive screening and provide tailored language healthcare resources for Hispanics
with limited English proficiency are needed, given that they had the highest prevalence of poor
self-rated memory among the groups examined in this study. These measures may help address
racial/ethnic disparities in cognitive health among older adults.
20
Table 1-1. Descriptive statistics by race/ethnicity, NHATS 2011
Non-Hispanic
whites
% or Mean
(SE)
Non-Hispanic
blacks
% or Mean (SE)
Hispanics
% or Mean (SE)
X
2
/t
test
Self-rated memory
(fair/poor)
14.54 25.19 38.04 85.48***
Age 3.49**
65-69 28.75 31.73 30.24
70-74 25.00 29.48 29.02
75-79 19.33 17.45 18.59
80-84 14.79 12.79 12.89
85+ 12.14 8.55 9.26
Gender
Female 56.40 59.36 55.40 0.79
Marital status 14.17***
Married/living
with a partner
60.31 39.82 54.45
Separated/divorced
10.85 22.18 16.83
Widowed 25.80 31.46 24.18
Never-married 3.04 6.54 4.53
Education level 81.38***
Less than high
school
15.43 37.61 57.91
High school
graduate
28.76 25.68 16.97
Some college 28.65 22.18 15.79
College or more 27.16 14.53 9.33
Annual household
income
a
10.47 (0.03) 9.73 (0.08) 9.49 (0.12) 79.76***
Number of chronic
conditions
b
1.05 (0.02) 1.41 (0.03) 1.22 (0.05) 33.28***
Depressive
symptoms
c
0.84 (0.03) 1.13(0.04) 1.37(0.13) 19.84***
Objective memory
status
d
8.66 (0.07) 6.99 (0.09) 7.05 (0.26) 62.59***
Functional
limitations
21
Note: All estimates are weighted to account for complex survey design.
Abbreviation: NHATS, National Health and Aging Trends Study; SE, standard error; ADLs,
activities of daily living; IADLs, instrumental activities of daily living.
*p<0.05; **p<0.01; ***p<0.001
a
Annual household income was log-transformed.
b
The number of chronic conditions is based on heart attack, hypertension, diabetes, and stroke.
c
Depressive symptoms was evaluated by the Patient Health Questionnaire-2 and values ranged
from 0 to 6, with a higher score indicating more severe depressive symptoms.
d
Objective memory status was based on immediate and delayed recall. Values ranged from 0 to
20, with a higher score indicating better memory status.
e
Difficulty with activities of daily living (eating, getting in/out of bed, getting in/out of chair,
walking, go outside, dressing, bathing, and toileting). Values ranged from 0 to 8, with a higher
score indicating more ADL limitations.
f
Difficulty with instrumental activities of daily living (meal preparation, laundry, light
housework, shopping for groceries, banking or paying bills, keeping track of medication, and
using phone calls). Values ranged from 0 to 7, with a higher score indicating more IADL
limitations.
g
Economic vulnerability is measured by whether respondents are covered by Medicaid.
h
Attending religious services was based on whether respondents participated in religious
services in the last month.
ADLs
e
0.35 (0.02) 0.49 (0.04) 0.75 (0.08) 22.28***
IADLs
f
0.55 (0.03) 0.82 (0.05) 1.09 (0.09) 28.10***
Economic
vulnerability
g
6.77 28.56 34.96 186.61***
Attending religious
services
h
57.08 70.28 58.90 15.51***
Limited English
proficiency
0.41 0.45 47.86 909.20***
Survey conducted in
Spanish
0.01 0.00 50.94 756.03***
N 4,753 1,442 388
22
Table 1-2. Odds Ratios from Logistic Regression Models for the Association Between Race/Ethnicity and Fair/Poor Self-rated
Memory
Variables Model 1
OR (95% CI)
Model 2
OR (95% CI)
Model 3
OR (95% CI)
Model 4
OR (95% CI)
Race/ethnicity (Ref: Non-Hispanic
whites)
Non-Hispanic blacks 2.03 (1.69-2.42)*** 1.57 (1.31-1.89)*** 1.33 (1.11-1.58)** 1.33 (1.11-1.60)**
Hispanics 3.78 (2.90-4.93)*** 2.48 (1.89-3.24)*** 2.12 (1.65-2.73)*** 1.56 (1.09-2.21)*
Age (Ref: 65-69)
70-74 1.28 (0.97-1.68) 1.19 (0.89-1.59) 1.18 (0.88-1.59) 1.17 (0.87-1.57)
75-79 1.48 (1.13-1.94)** 1.32 (1.00-1.75)* 1.14 (0.85-1.52) 1.15 (0.86-1.53)
80-84 1.99 (1.57-2.52)*** 1.80 (1.40-2.30)*** 1.36 (1.03-1.80)* 1.38 (1.03-1.83)*
85+ 2.42 (1.85-3.10)*** 2.17 (1.63-2.88)*** 1.42 (1.04-1.92)* 1.44 (1.06-1.96)*
Gender (Ref: male)
Female 0.75 (0.64-0.88)** 0.73 (0.62-0.86)*** 0.78 (0.66-0.92)** 0.79 (0.67-0.93)**
Marital status (Ref: married/living with a
partner
Separated/divorced 1.26 (1.00-1.59)* 1.15 (0.91-1.46) 0.99 (0.78-1.27) 0.97 (0.76-1.25)
Widowed 1.24 (1.00-1.53)* 1.04 (0.84-1.27) 0.93 (0.76-1.14) 0.93 (0.76-1.13)
Never-married 1.13 (0.69-1.83) 1.00 (0.62-1.62) 0.91 (0.55-1.49) 0.87 (0.53-1.42)
Education level (Ref: less than high
school)
High school graduate 2.91 (2.29-3.70)*** 1.79 (1.37-2.34)*** 1.69 (1.28-2.24)***
Some college 1.66 (1.32-2.09)*** 1.26 (0.98-1.63) 1.27 (0.98-1.64)
College or more 1.42 (1.11-1.81)** 1.16 (0.91-1.49) 1.17 (0.92-1.49)
Annual household income
a
0.90 (0.85-0.94)*** 0.93 (0.88-0.98)** 0.94 (0.88-0.99)*
Number of chronic conditions
b
1.09 (0.87-0.97)* 1.09 (1.01-1.18)
Depressive symptoms
c
1.38 (1.32-1.44)*** 1.37 (1.31-1.43)***
Objective memory status
d
0.89 (0.87-0.92)*** 0.89 (0.87-0.92)***
23
Note: All estimates are weighted to account for complex survey design.
Abbreviation: NHATS, National Health and Aging Trends Study; OR, odds ratios; ADLs, activities of daily living; IADLs,
instrumental activities of daily living.
*p<0.05; **p<0.01; ***p<0.001
a
Annual household income was log-transformed.
b
The number of chronic conditions is based on heart attack, hypertension, diabetes, and stroke.
c
Depressive symptoms was evaluated by the Patient Health Questionnaire-2 and values ranged from 0 to 6, with a higher score
indicating more severe depressive symptoms.
d
Objective memory status was based on immediate and delayed recall. Values ranged from 0 to 20, with a higher score indicating
better memory status.
e
Difficulty with activities of daily living (eating, getting in/out of bed, getting in/out of chair, walking, go outside, dressing, bathing,
and toileting). Values ranged from 0 to 8, with a higher score indicating more ADL limitations.
f
Difficulty with instrumental activities of daily living (meal preparation, laundry, light housework, shopping for groceries, banking or
paying bills, keeping track of medication, and using phone calls). Values ranged from 0 to 7, with a higher score indicating more
IADL limitations.
g
Economic vulnerability is measured by whether respondents are covered by Medicaid.
h
Attending religious services was based on whether respondents participated in religious services in the last month.
Functional limitations
ADLs
e
1.03 (0.98-1.09) 1.03 (0.97-1.09)
IADLs
f
1.10 (1.05-1.15)*** 1.09 (1.05-1.15)***
Economic vulnerability
g
1.16 (0.89-1.51)
Attending religious service
h
0.98 (0.78-1.51)
Limited English proficiency 1.81 (1.42-2.88)*
N 6,583 6,583 6,583 6,583
24
Table 1-3. Odds Ratios from Logistic Regression Models for the Association Between Race/Ethnicity and Fair/Poor Self-rated
Memory, Stratifying by Limited English Proficiency Among Hispanics
Variables Model 1
OR (95% CI)
Model 2
OR (95% CI)
Model 3
OR (95% CI)
Model 4
OR (95% CI)
Race/ethnicity (Ref: Non-Hispanic white)
Non-Hispanic black 2.03 (1.69-2.42)*** 1.60 (1.33-1.93)*** 1.34 (1.12-1.61)** 1.32 (1.10-1.59)**
Hispanics with LEP 7.14 (4.97-10.24)*** 3.89 (2.68-5.64)*** 3.16 (2.20-4.54)*** 3.02 (2.05-4.46)***
Hispanics without LEP 1.90 (1.40-2.59)*** 1.57 (1.14-2.16)** 1.42 (0.98-2.05) 1.42 (0.99-2.05)
Age (Ref: 65-69)
70-74 1.23 (0.93-1.63) 1.17 (0.87-1.57) 1.16 (0.86-1.56) 1.16 (0.86-1.56)
75-79 1.47 (1.11-1.93)*** 1.33 (1.00-1.76)* 1.13 (0.85-1.52) 1.14 (0.86-1.55)
80-84 1.96 (1.53-2.51)*** 1.79 (1.39-2.31)*** 1.35 (1.01-1.80)* 1.37 (1.02-1.83)*
85+ 2.38 (1.81-3.14)*** 2.17 (1.63-2.89)*** 1.41 (1.04-1.92)* 1.42 (1.04-1.94)*
Gender (Ref: male)
Female 0.75 (0.64-0.88)** 0.74 (0.62-0.86)*** 0.78 (0.66-0.92)** 0.79 (0.67-0.93)**
Marital status (Ref: married/living with a
partner
Separated/divorced 1.28 (1.03-1.60)* 1.17 (0.93-1.48) 1.01 (0.79-1.28) 0.97 (0.76-1.25)
Widowed 1.23 (1.00-1.53)* 1.05 (0.85-1.29) 0.94 (0.77-1.15) 0.92 (0.75-1.14)
Never-married 1.08 (0.67-1.73) 0.98 (0.61-1.58) 0.89 (0.53-1.45) 0.86 (0.52-1.42)
Education level (Ref: less than high
school)
High school graduate 2.78 (2.17-3.56)*** 1.72 (1.30-2.26)*** 1.67 (1.28-1.22)**
Some college 1.68 (1.33-2.12)*** 1.27 (0.98-1.65) 1.27 (0.98-1.63)
College or more 1.43 (1.12-1.83)** 1.16 (0.92-1.49) 1.16 (0.91-1.49)
Annual household income
a
0.90 (0.86-0.95)*** 0.93 (0.88-0.98)* 0.93 (0.88-0.99)*
Number of chronic conditions
b
1.09 (0.01-1.18)* 1.09 (1.01-1.18)
Depressive symptoms
c
1.37 (1.31-1.43)*** 1.37 (1.31-1.43)***
Objective memory status
d
0.89 (0.87-0.91)*** 0.89 (0.87-0.91)***
25
Note: All estimates are weighted to account for complex survey design.
Abbreviation: NHATS, National Health and Aging Trends Study; LEP, limited English proficiency; OR, odds ratios; ADLs, activities
of daily living; IADLs, instrumental activities of daily living.
*p<0.05; **p<0.01; ***p<0.001
a
Annual household income was log-transformed.
b
The number of chronic conditions is based on heart attack, hypertension, diabetes, and stroke.
c
Depressive symptoms was evaluated by the Patient Health Questionnaire-2 and values ranged from 0 to 6, with a higher score
indicating more severe depressive symptoms.
d
Objective memory status was based on immediate and delayed recall. Values ranged from 0 to 20, with a higher score indicating
better memory status.
e
Difficulty with activities of daily living (eating, getting in/out of bed, getting in/out of chair, walking, go outside, dressing, bathing,
and toileting). Values ranged from 0 to 8, with a higher score indicating more ADL limitations.
f
Difficulty with instrumental activities of daily living (meal preparation, laundry, light housework, shopping for groceries, banking or
paying bills, keeping track of medication, and using phone calls). Values ranged from 0 to 7, with a higher score indicating more
IADL limitations.
g
Economic vulnerability is measured by whether respondents are covered by Medicaid.
h
Attending religious services was based on whether respondents participated in religious services in the last month.
Functional limitations
ADLs
e
1.03 (0.97-1.09) 1.03 (0.97-1.09)
IADLs
f
1.09 (1.04-1.15)*** 1.09 (1.05-1.15)***
Economic vulnerability
g
1.16 (0.89-1.50)
Attending religious service
h
0.93 (0.78-1.11)
N 6,583 6,583 6,583 6,583
26
Chapter 3: The Relationship Between Multimorbidity and Types of Chronic Diseases and Self-
Rated Memory
Introduction
As the older adult population with dementia has increased, the attention on cognitive
aging has increased as well (Blazer et al., 2015). Approximately 6.2 million older adults are
currently living with Alzheimer’s disease in the U.S. (Alzheimer’s Association, 2021). Although
dementia prevalence and incidence are expected to decrease in high-income countries, the
absolute number of older adults with the disease is projected to increase unless potentially
modifiable risk factors are addressed (Larson et al., 2013; Norton et al., 2014). In line with this,
individuals’ perceived memory status is considered one of the early indicators of cognitive
decline and dementia onset.
Self-rated memory is often used in addition to objective cognitive status to assess
cognitive health in clinical and research settings (Marino et al., 2009). Although single questions
on self-rated memory lack psychometric properties, they are commonly used as a quick and easy
measure of cognitive screening (Cutler, 2015; Jang, et al., 2021). Self-rated memory is an
individual’s assessment of their own memory status as opposed to receiving an objective
diagnosis from a clinical evaluation, but subjective memory performance has been shown to be
related to objective memory (Dubois et al., 2016; Steinberg et al., 2013). Thus, self-rated
memory may help identify older adults who are at risk of cognitive impairment in the future
when there is no diagnosis of cognitive impairment in the present (Mitchell et al., 2014).
Furthermore, self-rating of poor memory is quite common, the prevalence among adults aged 65
and older being as high as 50% (Jonker et al., 2000; Mol et al., 2007). When it comes to
demographic characteristics and self-reported information, research suggests that demographic
27
characteristics (e.g., gender and education) are associated with self-rated health (Arezzo &
Giudici, 2017; Bamia et al., 2017). For example, women tend to rate their self-rated health
poorer than their men counterparts (Bamia et al., 2017). In line with this, we can postulate that
the respondents’ demographic characteristics would be associated with self-rated memory.
Along with cognition, chronic disease is another source of burden in later years of life.
The risk of chronic disease increases with age and may be linked to the development of poor
self-rated memory and cognitive impairment. It is well-documented that many risk factors for
chronic diseases are modifiable and linked to lifestyle factors. For instance, cigarette smoking
and obesity were observed to increase the risk of developing chronic diseases (Ng et al., 2020).
Having chronic disease may limit older adults’ independence, decrease quality of life, and
increase the cost of health care (Centers for Disease Control and Prevention [CDC], 2021).
Approximately, 60% of Americans aged 18 and older have at least one chronic disease, while
42% of adults have multiple chronic diseases (i.e., multimorbidity) (Buttorff et al., 2017). The
prevalence of chronic disease and multimorbidity increases with age. Among adults aged 65 and
older, it is common to have more than two chronic diseases. In fact, approximately 80% of older
adults have at least two chronic diseases (National Council on Aging, 2018). To summarize, the
prevalence of chronic diseases and negative self-rated memory is increasing, and it will
inevitably have a negative impact on cognitive health in later life.
Literature Review
Chronic Diseases and Cognitive Impairment and Dementia
The close linkages between chronic diseases and negative self-rated memory have been
identified in the aging population (Aarts et al., 2011; Yap et al., 2020). Given the high burden of
chronic diseases at older ages and their potential to undermine self-rated memory, it is important
28
to understand the relationship between chronic diseases and cognitive impairment and/or
dementia. Having multimorbidity may be a risk factor for worsening cognitive status and
dementia. Multimorbidity is an important factor for older adults since it is highly associated with
quality of life and life expectancy (Chudasama et al., 2019; Williams & Egede, 2016). In
addition, multimorbidity among older adults has been linked to higher costs and greater
utilization of health care, as well as to age-related cognitive decline (Buttorff et al. 2017; Fabbri
et al., 2015). In addition, care for multimorbidity is costly and challenging to manage for older
adults (Navickas et al., 2016). The relationship between multimorbidity and mild cognitive
impairment (MCI) and dementia was investigated in a study of older adults residing in
Minnesota. This study included 17 chronic conditions (e.g., hyperlipidemia, hypertension,
depression, diabetes, asthma, coronary artery diseases, substance abuse disorder). Although this
study did not investigate which individual chronic condition is an important risk factor for
developing MCI and dementia, the finding showed that risks of developing MCI and dementia
were significantly higher for those who had multimorbidity than those with no multimorbidity
(Vassilaki et al., 2015).
Moreover, previous studies have found that chronic diseases increase the risk of having
cognitive impairment in the future (Kurella et al., 2005; Strachan et al., 2008). For example,
Kurella and colleagues (2005) found that community-dwelling older adults with chronic kidney
disease had an increased risk of developing cognitive impairment. Also, people with diabetes,
particularly Type 2 diabetes, showed a significantly greater decline in cognition than those
without diabetes (Strachan et al., 2008). Moreover, people with diabetes had greater odds of
developing dementia. Thus, it can be hypothesized that having certain chronic diseases increases
the risk of developing cognitive impairment and dementia.
