International sex and age differences in physical function and disability

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

INTERNATIONAL SEX AND AGE DIFFERENCES IN PHYSICAL FUNCTION AND DISABILITY

by

Felicia Victoria Wheaton
____________________________________________________________________________

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)

August 2014

ii
DEDICATION

This dissertation is dedicated to my grandmother, Nancy Wiele, for her unconditional love
and unfaltering encouragement to always reach for my dreams. I love you into infinity.

iii
ACKNOWLEDGMENTS

I would like to thank my dissertation committee members, Dr. Jinkook Lee and Dr.
Aaron Hagedorn, for their valuable insights. Their probing questions helped me to step
back from the details of the results to see the big picture (and then zoom back in to those
details). Thank you to my mentor, Dr. Eileen Crimmins, for her guidance and support, from
my first semester to the last. Thank you for always encouraging me to focus on “the story
iv
TABLE OF CONTENTS

Dedication ii

Acknowledgments iii

List of Tables vi

List of Figures vii

Abstract viii

Chapter 1: Introduction 1

Chapter 2: Female Disability Disadvantage: A Global Perspective on Sex Differences 20
in Physical Function and Disability

Chapter 3: A Global Perspective on Age Differences in Physical Function and 45
Disability

Chapter 4: Rate of Change in Physical Function and Disability in Traditional Versus 73
Modern Contexts

Chapter 5: Conclusion 87

References 96

Appendix A: Question Wording and Grip Strength Protocol by Dataset 109

Appendix B: Characteristics of Included Versus Missing Respondents 113

Appendix C: OLS Models of Grips Strength Controlling for Height and Weight 114
Versus BMI

Appendix D: Comparison of OLS and Heckman Selection Models of Sex Differences 115
for Continuous Physical Performance Measures

Appendix E: Comparison of Models of Sex Differences with Data Pooled Across 116
Countries Versus Models Stratified by Country

Appendix F: Tsimane Longitudinal Data 120

Appendix G: Detailed Descriptive Data on Change by Age Group 121

Appendix H: Alternative Definition of Gait Speed Decline 122

v
Appendix I: Comparison of Different Time Intervals: Tsimane 123

Appendix J: Comparison of ADL Change: 2006-2008 Versus 2006-2010 (US) 128

Appendix K: Comparison of ADL and Gait Speed Change: 2004-2006 130
Versus 2004-2008 (US)

Appendix L: Random Subset of U.S. Data to Obtain Sample Sizes Equal to the 133
Tsimane

vi
LIST OF TABLES

Table 2.1: Characteristics of study countries 38

Table 2.2: Survey characteristics 39

Table 2.3: Survey sample characteristics 40

Table 2.4: Available measures by country 41

Table 2.5: Objective performance: Odds ratios and regression coefficients of the 42
effect of being female

Table 2.6: Self-reported measures: Odds ratios of the effect of being female 43

Table 2.7: Correlations between country-level indicators and the effect of sex 44

Table 3.1: Characteristics of study countries 63

Table 3.2: Survey characteristics 64

Table 3.3: Survey sample characteristics 65

Table 3.4: Available measures by country 66

Table 3.5: Physical performance: Odds ratios and regression coefficients of the 69
effect of age

Table 3.6: Self-reported measures: Odds ratios of the effect of age 70

Table 3.7: Correlations between country-level indicators and the effect of age 71

Table 3.8: Correlations between country-level indicators and prevalence/average 72
level of difficulties and performance

Table 4.1: Baseline characteristics and average change 85

Table 4.2: Hazard ratios from parametric survival models of ADL increase or 86
balance/gait speed decline

vii

LIST OF FIGURES

Figure 1.1: Trends in GNP per capita by country 9

Figure 1.2: Trends in infant mortality rate by country 10

Figure 1.3: Trends in age-standardized body mass index (BMI) for men and 11
women aged 20+ by country

Figure 1.4: Trends in the percentage of the labor force in agriculture and 12
forestry by country

Figure 1.5: Trends in access to improved water by country 13

Figure 1.6: Trends in average years of education among men aged 25+ by country 14

Figure 1.7: Trends in life expectancy at birth by country 15

Figure 1.8: Trends in life expectancy at age 60 by country 16

Figure 1.9: Trends in gender differences in average years of education among 17
adults 25+ by country

Figure 1.10: Trends in the ratio of female to male employment among those older 18
than 15 years by country

Figure 1.11: Prevalence of tobacco use among those aged 15+ by sex and country 19

Figure 3.1: Average physical performance by age and country 67

Figure 3.2: Prevalence of reported functioning and ADL difficulties by age 68
and country

Figure 5.1: Per capita GNP versus the effect of being female (regression coefficients) 91

Figure 5.2: Per capita GNP versus the effect of being female (odds ratios from 92
logistic regression)

Figure 5.3: Per capita GNP versus the effect of age 75-85 relative to 55-64 years 93
(regression coefficients)

Figure 5.4: Per capita GNP versus the effect of age 75-85 relative to 55-64 years 94
(odds ratios from logistic regression)

Figure 5.5: Per capita GNP versus prevalence of difficulty among those aged 55-64 95
viii
ABSTRACT

Worldwide population aging will undoubtedly be accompanied by an increase in the
number of disabled older adults. Female gender and increased age are two of the most
widely identified risk factors for poor physical functioning and disability. Yet the contexts
in which people are aging vary markedly across countries. Countries differ greatly in their
level of economic development, both past and present. Economic development is in turn
related to improvements in infrastructure, health care, public health, education, etc. that
are hypothesized to be related to improved physical function and less disability.
Therefore, this dissertation examined whether sex and age differences/changes in
both objectively-measured physical performance and reported difficulties with functional
tasks and activities of daily living (ADLs) were similar or varied across seven countries
whose per capita GNP ranged from $200 to $40,100 (United Sates, Taiwan, Korea, Mexico,
China, Indonesia, and the Tsimane of Bolivia). It also sought to determine if sex differences
and age differences varied systematically in terms of macro-level indicators including GDP,
life expectancy, and measures of gender equality.
Overall, sex differences were remarkably consistent across countries with very
different contexts. Sex differences in physical performance and functional limitations were
more pronounced than sex differences in difficulty with basic self-care tasks, but the
magnitude of differences did not vary systematically in relation to country-level measures
of development or gender equality. This may be because gender equality can be either
protective or detrimental, depending on the domain.
In terms of age differences, it was necessary to consider both the level of
performance/prevalence of difficulty at younger ages as well as age differences, since poor
ix
performance/high levels of difficulty among the young-old indicate that “aging” has already
occurred. Some populations did appear to be “aging” more rapidly, particularly those at the
lowest end of the development spectrum, however, there was no clear evidence for a linear
correlation between macro-level indicators of development and age differences.
Interestingly, findings showed that functioning in some domains could be fairly well
maintained despite declines in other domains, and these varied across countries. For
example, Indonesians appeared to be “aging” more rapidly in terms of upper body strength,
but showed relatively high levels of lower body function and less age-related decline. This
may be due to differences across populations in patterns of work, physical activity, the built
environment, etc.

1
CHAPTER 1: INTRODUCTION

In 2000, there were 606 million people aged 60 or older; by 2050, that number will
have climbed to nearly 2 billion (WHO World Population Ageing 1950-2050). Measured
physical performance declines with age and self-reported functional limitations and
disability become more common. Thus, as population aging progresses worldwide, there
will undoubtedly be an increase in the total number of disabled persons.
Functional impairments and disability are linked to increased risk for a variety of
negative outcomes. For example, disability has been linked both to loneliness (Drageset,
2004) and increased risk of falling (Shumway-Cook et al., 2009; Tromp et al., 2001). Those
with functional limitations and disability are also more likely to lose independence and
become institutionalized (Gaugler, Duval, Anderson, & Kane, 2007; Mor, Wilcox, Rakowski,
& Hiris, 1994). In a review of predictors of institutionalization in the elderly, 96% of studies
found a statistically significant positive association between functional impairment and
risk of nursing home placement (Luppa et al., 2010). In addition, not only are disabled
older adults more likely to need long term care, they are also more likely to be hospitalized
(Mor et al., 1994).
Higher acute and long-term care needs lead to higher costs among the disabled and
those with lower physical performance (Chan et al., 2002; Reuben et al., 2004). For
instance, a study by Liu and colleagues (1997) found that among community residents,
Medicare payments were three times higher among older adults with three or more ADL
limitations than for those with no disabilities. Finally, poor physical performance and ADL
limitations have been associated with greater risk of mortality (Beltran, Cuadrado, Martin,
Carbajal, & Moreiras, 2001; Cooper, Kuh, & Hardy, 2010; Fried et al., 1998; Guralnik, Fried,
2
& Salive, 1996; Keeler, Guralnik, Tian, Wallace, & Reuben, 2010; Scott, Macera, Cornman, &
Sharpe, 1997). Given the impact of disability on older adults’ lives and the number of
people affected, it is important to better understand the process of disability and how it
may vary across contexts.
To better understand the disablement process (Verbrugge & Jette, 1994), this
dissertation examines two of the most widely identified risk factors for poor physical
functioning and disability, age and sex, and examines the extent to which age and sex
differences are similar or different across a wide range of contexts. This dissertation seeks
to address the following questions: Do women always report greater disability and display
poorer performance on physical tasks? Are age-related declines in physical ability and
increases in disability consistent across countries or do individuals in some populations
“age” more quickly than others? Are longitudinal (intra-individual) changes in functioning
and disability consistent with cross-sectional age differences? Are findings consistent with
respect to different stages of the disablement process (functional impairments versus
disability) and methods of measurement (objective measures versus self-reports)? Do age
and sex differences vary systematically with respect to level of development?
Countries
To address these questions, data from seven populations were analyzed. These
include the United Sates, Taiwan, Korea, Mexico, China, Indonesia, and the Tsimane, a
traditional forager-horticulturalist population of Bolivia. These countries represent a wide
range in terms of economic development. In 1985, per capita GNP ranged from only $300 in
China to $14,300 in the United States (Figure 1) (CIA World Factbook, 2012). Although per
capita GNP has increased in all countries over time, large absolute differences remain.
3
Other characteristics and indicators that tend to change with economic
development differ across these countries, such as access to improved water, average life
expectancy, infant mortality rates, level of education, occupations, BMI, etc. These will be
described in greater detail below. Not only do differences exist at present, but countries
have also followed different trajectories of change over time. This is important because
conditions earlier in life are linked to poor functioning and disability in later life.
In addition, these countries also differ greatly in terms of normative gender roles,
social policies, culture, etc. Therefore, these populations serve as a natural experiment for
testing the universality of age and sex differences. The second chapter examines the
consistency of sex differences in disability and physical function across these seven
populations; the third chapter examines age differences across these same populations;
and the fourth chapter examines longitudinal change in these outcomes in the United States
and among the Tsimane. Theoretically, there is reason to hypothesize that age and sex
differences may vary across such contexts; this will be discussed in the next section.
In addition to theoretical motivations, from a practical standpoint, these seven
populations were selected because of the availability of recent, high-quality, population-
based survey data with comparable measures that included objective physical performance
measures. Surveys were collected between 2001 and 2011 and all either focused on older
adults or included substantial numbers of older adults. All included at least one physical
performance measure and considerable effort has been made to harmonize measures and
methodologies across these surveys. Self-reported items and their wording can be found in
Appendix A, as well as grip strength measurement protocol and equipment.
Age Differences
4
Current elderly populations in Asia and Latin America have lived through changing
epidemiologic conditions. Epidemiological transitions, characterized by a shift in mortality
from primarily infectious diseases to chronic ones, began earlier in developed countries but
are still underway in the developing world (Omran, 1971). As a result, lifetime exposure to
infection and inflammation may be higher in less developed regions where infectious
diseases are still quite common (Boutayeb, 2006). Elevated levels of inflammatory markers
such as C-reactive protein (CRP) and interleukin 6 have been linked to poor performance in
both cross-sectional and longitudinal studies (Cesari et al., 2004; Colbert et al., 2004;
Ferrucci et al., 1999). Figure 2 shows that infant mortality rates (which are strongly linked
to infectious disease) have historically been much higher in less developed countries such
as Indonesia and China, although they have declined over time (CIA World Factbook, 2012).
Older adults in such contexts may be particularly vulnerable to becoming disabled as many
have experienced both childhood infection and chronic disease in later life.
Chronic diseases are also associated with greater risk of poor physical functioning
and disability. In a study by Griffith and colleagues (2010), approximately 66% of ADL
disability was attributed to chronic conditions. Being overweight or obese and physical
inactivity are also associated with disability and poor function (Houston et al., 2009), both
directly, and indirectly because they are well-known risk factors chronic disease. Yet BMI
and physical activity vary across countries and over time. In general, age-standardized
average BMI among adults aged 20 and over tends to increase with economic development,
both at a single time point and over time (Figure 3) (Imperial College London, 2014).
However, BMI is somewhat lower among Asian countries relative to the U.S. and Mexico.
Although international data on physical activity over time is not available, Figure 4 depicts
5
the percentage of the labor force engaged in agriculture and forestry. These types of
occupations are physically demanding and are much more common in less developed
countries like China and Indonesia compared with the United States (CIA World Factbook,
2012). However, the proportion engaged in agriculture and forestry occupations has
declined over time (and with increasing development) in all countries. Overall, differences
in the prevalence of chronic diseases and their risk factors (e.g. high BMI and physical
inactivity), which are clearly patterned by level of development, may lead to differences
between countries in physical performance and disability.
In addition to epidemiologic context, there are several other characteristics that
vary systematically between countries that differ in levels of development. In general,
infrastructure is poorer in less developed countries, as can be seen by comparing access to
improved water. Nearly 100% of the U.S. population had access to improved water in 1990
compared to only 67% in China and 70% in Indonesia, although access has improved in
such countries in more recent years (Figure 5) (World Bank, 2014). Older adults in less
developed countries also have had fewer lifetime economic resources and live in societies
with few institutionalized safety nets. For instance, average education is lower among men
aged 25+ (and women, not shown) in lower-income countries (Figure 6) (Gapminder,
2014) and low education is a risk factor for poor functioning and disability (Jagger et al.,
2007). In addition, access to health care is limited or of lower quality than in more
developed countries; therefore, injuries and health conditions may be relatively more
disabling. In light of the disadvantages faced by the elderly in less developed countries, we
hypothesize that age-related declines in physical function and increases in disability would
6
occur earlier and be more pronounced in such contexts. However, selective survival of
healthier adults may counter this.
Average life expectancy at birth has varied considerably across countries
historically and is associated with economic development (Figure 7) (World Bank, 2014). A
similar pattern is observed for life expectancy at age 60 (Figure 8) (World Health
Organization (WHO), 2013). In countries with lower life expectancy, those in the worst
health may be less likely to survive, leading to better health at the population level.
Sex Differences
With respect to sex differences, one might expect male/female differences in
physical functioning and disability to be more pronounced in traditional cultures where
gender roles may be more demarcated. Greater socioeconomic status has been associated
with better functioning and lower risk of disability in numerous studies, but women
generally have lower average education and income. Yet there is variation in across
countries in male/female educational differences (Figure 9) (Gapminder, 2014). Gender
differences in years of schooling among adults aged 25 and older has been much smaller in
the U.S. compared with countries such as Korea, China, Indonesia and Taiwan, although
gaps have narrowed in recent years in nearly all countries. Sex differences in labor force
participation also vary considerably. The ratio of female to male employment among those
older than 15 tends to be higher in more developed countries such as the U.S., Taiwan, and
Korea (Figure 10) (International Labour Organization (ILO), 2014). However, China is an
exception to this pattern; female employment is very high, both historically and at present,
which may be due to China’s unique history, including communism.
7
In addition to differences in socioeconomic status, male/female differences in
physical function and disability may be related to epidemiologic transitions. For instance,
Myers and colleagues (2003) hypothesize that during epidemiologic transition, differences
in disability prevalence widens between men and women. Changes in chronic disease
incidence and survival likely influence the magnitude of male/female differences because
women tend to suffer more from disabling but non life-threatening conditions, whereas
men suffer disproportionately from diseases that have higher mortality rates (Verbrugge &
Wingard, 1987). Indeed, as infectious and childbirth-related deaths have dropped,
male/female differences in life expectancy have increased so that women outlive men in all
countries (Barford, 2006). However, female life expectancy at birth is not higher than that
of men among the Tsimane.
Differences in health behaviors between men and women may also contribute to
differences in diseases and ultimately disability. For example, the sex differences in
prevalence of tobacco use among adults 15 years and older varies across countries and is
much wider in Asia and to a lesser extent in Mexico, compared with the United States,
where the prevalence of tobacco use was almost as prevalent among women compared to
men in 2005 (Figure 11) (WHO, 2013).
Conclusion
Overall, not only does this set of papers contribute to our knowledge of age and sex
differences, it also adds to our understanding of the disablement process across developed
contexts. Since the average annual growth rate in the older population is much higher in
less developed regions of the world, 4 in 5 older adults will live in less developed regions
8
by 2050 (WHO World Population Ageing 1950-2050). The majority of older adults
vulnerable to becoming disabled will be increasingly concentrated in developing countries.