29
Chronic Diseases and Self-Rated Memory
We were able to identify a few studies on the association between chronic diseases and
self-rated memory with a single item (e.g., “Do you have problems with your memory?” or “Do
you have memory problems?”). Research shows that the respondents' cognitive complaints
commonly accompany chronic diseases and the complaints were elevated for those with
multimorbidity (Taylor et al., 2020). Similar findings were documented in a study conducted on
a Spanish population of older adults for whom negative subjective memory complaints increased
with the number of chronic diseases (Pedro et al., 2016). In line with this, prior cross-sectional
studies supported that there is a strong association between chronic diseases and subjective
cognitive complaints (Aarts et al., 2011; Pedro et al., 2016; Yap et al., 2020). This relationship
remains even after adjusting for potential factors including objective cognitive status, education
level, and psychological stress (Aarts et al., 2011; Begum et al., 2012).
Similarly, a study conducted with respondents aged 45 and older showed that there was a
positive relationship between subjective cognitive decline and chronic conditions (Taylor et al.,
2020). This study included eight chronic conditions (asthma, heart attack/heart disease, stroke,
cancer, chronic obstructive pulmonary disease [COPD], arthritis, kidney disease, and diabetes)
and demonstrated that having chronic diseases was associated with greater odds of subjective
cognitive decline. Particularly, respondents with stroke, heart disease, and COPD showed
significantly higher odds of reporting subjective cognitive decline compared to those without
these diseases. Nonetheless, one of the limitations of the study was that they included both
middle-aged and older adults; the relationship between subjective cognitive decline and chronic
diseases may differ for two age groups. For example, the prevalence of chronic diseases between
two age groups is significantly different; approximately 49.1% of middle-aged individuals have
30
multimorbidity, while about 80.1% of individuals who are aged 65 and older have
multimorbidity in the U.S. (Gerteis et al., 2014). Thus, it can be easily hypothesized that the
burden from chronic diseases is higher among older adults than the middle-aged group, which
may differently influence self-rated memory.
Although researchers recognize the importance of including objective memory status in
the study of self-rated memory, there is limited literature which measures the impact of objective
memory status on the association between chronic disease and self-rated memory. For example,
the aforementioned studies (e.g., Jacobs et al., 2019; Taylor et al., 2020) were not able to include
objective assessment of memory due to the conditions of the survey design. Also, a national
report demonstrated that chronic diseases impact subjective cognitive concerns (Centers for
Disease Control and Prevention & National Center for Chronic Disease Prevention and Health
Promotion, 2020). However, this report was not able to include objective memory status.
Therefore, this study will advance the literature by examining the association of self-rated
memory with multimorbidity and types of chronic diseases by including objective memory
status, as well as using a nationally representative sample of older adults in the U.S.
Purpose of This Study
Even if the importance of chronic diseases and self-rated memory is well-recognized,
there has been a lack of research on the impact of multimorbidity and types of chronic diseases
on self-rated memory in older adults. Although previous studies individually explored the
relationship between either multimorbidity or types of chronic diseases and subjective cognitive
concerns, there are few studies which have examined both the associations between
multimorbidity and types of chronic diseases and self-rated memory in the U.S. Hence, this study
first examines the association between multimorbidity and self-rated memory using a nationally
31
representative sample of older adults. We then explore which of eight individual chronic diseases
are linked to reporting poor self-rated memory. We control for a set of socio-demographic
characteristics (age, gender, race/ethnicity, marital status, annual household income, and
education level), lifestyle factors (smoking status and body mass index [BMI]), mental health
conditions (psychological well-being and depressive symptoms), prescribed medicine use, and
objective memory status. The proposed research questions are: “Does multimorbidity increase
the risk of reporting fair/poor self-rated memory?” and “Which chronic diseases are associated
with increased risk of reporting fair/poor self-rated memory?” Based on the aforementioned
literature, we hypothesize that older adults with multimorbidity would have a higher prevalence
of fair/poor self-rated memory. Given that there is a high prevalence of both poor self-rated
memory and chronic diseases with age, it is necessary to understand the association to provide
information on cognitive impairment when there is yet to be an objective diagnosis of dementia.
Research Design
Sample
We used data from the 2011 National Health and Aging Trends Study (NHATS). The
NHATS is nationally representative of Medicare beneficiaries aged 65 and older. The NHATS
contains information on multiple domains including demographics, socioeconomic status, and
health conditions of the respondents. The survey consisted of a two-hour in-person interview,
with proxy respondents used if older adults were unable to complete the interview. In 2011, a
total of 8,245 respondents were interviewed. Our study excluded proxy respondents since they
were not asked to rate the memory of the respondent (n = 583). In addition, we excluded
respondents who were residing in nursing homes as they were also not asked the question on
32
self-rated memory (n = 468). The final sample consisted of 6,481 individuals after excluding
respondents who were missing other covariates of interest (n = 713).
Measures
Self-rated memory. Participants were asked to rate their memory at present on a response
of excellent, very good, good, fair, or poor. We coded the response into a binary variable of
positive ratings (excellent/very good/good) and negative ratings (fair/poor). We’ve followed the
coding scheme to be in line with the existing literature on self-rated health (DeSalvo et al.,
2006), as well as how previous cognitive research classified the responses (Jang et al., 2021).
Chronic diseases. We included all eight chronic diseases (heart disease, heart attack,
hypertension, diabetes, stroke, osteoporosis, arthritis, and lung disease) that respondents were
asked about in the NHATS. For each chronic disease, the respondents were asked: Please tell me
if a doctor ever told you that you had (chronic disease). Respondents with two or more chronic
diseases were classified as having multimorbidity and those with no or one chronic disease as
having no multimorbidity. Then each individual type of chronic disease was examined in the
association.
Covariates. Control variables include demographic characteristics, socioeconomic status,
lifestyle factors, mental health conditions, prescribed medicine use, and objective memory status.
Demographic characteristics included age (65-69, 70-74, 75-79, 80-84, 85+), gender (male vs.
female), race/ethnicity (non-Hispanic whites, non-Hispanic blacks, Hispanics, and other), and
marital status (married/living with a partner, separated/divorced, widowed, and never married).
Information on socioeconomic status consisted of education level (less than high school, high
school graduate, some college, and college or more) and annual household income. We use the
NHATS imputed household income variable (Montaquila et al., 2015). The income variable was
33
log-transformed due to high skewness and kurtosis. Lifestyle factors included smoking status and
BMI. Smoking status was classified as never smoker, former smoker, and current smoker. BMI
was calculated using self-reported weight and height and defined using standard categories
(Centers for Disease Control and Prevention, 2021b): underweight (BMI<18.5 kg/m2), normal
weight (18.5≤BMI<25 kg/m2), overweight (25≤BMI<30 kg/m2), or obese (BMI≥30 kg/m2).
More than half of the overall sample was classified as overweight (38.4%) or obese (25.5%).
Research shows that individuals’ mood influences self-rated memory (Yates et al., 2017).
To provide information on the respondents’ mood, we included depressive symptoms and
psychological well-being. Depression was examined by using the total score of the Patient
Health Questionnaire 2 (PHQ-2) (Kroenke et al., 2003). ). The questions were “Over the last
month, how often have you had little interest or pleasure in doing things?”, as well as “Over the
last month, how often have you felt down, depressed, or hopeless?” The answers were classified
into 0 (not at all) 1 (several days), 2 (more than half the days), and 3 (nearly every day). The
total scores from the two questions ranged from 0 to 6, with a higher score indicating more
severe depressive symptoms. To capture the broad psychological well-being of the respondents,
we used a psychological well-being scale. The answers were coded into 0 (agree a lot) to 2
(agree not at all) based on the five statements: I gave up trying to improve my life a long time
ago; I like my living situation very much; I feel confident and good about myself; My life has
meaning and purpose, and; I have an easy time adjusting to change. The total scores ranged from
0 to 10, with a higher score indicating better psychological well-being following previous
research (Sol et al., 2020).
We also included prescribed medicine use since it can be assumed that individuals with
chronic disease may take prescribed medicine to manage their health conditions. When chronic
34
diseases are untreated or undertreated, older adults may experience subjective cognitive decline
(Taylor et al., 2020). At the same time, when older adults take medication to manage chronic
disease, it could also have a side effect which involves memory problems. Thus, we included
information on prescribed medicine use and objective memory status. The use of prescribed
medicine was asked by: In the last month, did you take any medicines prescribed by a doctor?
The responses were 1 (yes) and 0 (no).
The objective memory status was based on immediate and delayed word recall tests. For
both tests, a list of 10 nouns was read to respondents and they were asked to recall as many
nouns as possible. There were approximately five minutes between when the immediate and
delayed tests were administered. The measure ranged from 0 to 20 points, with higher scores
indicating better objective memory status.
Analytical Strategy
We performed descriptive analyses for respondents with and without multimorbidity.
Group differences in study variables were examined by using t-tests for continuous variables and
chi-squared tests for categorical variables. In the main analyses, we ran logistic regression
models predicting fair/poor self-rated memory. In the first model, the key independent variable
of interest was multimorbidity. In the second model, the key independent variables of interest
were the eight individual chronic diseases. Both models adjusted for the full set of covariates,
including socio-demographic characteristics (race/ethnicity, age, gender, marital status,
education level, and annual household income), lifestyle factors (smoking status and BMI),
mental health conditions (psychological well-being and depressive symptoms), prescribed
medicine use (yes or no), and objective memory status (based on immediate and delayed recall).
35
All analyses were performed using STATA 14.2 (Stata Corp., College Station, TX) and were
weighted to account for complex survey design.
Results
Descriptive characteristics of the sample
Table 2-1 shows weighted descriptive characteristics of the overall sample. More than
half of respondents reported having multimorbidity (N = 4,380; weighted percentage = 64.2%).
Most (82.7%) of respondents were non-Hispanic whites, 7.9% were non-Hispanic blacks, 6.4%
were Hispanics, and 2.8% were from other racial/ethnic groups. Approximately half of the
respondents were married or living with a partner (58.7%), and roughly a quarter were widowed.
In the overall sample, 46.9% were never smokers, 44.2% were former smokers, and only 8.85%
were current smokers. The most common chronic conditions were hypertension (63.6%) and
arthritis (53.2%), followed by diabetes (23.4%) and osteoporosis (20.5%) in the overall sample.
The chronic condition with the lowest prevalence was stroke (9.0%). For the overall sample,
having two chronic conditions was the modal category and having five and more chronic
conditions was the least common (7.0%). Figure 1 shows the distribution of chronic conditions
stratified by those who report excellent/very good/good self-rated memory and those who report
fair/poor self-rated memory. Among those who report fair/poor self-rated memory, 54.1% have
three or more chronic diseases. In contrast, this figure is 36.4% among those reporting
excellent/very good/good self-rated memory.
Approximately 69.7% of female respondents had multimorbidity, while 30.3% of female
respondents had no multimorbidity in our sample. Approximately 72.8% of respondents with
multimorbidity had a less than high school education level compared to 27.15% of respondents
without multimorbidity. In addition, those with multimorbidity were more likely to be classified
36
as obese (75.2%) than those without multimorbidity (24.7%). In addition, respondents with
multimorbidity scored lower on psychological well-being (5.1% versus 5.3%) and higher on
depressive symptoms (1.0% versus 0.5%). The prevalence of prescribed medicine use was
substantially higher among respondents with multimorbidity (69.2%) than respondents without
multimorbidity (30.8%). A higher percentage of those with multimorbidity reported having
fair/poor self-rated memory than respondents without multimorbidity (75.19% versus 24.81%).
Consistent with the differences in self-rated memory, the objective memory status score was
lower among those with multimorbidity compared to those without multimorbidity. In summary,
respondents with multimorbidity were more likely to be non-Hispanic blacks or Hispanic, older,
female, widowed, and less educated. In addition, they were more likely to have worse
psychological well-being, more depressive symptoms, and worse objective memory status.
Nearly all of them were on prescription medication.
Further information on the prevalence of combinations of chronic diseases in the overall
sample is provided in Appendix D. Appendix D shows what the most common combinations of
chronic diseases are for the overall sample. For example, among respondents with heart disease,
6.6% had a heart attack, 12.9% had hypertension, 6% had diabetes, 2.6% had stroke, 3.9% had
osteoporosis, 10.6% had arthritis, and 4.4% had lung disease. Having both hypertension and
arthritis was the most common condition (37.0%), followed by hypertension and diabetes
(18.4%). The least common combinations were stroke and lung disease (2.1%) and stroke and
osteoporosis (2.4%). In addition, Appendix E describes the bivariate association between each
type of chronic disease and fair/poor self-rated memory. The result showed that there are
differences in each type of chronic disease and fair/poor self-rated memory in the study.
37
Association between multimorbidity and types of chronic diseases and fair/poor self-rated
memory
The first set of results examines the association between having any multimorbidity and
fair/poor self-rated memory. Table 2 presents the results from logistic regression models
predicting fair/poor self-rated memory status. In Model 1, our main predictor was multimorbidity
and the findings showed that having two or more chronic diseases was a significant predictor of
fair/poor self-rated memory compared to those with none or one chronic disease, controlling for
the full set of covariates (95% confidence interval [CI] = 1.11-1.65). Being non-Hispanic black
or Hispanic was associated with 1.35 (95% CI = 1.12-1.63) and 2.21 (1.70-2.88) times higher
odds of reporting fair/poor self-rated memory compared to non-Hispanic whites, respectively. In
addition, none of the lifestyle variables (smoking status and BMI) were significant predictors of
fair/poor self-rated memory. Among mental health conditions, having a higher score on
depressive symptoms was associated with 1.41 (95% CI = 1.35-1.47) times higher odds of
reporting fair/poor self-rated memory, controlling for other variables. In addition, scoring lower
on objective memory status was a significant contributor of fair/poor self-rated memory (95% CI
= 0.85-0.90).
It is also informative to investigate the relationship between the types of chronic diseases
and self-rated memory status in the logistic regression model with the full set of covariates. In
Model 2, we investigated the associations between types of chronic diseases and fair/poor self-
rated memory separately in the logistic regression model—three out of eight chronic diseases
were significant predictors of fair/poor self-rated memory (i.e., stroke, osteoporosis, and
arthritis). For example, respondents with stroke had 1.41 times (95% CI = 1.12-1.76) higher odds
of reporting fair/poor self-rated memory. Respondents with osteoporosis or arthritis had 20%
38
(95% CI = 1.00-1.44) and 30% (95% CI = 1.19-1.60) higher odds of reporting fair/poor self-
rated memory, respectively. The findings for other predictors were very similar to those
described in Model 1. Based on both models, our findings suggested that being non-Hispanic
black, Hispanic, older age, female, having less than a high school education level, having low
annual household income, having higher levels of depressive symptoms, and having low
objective memory score were linked to significantly greater odds of reporting fair/poor self-rated
memory.
Discussion
As the prevalence and incidence of chronic diseases and cognitive impairment in aging
societies are projected to increase, the burden on public health with regards to medical costs and
disease management will inevitably increase. For example, the prevalence of poor self-rated
memory ranges from 25% to 50% among older adults and tends to increase with age (Jonker et
al., 2000). In addition, it is common for older adults aged 65 and older to have multimorbidity
(National Council on Aging, 2018). Thus, self-rated memory and chronic diseases are important
factors, both of which undermine older adults’ independence, mental health, and quality of life.
Moreover, having chronic diseases as well as poor self-rated memory may be a particularly
adverse combination. For example, having multimorbidity may require individuals to take
multiple medications for chronic disease management. Having poor self-rated memory may
make it more difficult to manage a complicated medication regimen. Also, older adults with poor
self-rated memory may be more likely to develop dementia in the future compared to older
adults without poor self-rated memory (Mitchell et al., 2014), which may worsen older adults’
quality of life and independence. Although the importance of both chronic diseases and self-rated
memory is well-documented, there are few studies on the association between chronic diseases
39
and self-rated memory. Thus, this study expanded the scope of the literature by investigating the
association of fair/poor self-rated memory with two important dimensions of chronic disease:
multimorbidity and types of diseases.
First, we found that respondents with multimorbidity had higher odds of reporting
fair/poor self-rated memory compared to those without multimorbidity, controlling for the full
set of covariates (demographic characteristics, socioeconomic status, lifestyle factors, mental
health conditions, prescribed medicine use, and objective memory status). This result is
consistent with a prior study conducted on the population aged 16 and older, which suggested
that there is a positive association between multimorbidity and subjective memory concerns
(Jacob et al., 2019). Our study contributes to literature by expanding the previous study with the
use of a nationally representative sample and focusing on older adults. Being non-Hispanic
blacks, Hispanics, older, and female were associated with significantly higher probabilities of
reporting fair/poor self-rated memory. Individuals with less than high school education, low
annual household income, more depressive symptoms, and low objective memory status were
significantly more likely to report fair/poor self-rated memory.
In addition, we further performed logistic regression analysis on the association between
types of chronic diseases and fair/poor self-rated memory, adjusted for our covariates. Among
the eight chronic diseases, three chronic diseases remained significant (i.e., stroke, osteoporosis,
and arthritis). Stroke was the chronic condition associated with the greatest elevation in the odds
of reporting fair/poor self-rated memory. Consistent with previous findings (Jacob et al., 2019;
Taylor et al., 2020), we were able to identify that specific chronic diseases are more likely to be
associated with fair/poor self-rated memory than those without that chronic disease. Our findings
also demonstrated that older adults with stroke, osteoporosis, or arthritis had a greater prevalence
40
of fair/poor self-rated memory than those without these chronic diseases. There are several
potential mechanisms on how specific diseases influence cognitive impairment. For example,
stroke may cause neuroanatomical lesions in areas of the brain including the hippocampus and
the white matter lesions, as well as cerebral microbleeds (Al-Qazzaz et al., 2014; Sun et al.,
2014). These in turn may contribute to the development of cognitive impairment following a
stroke. Low bone mineral density, which has been found to be highly associated with cognitive
impairment, may explain the relationship between osteoporosis or arthritis and cognitive
impairment (Kang et al., 2018). It is possible that bone mass loss increases inflammatory
markers (e.g., interleukin-6), which in turn increase the risk of developing Alzheimer’s disease
(Ershler, 1993; Lui et al., 2003).
A key strength of this study is the use of a nationally representative sample of older
adults that included information on self-rated memory status as well as objective memory status.