9

Figure 1.1. Trends in GNP per capita by country: CIA. (2012). The World Factbook.
Retrieved from https://www.cia.gov/library/publications/the-world-factbook/.
10

Figure 1.2. Trends in infant mortality rate (per 1,000 live births) by country: CIA. (2012).
The World Factbook. Retrieved from https://www.cia.gov/library/publications/the-world-
factbook/.
11

Figure 1.3. Trends in age-standardized body mass index (BMI) for men and women aged
20+ by country: Imperial College London. (2014). Global Burden of Metabolic Risk Factors
of Chronic Diseases. Retrieved from
http://www1.imperial.ac.uk/publichealth/departments/ebs/projects/eresh/majidezzati/
healthmetrics/metabolicriskfactors/
12

Figure 1.4. Trends in the percentage of the labor force in agriculture and forestry by
country: CIA. (2012). The World Factbook. Retrieved from
https://www.cia.gov/library/publications/the-world-factbook/.

13

Figure 1.5. Trends in access to improved water by country: World Bank. (2014). World
Development Indicators Database. Retrieved from http://data.worldbank.org/data-
catalog/world-development-indicators
14

Figure 1.6. Trends in average years of education among men aged 25+ by country:
Gapminder. (2014). Institute For Health Metrics and Evaluation Data. Retrieved from
http://www.gapminder.org/data/
15

Figure 1.7. Trends in life expectancy at birth by country: World Bank. (2014). World
Development Indicators Database. Retrieved from http://data.worldbank.org/data-
catalog/world-development-indicators
16

Figure 1.8. Trends in life expectacy at age 60 by country: World Health Organization
(WHO). (2013). Global Health Observatory Data Repository. Retrieved from
http://apps.who.int/gho/data/node.main?lang=en.
17

Figure 1.9. Trends in gender differences in average years of education among adults 25+ by
country: Gapminder. (2014). Institute For Health Metrics and Evaluation Data. Retrieved
from http://www.gapminder.org/data/
18

Figure 1.10. Trends in the ratio of female to male employment among those older than 15
years by country: International Labour Organization (ILO) (2014). Key Indicators of the
Labour Market (KILM). Retrieved from
http://www.ilo.org/empelm/what/WCMS_114240/lang--en/index.htm
19

Figure 1.11. Prevalence of tobacco use among those aged 15+ by sex and country: World
Health Organization (WHO). (2013). Global Health Observatory Data Repository. Retrieved
from http://apps.who.int/gho/data/node.main?lang=en.

20
CHAPTER 2: FEMALE DISABILITY DISADVANTAGE: A GLOBAL PERSPECTIVE ON SEX
DIFFERENCES IN PHYSICAL FUNCTION AND DISABILITY
Abstract
Objectives: To determine whether women always fare more poorly in terms of physical
function and disability across countries that vary widely in terms of their level of
development, epidemiologic context and level of gender equality.
Methods: Sex differences in self-reported and objective measures of disability and physical
function were compared among older adults aged 55-85 in the U.S., Taiwan, Korea, Mexico,
China, Indonesia, and among the Tsimane of Bolivia using population-based studies
collected between 2001 and 2011. Data were analyzed using logistic and OLS regression.
Confidence intervals were examined to see whether the effect of being female differed
significantly between countries.
Results: In all countries, women had consistently worse physical functioning (both self-
reported and objectively-measured). Women also tended to report more difficulty with
ADL tasks, although differences were not always significant. In general, sex differences
across measures were less pronounced in China. In Korea, women had significantly lower
grip strength, but sex differences in ADL difficulty were non-significant or even reversed.
Discussion: Overall, there was striking similarity in the magnitude and direction of sex
differences across countries despite considerable differences in context, although modest
variations in the effect of sex were observed.
21
Introduction
It is widely-accepted that older women suffer from higher rates of disability
compared to men. Women tend to live longer than men, however, they tend to suffer more
from disabling but non life-threatening conditions, whereas men suffer disproportionately
from diseases that have higher mortality rates (Verbrugge & Wingard, 1987). This pattern
has been documented in a number of countries, although to date, the majority of studies on
sex differences in disability have been conducted in developed countries such as the United
States, Japan, and European countries (Crimmins, Kim, & Solé-Auró, 2011; Minicuci et al.,
2004; Oksuzyan et al., 2010). More recently, researchers have begun to examine differences
between men and women in less developed countries. For instance, women have been
found to report significantly higher levels of disability in both Mexico (Wong et al., 2010)
and Guatemala (Yount, Hoddinott, & Stein, 2010). Disability prevalence is also higher
among women in India (Roy & Chaudhuri, 2008), Indonesia (Kaneda & Zimmer, 2007),
China (Kaneda et al., 2009; Wang et al., 2009), and other Asian countries (Ofstedal et al.,
2007; Zimmer, Linda, & Chang, 2002). Yount and Agree (2005) found a similar pattern in
Egypt and Tunisia. However, given the large range of contexts in which adults are ageing
worldwide, particularly in terms of level of development and with respect to gender roles,
it is likely that sex differences in disability are more pronounced in some countries than
others.
Yet little is known about under what conditions the gender gap is smaller, non-
existent, or even reversed. Countries differ in many characteristics and therefore,
examining gender differences across different populations can serve as a “natural
experiment” to test whether sex differences are universal or are specific to time and place.
22
Gender differences in physical function and disability may vary across countries because
men and women are exposed to protective and risk factors differentially in different
contexts.
One major way in which countries differ is in terms of level of development.
Increasing economic development has been linked to changes in many well-known
proximal risk factors for poor physical function and disability including the amount of
physical labor, obesity, and increased importance of chronic conditions such as arthritis
and cardiovascular disease. With increasing levels of development and urbanization, there
is a shift from under- to over-nutrition and towards more sedentary lifestyles (Popkin,
1999). It is hard to say whether these changes will affect men or women more. On the other
hand, greater economic development and higher standards of living are linked with a larger
gap in life expectancy between men and women (Kinsella & Velkoff, 2001). Women may be
at a greater risk of disability as chronic diseases become more prevalent, the survival of
both men and women increases and as the survival gap between the sexes increases
(Myers, Lamb, & Agree, 2003).
Countries also differ markedly in cultural attitudes concerning normative gender
roles, as gender is not only biologically-rooted, but is also socially constructed. According to
the differential exposure hypothesis of health, women are exposed to more risk factors and
fewer protective factors (Denton, Prus, & Walters, 2004). Compared with men, women tend
to experience greater disadvantage in nearly all societies – for example, women are more
likely to live in poverty, work in low-paying occupations, and become widowed. These
disadvantages in upstream or distal factors may translate into poorer functioning and
greater disability among women (Kaneda et al., 2009). On the other hand, according to the
23
differential vulnerability hypothesis, women and men respond differently to protective and
risk factors (Denton, Prus, & Walters, 2004). Thus, it is possible that women may
experience poorer physical function and greater disability even in contexts where women
and men live similar lives, and are exposed to similar risk factors. There is some evidence
to support this hypothesis. For instance, women with arthritis are more likely than men
with arthritis to have an ADL or IADL disability (Peek & Coward, 1999). The association
between body mass index and mobility difficulty is also stronger for women versus men
(Wray & Blaum, 2001).
However, the magnitude of sex differences in disability likely varies depending on
the specific nature of gender roles, which vary both geographically and temporally. In more
traditional or less developed societies, gender roles tend to be more pronounced. Women
in such contexts may be much less likely to engage in risky behaviors such as drinking or
smoking, but they may also have less education and fewer economic resources, and have
experienced greater physiological wear and tear due to pregnancy and childbearing;
factors that could lead to greater risk of disability. In contrast, women’s behaviors (e.g.
smoking), environmental exposures (e.g. work stress, pollution), and access to resources
tend to be more similar to men’s in industrialized developed nations with strong social
protections. In contexts where gender roles are similar, differences in disability may also
be smaller. For example, Chun and colleagues (2012) suggest that the recent narrowing of
the gender gap in self-rated health in Korea is attributable to increasing gender equality.
Findings from individual studies hint at universal patterns of sex differences, but
few studies have attempted to synthesize differences between men and women across
countries, especially less developed countries. Previously, Rahman and colleagues (1994)
24
examined sex differences in self-reported functional limitations and disability in Jamaica,
Malaysia, Bangladesh and the United States but did not formally compare sex differences
between countries. Another study compared sex differences across seven Latin American
cities (Alvarado, Guerra, & Zunzunegui, 2007) and found that social and health factors
accounted for sex differences in some cities, but not in others. However, previous studies
have tended to focus on contexts that are relatively similar and tend to rely on self-reports
of disability and physical function. Only a few comparative studies include objectively
measured physical performance, and those that do focus exclusively on developed
countries.
To address these gaps in the literature, this study examines gender differences
among older adults in both objective and self-reported measures across seven
countries/populations, including the Tsimane, an indigenous forager-horticulturalist group
living in the Bolivian Amazon to the United States. These seven populations (United States,
Taiwan, Korea, Mexico, China, Indonesia, and the Tsimane of Bolivia) span a wide range in
terms of level of economic development and gender roles and equality. In addition, we
examine not only self-reported difficulty with activities of daily living (ADLs), but also
compare self-reported functioning and objectively measured physical performance.
Physical performance measures may be less prone to cultural bias (Guralnik et al., 1989;
Kempen et al., 1996) and self-reported functioning measures aim to assess abilities in
“situation free” tasks (Verbrugge & Jette, 1994).
For all outcomes, we hypothesize that sex differences will be larger in developing
countries, which tend to have more pronounced gender role differences. Additionally, if
gender differences are consistent across both reported and measured outcomes, it would
25
suggest that gender differences are real and are not merely due to reporting differences
between men and women.
Methods
Data
Data for the U.S., Taiwan, Korea, Mexico, China, Indonesia, and Bolivia come from
seven population-based surveys collected between 2001 and 2011. These datasets include
the 2006 wave of the Health and Retirement Study (HRS) in the United States, the 2006
wave of the Social Environment and Biomarkers of Ageing Study (SEBAS) in Taiwan, and
the 2006 wave of the Korean Longitudinal Study of Aging (KLoSA), the 2001 wave of the
Mexican Health and Aging Study (MHAS), the 2011/2012 China Health and Retirement
Longitudinal Study (CHARLS), the 2007/2008 wave of the Indonesian Family Life Study
(IFLS-4), and the UNM-UCSB Tsimane Health & Life History Project (THLHP). These seven
datasets were selected to represent the range of high, middle and low-income countries;
they also include both Western and Asian countries. The traditional forager-horticulturalist
population, the Tsimane, is included in order to extend the range of development to a
traditional society that is rare in the contemporary world. All of these studies contain
comparable measures, including at least one objective physical performance measure.
Each survey collected data on both self-reported and objective measures of disability and
physical function and either focused the older population or included substantial numbers
of older adults. In addition, considerable effort has been made to harmonize measures and
methodologies across these surveys. . For this study, we focus on older adults aged 55 to 85
years because of small number of very old individuals among the Tsimane, topcoding of age
26
among the Taiwanese, and differential use of nursing homes and thus inclusion in the