Moreover, our study included all chronic diseases that were asked about by the survey, which are
common among the older population. Nonetheless, there are limitations to this study. First,
whether a respondent had a chronic disease was ascertained through self-reports. In addition,
respondents were asked only whether they had a chronic disease, lacking information on time
since diagnosis, which is associated with severity. The severity of the chronic disease may be
linked to self-rated memory since greater severity has been found to be associated with a lower
quality of life and worse physical and mental health (Pizzi et al., 2006). In addition, individuals’
mood is highly associated with self-rated memory among older adults (Yates et al., 2017).
Although the survey did not ask about mood specifically, we did include measures of
psychological well-being and depressive symptoms, which have been found to be highly
associated with individuals’ mood in prior studies (Polak et al., 2015). Moreover, self-rated
41
memory was not asked about to older adults who were institutionalized, therefore, our findings
are restricted to older adults who are living in community settings. Lastly, it is unclear whether
the presence of certain chronic diseases may invalidate how respondents rate their memory even
if we included objective memory status in the analysis.
Despite these limitations, this study contributes to the literature by expanding two
measurements of chronic disease on self-rated memory through the use of a nationally
representative sample of older adults. The study’s findings highlight the importance of the
prevention of chronic diseases and how multimorbidity negatively impacts respondents’ self-
rated memory, which may be linked to cognitive impairment in the long term. Considering
projections for both the absolute number of older adults with cognitive impairment and with
chronic disease are increasing, it is important that health professionals recognize that poor self-
rated memory is highly prevalent among older adults with multimorbidity. Not only health
professionals but also patients need to understand the impact of multimorbidity on self-rated
memory and factor the relationship into disease management and treatment. Our results highlight
the importance of monitoring self-rated memory among older adults with stroke, osteoporosis,
and arthritis in clinical settings. A prior study has found that older adults are less likely to initiate
discussions about memory problems with their health care providers even when they believe
their memory is fair/poor (Waldorff et al., 2008). Our findings demonstrate that older adults with
stroke, osteoporosis, and arthritis are more likely to report fair/poor self-rated memory than those
without these diseases. Thus, clinicians who treat patients with these chronic conditions should
pay close attention to patients’ self-rated memory status.
There are mixed trends in lifestyle factors in the U.S. First, the increasing prevalence of
obesity may lead to increased risk of developing chronic diseases among older adults in the
42
future. This has an important implication, since this may lead to elevated prevalence of fair/poor
self-rated memory and lead to cognitive decline among older adults. At the same time, smoking
trends are declining, which may potentially lead to reduced prevalence of chronic diseases. It
would be interesting to witness how the future trends of multimorbidity play out among older
adults and how they affect the individuals’ self-rated memory based on these mixed trends in
lifestyle factors.
Considering this study was not able to include severity of chronic disease, future work is
needed on how this factor influences self-rated memory. Does severity of chronic disease
worsen self-rated memory among older adults? Given that disease management and treatments
are complex as chronic disease, more studies are needed to explore whether severity of disease
influences self-rated memory.
43
Table 2-1. Descriptive statistics by multimorbidity, NHATS 2011
Overall
sample
(N=6,481)
% or Mean
(SE)
No
multimorbidity
f
(N=2,101)
% or Mean (SE)
Multimorbidity
g
(N=4,380)
% or Mean (SE)
X
2
/t test
Self-rated memory
(fair/poor)
16.80
24.81 75.19
58.84***
Race/ethnicity 8.21**
Non-Hispanic
whites
82.71
36.61 63.39
Non-Hispanic
blacks
7.99
26.67 73.33
Hispanics 6.48 34.11 65.89
Other 2.82 39.02 60.98
Age 31.51***
65-69 29.35 45.19 54.81
70-74 25.97 35.65 64.35
75-79 19.32 30.74 69.26
80-84 14.36 27.98 72.02
85+ 11.00 29.47 70.53
Gender
Female 55.87 30.30 69.70 73.34***
Marital status 22.79***
Married/living
with a partner
58.72
40.09 59.91
Separated/divorced 12.34 33.33 66.67
Widowed 25.72 26.40 73.60
Never-married 3.22 39.57 60.43
Education level 20.79***
Less than high
school
19.78
27.15 72.85
High school
graduate
27.47
34.26 65.74
Some college 27.25 35.31 64.69
College or more 25.50 44.38 55.62
Annual household
income
a
10.33 (0.02)
10.50 (0.04) 10.24 (0.02)
28.87***
Smoking status 1.65
Never smoker 46.95 37.12 62.88
44
Former smoker 44.21 34.19 65.81
Current smoker 8.85 35.92 64.08
BMI 35.92***
Underweight 1.79 44.43 55.57
Normal 31.25 41.01 58.99
Overweight 38.40 39.15 60.85
Obese 28.56 24.77 75.23
Psychological well-
being
b
5.23 (0.01)
5.33 (0.02) 5.18 (0.01)
25.61***
Depressive symptoms
c
0.89 (0.02)
0.57 (0.02) 1.06 (0.02)
207.72***
Prescribed medicine
use
d
90.71
30.80 69.20
447.05***
Objective memory
status
e
8.43 (0.06)
8.89 (0.08) 8.17 (0.06)
79.25***
Types of chronic
diseases
__ __ __
Heart disease 16.96
Heart attack 13.76
Hypertension 63.69
Diabetes 23.45
Stroke 9.06
Osteoporosis 20.51
Arthritis 53.23
Lung disease 15.58
Number of chronic
diseases
__ __ __
0 or 1 35.72
2 27.33
3 19.76
4 10.13
5 or more 7.06
Note: All estimates are weighted to account for complex survey design.
Abbreviation: NHATS, National Health and Aging Trends Study; SE, standard error.
*p<0.05; **p<0.01; ***p<0.001
a
Annual household income was log-transformed.
b
Psychological well-being values ranged from 0 to 10, with a higher score corresponding to
greater psychological well-being
c
Depressive symptoms was evaluated by the Patient Health Questionnaire-2 and values ranged
from 0 to 6, with a higher score indicating more severe depressive symptoms.
d
The use of prescribed medicine was asked by: In the last month, did you take any medicines
prescribed by a doctor? The responses were 1 (yes) and 0 (no).
45
e
Objective memory status was based on immediate and delayed recall. Values ranged from 0 to
20, with a higher score indicating better memory status.
f
No multimorbidity means the individual has 0 or 1 chronic condition
g
Multimorbidity means that the individual has 2+ chronic conditions
46
Table 2-2. Odds Ratios from Logistic Regression Models for the Association Between
Multimorbidity and Types of Chronic Diseases and Fair/Poor Self-rated Memory
Variables Model 1
OR (95% CI)
Model 2
OR (95% CI)
Multimorbidity (Ref: None or one)
Two or more 1.35 (1.11-1.65)** ⸺
Types of chronic diseases
Heart disease ⸺ 1.13 (9.12-1.40)
Heart attack ⸺ 1.01 (0.80-1.25)
Hypertension ⸺ 0.97 (0.81-1.17)
Diabetes ⸺ 1.04 (0.88-1.24)
Stroke ⸺ 1.41 (1.12-1.76)**
Osteoporosis ⸺ 1.20 (1.00-1.44)*
Arthritis ⸺ 1.30 (1.19-1.60)***
Lung disease ⸺ 1.18 (0.96-1.45)
Race/ethnicity (Ref: Non-Hispanic whites)
Non-Hispanic blacks 1.35 (1.12-1.63)** 1.42 (1.16-1.72)**
Hispanics 2.21 (1.70-2.88)*** 2.29 (1.74-3.02)***
Other 1.19 (0.73-1.92) 1.20 (0.76-1.90)
Age (Ref: 65-69)
70-74 1.15 (0.86-1.53) 1.15 (0.86-1.54)
75-79 1.10 (0.83-1.46) 1.09 (0.82-1.45)
80-84 1.32 (1.01-1.73)* 1.30 (0.99-1.72)
85+ 1.47 (1.08-2.01)* 1.45 (1.06-1.97)*
Gender (Ref: male)
Female 0.79 (0.67-0.93)** 0.69 (1.57-1.83)***
Marital status (Ref: married/living with a
partner
Separated/divorced 0.96 (0.75-1.23) 0.94 (0.74-1.21)
Widowed 0.88 (0.72-1.07) 0.86 (0.71-1.05)
Never-married 0.79 (0.48-1.30) 0.79 (0.49-1.29)
Education level (Ref: less than high
school)
High school graduate 1.79 (1.39-2.30)*** 1.74 (1.35-2.25)***
Some college 1.28 (0.99-1.67) 1.29 (0.99-1.68)
College or more 1.15 (0.88-1.50) 1.14 (0.87-1.50)
Annual household income
a
0.91 (0.86-0.97)** 0.92 (0.87-0.97)**
Smoking status (Ref: never smoker)
Former smoker 0.91 (0.78-1.06) 0.89 (0.76-1.04)
Current smoker 0.80 (0.60-1.07) 0.77 (0.58-1.04)
BMI (Ref: normal)
Underweight 1.19 (0.83-1.69) 1.19 (0.83-1.70)
Overweight 0.87 (0.72-1.06) 0.86 (0.71-1.06)
Obese 0.97 (0.80-1.17) 0.96 (0.79-1.17)
Psychological well-being
b
0.94 (0.86-1.02) 0.94 (0.86-1.02)
Depressive symptoms
c
1.41 (1.35-1.47)*** 1.38 (1.32-1.44)***
47
Prescribed medicine use 1.08 (0.75-1.57) 1.10 (0.75-1.60)
Objective memory status
d
0.88 (0.85-0.90)*** 0.88 (0.85-0.90)***
N 6,481 6,481
Note: All estimates are weighted to account for complex survey design.
Abbreviation: OR, odds ratio; CI, confidence interval.
*p<0.05; **p<0.01; ***p<0.001
a
Annual household income was log-transformed.
b
Psychological well-being values ranged from 0 to 10, with a higher score corresponding to
greater psychological well-being
c
Depressive symptoms was evaluated by the Patient Health Questionnaire-2 and values ranged
from 0 to 6, with a higher score indicating more severe depressive symptoms.
d
The use of prescribed medicine was asked by: In the last month, did you take any medicines
prescribed by a doctor? The responses were 1 (yes) and 0 (no).
e
Objective memory status was based on immediate and delayed recall. Values ranged from 0 to
20, with a higher score indicating better memory status.
48
Chapter 4: Gender Differences in Survival with Cognitive Impairment and Dementia
Introduction
Life expectancy is an important summary measure of population health. In 2018, the life
expectancy at birth was 76.2 and 81.2 for males and female in the U.S. (Arias & Xu, 2020b). In
recent decades, the number of deaths caused by Alzheimer’s disease, which is the most common
cause of dementia, has increased and it is the 6
th
leading cause of death in the U.S. (Alzheimer’s
Association, 2021). Dementia is one type of brain disorder and it substantially reduces one’s life
span (Todd et al., 2013). The deaths contributed by dementia have increased 123% between 2000
and 2015 (Alzheimer’s Association, 2021). Although recent studies suggested a reduction in the
prevalence of dementia among older adults (Crimmins et al., 2011; Langa et al., 2017), the
projection of the U.S. population shows that the absolute number of the older population with
dementia will continue to increase, while deaths caused by other causes will decline (Prince et
al., 2013). Understanding mortality risk after developing dementia and cognitively impaired but
not dementia (CIND) is important for patients, caregivers, and public health (Mayeda et al.,
2017). Gender differences in mortality risk of dementia and CIND are plausible, considering
differences in life expectancy, sociodemographic characteristics, health conditions, and lifestyle
factors (Arias & Xu, 2020; Alzheimer’s Association, 2021; Carr & Bodnar-Deren, 2009;
Clarkson et al., 2019; Neviani et al., 2017). Thus, the burden of dementia is expected to increase,
making it critical to understand how dementia mortality differs by gender.
Literature Review
Sociodemographic Factors and Mortality Risk by Gender
Understanding gender differences in mortality risk associated with cognitive impairment
and dementia is important not only for patients, but also for families and caregivers so that
49
individuals can better allocate their financial resources. There are various factors that may
contribute to different mortality risks of dementia by gender. As mentioned above, there are
differences in life expectancy by gender. Also, two-third of dementia cases are women in the
U.S. (Alzheimer’s Association, 2021), due to their greater longevity and the increasing risk of
developing dementia with age (Mielke et al., 2014). Thus, it is important to investigate how
survival with dementia may differ by gender, and which factors influence survival among men
and women with impaired cognitive functioning. There are various factors that may contribute to
the differences in survival for individuals with cognitive impairment or dementia across genders.
First, the incidence of dementia increases with age, and age is the strongest risk factor for
developing dementia (Alzheimer’s Association, 2021). The risk of developing dementia doubles
every five years after age 65. Women tend to have dementia at later ages due to a longer life
expectancy than men (Beam et al., 2018). Also, men were more likely to have shorter survival
with dementia than their women counterparts (Dal Forno et al., 2002). In line with this, it is
possible that women, compared to men, are more likely to live for a longer amount of time with
cognitive impairment.
Not only life expectancy, but also the characteristics of marital status may play a role in
differences in mortality of dementia or cognitive impairment by gender. For example, a high
proportion of older women tend to live alone compared to their men counterparts. In developed
countries, older women are more likely to live alone or live in an institution (e.g., nursing homes,
assisted living facilities, etc.), while older men tend to live in married couple households (Carr &
Bodnar-Deren, 2009). Unmarried older adults tend to have a higher mortality risk than their
married counterparts and there were also gender differences in the association (Goldman et al.,
1995).
50
When we look into education level and dementia, it is well-established that a low level of
education is highly associated with cognitive impairment and dementia in later years (Crimmins
& Saito, 2001; Gatz et al., 2001). At the same time, there was an education gap by gender.
Although the gap has decreased in recent years, it remains sizeable for the older adult population.
For instance, approximately 33% of older men (aged 65 and older) and 24% of older women
counterpart had completed college or more education in 2019 (U.S. Census Bureau, 2019). In
addition, several studies have shown that higher education is associated with lower mortality
(Hendi, 2015, 2017; Rogers et al., 2010). Given educational differences between men and
women, it is possible that mortality risk may vary by education level and gender.
Lifestyle Factors and Mortality Risk by Gender
In addition to sociodemographic factors, health behaviors are closely linked to cognitive
health, and positive health behaviors are associated with lower incidence of chronic conditions
and better cognitive health. For example, individuals who exercise frequently are less likely to
develop chronic disease, disability, and even cognitive impairment at older ages (Clarkson et al.,
2019; Neviani et al., 2017). Older women tend to exercise less frequently than older men in later
life (M. S. Kaplan et al., 2001). Smoking status and alcohol consumption are also highly
associated with cognitive impairment and dementia (Anstey et al., 2007; Luchsinger et al., 2004).
Smoking is the main contributor to preventable morbidity and mortality in the U.S. (CDC, 2008),
and men are more likely to be current smokers than women (Cornelius et al., 2020). In line with
this, a higher percentage of men are drinkers, drink more frequently, and more heavily compared
to women (White et al., 2015).
Chronic Diseases and Mortality Risk by Gender
51
In addition, there are also differences in the prevalence of chronic diseases and lifestyle
factors by gender, which may be associated with mortality risk. Chronic diseases are considered
another important component of health and well-being at older ages, and they negatively impact
cognition with age. Chronic diseases may limit older adults’ activities of daily living, reduce
quality of life, and increase the cost of health care (CDC, 2020). Considering the importance of
chronic diseases, there is a discrepancy in the prevalence of chronic diseases between men and
women. For example, older women have a higher prevalence of hypertension and stroke than
older men (Martins et al., 2001; Reeves et al., 2008); while among 70-year-old older men had
higher prevalence of diabetes than older women counterparts in a population-based prospective
study (Nordström et al., 2016). Therefore, chronic diseases may have different impacts on
cognitive impairment and dementia and even mortality risk for men and women.
Purpose of the Study
The burden of cognitive impairment and dementia is increasing in the U.S., and it is
critical to understand the mortality risk of individuals with cognitive impairment or dementia
since these conditions can lead to a loss of independent living, increase the cost of healthcare,
and place high burdens on public health. As mentioned above, the burden of dementia and CIND
is expected to increase in the future. Knowledge of differences in the mortality risk associated
with cognitive impairment or dementia by gender could be informative for family members,
patients, and caregivers so that they can to allocate their resources accurately. This study uses a
nationally representative sample of community-dwelling population to investigate the differences
in mortality risk by gender and gender differences in cognitive impairment or dementia. The
research questions of this project are: “How do cognitive impairment or dementia affect
mortality for men and women?” and “Are there gender differences in survival with CIND or
52
dementia?” With the first research question, we investigate how these two cognitive statuses
impact mortality risk on women, as well as the impact on men. Meanwhile, the second question
directly compared the two genders to investigate how mortality risk impacts women differently
from men. It also aims to identify potential factors that contribute to gender differences.
Method
Data
The data from the present study is retrieved from the Health and Retirement Study
(HRS). The HRS is a nationally representative longitudinal panel study of older adults in the
U.S. The HRS was established to provide detailed information on the demographics,
socioeconomic status, and health status of individuals who were 50 years of age or older. The
survey has been conducted every 2 years beginning in 1992, and the response rate at every
survey has exceeded 86%. Following numerous prior studies on cognition, we decided to use the
HRS 2000 since cognitive measurements were updated in the survey from 2000 (Langa et al.,
2017; Suemoto et al., 2015). Thus, we included respondents who were interviewed between
January and December 2000 and these respondents were followed up for 16 years. We have two
analytic samples. First, we examined respondents who were observed for at least 10 years or die
within 10 years to determine the probability of dying within 10 years. This sample consisted of
11,387 respondents. For the second analytic sample, we did not impose this restriction and
assessed mortality over 16 years. This sample consisted of 18,913 respondents. Among the
18,913 respondents, 7,351 (38.8%) were documented to be alive at the end of the follow-up
period; 8,934 (47.2%) had died; 2,628 (13.8%) were lost to follow-up and censored at their last
interview date.