studies among the very old.
The range in level of development, mortality and life expectancy, and equality
between the sexes is shown in Table 1. For example, per capita GDP ranges from nearly
$50,000 in the United States to only $200 among the Tsimane of Bolivia. The male/female
ratio of labor force participation (LFP), and male/female differences in years of education
and the prevalence of current tobacco use also varied widely, with the smallest differences
generally found in the U.S. with the exception of LFP.
< Insert Table 1 about here >
Table 2 gives the characteristics of each dataset; the year of the survey, eligible
sample universe, whether it was nationally-representative, and the method of interview.
Most surveys focused on older adults, but two included individuals of all ages. All surveys
were conducted using in-person interviews, except for the HRS, which used a combination
of in-person and telephone interviews. Finally, most surveys were nationally
representative. The Tsimane are not representative of the national population, but are an
isolated indigenous forager-horticulturist group comprised of approximately 9,000
individuals who live in the Beni region of Bolivia (Beheim n.d.). They reside in 80+ small
villages of 50-150 people living in extended family clusters. In this traditional population,
average life expectancy is very low.
< Insert Table 2 about here >
Table 3 shows the total sample size, age-eligible sample size (55-85 years), and
analytic sample size for each dataset. The analytic sample size excludes individuals missing
data on sex or ADL disability, and those interviewed via proxy. In some cases, sample sizes
27
for specific outcomes may be smaller, particularly for physical performance measures.
Characteristics of those included in analyses versus those excluded because of missing data
for each outcome are presented in Appendix B. In general, those who were excluded tended
to be somewhat older and were often more likely to be female. On average, respondents
were in their mid to upper 60s. The proportion of women ranged from 46.8 per cent among
the Tsimane to 55.4 per cent in the United States.
< Insert Table 3 about here >
Outcomes
Outcomes include measured physical performance, and self-reported functional
limitations and difficulty with activities of daily living (ADLs). ADLs were available for all
countries, but not all of the physical performance or functional limitation measures were
available for all countries. Table 4 summarizes the measure available across each dataset.
< Insert Table 4 about here >
Physical Performance Measures
Physical performance measures included tests of grip strength, lower extremity
function, and balance. Grip strength is a measure of upper body strength and is the average
of two or three trials using the dominant hand, measured in kilograms. Those who could
not perform the test due to injury, surgery, pain, or other health/safety reasons were coded
as missing. This measure was available for all but two countries (Mexico and the Tsimane).
Lower extremity function was assessed with a measure of gait speed and time to
complete five chair stands. Gait speed was measured with a timed walk (U.S.: 98.5 inches,
Tsimane and Taiwan: 3 meters; China: 2.5 meters) averaged over two trials. Gait speed was
calculated as meters per second. Those who tried but were unable or refused for
28
health/safety reasons, and those who used a walking aid were coded as missing. Since this
test was only administered to U.S. adults aged 65+, analysis of this measure was restricted
to those 65-85. Lower extremity function was also assessed using performance on the chair
stand test as the number of seconds taken to stand up from a sitting position five times,
top-coded at 25 seconds. Those who were unable to complete 5 repetitions, tried but were
unable, or did not try for health reasons were coded as missing.
Balance was assessed with two tests: the time a person was able to hold the tandem
position and balance on one leg. The full tandem test was administered to the Tsimane and
older adults in the U.S. and China. Among the Tsimane, the tandem test measured the
seconds an individual could hold the tandem position, for a maximum of ten seconds. U.S.
respondents were first asked to complete a semi-tandem test. If individuals younger and
older than 65 could hold the semi-tandem position for 60 seconds and 30 seconds,
respectively, then they were asked to hold the tandem position. Respondents in China were
administered the full tandem test if they could hold the semi-tandem position for ten
seconds. Performance on the tandem stand was dichotomized and poor balance was
defined as being unable to hold the tandem position for ten seconds. Individuals who did
not try because of safety reasons or who tried and were unable to do the test were coded as
missing, as were U.S. and Chinese respondents who could not complete the semi-tandem
stand. Tsimane and Mexican older adults were asked to stand on one leg for a maximum of
ten seconds. Those who were unable to hold the position for ten seconds were considered
to have poor balance, as were those who did not try for safety reasons or who tried and
were unable.
Self-Reported Disability & Physical Function
29
Disability was measured using self-reported difficulty with three activities of daily
living (ADLs) that were available in all seven countries (bathing, dressing, and going to the
toilet). Self-reported measures of physical function were available for some, but not all
countries, and included ability to squat, climb stairs, and carry a heavy load. Each item was
dichotomized to indicate a person either “had no difficulty” or “had difficulty/could not
perform the task.”
Analysis
Logistic regression was used to determine whether odds of ADL difficulty, functional
limitations and poor performance were significantly higher among women in each country,
controlling for categorical age (55-64, 65-74, and 75-85 years). OLS regression was used to
analyze the effect of sex for continuous measures of functioning. Models of grip strength
also included BMI to control for the effect of body size. We performed alternative analyses
controlling for height and weight (Appendix C) because separately these explain more of
the effect of sex, however, because the relative differences between countries in the effect
of sex remained similar, we proceeded to control for BMI in models of grip strength. We
also analyzed continuous measures using Heckman selection models to account for
potential selection bias (Appendix D), however, in nearly all cases, results were very
similar to those obtained with OLS regression. The one exception was in the analysis of
chair stands among the Taiwanese; the effect of sex was smaller in the Heckman model that
accounted for selection compared with the OLS regression model, although the effect
remained significant.
Stata’s survey (SVY) command was used to account for the effect of stratified and/or
clustered sampling, however, when compared to models run without the SVY command,
30
results were nearly identical (Appendix E). Data were analyzed separately for each country
and 95% confidence intervals were examined to assess differences in the effect of sex
across countries. For comparison, we also analyzed data pooled across countries in models
including a sex*country interaction term, however, this approach yielded very similar
findings and are also presented as supplemental material in Appendix E. Data were
analyzed using Stata (v. 11). Finally, we displayed countries ranked according to GNP in our
tables to examine whether there was a trend in the effect of sex with increasing economic
development. We also calculated correlation coefficients between country-level
characteristics and the effect of sex to assess whether there was evidence for an
association.
Results
Physical Performance Measures
Overall, results from OLS and logistic regression, show that women had significantly
worse physical performance relative to men for nearly all measures in all countries (Table
5). Results from OLS regression adjusted for age and BMI indicate that grip strength was
significantly lower among women in all countries. Women had significantly slower gait
speed than men in all four countries with data. Sex differences were greatest among the
Taiwanese, where average gait speed was .14 m/s slower among women. The effect of sex
appeared larger in Taiwan relative to China and the Tsimane. Sex differences in gait speed
were also larger in the U.S. than among the Tsimane.
With respect to chair stands, women took significantly longer in all four countries
where this test was administered. The effect of being female did not differ by country,
except that the effect of being female was smallest in China.
31
Turning to balance, results from logistic regression show that women had
significantly higher odds of being unable to hold the tandem position for ten seconds in all
three countries with data, however, the magnitude of sex differences did not vary
significantly. For the one-leg stand, Tsimane women had higher odds of poor balance but
no significant sex differences were observed in Mexico.
While women had worse physical performance than men in all but one measure for
one country, there was little evidence that the magnitude of sex differences varied
consistently with per capita GDP for most measures. Grip strength is the exception; the
difference between men and women increased with increasing GDP.
< Insert Table 5 about here >
Functional Tasks
Logistic regression results controlling for age indicate that in all countries
examined, women were more likely to report difficulty with each functional task (Table 6).
Although there was some overlap in confidence intervals, sex differences in these
functional tasks were most pronounced in Taiwan.
The magnitude of sex differences in difficulty squatting was fairly consistent across
the five countries, ranging from 1.2 times higher odds of difficulty in China to 2.0 times
higher odds in Taiwan. For difficulty climbing stairs, the effect of sex was also smallest in
China (OR=1.3) compared with the U.S., Taiwan, and Mexico where the odds of difficulty
were 2.0 to 2.6 times higher among women. Finally, for difficulty carrying, the odds ratios
associated with being female were quite consistent in the U.S., Mexico, China and Indonesia
but the effect of sex was much larger in Taiwan (OR=5.1).
32
There was no clear pattern of sex differences with respect to GNP. However, sex
differences for all three tasks were the least pronounced in China and the most pronounced
in Taiwan.
< Insert Table 6 about here >
ADLs
Results from logistic regression models controlling for age indicate that where there
was a significant difference between the sexes, women generally reported more difficulty