53
Cognitive status was evaluated using a modified version of the Telephone Interview for
Cognitive Status (TICS-M) (Brandt et al., 1988). The TICS-M shows high sensitivity and
accuracy for cognition in community-dwelling samples. This study used the 27-point scale of
cognition following previous studies (Crimmins et al., 2011). The scale ranges from 0 to 27 and
includes immediate and delayed word recall, serial 7 subtractions, and backwards counting.
Respondents who scored from 0 to 6 points were classified as having dementia; 7 to 11 as
cognitively impaired with no dementia (CIND), and 12 to 27 as normal at baseline.
Control variables include demographic characteristics, socioeconomic status, health
conditions, and lifestyle factors at baseline. Demographic characteristics included age at baseline
(50-54, 55-59, 60-64, 65-69, 70-74, 75-79, 80-84, 85+), race/ethnicity (non-Hispanic white, non-
Hispanic black, and other race/ethnicity), and marital status (married/living with a partner,
separated/divorced/widowed, and never married). Information on socioeconomic status consisted
of education level (less than high school, high school graduate, some college, and college or
more). Respondents were asked whether a doctor had ever told them they had diabetes,
hypertension, or stroke. The lifestyle factors included exercise, smoking status, and alcohol
intake. For exercise, respondents were asked, “On average over the last 12 months have you
participated in vigorous physical activity or exercise three times a week or more? By vigorous
physical activity, we mean things like sports, heavy housework, or a job that involves physical
labor”. The answers were classified into 1 = yes and 0 = no. For smoking, respondents were
classified as never or former/current smokers. Lastly, individuals were asked about any alcohol
consumption using the question, “Do you ever drink any alcoholic beverages such as beer, wine,
or liquor?” The response categories were 1 (yes) or 0 (no).
Methods
54
First, we investigated the probability of dying within 10 years of baseline by age group,
gender, and cognitive status. Second, we examined the Kaplan-Meier curves for all the
categorical predictors and performed the tests of equality across strata to explore whether the
predictors were meaningful. Using Cox proportional hazards regressions, we first estimated
hazard ratios (HRs) to show examine whether cognitive impairment and dementia affect
mortality by gender. We model the predictors in three steps. Model 1 tests whether cognitive
status is linked to mortality risk when controlling only for age at baseline and race/ethnicity.
Model 2 controls for marital status and education level, and Model 3 additionally controls for
lifestyle factors (exercise, smoking status, and alcohol consumption) and health conditions
(hypertension, diabetes, and stroke). Lastly, we used Cox proportional hazards regression to
estimate the hazard ratios to investigate whether there are gender differences in survival with
CIND or dementia. All analyses were performed using STATA 14.2 (Stata Corp., College
Station, TX) and weighted to account for complex survey design.
Results
Table 3-1 describes the probability of dying within 10 years by age group, gender, and
cognitive status (N = 11,387). We were able to identify that the probability of dying is higher for
the respondents with CIND or dementia groups than the respondents with normal cognition for
both genders. For example, the percentages of women respondents with CIND and dementia who
died within 10 years were 89.6% and 91.8%, respectively. In contrast, the percentage of women
respondents aged 85+ with normal cognition who died within 10 years was 76.6%. Similar
patterns were also observed for men. The percentages of men respondents aged 85+ with CIND
and dementia who died within 10 years were 96.2% and 99.2%, respectively, compared to 84.5%
for men respondents with normal cognition. The probability of dying increases with age for both
55
genders across cognitive states. The percentage of women respondents with normal cognition
who died within 10 years was 7.3% among those aged 50-54 at baseline compared to 76.6%
among those aged 85+ at baseline. Among men, these figures were 6.1% and 84.5%.
The baseline sample characteristics are shown in Table 3-2 (N = 18,913). Among both
men and women, a higher proportion of those who died had dementia compared to those who
were alive or censored (women: 10.8% vs. 1.6%; men: 9.6% vs. 1.4%). More than 80% of
respondents were non-Hispanic white, followed by non-Hispanic black and other race/ethnicity
for both genders. Approximately 81.2% of the alive or censored men respondents were
married/living with a partner, while 39.5% of the alive or censored women group was
married/living with a partner. For all health conditions, respondents who died had a higher
proportion of having chronic health conditions (i.e., hypertension, diabetes, and stroke) at
baseline compared to those in the alive or censored groups. For lifestyle factors, a higher
proportion of the respondents in the alive or censored groups exercised compared to those who
died for both genders (women: 29.2% vs. 45.9%; men: 39.4% vs. 57.7%). Lastly, approximately
16.5% and 19.2% of respondents who died were former/current smokers among women and
men, respectively. We used Kaplan-Meier survival curves to compare differences in survival by
gender and cognitive status (Appendix F). In each cognitive status category, women have better
survival than men counterparts.
To assess within-gender differences, we estimated Cox proportional hazards models for
men and women (Table 3-3). The hazard ratio for cognitive status gradually decreased for both
genders when socio-demographic characteristics, health conditions, and lifestyle factors were
introduced, but remained significant throughout the models. Starting with women, Model 1
shows that respondents with CIND and dementia had higher risks of dying than respondents with
56
normal cognition, adjusting for age and race/ethnicity. In fact, respondents with dementia had a
125% (95% CI 1.99-2.54) higher risk of dying than respondents with normal cognition, adjusting
for age and race/ethnicity. Model 2 adds marital status and education level to Model 1 to assess
the differences in mortality risks due to socioeconomic status. The hazard ratios for cognitive
status are somewhat smaller than in Model 1 but remain significant. Respondents with less than
high school or high school graduate education had higher risks of dying than those with a college
or more degree; these were 27% (95% 1.11-1.45) and 23% (95% CI 1.09-1.38) higher,
respectively. Model 3 adds health conditions and lifestyle factors. Those with CIND or dementia
continued to have significantly higher risks of dying than those with normal cognition; these
were 40% (95% CI 1.29-1.52) and 91% (95% CI 1.65-2.22), respectively. Respondents aged 75
and older at baseline had a significantly higher risk of dying than respondents aged 50-54 at
baseline. Among women, education was no longer a significant predictor of mortality risk after
controlling for health conditions and lifestyle factors. Furthermore, respondents with
hypertension, diabetes, or stroke had a 25% (95% CI 1.14-1.36), 80% (95% CI 1.59-2-.04), and
51% (95% CI 1.33-1.72) higher risks of dying than respondents without those health conditions,
respectively. Women who were former/current smokers had a 132% (95% CI 2.04-2.63) higher
risk of dying compared to women who were never smokers.
Turning our attention from women to men respondents, we were able to identify the
differences in mortality risk by characteristics. Among men, Model 1 shows that respondents
who have CIND or dementia have higher risks of dying compared to respondents with normal
cognition controlling for age and race/ethnicity, 61% (95% CI 1.48-1.74) and 176% (2.41-3.16),
respectively. When we add marital status and education level in Model 2, the results for
cognitive status remained significant and fairly similar to Model 1. Respondents who were
57
separated/divorced/widowed, or never married had 27% (95% CI 1.15-1.41) and 33% (95% CI
1.05-1.69) higher risks of mortality compared to those who were married/living with a partner.
Also, men in all three lower education categories had significantly higher risks of dying than
men with a college or more education level. Model 3 adds health conditions and lifestyle factors.
Respondents with CIND had a 32% (95% CI 1.22-1.43) higher mortality risk compared to
respondents with normal cognition, while those with dementia had 105% (95% CI 1.75-2.40)
higher mortality risks compared to their counterparts, controlling for the full set of covariates. In
addition, respondents with hypertension, diabetes, or stroke had a 12% (95% CI 1.03-1.23), 54%
(95% CI 1.40-1.69), and 44% (95% CI 1.24) higher risks of dying than respondents without
those health conditions, respectively. Furthermore, respondents who reported consuming any
alcohol had a 14% (95% CI 0.80-0.93) lower risk of dying compared to those reported no alcohol
consumption. Lastly, men who were former/current smokers had a 86% higher risk of dying than
never smokers (95% CI 1.71-2.02).
To assess gender differences in survival with CIND or dementia, we carried out a Cox
proportional hazards model. Table 3-4 demonstrated the estimates of the hazard risk among the
respondents with CIND or dementia at baseline. women with dementia had a lower hazard risk
of death than the men with dementia. The estimated hazard ratio was 0.71 indicating that women
with dementia had a 29% lower hazard risk of death than the men respondents with CIND or
dementia (95% CI 0.66-0.76). Also, being other race/ethnicity had a 19% (95% 0.69-0.94) lower
hazard risk of death compared to non-Hispanic whites. When we add marital status and
education level in Model 2, the coefficient for women stayed the similar as Model 1. Being
women with dementia had a 31% (95% CI 0.63-0.75) lower hazard risk of death than the men
respondents with CIND or dementia. In the final model, the magnitude of coefficient for women
58
has reduced by it remained significant (95% CI 0.61-0.75). In terms of health conditions, the
respondents with hypertension or stroke showed 44% (95% CI 1.32-1.57) and 47% (95% CI
1.29-1.66) higher hazard risks than the respondents without those health conditions.
Furthermore, respondents who reported consuming any alcohol had a 22% (95% CI 0.70-0.86)
lower risk of dying compared to those reported no alcohol consumption. Lastly, former/current
smokers had a 84% higher risk of dying than never smokers (95% CI 1.61-2.11).
Discussion
The burden of cognitive impairment and dementia among older adults is increasing and it
is important to understand how mortality risk differs with cognitive impairment and dementia by
gender. More women are living with dementia than men because women tend to have a longer
life expectancy than men. However, it is still unknown how the mortality risk differs by gender
after individuals develop dementia or CIND. Understanding the mortality risk by gender could
be beneficial to family members, caregivers, and individuals with CIND or dementia so that they
can allocate their resources before entering severe stages of the disease. This study contributes to
the existing literature by investigating difference in mortality risk by gender using a nationally
representative sample.
First, our study established that individuals who have CIND or dementia had higher a
probability of dying within 10 years of baseline compared to the respondents with normal
cognition for both men and women. Also, the probability of dying within 10 years gradually
increased with age for both men and women across cognitive status. It is not surprising that the
probability of dying increases for the respondents with CIND and dementia. As dementia is a
progressive disease, the symptoms gradually worsen over time and eventually lead to death.
Thus, it is likely that individuals with cognitive impairment and dementia will have a shorter life
59
span than individuals with normal cognition. In line with this, a previous study found that years
of life lost were approximately 9.3 years for dementia and 5.2 years for mild cognitive
impairment (Strand et al., 2018). Our study also demonstrated that there is higher probability of
dying within 10 years for CIND and with dementia groups than normal in both genders across
age groups. This is consistent with a previous study that mortality risk was higher for individuals
with cognitive impairment or dementia than individuals with normal cognitive status (Wang et
al., 2020).
Second, this study demonstrated that having CIND and dementia increases mortality risk
for both men and women, even after adjusting for the full set of covariates. Compared with
women respondents with normal cognition, those with CIND had a 40% higher mortality risk,
whereas those with dementia had a 91% higher mortality risk. Similarly, men respondents with
CIND or dementia had 32% and 105% higher mortality risks than men respondents with normal
cognition, respectively. Thus, we were able to identify a similar pattern of mortality risk by
gender adjusting for the covariates. Our finding suggest that health conditions and lifestyle
factors were strong predictors of mortality risk in both groups. For example, women respondents
with diabetes had an 80% higher mortality risk than their counterparts. Meanwhile, the men
respondents with diabetes had a 54% higher mortality risk than men respondents without
diabetes. This is consistent with prior studies that diabetes is associated with increased mortality
risk (Bertoni et al., 2002; Li et al., 2019).
Third, our study demonstrated that there was gender differences in survival with CIND or
dementia. Although the magnitude of gender reduced throughout the models, it remained
significant. Compared to men with CIND or dementia, women with CIND or dementia showed a
32% lower mortality risk adjusting for the covariates. In addition, all health conditions and
60
lifestyle factors were significant predictors of mortality in our sample. For example, smoking
status was highly associated with mortality risk. As noted earlier, smoking is the leading cause of
preventable mortality in the U.S. (CDC, 2008). Continued efforts to reduce smoking may
contribute to reductions in both cognitive impairment and mortality. On the other hand, alcohol
consumption was associated with a lower risk of mortality in our study. Similar to our study
findings, prior study suggested that moderate alcohol consumption reduced overall mortality risk
for both middle-aged and older adult population (Thun et al., 1997).
Our study has several limitations. First, the assessment of cognitive status was not based
on a clinical neuropsychological evaluation of respondents. However, a recent study that
compared the HRS and Medicare claims data from 2000 to 2008 demonstrated that these two
data sources showed similar dementia prevalence based on cognition classification of the HRS
and dementia diagnosis from the Medicare claims data (Chen et al., 2019). Second, we were not
able to identify the dementia stage. Given that dementia is a progressive disease, the symptoms
of dementia could differ considerably depending on the stage. It is possible that the respondents
with late dementia may have a higher mortality risk compared to respondents with early
dementia. Third, without information on neuroimaging, our study could not investigate dementia
subtypes. Future research with information on neuropathology with dementia subtypes is needed.
Lastly, we did not have clinical diagnosis of chronic conditions, rather they were based on self-
reports. Although we do not have information based on clinical diagnosis, the reliability of self-
reporting of chronic disease has been well established (Najafi et al., 2019).
Despite these limitations, our study contributes to the literature on survival of cognitive
impairment and dementia by gender. Our study shows how mortality risks differ based on
cognitive status and gender. In addition, although there are many risk factors that increase
61
mortality among older adults, our study suggests that health conditions and lifestyle factors seem
to be the strongest predictors of mortality for both genders. Interventions targeted at improving
the management of chronic conditions, preventing chronic conditions, and promoting healthy
behaviors should be considered as they may be strongly associated with mortality in the aging
population. Also, not limited to men and women, there are various ways that individuals label
their genders these days (e.g., transgender, gender neutral, etc.). It is necessary to explore how
other labels of genders with cognitive impairment or dementia contribute to mortality risk.
62
Table 3-1. The probability of dying within 10 years of baseline by age group and gender, HRS
2000
Women (%) Men (%)
Baseline
age group
Normal CIND With
dementia
Normal CIND With
dementia
50-54 7.32 8.15 24.42 6.10 22.14 0.00
55-59 7.90 17.43 14.19 12.65 19.84 47.22
60-64 10.61 23.65 39.13 17.37 27.24 38.76
65-69 16.75 34.64 52.24 23.68 39.53 72.04
70-74 24.99 43.73 63.77 33.54 54.14 74.49
75-79 38.59 56.68 69.74 51.35 60.87 90.51
80-84 58.94 67.38 87.74 64.39 83.34 90.09
85+ 76.68 89.61 91.81 84.59 96.21 99.29
Total 20.32 48.38 71.78 24.05 50.87 76.87
63
Table 3-2. Characteristics of the sample, HRS 2000
Women Men
% Dead Alive or Censored Dead Alive or Censored
Cognitive status
Normal 65.86 90.14 66.53 90.07
CIND 23.30 8.55 23.83 8.88
With dementia
10.84 1.30 9.64 1.05
Age group
50-54 3.44 18.24 4.04 21.21
55-59
7.74 27.42 11.11 32.62
60-64
8.69 19.95 12.62 21.09
65-69 12.54 15.62 15.01 14.67
70-74
17.43 10.73 18.69 6.98
75-79
20.01 5.39 18.33 3.07
80-84 16.62 1.99 12.43 0.29
85+
13.54 0.66 7.76 0.07
Race/Ethnicity
Non-Hispanic white 86.55 86.15 87.57 88.21
Non-Hispanic black
10.91 9.39 9.41 7.92
Other race/ethnicity
2.53 4.47 3.02 3.87
Marital status
Married/living with a
partner
39.51 62.30 70.94 81.22
Separated/divorced/widowed
56.71 34.08 25.20 15.05
Never married 3.78 3.61 3.86 3.73
Education level
Less than high school 33.81 20.05 34.14 18.01
High school graduate 37.80 37.35 29.40 30.47
64
Note: All estimates are weighted to account for complex survey design. The question for health conditions was
“
Has a doctor ever
told you that you have diabetes or high blood sugar?
”
Based on the respondents' self-report, the answer was yes (1) and no (0). The
question for exercise was
“
On average over the last 12 months have you participated in vigorous physical activity or exercise three
times a week or more? By vigorous physical activity, we mean things like sports, heavy housework, or a job that involves physical
labor
”
. Based on the question, the answer was classified into 1 (yes) and 0 (no). Alcohol intake was asked by: Do you ever drink any
alcoholic beverages such as beer, wine, or liquor? The answer was yes (1) or no (0).
Some college 17.08 23.24 17.08 20.06
College or more 11.32 19.36 19.38 31.46
Health conditions
Hypertension 60.23 40.16 53.04 40.95
Diabetes 20.46 8.24 21.32 10.49
Stroke 10.56 2.69 11.54 2.87
Exercise 29.21 45.91 39.43 57.75
Smoking status
Never smoker 83.43 87.11 80.73 84.42
Former/current smoker 16.57 12.89 19.27 15.58
Alcohol consumption 32.30 47.71 50.37 63.15
N 4,845 6,103 4,089 3,876
65
Table 3-3. Cox proportional hazard models by gender
Women Men
Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
Hazard Ratio (P value) Hazard Ratio (P value)
Cognitive status (Ref: normal)
CIND 1.56*** 1.47*** 1.40*** 1.61*** 1.49*** 1.32***
With dementia 2.25*** 2.12*** 1.91*** 2.76*** 2.45*** 2.05***
Age group (Ref: 50-54)
55-59 0.99 0.97 0.94 1.22 1.20 1.23
60-64
0.86 0.84 0.82 1.23 1.20 1.17
65-69
0.88 0.84 0.78 1.17 1.13 1.10
70-74
0.82 0.77 0.71* 1.18 1.15 1.05
75-79 0.75 0.69* 0.61** 1.04 0.99 0.91
80-84 0.64* 0.57** 0.48*** 0.91 0.84 0.75
85+
0.48*** 0.42*** 0.36*** 0.71 0.66 0.54**
Race/Ethnicity (Ref: Non-Hispanic white)
Non-Hispanic black 0.99 0.94 0.85 1.04 0.97 0.89*
Other race/ethnicity 0.78** 0.75** 0.76** 0.87 0.86 0.78*
Marital status (Ref: married/living with a partner)
Separated/divorced/widowed
1.20*** 1.10**
1.27*** 1.25
Never married
1.49** 1.49**
1.33* 1.33*
Education level (Ref: college or more)
Less than high school
1.27** 1.00
1.34*** 1.19**
High school graduate
1.23** 1.07
1.29*** 1.18**
Some college
1.08 1.01
1.27*** 1.18**
Health conditions
Hypertension
1.25***
1.12**
66
p<0.05; **p<0.01; ***p<0.001
Note: All estimates are weighted to account for complex survey design. The question for health conditions was
“
Has a doctor ever
told you that you have diabetes or high blood sugar?
”
Based on the respondents' self-report, the answer was yes (1) and no (0). The
question for exercise was
“
On average over the last 12 months have you participated in vigorous physical activity or exercise three
times a week or more? By vigorous physical activity, we mean things like sports, heavy housework, or a job that involves physical
labor
”
. Based on the question, the answer was classified into 1 (yes) and 0 (no). Alcohol intake was asked by: Do you ever drink any
alcoholic beverages such as beer, wine, or liquor? The answer was yes (1) or no (0).
Diabetes
1.80***
1.54***
Stroke
1.51***
1.44***
Exercise
0.74***
0.67***
Smoking status (Ref: never smoker)
Former/current smoker
2.32***
1.86***
Alcohol consumption
0.82***
0.86***
N 10,948 10,948 10,948 7,965 7,965 7,965
67
Table 3-4. Cox proportional hazard models among the respondents with CIND and dementia
Model 1 Model 2 Model 3
Hazard Ratio (P value)
Gender (Ref: men)
Women 0.71*** 0.69*** 0.68***
Age group (Ref: 50-54)
55-59 1.39 1.40 1.39
60-64 1.48 1.49 1.45
65-69
1.54 1.54 1.41
70-74 1.51 1.52 1.38
75-79 1.53 1.52 1.28
80-84
1.45 1.42 1.17
85+ 1.27 1.24 0.98
Race/ethnicity (Ref: Non-Hispanic whites)
Non-Hispanic blacks
1.99 0.99 0.92
Other 0.81** 0.81** 0.77**
Marital status (Ref: married/living with a partner)
Separated/divorced/widowed 1.10* 1.07
Never married
1.16 1.22*
Education level (Ref: college or more)
Less than high school
0.99 0.88
High school graduate
1.02 0.97
Some college
1.03 1.02
Health conditions
Hypertension
1.44***
Diabetes
1.09*
Stroke
1.47***
Exercise
0.69***
Smoking status (Ref: never smoker)
Former/current
smoker
1.84***
Alcohol consumption
0.78***
N 4,742 4,742 4,742
*p<0.05; **p<0.01; ***p<0.001
Note: All estimates are weighted to account for complex survey design. The question for health
conditions was “Has a doctor ever told you that you have diabetes or high blood sugar?” Based
on the respondents' self-report, the answer was yes (1) and no (0). The question for exercise was
“On average over the last 12 months have you participated in vigorous physical activity or
exercise three times a week or more? By vigorous physical activity, we mean things like sports,
heavy housework, or a job that involves physical labor”. Based on the question, the answer was
68
classified into 1 (yes) and 0 (no). Alcohol intake was asked by: Do you ever drink any alcoholic
beverages such as beer, wine, or liquor? The answer was yes (1) or no (0).
69
Chapter 5: Summary & Discussion
As life expectancy has increased and populations age, a greater number of older adults
are experiencing changes in memory as they age. The prevalence of chronic disease remains high
and the burden of cognitive decline among older adults is projected to increase in the future.
Thus, the cost of medical care and disease management will likely increase and put a burden on
public health. At the same time, demographic trends indicate that the aging population is
becoming more diverse as the proportion of minority individuals is steadily increasing in the
U.S. Generally speaking, older minority individuals tend to have worse cognitive functioning
compared to non-Hispanic whites (Díaz-Venegas et al., 2016). We posit that there are differences
in individuals’ perceived memory (commonly referred to as self-rated memory in the literature)
by race/ethnicity due to between-group differences in social and cultural characteristics (Segel-
Karpas & Palgi, 2019).
Self-rated memory can be an important tool for detecting early signs of cognitive decline
and dementia onset. Therefore, self-rated memory is widely used in clinical settings to assess
individuals’ cognitive status (Marino et al., 2009). Researchers have demonstrated that self-rated
memory is associated with many dimensions of health including functional status, disability, and
mental health (Cordier et al., 2019; Ficker et al., 2014; Hill et al., 2016). For example, older
adults with poor self-rated memory had a higher likelihood of needing assistance with activities
of daily living compared to their counterparts (Cordier et al., 2019). This dissertation expands the
literature on self-rated memory by race/ethnicity and multimorbidity.
Study 1
The first study adds to the little empirical evidence on differences in self-rated memory
across non-Hispanic white, non-Hispanic black, and Hispanic older adults in the United States.
70
We used data on adults aged 65 and older from the 2011 (Round 1) of the National Health and
Aging Trends Study (NHATS). We performed weighted logistic regressions to examine the
association between having fair/poor self-rated memory and race/ethnicity, controlling for
sociodemographic characteristics (i.e., age, gender, and marital status), health conditions (i.e., the
number of chronic diseases, depressive symptoms, objective memory status, and functional
limitations), and sociocultural factors (i.e., economic vulnerability, limited English proficiency,
and religious services). We found that the older minority adults had greater odds of reporting
fair/poor self-rated memory than non-Hispanic whites. Moreover, we were able to identify that
Hispanics with limited English proficiency had higher odds of reporting fair/poor self-rated
memory than non-Hispanic whites, whereas there were no significant differences between
Hispanics without limited English proficiency and non-Hispanic whites. This chapter adds to the
literature on differences in self-rated memory by race/ethnicity, as well as demonstrating how
limited English proficiency plays a role among the Hispanic respondents. This study has
important implications, particularly for health clinicians who are in charge of examining and
interpreting assessments of cognition. Health professionals should understand that minority older
adults are more likely to report fair/poor self-rated memory, especially those with limited English
proficiency. Tailored cognitive screening will efficiently respond to the specific needs of
minority older adults and may help address disparities in cognitive functioning among older
adults.
Study 2
The second study investigated the impact of multimorbidity and types of chronic disease
on self-rated memory in older adults in the United States. We used the data from the 2011 wave
of the NHATS (Round 1). We performed logistic regressions to investigate the association
71
between multimorbidity and types of chronic diseases and fair/poor self-rated memory,
controlling for sociodemographic characteristics (age, gender, race/ethnicity, marital status,
annual household income, and education level), lifestyle factors (smoking status and body mass
index), mental health conditions (psychological well-being and depressive symptoms),
prescribed medicine use and objective memory status. We found that respondents with
multimorbidity had 35% higher odds of reporting fair/poor self-rated memory compared to those
with no or one chronic disease. When we examined the associations between specific types of
chronic diseases and fair/poor self-rated memory, three out of eight chronic disease were
significant predictors. For example, stroke, osteoporosis, and arthritis were identified as
increasing the odds of reporting fair/poor self-rated memory by 41%, 20%, and 30%,
respectively. This study adds to literature by expanding on how multimorbidity impacts self-
rated memory as well as also determining what types of chronic diseases have the greatest
impact. These findings demonstrate the importance of chronic disease prevention and how
multimorbidity negatively influences individuals’ perceived memory, which may be associated
with cognitive decline in the long term. This research suggests the need to educate older adults
with multimorbidity and certain types of diseases regarding negative self-rated memory and its
consequences.
Chapter 3
Our last chapter of the dissertation investigated differences in mortality risk associated
with cognitive impairment and dementia by gender. We used the data drawn from the Health and
Retirement Study 2000 with 16 years of mortality follow up. We used Cox proportional hazard
models to examine the association between cognitive status and mortality risk by gender,
adjusting for sociodemographic characteristics (i.e., age, gender, race/ethnicity, marital status,
72
education level), health conditions (i.e., hypertension, diabetes, stroke), and lifestyle factors (i.e.,
exercise, smoking status, and alcohol consumption). Our study demonstrated that having CIND
or dementia increases mortality risk for both men (HR: 2.05, 95% CI 1.75-2.40) and women
(HR: 1.91, 95% CI 1.65-2.22), even after adjusting for the control variables. In addition, when
we investigated gender differences in mortality among respondents with CIND and dementia,
women had a 32% lower mortality risk compared to men, controlling for the full set of the
covariates (HR: 0.68, 95% CI 0.61-0.75). This study contributes to the literature on survival with
cognitive impairment by gender and suggests that health conditions and lifestyle factors play an
important role in mortality for both genders. Targeted interventions for the management and
prevention of chronic conditions and the promotion of healthy behaviors may reduce mortality
risk among older adults.
Discussion
The three chapters of this dissertation give us broader perspectives on cognition (i.e., self-
rated memory, cognitive impairment no dementia, and dementia). Investigating self-rated
memory by race/ethnicity allowed us to understand how sociocultural factors play a role in
differences across groups. Indeed, we found that limited English proficiency is an important
factor linked to poor self-rated memory among older Hispanics. Language can be an important
component for the social adjustment of minority older adults, since it is a critical tool for
communication (Chiswick & Miller, 1994). It has been noted that language and cultural barriers
are the main hardship for certain minority populations (Ponce et al., 2006b). Minority older
adults who have limited English proficiency were more likely to have poorer mental health and
general health, yet are less likely to have access to healthcare (Ponce et al., 2006b; Tam & Page,
2016). Minority older adults with limited English proficiency may have a hard time
73
understanding written medical forms and understanding a doctor’s note (Shi et al., 2009; Smith,
2009). Therefore, it is critical to provide tailored language healthcare resources when evaluating
individuals’ self-rated memory among the minority older population.
Although this research has contributed novel information about self-rated memory and
race/ethnicity, there are limitations that should be addressed. First, the variables used in this
research were based on self-reports. Although we did not have information with regards to
clinical diagnosis, reliability of self-reported information on chronic disease and healthcare
utilization is well documented in prior research (Najafi et al., 2019; Short et al., 2009). Second,
the NHATS is based on the Medicare population who are older than 65. Thus, the findings from
this research may not be generalized to the aging population. Nonetheless, considering more than
83% of older adults are covered by Medicare (Kaiser Family Foundation, 2016), the sample used
in this research covers a large share of older adults in the U.S. Lastly, although we used a
nationally representative sample, the number of Hispanic respondents was small (N=388). Future
research would benefit from access to samples including larger numbers of minority older adults.
Despite the aforementioned limitations, this research adds to literature on self-rated memory and
race/ethnicity.
Second, examining the relationship between multimorbidity and self-rated memory
provides insight that having multiple chronic diseases can contribute to fair/poor self-rated
memory. Also, we found that three chronic diseases (stroke, osteoporosis, and arthritis) are
closely linked to fair/poor self-rated memory. This study suggests that prevention of chronic
diseases is critical for older adults. Having chronic diseases requires constant care and
management, is a challenge to manage, and costly for older individuals (Navickas et al., 2016).
Given that more than half of Americans aged 18 and older already have at least one chronic
74
disease (Buttorff et al., 2017), it is possible that a higher percentage of older individuals will
have multimorbidity in the future. Having both multimorbidity and fair/poor self-rated memory
will negatively impact older adults’ independence, mental health, and quality of life. Therefore,
not only health professionals, but also older adults would benefit from understanding the
negative consequences of having multimorbidity and certain types of chronic diseases.
The limitations of this study include a lack of information on the duration and severity of
chronic disease. Severity of disease could be an important factor because it has been suggested
that greater severity is linked to poorer mental health and a lower quality of life (Pizzi et al.,
2006). Moreover, it was found that the respondents’ mood plays a role in individuals’ self-rated
memory (Yates et al., 2017). Although we weren’t able to include the respondents’ mood as a
covariate, we were able to include psychological well-being, which is linked to individuals’
mood (Polak et al., 2015). Lastly, this research only included respondents who were community
dwelling older adults, thus, our results are limited to those who are living in community settings.
However, a key strength of this research is that it used nationally representative sample of older
adults and it included self-rated memory as well as objective memory status.
Our final chapter provides information on how cognitive impairment or dementia affects
mortality among men and women and whether there are gender differences in survival with
cognitive impairment or dementia. The study suggests that having cognitive impairment or
dementia increases mortality risk for both genders, even after controlling for the full set of
covariates. In addition, when we investigated the mortality risk for the respondents with
cognitive impairment or dementia by gender, we found that women had a lower mortality risk
compared to men. Our findings demonstrated that certain lifestyle factors (smoking, exercise,
alcohol consumption) are the strongest predictors of mortality. Health behaviors are modifiable
75
and improve quality of life for older adults (Thompson et al., 2012). Moreover, maintaining
healthy behaviors could have a positive effect on brain health as well (Mintzer et al., 2019).
Therefore, promoting healthy behavior interventions could benefit the aging population.
This research has several limitations. First, the evaluation of cognition was not based on
clinical evaluation. Nonetheless, the validity of the cognitive status measure is supported by a
recent study that found similar dementia prevalence comparing the HRS and Medicare claims
data (Chen et al., 2019). Second, there are multiple stages of dementia as it is a progressive
disease. However, we were not able to include information on respondents’ stage of dementia. It
is possible that the respondents with late stage dementia have a higher mortality risk than the
respondents with early stage of dementia. In addition, there are various dementia subtypes, but
we were not able to identify dementia subtypes. Lastly, our study was only able to include self-
reports of chronic diseases. Despite these limitations, our research adds to the literature on
cognitive impairment and dementia mortality by gender.
In conclusion, this dissertation provides important findings on self-rated memory and
differences in survival by cognitive status among the older adults. These studies show that
race/ethnicity and multimorbidity influence an individual’s self-rated memory. They also show
how cognitive impairment or dementia can increase mortality risk by gender. There is more to
learn about cognitive health among older adults, and this work provides several future directions
for research on cognition and health. Future studies may also consider whether the findings from
our research hold when larger samples the minority populations are used. In addition, future
studies incorporating information on whether the severity and duration of chronic disease impact
self-rated memory would be informative. Lastly, future work with information on
76
neuropathology with dementia subtypes or clinical diagnosis of dementia could provide greater
insight into the linkages between cognitive impairment and mortality.
77
References
Aarts, S., van den Akker, M., Hajema, K. J., van Ingen, A. M., Metsemakers, J. F. M., Verhey, F.
R. J., & van Boxtel, M. P. J. (2011). Multimorbidity and its relation to subjective memory
complaints in a large general population of older adults. International Psychogeriatrics,
23(4), 616–624. https://doi.org/10.1017/S1041610210002024
Alley, D., Suthers, K., & Crimmins, E. (2007). Education and Cognitive Decline in Older
Americans: Results From the AHEAD Sample. Research on Aging, 29(1), 73–94.
https://doi.org/10.1177/0164027506294245
Al-Qazzaz, N., Ali, S., Ahmad, S. A., Islam, S., & Mohamad, K. (2014). Cognitive impairment
and memory dysfunction after a stroke diagnosis: A post-stroke memory assessment.
Neuropsychiatric Disease and Treatment, 1677. https://doi.org/10.2147/NDT.S67184
Alzheimer’s Association. (2019). 2019 Alzheimer’s Disease Facts and Figures.
https://www.alz.org/media/Documents/alzheimers-facts-and-figures-2019-r.pdf
Alzheimer’s Association. (2021). New Alzheimer’s Association Report Examines Racial and
Ethnic Attitudes on Alzheimer’s and Dementia Care.
https://www.alz.org/news/2021/new-alzheimers-association-report-examines-racial
Anstey, K. J., von Sanden, C., Salim, A., & O’Kearney, R. (2007). Smoking as a Risk Factor for
Dementia and Cognitive Decline: A Meta-Analysis of Prospective Studies. American
Journal of Epidemiology, 166(4), 367–378. https://doi.org/10.1093/aje/kwm116
Arezzo, M. F., & Giudici, C. (2017). Social Capital and Self Perceived Health Among European
Older Adults. Social Indicators Research, 130(2), 665–685.
https://doi.org/10.1007/s11205-015-1195-z
78
Arias, E., & Xu, J. (2020a). United States life tables, 2018. National Vital Statistics Reports,
69(12). https://www.cdc.gov/nchs/data/nvsr/nvsr69/nvsr69-12-508.pdf
Arias, E., & Xu, J. Q. (2020b). United States life tables, 2018. National Vital Statistics Reports,
69(12). https://www.cdc.gov/nchs/data/nvsr/nvsr69/nvsr69-12-508.pdf
Assari, S. (2018). Parental Educational Attainment and Mental Well-Being of College Students;
Diminished Returns of Blacks. Brain Sciences, 8(11), E193.
https://doi.org/10.3390/brainsci8110193
Bamia, C., Orfanos, P., Juerges, H., Schöttker, B., Brenner, H., Lorbeer, R., Aadahl, M.,
Matthews, C. E., Klinaki, E., Katsoulis, M., Lagiou, P., Bueno-de-mesquita, H. B.,
Eriksson, S., Mons, U., Saum, K.-U., Kubinova, R., Pajak, A., Tamosiunas, A.,
Malyutina, S., … Trichopoulos, D. (2017). Self-rated health and all-cause and cause-
specific mortality of older adults: Individual data meta-analysis of prospective cohort
studies in the CHANCES Consortium. Maturitas, 103, 37–44.
https://doi.org/10.1016/j.maturitas.2017.06.023
Barnes, L. L., Schneider, J. A., Boyle, P. A., Bienias, J. L., & Bennett, D. A. (2006). Memory
complaints are related to Alzheimer disease pathology in older persons. Neurology, 67(9),
1581–1585. https://doi.org/10.1212/01.wnl.0000242734.16663.09
Beam, C. R., Kaneshiro, C., Jang, J. Y., Reynolds, C. A., Pedersen, N. L., & Gatz, M. (2018).