with ADL tasks, although there was one exception. In contrast to the physical performance
measures and the functional tasks examined above, there is no gender difference in having
difficulty with ADL tasks in many cases. In 9 of the 21 observations, there is no significant
gender difference in reported difficulty. In terms of dressing, women had higher odds of
difficulty in all countries except China and Korea. The effect of sex was most pronounced
among the Tsimane, Taiwanese and Indonesians, although based on the overlap in
confidence intervals, these effects were not significantly greater than that observed in the
U.S. and Mexico. For bathing, women reported greater difficulty in the U.S., Taiwan,
Indonesia, and among the Tsimane but not in Korea, China, and Mexico. Yet among the
former, there was little variation in the effect of sex, except that it appear smaller in the U.S.
compared with Taiwan. Finally, in terms of using the toilet, women were more likely to
report difficulty in the U.S., Taiwan and China. Yet the effect of sex appears significantly
smaller in China compared with the U.S. Among the Tsimane, the lack of significant sex
differences may be due to the small sample size, and among Koreans, women were actually
significantly less likely to report difficulty using the toilet (OR=0.5).
33
Taken together results for reported disability were consistent with findings for all
other outcomes besides grip strength in that there was no clear trend seen in the
relationship between sex differences and GDP. However, sex differences in ADLs were
consistently smallest among older adults in Korea and China for all three outcomes.
Finally, table 7 shows the correlation coefficients between country-level indicators
and the effect of sex for each measure for a more formal test of hypotheses. In general,
there was little evidence for associations, except that higher GNP was associated with
larger male/female differences in grip strength, and larger differences in education and
tobacco use prevalence were associated with smaller effects of sex. However, these findings
must be interpreted with caution due to the small number of points.
< Insert Table 7 about here >
Discussion
Overall, results indicate that sex differences were almost universal for both
objective physical performance measures and for self-reported physical function. Women
had weaker grip strength, slower gait speed, took longer to rise five times from a sitting
position and had worse balance, and they also reported significantly more difficulty
squatting, climbing stairs, and carrying a heavy load. This is consistent with findings from
previous studies of sex differences in physical function (Merrill et al., 1997; Oksuzyan et al.,
2010; Steffen, Hacker, & Mollinger, 2002). The female disadvantage in physical ability
appears to be remarkably consistent across markedly different contexts, and it is likely that
a large part of this is due to differences in body composition between men and women.
Women of all ages generally have lower muscle mass and strength and a greater
34
percentage of body fat (Leveille et al., 2000). They may also be less physically active
(Caspersen, Pereira, & Curran, 2000).
Yet relatively few comparative studies or studies of older adults in developing
countries have examined objective measures of performance. Many researchers have
suggested that sex differences in subjective measures are in part due to a greater tendency
to report poor health among women (because of greater sensitivity to symptoms or greater
contact with the health care system), although findings have not been conclusive (Ferrer et
al., 1999; Macintyre, Hunt, & Sweeting ,1996). Furthermore, less is known about reporting
differences between men and women in non-Western populations. Here, concordance in
findings across both self-reported and objective measures suggests that sex differences
cannot be explained by reporting differences alone. Despite the fact that there was not one-
to-one correspondence between physical and self-reported functioning in this study, the
physical performance tasks assess various aspects of strength in major muscle groups and
balance that would be needed to be able to perform tasks such as climbing stairs, carrying,
squatting, and raising arms that are assessed by self-reported items.
Compared with physical function outcomes, findings for ADL tasks were somewhat
less consistent – women usually reported greater difficulty although sex differences were
not significant in all cases. In Korea, odds of difficulty using the toilet were actually lower
among women. Yet this is consistent with previous studies (Lee & Lee, 2013). For example,
Park and colleagues (2009) found that the prevalence of ADL limitations was higher among
men aged 45-64 years in Korea and was similar among men and women aged 65 and older.
Results for other countries were more consistent with previous findings. For
instance, results from a comparative study of 11 Western developed countries found that
35
women had greater odds of ADL difficulties in all countries, although differences were not
always significant (Crimmins et al., 2011). Similarly, a study by Ofstedal and colleagues
(2007) found significant sex differences in Nagi (functional) limitations across all five Asian
settings examined, but sex differences in ADL limitations were only significant in one out of
four countries.
Yet, while female disadvantage was observed for nearly all measures, sex
differences did not appear to show a clear association with level of development or gender
equality indicators such as male-female differences in average education. It is difficult to
quantify overall gender equality because it varies across different domains within a
country. For example, in Table 1, gender differences are relatively small in the U.S. for all
three indicators. However, in China, female-to-male labor participation is quite high, but
there are large sex differences in the prevalence of smoking and in average years of
education. In addition, gender equality is desirable in some areas, such as income or
political participation, but detrimental if women start smoking or drinking at the same
rates as men. Therefore, it is likely that with increasing gender equality and development,
women are still exposed to both protective and risk factors. In other words, gains in some
areas may be offset by exposure to other types of risk factors. For example, in the United
States, women’s higher labor force participation may have led to greater economic equality,
but it also increases women’s exposure to stress and changes the nature of familial and
social networks.
Finally, although sex differences tended to be quite consistent with little clear
pattern by country, sex differences were often found to be the least pronounced among
Chinese older adults. This was observed for both objective and subjective measures. The
36
context in China is considerably different compared with other countries, even in Asia. For
example, under Communism, women in China had more similar labor force participation
rates as men (Yu & Sarri, 1997). Men and women may also have more similar levels of
physical activity since much of the elderly labor force has worked in agriculture (Kaneda et
al., 2009). In addition there is a strong culture of familialism, whereby older adults receive
instrumental and financial assistance from their children, and are much less likely to live
alone (Chuanyi & Qin, 1992; Leung, 1997; Zimmer, 2005). While increasing development
has been associated with an erosion of family-based support in many countries, filial
responsibility has remained strong in China, and has even been codified into law (Zimmer,
2005). Finally, although Wang and colleagues (2009) found evidence to suggest that sex
differences in China increase with age, this could also signify a narrowing of gender
differences in younger cohorts, as has been seen in China (Chun et al., 2012).
Limitations
Several limitations merit discussion. First, the use of cross-sectional data means that
it is only possible to compare sex differences in the prevalence of disability and poor
physical function, and average performance at a single point in time. Since prevalence of
difficulty is a function of both onset and duration, the female disadvantage in disability and
poor function may be due to earlier onset, a lower likelihood of recovery, increased
survival among those in poorer health, or a combination of these. However, it is impossible
to identify which aspects of the process are responsible for the observed gender disparities.
In many cases, longitudinal data are not currently available. However, follow-up waves
now exist for many of these studies. Therefore, future work could examine gender
differences in disability/functioning transitions.
37
Other issues arise when comparing findings across countries. For example, it is
possible that individuals with similar levels of disability or physical function respond
differently across countries, due to differences in question wording, cultural differences,
and other sources of measurement error. However, since we compare men and women
within each country, this may be less problematic. In addition, this study compares not only
self-reported difficulties, but also objectively-measured physical performance. Although
caution must be exercised in comparing physical performance across countries due to
some variation in the measurement protocol for objective measures (e.g. hand
dynamometers used to measure grip strength), this is less problematic for between-sex
comparisons within countries. Physical performance measures may be less prone to
cultural bias and we find good concordance of gender differences between objective and
subjective functioning. Although not all surveys collected data on all measures, each survey
examined contained at least one objective test of physical performance. Finally, surveys
varied in terms of their sample size, therefore, care must be taken to examine both the
magnitude and significance of effect sizes. In some cases, lack of statistical significance may
be due to low power to detect differences as a result of small sample size.
Conclusion
Overall, there was striking similarity in the magnitude and direction of sex
differences across countries despite considerable differences in context, although modest
variations in the effect of sex were observed. In sum although sex differences were quite
consistent, it is important to note that they may be similar for different reasons. Thus,
future work should examine how various protective and risk factors interact and offset
each other.
38
Table 2.1. Characteristics of study countries (CIA World Factbook, 2005)
U.S. Taiwan Korea Mexico China Indones. Tsimane
1