Differences Between Women and Men in Incidence Rates of Dementia and Alzheimer’s
Disease. Journal of Alzheimer’s Disease, 64(4), 1077–1083. https://doi.org/10.3233/JAD-
180141
79
Begum, A., Morgan, C., Chiu, C.-C., Tylee, A., & Stewart, R. (2012). Subjective memory
impairment in older adults: Aetiology, salience and help seeking. International Journal of
Geriatric Psychiatry, 27(6), 612–620. https://doi.org/10.1002/gps.2760
Bertoni, A. G., Krop, J. S., Anderson, G. F., & Brancati, F. L. (2002). Diabetes-Related
Morbidity and Mortality in a National Sample of U.S. Elders. Diabetes Care, 25(3), 471–
475. https://doi.org/10.2337/diacare.25.3.471
Betancourt, J. R., Renfrew, M. R., Green, A. R., Lopez, L., & Wasserman, M. (2012). Improving
patient safety systems for patients with limited English proficiency: A guide for hospitals.
Agency for Healthcare Research and Quality.
Borrell, L. N., & Dallo, F. J. (2008). Self-Rated Health and Race Among Hispanic and Non-
Hispanic Adults. Journal of Immigrant and Minority Health, 10(3), 229–238.
https://doi.org/10.1007/s10903-007-9074-6
Brandt, J., Spencer, M., & Folstein, M. (1988). The telephone interview for cognitive status.
Neuropsychiatry, Neuropsychology and Behavioral Neurology, 1(2), 111–117.
Brewster, P. W. H., Melrose, R. J., Marquine, M. J., Johnson, J. K., Napoles, A., MacKay-
Brandt, A., Farias, S., Reed, B., & Mungas, D. (2014). Life experience and demographic
influences on cognitive function in older adults. Neuropsychology, 28(6), 846–858.
https://doi.org/10.1037/neu0000098
Buttorff, C., Ruder, T., & Bauman, M. (2017). Multiple chronic conditions in the United States.
Rand Corporation. https://www.rand.org/blog/rand-review/2017/07/chronic-conditions-
in-america-price-and-prevalence.html
80
Carr, D., & Bodnar-Deren, S. (2009). Gender, Aging and Widowhood. In P. Uhlenberg (Ed.),
International Handbook of Population Aging (pp. 705–728). Springer Netherlands.
https://doi.org/10.1007/978-1-4020-8356-3_32
CDC. (2020, September 11). Older Adults. https://www.cdc.gov/coronavirus/2019-ncov/need-
extra-precautions/older-adults.html
Centers for Disease Control and Prevention. (2021a). About chornic diseases.
https://www.cdc.gov/chronicdisease/about/index.htm#:~:text=Chronic%20diseases%20ar
e%20defined%20broadly,disability%20in%20the%20United%20States.
Centers for Disease Control and Prevention. (2021b). Healthy weight, nutrition, and physical
activity. https://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/index.html
Centers for Disease Control and Prevention (CDC). (2008). Smoking-attributable mortality,
years of potential life lost, and productivity losses—United States, 2000-2004. Morbidity
and Mortality Weekly Report, 57(45), 1226–1228.
Centers for Disease Control and Prevention, & National Center for Chronic Disease Prevention
and Health Promotion. (2020). Chronic Diseases and Cognitive Decline—A Public
Health Issue. https://www.cdc.gov/aging/pdf/20-03-Chronic-Diseases-and-Cognitive-
Decline-Pages-h.pdf
Chen, Y., Tysinger, B., Crimmins, E., & Zissimopoulos, J. M. (2019). Analysis of dementia in
the US population using Medicare claims: Insights from linked survey and administrative
claims data. Alzheimer’s & Dementia (New York, N. Y.), 5, 197–207.
https://doi.org/10.1016/j.trci.2019.04.003
81
Chiswick, B. R., & Miller, P. W. (1994). Language choice among immigrants in a multi-lingual
destination. Journal of Population Economics, 7(2), 119–131.
https://doi.org/10.1007/BF00173615
Chudasama, Y. V., Khunti, K. K., Zaccardi, F., Rowlands, A. V., Yates, T., Gillies, C. L.,
Davies, M. J., & Dhalwani, N. N. (2019). Physical activity, multimorbidity, and life
expectancy: A UK Biobank longitudinal study. BMC Medicine, 17(1), 108.
https://doi.org/10.1186/s12916-019-1339-0
Clarkson, M. J., May, A. K., & Warmington, S. A. (2019). Chronic Blood Flow Restriction
Exercise Improves Objective Physical Function: A Systematic Review. Frontiers in
Physiology, 10, 1058. https://doi.org/10.3389/fphys.2019.01058
Cooper, C., Tandy, A. R., Balamurali, T. B. S., & Livingston, G. (2010). A systematic review
and meta-analysis of ethnic differences in use of dementia treatment, care, and research.
The American Journal of Geriatric Psychiatry: Official Journal of the American
Association for Geriatric Psychiatry, 18(3), 193–203.
https://doi.org/10.1097/JGP.0b013e3181bf9caf
Cordier, R., Chen, Y., Clemson, L., Byles, J., & Mahoney, N. (2019). Subjective memory
complaints and difficulty performing activities of daily living among older women in
Australia. Australian Occupational Therapy Journal, 66(2), 227–238.
https://doi.org/10.1111/1440-1630.12548
Cornelius, M. E., Wang, T. W., Jamal, A., Loretan, C. G., & Neff, L. J. (2020). Tobacco Product
Use Among Adults—United States, 2019. MMWR. Morbidity and Mortality Weekly
Report, 69(46), 1736–1742. https://doi.org/10.15585/mmwr.mm6946a4
82
Crimmins, E. M., Kim, J. K., Langa, K. M., & Weir, D. R. (2011). Assessment of Cognition
Using Surveys and Neuropsychological Assessment: The Health and Retirement Study
and the Aging, Demographics, and Memory Study. The Journals of Gerontology Series
B: Psychological Sciences and Social Sciences, 66B(Supplement 1), i162–i171.
https://doi.org/10.1093/geronb/gbr048
Crimmins, E. M., & Saito, Y. (2001). Trends in healthy life expectancy in the United States,
1970–1990: Gender, racial, and educational differences. Social Science & Medicine,
52(11), 1629–1641. https://doi.org/10.1016/S0277-9536(00)00273-2
Crimmins, E. M., & Zhang, Y. S. (2019). Aging Populations, Mortality, and Life Expectancy.
Annual Review of Sociology, 45(1), 69–89. https://doi.org/10.1146/annurev-soc-073117-
041351
Cutler, S. J. (2015). Worries About Getting Alzheimer’s: Who’s Concerned? American Journal
of Alzheimer’s Disease & Other Dementiasr, 30(6), 591–598.
https://doi.org/10.1177/1533317514568889
Dal Forno, G., Carson, K. A., Brookmeyer, R., Troncoso, J., Kawas, C. H., & Brandt, J. (2002).
APOE genotype and survival in men and women with Alzheimer’s disease. Neurology,
58(7), 1045–1050. https://doi.org/10.1212/WNL.58.7.1045
DeSalvo, K. B., Bloser, N., Reynolds, K., He, J., & Muntner, P. (2006). Mortality prediction
with a single general self-rated health question: A meta-analysis. Journal of General
Internal Medicine, 21(3), 267–275. https://doi.org/10.1111/j.1525-1497.2005.00291.x
Díaz-Venegas, C., Downer, B., Langa, K. M., & Wong, R. (2016). Racial and ethnic differences
in cognitive function among older adults in the USA: Cognition of US older adults by
83
race/ethnicity. International Journal of Geriatric Psychiatry, 31(9), 1004–1012.
https://doi.org/10.1002/gps.4410
DuBard, C. A., & Gizlice, Z. (2008). Language Spoken and Differences in Health Status, Access
to Care, and Receipt of Preventive Services Among US Hispanics. American Journal of
Public Health, 98(11), 2021–2028. https://doi.org/10.2105/AJPH.2007.119008
Dubois, B., Hampel, H., Feldman, H. H., Scheltens, P., Aisen, P., Andrieu, S., Bakardjian, H.,
Benali, H., Bertram, L., Blennow, K., Broich, K., Cavedo, E., Crutch, S., Dartigues, J.-F.,
Duyckaerts, C., Epelbaum, S., Frisoni, G. B., Gauthier, S., Genthon, R., … Proceedings
of the Meeting of the International Working Group (IWG) and the American Alzheimer’s
Association on “The Preclinical State of AD”; July 23, 2015; Washington DC, USA.
(2016). Preclinical Alzheimer’s disease: Definition, natural history, and diagnostic
criteria. Alzheimer’s & Dementia, 12(3), 292–323.
https://doi.org/10.1016/j.jalz.2016.02.002
Ershler, W. B. (1993). Interleukin-6: A Cytokine for Gerontolgists. Journal of the American
Geriatrics Society, 41(2), 176–181. https://doi.org/10.1111/j.1532-5415.1993.tb02054.x
Fabbri, E., An, Y., Zoli, M., Simonsick, E. M., Guralnik, J. M., Bandinelli, S., Boyd, C. M., &
Ferrucci, L. (2015). Aging and the Burden of Multimorbidity: Associations With
Inflammatory and Anabolic Hormonal Biomarkers. The Journals of Gerontology: Series
A, 70(1), 63–70. https://doi.org/10.1093/gerona/glu127
Ficker, L. J., Lysack, C. L., Hanna, M., & Lichtenberg, P. A. (2014). Perceived Cognitive
Impairment among African American elders: Health and functional impairments in daily
life. Aging & Mental Health, 18(4), 471–480.
https://doi.org/10.1080/13607863.2013.856859
84
Fornazzari, L., Fischer, C., Hansen, T., & Ringer, L. (2009). Knowledge of Alzheimer’s disease
and subjective memory impairment in Latin American seniors in the Greater Toronto
Area. International Psychogeriatrics, 21(05), 966.
https://doi.org/10.1017/S1041610209990433
Franco, Y., & Choi, E. Y. (2020). The Relationship Between Immigrant Status and Undiagnosed
Dementia: The Role of Limited English Proficiency. Journal of Immigrant and Minority
Health, 22(5), 914–922. https://doi.org/10.1007/s10903-019-00963-w
Gatz, M., Svedberg, P., Pedersen, N. L., Mortimer, J. A., Berg, S., & Johansson, B. (2001).
Education and the Risk of Alzheimer’s Disease: Findings From the Study of Dementia in
Swedish Twins. The Journals of Gerontology Series B: Psychological Sciences and
Social Sciences, 56(5), P292–P300. https://doi.org/10.1093/geronb/56.5.P292
Gerteis, J., Izrael, D., Deitz, D., LeRoy, L., Ricciardi, R., Miller, T., & Basu, J. (2014). Multiple
chronic conditions chartbook (AHRQ Publications No, Q14-0038).
https://www.ahrq.gov/sites/default/files/wysiwyg/professionals/prevention-chronic-
care/decision/mcc/mccchartbook.pdf
Goldman, N., Korenman, S., & Weinstein, R. (1995). Marital status and health among the
elderly. Social Science & Medicine, 40(12), 1717–1730. https://doi.org/10.1016/0277-
9536(94)00281-W
Graham, E. A., Jacobs, T. A., Kwan-Gett, T. S., & Cover, J. (2008). Health Services Utilization
by Low-Income Limited English Proficient Adults. Journal of Immigrant and Minority
Health, 10(3), 207–217. https://doi.org/10.1007/s10903-007-9069-3
Harada, C. N., Natelson, M. C., & Triebel, K. L. (2013). Normal Cognitive Aging. Clinics in
Geriatric Medicine, 29(4), 737–752. https://doi.org/10.1016/j.cger.2013.07.002
85
Hendi, A. S. (2015). Trends in U.S. life expectancy gradients: The role of changing educational
composition. International Journal of Epidemiology, 44(3), 946–955.
https://doi.org/10.1093/ije/dyv062
Hendi, A. S. (2017). Trends in Education-Specific Life Expectancy, Data Quality, and Shifting
Education Distributions: A Note on Recent Research. Demography, 54(3), 1203–1213.
https://doi.org/10.1007/s13524-017-0574-2
Hill, N. L., Mogle, J., Wion, R., Munoz, E., DePasquale, N., Yevchak, A. M., & Parisi, J. M.
(2016). Subjective Cognitive Impairment and Affective Symptoms: A Systematic
Review. The Gerontologist, 56(6), e109–e127. https://doi.org/10.1093/geront/gnw091
Holahan, J., & Kim, J. (2000). Why Does The Number Of Uninsured Americans Continue To
Grow?: Gains in employer coverage, caused by workers’ shifting into higher-paying jobs,
have not offset declines in Medicaid coverage. Health Affairs, 19(4), 188–196.
https://doi.org/10.1377/hlthaff.19.4.188
Idler, E. L., & Benyamini, Y. (1997). Self-rated health and mortality: A review of twenty-seven
community studies. Journal of Health and Social Behavior, 38(1), 21–37.
Jacob, L., Haro, J. M., & Koyanagi, A. (2019). Physical multimorbidity and subjective cognitive
complaints among adults in the United Kingdom: A cross-sectional community-based
study. Scientific Reports, 9(1), 12417. https://doi.org/10.1038/s41598-019-48894-8
Jang, Y., Choi, E. Y., Franco, Y., Park, N. S., Chiriboga, D. A., & Kim, M. T. (2021). Racial and
ethnic differences in cognitive health appraisals: A comparison of non-Hispanic White,
non-Hispanic Black, and Hispanic older adults. Aging & Mental Health, 1–7.
https://doi.org/10.1080/13607863.2021.1899132
86
Jang, Y., Haley, W. E., Choi, E. Y., & Franco, Y. (2021). Racial/Ethnic Differences in
Correspondence between Subjective Cognitive Ratings and Cognitive Impairment. The
American Journal of Geriatric Psychiatry, S1064748121005248.
https://doi.org/10.1016/j.jagp.2021.10.015
Jang, Y., & Kim, M. T. (2019). Limited English Proficiency and Health Service Use in Asian
Americans. Journal of Immigrant and Minority Health, 21(2), 264–270.
https://doi.org/10.1007/s10903-018-0763-0
Jessen, F., Feyen, L., Freymann, K., Tepest, R., Maier, W., Heun, R., Schild, H.-H., & Scheef, L.
(2006). Volume reduction of the entorhinal cortex in subjective memory impairment.
Neurobiology of Aging, 27(12), 1751–1756.
https://doi.org/10.1016/j.neurobiolaging.2005.10.010
Jonker, C., Geerlings, M. I., & Schmand, B. (2000). Are memory complaints predictive for
dementia? A review of clinical and population-based studies. International Journal of
Geriatric Psychiatry, 15(11), 983–991. https://doi.org/10.1002/1099-
1166(200011)15:11<983::aid-gps238>3.0.co;2-5
Kaiser Family Foundation. (2016). Profile of Medicare beneficiaries by race and ethnicity:A
chartpack.
Kang, H. G., Park, H. Y., Ryu, H. U., & Suk, S.-H. (2018). Bone mineral loss and cognitive
impairment: The PRESENT project. Medicine, 97(41), e12755.
https://doi.org/10.1097/MD.0000000000012755
Kaplan, G. A., Goldberg, D. E., Everson, S. A., Cohen, R. D., Salonen, R., Tuomilehto, J., &
Salonen, J. (1996). Perceived Health Status and Morbidity and Mortality: Evidence from
87
the Kuopio Ischaemic Heart Disease Risk Factor Study. International Journal of
Epidemiology, 25(2), 259–265. https://doi.org/10.1093/ije/25.2.259
Kaplan, M. S., Newsom, J. T., McFarland, B. H., & Lu, L. (2001). Demographic and
psychosocial correlates of physical activity in late life. American Journal of Preventive
Medicine, 21(4), 306–312. https://doi.org/10.1016/S0749-3797(01)00364-6
Kim, G., Worley, C. B., Allen, R. S., Vinson, L., Crowther, M. R., Parmelee, P., & Chiriboga, D.
A. (2011). Vulnerability of Older Latino and Asian Immigrants with Limited English
Proficiency: VULNERABILITY OF LEP ELDERS. Journal of the American Geriatrics
Society, 59(7), 1246–1252. https://doi.org/10.1111/j.1532-5415.2011.03483.x
Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2003). The Patient Health Questionnaire-2:
Validity of a two-item depression screener. Medical Care, 41(11), 1284–1292.
https://doi.org/10.1097/01.MLR.0000093487.78664.3C
Kurella, M., Chertow, G. M., Fried, L. F., Cummings, S. R., Harris, T., Simonsick, E.,
Satterfield, S., Ayonayon, H., & Yaffe, K. (2005). Chronic kidney disease and cognitive
impairment in the elderly: The health, aging, and body composition study. Journal of the
American Society of Nephrology, 16(7), 2127–2133.
https://doi.org/10.1681/ASN.2005010005
Langa, K. M., Larson, E. B., Crimmins, E. M., Faul, J. D., Levine, D. A., Kabeto, M. U., &
Weir, D. R. (2017). A Comparison of the Prevalence of Dementia in the United States in
2000 and 2012. JAMA Internal Medicine, 177(1), 51.
https://doi.org/10.1001/jamainternmed.2016.6807
88
Larson, E. B., Yaffe, K., & Langa, K. M. (2013). New insights into the dementia epidemic. New
England Journal of Medicine, 369(24), 2275–2277.
https://doi.org/10.1056/NEJMp1311405
Lawton, M. P., & Brody, E. M. (1969). Assessment of older people: Self-maintaining and
instrumental activities of daily living. The Gerontologist, 9(3–1), 179–186.
https://doi.org/10.1093/geront/9.3_Part_1.179
Lehning, A. J., Mattocks, N., Smith, R. J., Kim, K., & Cheon, J. H. (2020). Neighborhood age
composition and self-rated health: Findings from a nationally representative study.
Journal of Gerontological Social Work, 1–17.
https://doi.org/10.1080/01634372.2020.1866731
Li, S., Wang, J., Zhang, B., Li, X., & Liu, Y. (2019). Diabetes Mellitus and Cause-Specific
Mortality: A Population-Based Study. Diabetes & Metabolism Journal, 43(3), 319–341.
https://doi.org/10.4093/dmj.2018.0060
Luchsinger, J. A., Tang, M.-X., Siddiqui, M., Shea, S., & Mayeux, R. (2004). Alcohol Intake and
Risk of Dementia: ALCOHOL INTAKE AND RISK OF DEMENTIA. Journal of the
American Geriatrics Society, 52(4), 540–546. https://doi.org/10.1111/j.1532-
5415.2004.52159.x
Lui, L.-Y., Stone, K., Cauley, J. A., Hillier, T., & Yaffe, K. (2003). Bone loss predicts
subsequent cognitive decline in older women: The study of osteoporotic fractures.
Journal of the American Geriatrics Society, 51(1), 38–43. https://doi.org/10.1034/j.1601-
5215.2002.51007.x
89
Lukoff, D., Lu, F. G., & Turner, R. (1995). Cultural considerations in the assessment and
treatment of religious and spiritual problems. Psychiatric Clinics of North America,
18(3), 467–485. https://doi.org/10.1016/S0193-953X(18)30035-2
Maki, Y., Yamaguchi, T., Yamagami, T., Murai, T., Hachisuka, K., Miyamae, F., Ito, K., Awata,
S., Ura, C., Takahashi, R., & Yamaguchi, H. (2014). The impact of subjective memory
complaints on quality of life in community-dwelling older adults: Memory complaints
and QOL in the elderly. Psychogeriatrics, 14(3), 175–181.
https://doi.org/10.1111/psyg.12056
Marino, S. E., Meador, K. J., Loring, D. W., Okun, M. S., Fernandez, H. H., Fessler, A. J.,
Kustra, R. P., Miller, J. M., Ray, P. G., Roy, A., Schoenberg, M. R., Vahle, V. J., &
Werz, M. A. (2009). Subjective perception of cognition is related to mood and not
performance. Epilepsy & Behavior, 14(3), 459–464.
https://doi.org/10.1016/j.yebeh.2008.12.007
Martins, D., Nelson, K., Pan, D., Tareen, N., & Norris, K. (2001). The effect of gender on age-
related blood pressure changes and the prevalence of isolated systolic hypertension
among older adults: Data from NHANES III. The Journal of Gender-Specific Medicine:
JGSM: The Official Journal of the Partnership for Women’s Health at Columbia, 4(3),
10–13, 20.
Mather, M., Jacobsen, L. A., & Pollard, K. M. (2019). Population bulletin. Population Reference
Bureau, 70(2). https://www.prb.org/wp-content/uploads/2019/07/population-bulletin-
2015-70-2-aging-us.pdf
90
Mayeda, E. R., Glymour, M. M., Quesenberry, C. P., Johnson, J. K., Pérez-Stable, E. J., &
Whitmer, R. A. (2017). Survival after dementia diagnosis in five racial/ethnic groups.
Alzheimer’s & Dementia, 13(7), 761–769. https://doi.org/10.1016/j.jalz.2016.12.008
McMullen, C. K., & Luborsky, M. R. (2006). Self-Rated Health Appraisal as Cultural and
Identity Process: African American Elders’ Health and Evaluative Rationales. The
Gerontologist, 46(4), 431–438. https://doi.org/10.1093/geront/46.4.431
Mielke, M. M., Vemuri, P., & Rocca, W. A. (2014). Clinical epidemiology of Alzheimer’s
disease: Assessing sex and gender differences. Clinical Epidemiology, 6, 37–48.
https://doi.org/10.2147/CLEP.S37929
Mintzer, J., Donovan, K. A., Kindy, A. Z., Lock, S. L., Chura, L. R., & Barracca, N. (2019).
Lifestyle Choices and Brain Health. Frontiers in Medicine, 6, 204.
https://doi.org/10.3389/fmed.2019.00204
Mitchell, A. J., Beaumont, H., Ferguson, D., Yadegarfar, M., & Stubbs, B. (2014). Risk of
dementia and mild cognitive impairment in older people with subjective memory
complaints: Meta-analysis. Acta Psychiatrica Scandinavica, 130(6), 439–451.
https://doi.org/10.1111/acps.12336
Mol, M., Carpay, M., Ramakers, I., Rozendaal, N., Verhey, F., & Jolles, J. (2007). The effect of
perceived forgetfulness on quality of life in older adults; a qualitative review.
International Journal of Geriatric Psychiatry, 22(5), 393–400.
https://doi.org/10.1002/gps.1686
Molarius, A., & Janson, S. (2002). Self-rated health, chronic diseases, and symptoms among
middle-aged and elderly men and women. Journal of Clinical Epidemiology, 55(4), 364–
370. https://doi.org/10.1016/s0895-4356(01)00491-7
91
Montaquila, J., Freedman, V. A., & Kasper, J. D. (2015). National Health and Aging Trends
Study Round 3 income imputation (NHATS Technical Paper #10).
Montejo, P., Montenegro, M., Fernández, M. A., & Maestú, F. (2011). Subjective memory
complaints in the elderly: Prevalence and influence of temporal orientation, depression
and quality of life in a population-based study in the city of Madrid. Aging & Mental
Health, 15(1), 85–96. https://doi.org/10.1080/13607863.2010.501062
Morley, J. E., Morris, J. C., Berg-Weger, M., Borson, S., Carpenter, B. D., del Campo, N.,
Dubois, B., Fargo, K., Fitten, L. J., Flaherty, J. H., Ganguli, M., Grossberg, G. T.,
Malmstrom, T. K., Petersen, R. D., Rodriguez, C., Saykin, A. J., Scheltens, P., Tangalos,
E. G., Verghese, J., … Vellas, B. (2015). Brain health: The importance of recognizing
cognitive impairment: An IAGG consensus conference. Journal of the American Medical
Directors Association, 16(9), 731–739. https://doi.org/10.1016/j.jamda.2015.06.017
Mukadam, N., Cooper, C., & Livingston, G. (2013). Improving access to dementia services for
people from minority ethnic groups: Current Opinion in Psychiatry, 26(4), 409–414.
https://doi.org/10.1097/YCO.0b013e32835ee668
Najafi, F., Moradinazar, M., Hamzeh, B., & Rezaeian, S. (2019). The reliability of self-reporting
chronic diseases: How reliable is the result of population-based cohort studies. Journal of
Preventive Medicine and Hygiene, 60(4), E349–E353. https://doi.org/10.15167/2421-
4248/jpmh2019.60.4.1118
National Council on Aging. (2018). Healthy aging facts. National Council on Aging.
https://d2mkcg26uvg1cz.cloudfront.net/wp-content/uploads/2018-Healthy-Aging-Fact-
Sheet-7.10.18-1.pdf
92
Navickas, R., Petric, V.-K., Feigl, A. B., & Seychell, M. (2016). Multimorbidity: What do we
know? What should we do? Journal of Comorbidity, 6(1), 4–11.
https://doi.org/10.15256/joc.2016.6.72
Neviani, F., Belvederi Murri, M., Mussi, C., Triolo, F., Toni, G., Simoncini, E., Tripi, F.,
Menchetti, M., Ferrari, S., Ceresini, G., Cremonini, A., Bertolotti, M., Neri, G., Squatrito,
S., Amore, M., Zanetidou, S., & Neri, M. (2017). Physical exercise for late life
depression: Effects on cognition and disability. International Psychogeriatrics, 29(7),
1105–1112. https://doi.org/10.1017/S1041610217000576
Ng, R., Sutradhar, R., Yao, Z., Wodchis, W. P., & Rosella, L. C. (2020). Smoking, drinking, diet
and physical activity—Modifiable lifestyle risk factors and their associations with age to
first chronic disease. International Journal of Epidemiology, 49(1), 113–130.
https://doi.org/10.1093/ije/dyz078
Nordström*, A., Hadrévi, J., Olsson, T., Franks, P. W., & Nordström, P. (2016). Higher
prevalence of type 2 diabetes in men than in women is associated with differences in
visceral fat mass. The Journal of Clinical Endocrinology & Metabolism, 101(10), 3740–
3746. https://doi.org/10.1210/jc.2016-1915
Norton, S., Matthews, F. E., Barnes, D. E., Yaffe, K., & Brayne, C. (2014). Potential for primary
prevention of Alzheimer’s disease: An analysis of population-based data. The Lancet
Neurology, 13(8), 788–794. https://doi.org/10.1016/S1474-4422(14)70136-X
Pedro, M. C., Mercedes, M.-P., Ramón, L.-H., & Borja, M. R. (2016). Subjective memory
complaints in elderly: Relationship with health status, multimorbidity, medications, and
use of services in a population-based study. International Psychogeriatrics, 28(11),
1903–1916. https://doi.org/10.1017/S104161021600106X
93
Pizzi, L. T., Weston, C. M., Goldfarb, N. I., Moretti, D., Cobb, N., Howell, J. B., Infantolino, A.,
Dimarino, A. J., & Cohen, S. (2006). Impact of chronic conditions on quality of life in
patients with inflammatory bowel disease. Inflammatory Bowel Diseases, 12(1), 47–52.
https://doi.org/10.1097/01.mib.0000191670.04605.e7
Polak, M. A., Richardson, A. C., Flett, J. A. M., Brookie, K. L., & Conner, T. S. (2015).
Measuring mood: Considerations and innovations for nutrition science (Nutrition for
Brain Health and Cognitive Performance (). Taylor and Francis.
Ponce, N. A., Hays, R. D., & Cunningham, W. E. (2006a). Linguistic disparities in health care
access and health status among older adults. Journal of General Internal Medicine, 21(7),
786–791. https://doi.org/10.1111/j.1525-1497.2006.00491.x
Ponce, N. A., Hays, R. D., & Cunningham, W. E. (2006b). Linguistic disparities in health care
access and health status among older adults. Journal of General Internal Medicine, 21(7),
786–791. https://doi.org/10.1111/j.1525-1497.2006.00491.x
Prince, M., Bryce, R., Albanese, E., Wimo, A., Ribeiro, W., & Ferri, C. P. (2013). The global
prevalence of dementia: A systematic review and metaanalysis. Alzheimer’s & Dementia,
9(1), 63-75.e2. https://doi.org/10.1016/j.jalz.2012.11.007
Reeves, M. J., Bushnell, C. D., Howard, G., Gargano, J. W., Duncan, P. W., Lynch, G.,
Khatiwoda, A., & Lisabeth, L. (2008). Sex differences in stroke: Epidemiology, clinical
presentation, medical care, and outcomes. The Lancet. Neurology, 7(10), 915–926.
https://doi.org/10.1016/S1474-4422(08)70193-5
Reisberg, B., Shulman, M. B., Torossian, C., Leng, L., & Zhu, W. (2010). Outcome over seven
years of healthy adults with and without subjective cognitive impairment. Alzheimer’s &
Dementia, 6(1), 11–24. https://doi.org/10.1016/j.jalz.2009.10.002
94
Rogers, R. G., Everett, B. G., Zajacova, A., & Hummer, R. A. (2010). Educational degrees and
adult mortality risk in the United States. Biodemography and Social Biology, 56(1), 80–
99. https://doi.org/10.1080/19485561003727372
Segel-Karpas, D., & Palgi, Y. (2019). ‘It is nothing more than a senior moment’: The moderating
role of subjective age in the effect of change in memory on self-rated memory. Aging &
Mental Health, 23(2), 272–276. https://doi.org/10.1080/13607863.2017.1399350
Sen, A. (2002). Health: Perception versus observation. BMJ, 324(7342), 860–861.
https://doi.org/10.1136/bmj.324.7342.860
Shi, L., Lebrun, L. A., & Tsai, J. (2009). The influence of English proficiency on access to care.
Ethnicity & Health, 14(6), 625–642. https://doi.org/10.1080/13557850903248639
Short, M. E., Goetzel, R. Z., Pei, X., Tabrizi, M. J., Ozminkowski, R. J., Gibson, T. B., DeJoy,
D. M., & Wilson, M. G. (2009). How accurate are self-reports? Analysis of self-reported
health care utilization and absence when compared with administrative data. Journal of
Occupational and Environmental Medicine, 51(7), 786–796.
https://doi.org/10.1097/JOM.0b013e3181a86671
Smith, D. L. (2009). Health Care Disparities for Persons with Limited English Proficiency:
Relationships from the 2006 Medical Expenditure Panel Survey (MEPS). Journal of
Health Disparities Research and Practice, 3(4).
Sol, K., Zaheed, A. B., Kraal, A. Z., Sharifian, N., Arce Rentería, M., & Zahodne, L. B. (2020).
Psychological predictors of memory decline in a racially and ethnically diverse
longitudinal sample of older adults in the United States. International Journal of
Geriatric Psychiatry, 35(2), 204–212. https://doi.org/10.1002/gps.5236
95
Strachan, M. W. J., Price, J. F., & Frier, B. M. (2008). Diabetes, cognitive impairment, and
dementia. BMJ, 336(7634), 6–6. https://doi.org/10.1136/bmj.39386.664016.BE
Strand, B. H., Knapskog, A.-B., Persson, K., Edwin, T. H., Amland, R., Mjørud, M., Bjertness,
E., Engedal, K., & Selbæk, G. (2018). Survival and years of life lost in various
aetiologies of dementia, mild cognitive impairment (MCI) and subjective cognitive
decline (SCD) in Norway. PLOS ONE, 13(9), e0204436.
https://doi.org/10.1371/journal.pone.0204436
Suemoto, C. K., Gilsanz, P., Mayeda, E. R., & Glymour, M. M. (2015). Body mass index and
cognitive function: The potential for reverse causation. International Journal of Obesity,
39(9), 1383–1389. https://doi.org/10.1038/ijo.2015.83
Sun, J.-H., Tan, L., & Yu, J.-T. (2014). Post-stroke cognitive impairment: Epidemiology,
mechanisms and management. Annals of Translational Medicine, 2(8), 80.
https://doi.org/10.3978/j.issn.2305-5839.2014.08.05
Tam, K. W., & Page, L. (2016). Effects of language proficiency on labour, social and health
outcomes of immigrants in Australia. Economic Analysis and Policy, 52, 66–78.
https://doi.org/10.1016/j.eap.2016.08.003
Taylor, C. A., Bouldin, E. D., Greenlund, K. J., & McGuire, L. C. (2020). Comorbid Chronic
Conditions Among Older Adults with Subjective Cognitive Decline, United States, 2015–
2017. Innovation in Aging, 4(1), igz045. https://doi.org/10.1093/geroni/igz045
Thompson, W. W., Zack, M. M., Krahn, G. L., Andresen, E. M., & Barile, J. P. (2012). Health-
related quality of life among older adults with and without functional limitations.
American Journal of Public Health, 102(3), 496–502.
https://doi.org/10.2105/AJPH.2011.300500
96
Thun, M. J., Peto, R., Lopez, A. D., Monaco, J. H., Henley, S. J., Heath, C. W., & Doll, R.
(1997). Alcohol Consumption and Mortality among Middle-Aged and Elderly U.S.
Adults. New England Journal of Medicine, 337(24), 1705–1714.
https://doi.org/10.1056/NEJM199712113372401
Todd, S., Barr, S., Roberts, M., & Passmore, A. P. (2013). Survival in dementia and predictors of
mortality: A review. International Journal of Geriatric Psychiatry, 28(11), 1109–1124.
https://doi.org/10.1002/gps.3946
U.S. Census. (2014). Table 1: Educational Attainment of the Population 18 year, and over, by
age, sex, race, and Hispanic Origin: 2014. (Current Population Survey 2014: Annual
Social and Economic Supplement.).
http://www.census.gov/hhes/socdemo/education/data/cps/2014/tables.html
U.S. Census Bureau. (2019). Sex by age by educational attainment for the population 18 years
and over.
https://data.census.gov/cedsci/table?q=&t=Age%20and%20Sex%3AEducational%20Atta
inment&tid=ACSDT1Y2019.B15001
Vassilaki, M., Aakre, J. A., Cha, R. H., Kremers, W. K., St Sauver, J. L., Mielke, M. M., Geda,
Y. E., Machulda, M. M., Knopman, D. S., Petersen, R. C., & Roberts, R. O. (2015).
Multimorbidity and risk of mild cognitive impairment. Journal of the American
Geriatrics Society, 63(9), 1783–1790. https://doi.org/10.1111/jgs.13612
Waldorff, F. B., Rishoj, S., & Waldemar, G. (2008). If you don’t ask (about memory), they
probably won’t tell. The Journal of Family Practice, 57(1), 41–44.
Wang, M.-C., Li, T.-C., Li, C.-I., Liu, C.-S., Lin, C.-H., Lin, W.-Y., Yang, C.-W., Yang, S.-Y.,
& Lin, C.-C. (2020). Cognitive function and its transitions in predicting all-cause
97
mortality among urban community-dwelling older adults. BMC Psychiatry, 20(1), 203.
https://doi.org/10.1186/s12888-020-02618-9
White, A., Castle, I.-J. P., Chen, C. M., Shirley, M., Roach, D., & Hingson, R. (2015).