Development

GDP per cap ($)
40,000 25,300 19,200 9,600 5,600 3,600 200
Infant mort rate 6.5 6.4 7.1 20.9 24.2 35.6 105.0
Life exp at birth
Males 74.9 74.5 73.4 72.4 70.7 67.1 54.3
Females 80.7 80.3 80.6 78.1 74.1 72.1 54.0
Diff (F-M) 5.8 5.8 7.2 5.7 3.4 5.0 -0.3
Gender Equality

Bolivia
M/F labor force
participation
2

0.81
0.73
0.69 0.50 0.90 0.55
0.78
Diff in yrs educ
(M-F)
3

0.0
0.9
1.2 0.7 1.3 1.2
1.7
Diff in tobacco
use (M-F)
4

5% --- 48% 25% 56% 61% 5%
1. Gurven, Kaplan and Supa 2007
2. >15yrs, 2005, International Labour Organization
3. Adults 25+, 2009, Gakidou et al. 2010
4. Adults 15+, 2005, World Health Organization (WHO)

39

Table 2.2. Survey characteristics
Country Dataset Year Eligible
Population
Nationally
Rep?
Method of
Interview
U.S. Health & Retirement
Study (HRS)
2006 50+ (and
spouses)
Yes Phone & in-
person
Taiwan Social Environment &
Biomarkers of Ageing
Study (SEBAS)
2006 53+ Yes In-person
Korea Korean Longitudinal
Study of Aging
(KLoSA)
2006 45+ Yes In-person
Mexico Mexican Health &
Aging Study (MHAS)
2001 50+ (and
spouses)
Yes In-person
China China Health &
Retirement
Longitudinal Study
(CHARLS)
2011-2012 45+ No In-person
Indonesia Indonesian Family
Life Study (IFLS4)
2007-2008 All ages 83% of
population
In-person
Tsimane Tsimane Health &
Life History Project
Baseline
(ongoing)
40+ No In-person

40

Table 2.3. Survey sample characteristics

U.S. Taiwan Korea Mexico China Indones. Tsimane
Total Sample 18,469 1,284 10,254 15,402 17,708 44,103 ~3,000
Age-Eligible
Sample (55-85)
15,038 1,092 6,747 9,722 11,071 4,220 596
Analytic Sample* 14,125 1,051 6,532 8,846 10,227 4,196 449
% Women 55% 48% 54% 53% 51% 51% 47%
% Age 55-64 33% 50% 42% 53% 56% 54% 53%
% Age 65-74 42% 27% 40% 32% 30% 35% 34%
% Age 75-85 25% 23% 18% 14% 14% 12% 13%
*Excludes those missing data on sex or ADL disability and those interviewed via proxy

41

Table 2.4. Available measures by country

Objective Measures Self-Reported Measures

Physical Performance Functional Tasks ADLs

Grip Gait Chair Tand 1 Leg Squat Stair Carry Dress Bathe Toilet
U.S. X X

X

X X X X X X
Taiwan X X X

X X X X X X
Korea X

X X X
Mexico

X X X X X X X
China X X X X

X X X X X X
Indon. X

X

X

X X X X
Tsim.

X X X X

X X X
42
Table 2.5. Objective performance: Odds ratios and regression coefficients of the effect of being female, controlling for
age
U.S. Taiwan Korea Mexico China Indonesia Tsimane
B B B B B B B
Grip strength -15.9 *** -14.5 *** -12.3 *** ---

-11.1 *** -10.3 *** ---
(kg) [-16.4,-15.5] [-15.6,-13.4] [-12.5,-11.9]

[-11.6,-10.7] [-10.9,-9.7]

Gait speed -.07 *** -.14 *** ---

---

-.04 *** ---

-.03 **
(m/s, 65+) [-.09, -.06] [-.18,-.09]

[-.06,-.02]

[-.05,-.01]
Chair Stands ---

1.30 *** ---

---

0.87 *** 1.44 *** 1.36 ***
(seconds)

[.74,1.86]

[.67,1.07] [1.21,1.67] [.75,1.98]

OR

OR

OR

OR

OR

OR

OR
Tandem Stand 1.59 *** ---

---

---

2.01 *** ---

2.45 *
(<10sec) [1.37,1.86]

[1.70,2.37]

[1.24,4.84]
One-Leg Stand ---

---

---

1.14

---

---

2.59 ***
(<10sec)

[.79,1.65]

[1.62,4.13]
*p<.05 **p<.01 ***p<.001
43
Table 2.6. Self-reported measures: Odds ratios of the effect of being female, controlling for age
U.S. Taiwan Korea Mexico China Indonesia Tsimane
Func Tasks OR OR OR OR OR OR OR
Squatting 1.61 *** 2.04 ** ---

1.84 *** 1.22 *** 1.37 * ---

[1.48,1.75] [1.42,2.93]

[1.55,2.20] [1.10,1.36] [1.05,1.79]

Stairs 1.96 *** 2.63 *** ---

1.97 *** 1.33 *** ---

---

[1.80,2.13] [1.93,3.57]

[1.66,2.34] [1.17,1.50]

Carrying 2.66 *** 5.13 *** ---

2.62 *** 2.28 *** 2.66 *** ---

[2.37,2.97] [3.09,8.54]

[2.03,3.38] [1.97,2.63] [2.23,3.16]

ADLs OR

OR

OR

OR

OR

OR

OR
Dressing 1.58 *** 2.16 * .74

1.43 * 1.04

2.00 *** 2.42 **

[1.36,1.84] [1.12,4.19] [.52,1.06] [1.07,1.90] [.85,1.26] [1.58,2.53] [1.40,4.19]
Bathing 1.33 ** 2.15 *** .96

1.29

1.09

1.96 *** 2.09 **

[1.10,1.62] [1.67,2.78] [.68,1.37] [.88,1.90] [.92,1.30] [1.42,2.70] [1.20,3.64]
Toilet 2.09 *** 2.29 ** .53 * 1.50

1.22 ** 1.26

2.15

[1.64,2.66] [1.30,4.02] [.31,.92] [.98,2.30] [1.06,1.40] [.74,2.16] [.99,4.67]
*p<.05 **p<.01 ***p<.001

44
Table 2.7. Correlations between country-level indicators (2005) and the effect of sex

GNP
Infant Mortality
Rate
Educ Diff
(yrs, M-F)
Tobacco Use Diff
(M-F)
Employment
Ratio (F/M)
Life Exp Diff
(yrs, F-M)
OR N Corr p N Corr p N Corr p N Corr p N Corr p N Corr p
Dress 7 -.17 .717 7 .59 .164
7 -.44 .318 5 -.11 .865 6 -.28 .588
7
.55 .198
Bathe 7
-.21
.659 7 .50 .256
7 -.22 .630 5 .21 .741 6 -.28 .587
7
.45 .314
Toilet 7 .27 .557 7 .29 .526
7 -.73 .062 5 -.73 .162 6 .16 .762
7
.41 .363
Squat 5
.46 .437 5 -.66 .227 5 -.45 .449 4 -.74 .257 5 -.33 .587 5 -.82 .090
Stairs 4
.50 .502 4 -.77 .233 4 -.34 .665 3 -.92 .259 4 -.40 .598 4 -.82 .180
Carry 5
.35 .561 5 -.56 .324 5 -.04 .951 4 -.46 .540 5 .01 .984 5 -.48 .415
Tandem 3
-.92 .262 3 .49 .217 3 .21 .864 2 --- --- 2 --- --- 3 .99 .076
One-Leg 2
--- --- 2 --- --- 2 --- --- 1 --- --- 1 --- --- 2 --- ---
Coeff