Converging Patterns of Alcohol Use and Related Outcomes Among Females and Males
in the United States, 2002 to 2012. Alcoholism: Clinical and Experimental Research,
39(9), 1712–1726. https://doi.org/10.1111/acer.12815
Williams, J. S., & Egede, L. E. (2016). The Association Between Multimorbidity and Quality of
Life, Health Status and Functional Disability. The American Journal of the Medical
Sciences, 352(1), 45–52. https://doi.org/10.1016/j.amjms.2016.03.004
World Health Organization. (2021). Ageing and health. https://www.who.int/en/news-room/fact-
sheets/detail/ageing-and-health
Yap, K. H., Warren, N., Allotey, P., & Reidpath, D. D. (2020). Chronic disease profiles of
subjective memory complaints: A latent class analysis of older people in a rural
Malaysian community. Aging & Mental Health, 24(5), 709–716.
https://doi.org/10.1080/13607863.2018.1550632
Yates, J. A., Clare, L., Woods, R. T., & MRC CFAS. (2017). Subjective memory complaints,
mood and MCI: A follow-up study. Aging & Mental Health, 21(3), 313–321.
https://doi.org/10.1080/13607863.2015.1081150
Zhang, W., Hong, S., Takeuchi, D. T., & Mossakowski, K. N. (2012). Limited English
proficiency and psychological distress among Latinos and Asian Americans. Social
Science & Medicine, 75(6), 1006–1014. https://doi.org/10.1016/j.socscimed.2012.05.012
98
Zsembik, B. A., & Peek, M. K. (2001). Race differences in cognitive functioning among older
adults. The Journals of Gerontology Series B: Psychological Sciences and Social
Sciences, 56(5), S266–S274. https://doi.org/10.1093/geronb/56.5.S266
99
Appendices
Appendix A. Odds Ratios from Logistic Regression Models for the Association Between Race/Ethnicity and Good/Fair/Poor Self-
rated Memory
Note: All estimates are weighted to account for complex survey design.
Abbreviation: NHATS, National Health and Aging Trends Study; OR, odds ratios; ADLs, activities of daily living; IADLs,
instrumental activities of daily living.
*p<0.05; **p<0.01; ***p<0.001
a
Annual household income was log-transformed.
b
The number of chronic conditions is based on heart attack, hypertension, diabetes, and stroke.
c
Depressive symptoms was evaluated by the Patient Health Questionnaire-2 and values ranged from 0 to 6, with a higher score
indicating more severe depressive symptoms.
d
Objective memory status was based on immediate and delayed recall. Values ranged from 0 to 20, with a higher score indicating
better memory status.
e
Difficulty with activities of daily living (eating, getting in/out of bed, getting in/out of chair, walking, go outside, dressing, bathing,
and toileting). Values ranged from 0 to 8, with a higher score indicating more ADL limitations.
f
Difficulty with instrumental activities of daily living (meal preparation, laundry, light housework, shopping for groceries, banking or
paying bills, keeping track of medication, and using phone calls). Values ranged from 0 to 7, with a higher score indicating more
IADL limitations.
g
Economic vulnerability is measured by whether respondents are covered by Medicaid.
h
Attending religious services was based on whether respondents participated in religious services in the last month.
Variables Model 1
OR (95% CI)
Model 2
OR (95% CI)
Model 3
OR (95% CI)
Model 4
OR (95% CI)
Race/ethnicity (Ref: Non-Hispanic
whites)
Non-Hispanic blacks 1.71 (1.50-1.96)*** 1.43 (1.24-1.65)*** 1.25 (1.08-1.44)** 1.27 (1.11-1.60)**
Hispanics 2.57 (2.03-3.25)*** 1.89 (1.47-2.43)*** 1.70 (1.34-2.16)*** 1.29 (0.94-1.77)
Age (Ref: 65-69)
70-74 1.14 (0.96-1.36) 1.09 (0.91-0.31) 1.04 (0.87-1.26) 1.03 (0.85-1.24)
75-79 1.34 (1.13-1.59)** 1.22 (1.02-1.46) 1.08 (0.90-1.31) 1.08 (0.89-1.30)
80-84 1.84 (1.53-2.21)*** 1.69 (1.40-2.04)*** 1.40 (1.14-1.73)** 1.39 (1.12-1.72)**
85+ 1.86 (1.49-2.33)*** 1.72 (1.36-2.17)*** 1.29 (1.01-1.67)* 1.30 (1.01-1.68)*
Gender (Ref: male)
Female 0.96 (0.86-1.07) 0.92 (0.82-1.03) 0.99 (0.89-1.12) 1.00 (0.89-1.12)
Marital status (Ref: married/living with a
partner
Separated/divorced 1.16 (0.98-1.38) 1.06 (0.89-1.25) 0.98 (0.83-1.16) 0.99 (1.29-1.89)***
Widowed 1.31 (1.11-1.54)** 1.12 (0.95-1.33) 1.06 (0.90-1.26) 1.07 (0.90-1.27)
Never-married 0.79 (0.57-1.09) 0.73 (0.52-1.01) 0.69 (0.50-1.95)* 0.69 (0.51-0.95)*
Education level (Ref: less than high
school)
High school graduate 2.15 (1.77-2.61)*** 1.59 (1.31-1.92)*** 1.56 (1.29-1.89)***
Some college 1.73 (1.47-2.04*** 1.47 (1.23-1.74)*** 1.48 (1.25-1.76)***
College or more 1.30 (1.12-1.50)** 1.15 (0.99-1.33) 1.15 (0.99-1.33)
Annual household income
a
0.89 (0.84-0.95)** 0.93 (0.77-0.99)** 0.93 (0.88-0.98)*
Number of chronic conditions
b
1.11 (1.04-1.19)** 1.11 (1.04-1.19)***
100
Note: All estimates are weighted to account for complex survey design.
Abbreviation: NHATS, National Health and Aging Trends Study; LEP, limited English proficiency; OR, odds ratios; ADLs, activities
of daily living; IADLs, instrumental activities of daily living.
*p<0.05; **p<0.01; ***p<0.001
a
Annual household income was log-transformed.
b
The number of chronic conditions is based on heart attack, hypertension, diabetes, and stroke.
c
Depressive symptoms was evaluated by the Patient Health Questionnaire-2 and values ranged from 0 to 6, with a higher score
indicating more severe depressive symptoms.
d
Objective memory status was based on immediate and delayed recall. Values ranged from 0 to 20, with a higher score indicating
better memory status.
e
Difficulty with activities of daily living (eating, getting in/out of bed, getting in/out of chair, walking, go outside, dressing, bathing,
and toileting). Values ranged from 0 to 8, with a higher score indicating more ADL limitations.
f
Difficulty with instrumental activities of daily living (meal preparation, laundry, light housework, shopping for groceries, banking or
paying bills, keeping track of medication, and using phone calls). Values ranged from 0 to 7, with a higher score indicating more
IADL limitations.
g
Economic vulnerability is measured by whether respondents are covered by Medicaid.
h
Attending religious services was based on whether respondents participated in religious services in the last month.
Depressive symptoms
c
1.24 (1.18-1.29)*** 1.23 (1.18-1.29)***
Objective memory status
d
0.93 (0.91-0.95)*** 0.93 (0.91-0.95)***
Functional limitations
ADLs
e
1.01 (0.95-1.06) 1.01 (0.96-1.06)
IADLs
f
1.04 (1.09-1.08)* 1.04 (0.99-1.08)
Economic vulnerability
g
0.95 (0.78-1.16)
Attending religious service
h
1.02 (0.89-1.16)
Limited English proficiency 2.13 (1.29-3.52)**
N 6,583 6,583 6,583 6,583
101
Appendix B. Odds Ratios from Logistic Regression Models for the Association Between Race/Ethnicity and Good/Fair/Poor Self-
rated Memory, Stratifying by Limited English Proficiency
Variables Model 1
OR (95% CI)
Model 2
OR (95% CI)
Model 3
OR (95% CI)
Model 4
OR (95% CI)
Race/ethnicity (Ref: Non-Hispanic
whites)
Non-Hispanic blacks 1.71 (1.50-1.96)*** 1.44 (1.25-1.66)*** 1.26 (1.09-1.44)** 1.26 (1.09-1.45)**
Hispanics with LEP 5.17 (3.33-8.04)*** 3.25 (2.09-5.07)*** 2.84 (1.85-4.35)*** 2.87 (1.87-4.41)***
Hispanics without LEP 1.56 (1.13-2.16)** 1.35 (0.97-1.89) 1.25 (0.88-1.77) 1.25 (0.88-1.76)
Age (Ref: 65-69)
70-74 1.13 (0.95-1.34) 1.08 (0.89-1.30) 1.03 (0.85-1.24) 1.03 (0.85-1.24)
75-79 1.33 (1.12-1.58)** 1.22 (1.02-1.45)* 1.08 (0.89-1.31) 1.08 (0.89-1.30)
80-84 1.83 (1.51-2.20)*** 1.69 (1.40-2.04)*** 1.39 (1.13-1.72)** 1.39 (1.12-1.71)**
85+ 1.85 (1.48-2.32)*** 1.71 (1.36-2.17)*** 1.29 (1.00-1.67)* 1.29 (1.00-1.67)*
Gender (Ref: male)
Female 0.96 (0.86-1.07) 0.92 (0.82-1.03) 1.00 (0.89-1.12) 1.00 (0.88-1.12)**
Marital status (Ref: married/living with a
partner
Separated/divorced 1.17 (0.99-1.39) 1.07 (0.90-1.26) 0.98 (0.83-1.17) 0.99 (0.83-1.16)
Widowed 1.31 (1.12-1.53)** 1.13 (0.96-1.34) 1.07 (0.90-1.26) 1.07 (0.91-1.27)
Never-married 0.77 (0.56-1.06) 0.72 (0.52-0.99)* 0.68 (0.50-0.93)* 0.69 (0.50-0.94)
Education level (Ref: less than high
school)
High school graduate 2.08 (1.71-2.52)*** 1.53 (1.27-1.85)*** 1.54 (1.28-1.87)***
Some college 1.75 (1.49-2.05)*** 1.47 (1.24-1.74)*** 1.47 (0.24-1.74)
College or more 1.30 (1.12-1.51)** 1.15 (0.99-1.32) 1.15 (0.99-1.32)
Annual household income
a
0.90 (0.85-0.96)** 0.93 (0.88-0.98)* 0.93 (0.88-0.99)*
Number of chronic conditions
b
1.11 (1.04-1.19)*** 1.11 (1.04-1.19)**
Depressive symptoms
c
1.23 (1.17-1.20)*** 1.24 (1.18-1.29)***
102
Note: All estimates are weighted to account for complex survey design.
Abbreviation: NHATS, National Health and Aging Trends Study; LEP, limited English proficiency; OR, odds ratios; ADLs, activities
of daily living; IADLs, instrumental activities of daily living.
*p<0.05; **p<0.01; ***p<0.001
a
Annual household income was log-transformed.
b
The number of chronic conditions is based on heart attack, hypertension, diabetes, and stroke.
c
Depressive symptoms was evaluated by the Patient Health Questionnaire-2 and values ranged from 0 to 6, with a higher score
indicating more severe depressive symptoms.
d
Objective memory status was based on immediate and delayed recall. Values ranged from 0 to 20, with a higher score indicating
better memory status.
e
Difficulty with activities of daily living (eating, getting in/out of bed, getting in/out of chair, walking, go outside, dressing, bathing,
and toileting). Values ranged from 0 to 8, with a higher score indicating more ADL limitations.
f
Difficulty with instrumental activities of daily living (meal preparation, laundry, light housework, shopping for groceries, banking or
paying bills, keeping track of medication, and using phone calls). Values ranged from 0 to 7, with a higher score indicating more
IADL limitations.
g
Economic vulnerability is measured by whether respondents are covered by Medicaid.
h
Attending religious services was based on whether respondents participated in religious services in the last month.
Objective memory status
d
0.93 (0.91-0.94)*** 0.93 (0.91-0.94)***
Functional limitations
ADLs
e
1.00 (0.95-1.06) 1.01 (0.96-1.06)
IADLs
f
1.03 (0.99-1.08) 1.04 (0.99-1.08)
Economic vulnerability
g
0.95 (0.74-1.16)
Attending religious service
h
1.01 (0.89-1.15)
N 6,583 6,583 6,583 6,583
103
Appendix C. Distribution of Chronic Disease Stratified by Those Who Report Excellent/Very
Good/Good Self-rated Memory and Those Who Report Fair/Poor Self-rated Memory
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
0 or 1 2 3 4 5 or more
Percent
Number of Chronic Diseases
The Prevalence of Self-rated Memory by Number of Chronic
Diseases
Excellent/Very Good/Good Fair/Poor
104
Appendix D. Combinations of Chronic Diseases
Types of chronic
diseases
Heart
disease
Heart
attack
Hypertension Diabetes Stroke Osteoporosis Arthritis
Lung
disease
Heart disease
__ 6.62 12.96 6.00 2.68 3.99 10.68 4.41
Heart attack __ __ 10.07 4.89 2.74 2.92 8.25 3.33
Hypertension
__ __ __ 18.42 7.05 13.53 37.04 11.06
Diabetes
__ __ __ __ 3.41 4.45 13.58 4.07
Stroke __ __ __ __ __ 2.48 5.80 2.11
Osteoporosis __ __ __ __ __ __ 14.71 4.79
Arthritis __ __ __ __ __ __ __ 10.49
Lung disease __ __ __ __ __ __ __ __
105
Appendix E. The Bivariate Associations Between Each Chronic Disease and Fair/Poor Self-Rated Memory
Self-Rated Memory
X
2
Types of chronic disease
Excellent/Very good/good Fair/Poor
Heart disease 13.16 3.80 21.01***
Heart attack
10.57 3.19 28.08***
Hypertension
52.06 11.64 15.18**
Diabetes 18.30 5.15 38.70***
Stroke 6.45 2.61 61.70***
Osteoporosis
16.44 4.07 14.08***
Arthritis 42.45 10.78 53.23***
Lung disease 12.12 3.47 21.43***
106
Appendix F. Survival curves by gender and cognitive status
Abstract (if available)
Abstract
As older adults live longer and the number of older adults is increasing, it is critical to maintain healthy cognition to live healthy lives. Cognitive health has been an important field of research in gerontology and the study of aging. Researchers have shown the importance of cognitive health at older ages, documenting that cognitive health declines with age and is closely associated with older adults’ quality of life and mortality. Needless to say, maintaining healthy cognition is important not only for older individuals but also for their caregivers and public health planning. Self-rated memory has been highlighted in recent years due to its usefulness as an indicator of general cognitive health and its ease of administration. Self-rated memory is commonly utilized in clinical settings to assess cognitive health (Marino et al., 2009) because perceived memory change can be an early indication of memory decline as well as dementia onset. Older adults with negative self-rated memory tend to experience various types of negative health outcomes (e.g., functional limitations, physical disabilities, psychological distress, etc.). This dissertation fills a gap in the existing literature by exploring how self-rated memory varies by race/ethnicity, how multimorbidity is linked to self-rated memory, and how cognitive impairment impacts the mortality of older adults by gender. ❧ The aim of this dissertation is to further our understanding of the cognitive status of older adults. The first study explores differences in self-rated memory across non-Hispanic white, non-Hispanic black, and Hispanic older adults in the U.S., including how limited English proficiency impacts self-rated memory by race/ethnicity (Chapter 2). This chapter has important implications for individuals who are responsible for administering cognitive testing. Also, the findings from this chapter may contribute to timely or early diagnosis of cognitive impairment among older adults. The second chapter of this dissertation explores whether multimorbidity and specific types of chronic diseases are associated with poor self-rated memory (Chapter 3). This chapter explores which of eight individual chronic diseases are linked to reporting poor self-rated memory. Given that the prevalence of chronic disease and poor self-rated memory increase with age, this study provides evidence for the link between multimorbidity and self-rated memory among older adults at earlier stages of cognitive impairment. Finally, the dissertation investigates how cognitive impairment contributes to the mortality of older adults by gender (Chapter 4), and it aims to identify possible factors that may contribute to gender differences in mortality. This dissertation consists of three studies which used secondary data from the National Health and Aging Trends Study (NHATS) and the Health and Retirement Study (HRS), which are nationally representative samples of older adults in the United States.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Self-perceptions of Aging in the Context of Neighborhood and Their Interplay in Late-life Cognitive Health
PDF
Life course implications of adverse childhood experiences: impacts on elder mistreatment, subjective cognitive decline, and caregivers' health
PDF
Three essays on modifiable determinants of shingles: risk factors for shingles incidence and factors affecting timing of vaccine uptake
PDF
Biomarkers of age-related health changes: associations with health outcomes and disparities
PDF
The effects of time and space on health status of older adults in China
PDF
Health differences between the elderly in Japan and the United States
PDF
Palliative care: breaking the knowledge barrier among community-based older adults
PDF
How emotional arousal influences memory and learning in younger and older adults
PDF
Structure and function of the locus coeruleus across the lifespan
PDF
Predicting cognitive decline and dementia in elderly twins from indicators of early life oral health
PDF
Mechanisms of stress effects on learning and decision making in younger and older adults
PDF
Age integration in late life: sociodemographic & psychosocial correlates of intergenerational-only, peer-only, and age-integrated social networks
PDF
Social determinants of physiological health and mortality in China
PDF
A biodemographic approach to understanding sociodemographic disparities in kidney functioning on three dimensions: individual, population, and cross-national
PDF
Lifting up ourselves by lifting up others: examining cognitive-affective states linking generativity to well-being
PDF
Is stress exposure enough? Race/ethnic differences in the exposure and appraisal of chronic stressors among older adults
PDF
The impact of emotion, valence, and arousal on differential memory processes in younger and older people
PDF
Estimating survival in the face of pain: evidence from the health and retirement study
PDF
Neuroinflammation and the behavioral consequences of air pollution over the life course
PDF
Thinking generatively versus acting generatively: exploring the associations of generative self-concept and generative activity with cognitive function among older adults
Asset Metadata
Creator
Jeong, Yu Jin
(author)
Core Title
Cognitive health and self-rated memory in later life: linkages to race/ethnicity, multimorbidity, and survival
School
Leonard Davis School of Gerontology
Degree
Doctor of Philosophy
Degree Program
Gerontology
Degree Conferral Date
2022-05
Publication Date
03/01/2022
Defense Date
02/07/2022
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cognitive health,multimorbidity,OAI-PMH Harvest,race/ethnicity,self-rated memory,Survival
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Ho, Jessica Y. (
committee chair
), Jang, Yuri (
committee member
), Saenz, Joseph L. (
committee member
)
Creator Email
yujinfranco@gmail.com,yujinjeo@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC110765148
Unique identifier
UC110765148
Legacy Identifier
etd-JeongYuJin-10410
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Jeong, Yu Jin
Type
texts
Source
20220301-usctheses-batch-914
(batch),
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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
cognitive health
multimorbidity
race/ethnicity
self-rated memory