Grip (kg) 5
-.98 .004 5 .83 .079 5 .89 .042 4 .99 .007 5 -.35 .566 5 .42 .485
Gait (m/s) 4
-.58 .417 4 .64 .362 4 -.08 .915 2 --- --- 3 .97 .166 4 .73 .266
Chair (sec) 4
.01 .988 4 .33 .675 4 -.47 .531 2 --- --- 3 -.95 .192 4 .00 .998

45
CHAPTER 3: A GLOBAL PERSPECTIVE ON AGE DIFFERENCES IN PHYSICAL FUNCTION
AND DISABILITY

Abstract
Objectives: To determine whether populations age differently across a range of countries in
terms of physical function and disability.
Methods: Age differences in self-reported and objective measures of physical function and
disability were compared among older adults aged 55-85 in the U.S., Taiwan, Korea,
Mexico, China, Indonesia, and among the Tsimane of Bolivia using population-based studies
collected between 2001 and 2011. Data were analyzed using logistic and OLS regression to
determine whether older adults aged 65-74 and 75-85 had significantly higher odds of
disability or poor performance or worse average performance compared with those aged
55-64, controlled for age.
Results: Significant age differences were observed for all measures in all countries. While
age differences in physical performance measures were fairly consistent across most
countries, age differences in reported difficulty with functional tasks were more variable.
For example, age differences among the Taiwanese were more pronounced relative to
many other countries, but this was in part due to the low prevalence of difficulties among
the youngest age group. Finally, although absolute differences in the prevalence of ADL
difficulties were quite variable, age differences were quite consistent. Age differences were
somewhat smaller in the U.S. and China, however, the prevalence of difficulties at younger
ages was already relatively high in these countries.
Discussion: Across countries, both differences in performance and prevalence of difficulties
among younger individuals and differences in age-related decline indicate that some
populations are “aging” faster than others.
46
Introduction
While aging is a universal process and age is one of the most important risk factors
for decline in physical ability and disability, the pace of aging varies across individuals;
some people age faster than other (Crimmins, Kim, & Seeman, 2009). In some countries,
decline in physical ability and the onset of disability may occur at younger ages than in
other countries, and changes with age may differ. It is important to understand how
physical ability and disability vary across populations because these have a profound
impact on individuals’ quality of life, need for care, and future health and survival. Physical
limitations and disability are also costly to societies, and will grow more costly as
populations age around the world. By 2050 it is projected that there will be nearly 2 billion
people aged 60 and older in the world, thus, there will undoubtedly be an increase in the
total number of disabled persons (United Nations (UN), 2002). Yet there is also a need to
understand the role of aging explicitly.
Older populations worldwide show markedly different levels of disability burden
(WHO, 2004). For example, findings from the SAGE study including 6 countries found that
ADL disability among older adults aged 70+ ranged from 27% in China to 68% in India (He,
Muenchrath & Kowal, 2012). However, much less is known about how age differences vary
across contexts. This is important because earlier and faster aging are more detrimental to
individuals and society. Therefore, in this paper we address the issue of whether
populations age differently across a number of countries in terms of disability.
Background
There is a great deal of variation across countries in the contexts in which current
cohorts of older adults have aged. For example, older adults in developing countries have
47
been exposed to less favorable conditions both in childhood and throughout the life course.
Current elderly populations in Asia and Latin America have lived through changing
epidemiologic conditions. Earlier in life, most mortality among these cohorts was due to
infectious disease, but now chronic diseases are the leading causes of death. Older adults in
these countries have also had poorer nutrition, fewer economic resources and live in
societies with few institutionalized safety nets. This has led many researchers to predict
that older adults in developing countries may have a much higher burden of disability and
poor health than similarly-aged individuals in developed countries (Boutayeb & Boutayeb,
2005; Monteverde, Noronha, & Palloni, 2009; Popkin, 2002). On the other hand, if mortality
rates are higher among the disabled and those in poor health in developing countries, the
prevalence of disability may be lower than might otherwise be expected. For example,
Meyers and colleagues (2003) suggest that at earlier stages of “disability transition,”
disability incidence is high but disability prevalence is low because of high mortality among
the disabled.
To understand whether some populations are “aging” more quickly than others, it is
necessary to consider the level of physical function and disability among younger age
groups in conjunction with change across age groups. For example, it is possible for age
differences to appear smaller if the prevalence of poor functioning and disability was
already high among younger adults. However, individuals in such populations can be
considered to have aged more quickly, since onset occurred at younger ages prior to the
study.
In this paper, age differences in physical function and disability are compared across
7 countries/populations. These seven countries (United States, Taiwan, Korea, Mexico,
48
China, Indonesia and the Tsimane of Bolivia) span a wide range in terms of level of
economic development, health care systems, life expectancy etc. Unlike most previous
comparative studies, we examine both objectively-measured physical performance and
self-reported functioning. Although self-reported functioning measures aim to assess
abilities in “situation free” tasks, physical performance measures are thought to be less
prone to cultural influence (Guralnik, Branch, Cummings, & Curb, 1989; Kempen et al.,
1996). Performance-based and self-reported measures can also be compared to assess
whether self-reports are consistent with objective measures. Finally, we compare disability
in activities of daily living (ADLs). Compared with physical function, ADL tasks such as
bathing or dressing have a larger environmental component and are often conceptualized
as “a lack of fit between environmental demand and personal capability” (Verbrugge &
Jette, 1994). Therefore, the prevalence of difficulty with such tasks may be more variable
across contexts due to differences in home environment, assistive devices, the availability
of care, etc.
Methods
Data
Data for the U.S., Taiwan, Korea, Mexico, China, Indonesia and Bolivia come from 7
population-based surveys collected between 2001 and 2011. These datasets include the
2006 wave of the Health and Retirement Study (HRS) in the United States, the 2006 wave
of the Social Environment and Biomarkers of Ageing Study (SEBAS) in Taiwan, the 2006
wave of the Korean Longitudinal Study of Aging (KLoSA), the 2001 wave of the Mexican
Health and Aging Study (MHAS), the 2011/2012 China Health and Retirement Longitudinal
49
Study (CHARLS), the 2007/2008 wave of the Indonesian Family Life Study (IFLS-4), and
the UNM-UCSB Tsimane Health & Life History Project (THLHP).
These seven datasets were selected to represent the range of high, middle and low-
income countries; they also include both Western and Asian countries. The traditional
forager-horticulturalist population, the Tsimane, is included in order to extend the range of
development to a traditional society that is rare in the contemporary world. These surveys
span a wide range in terms of level of development (Table 1). For example, per capita GNP
ranges from roughly $40,000 in the United States to only $200 among the Tsimane of
Bolivia.
< Table 1 about here >
All of these studies contain comparable measures, including at least one objective
physical performance measure. Each survey collected data on both self-reported and
objective measures of disability and physical function and either focused the older
population or included substantial numbers of older adults. In addition, considerable effort
has been made to harmonize measures and methodologies across these surveys. . For this
study, we focus on older adults aged 55 to 85 years because of small number of very old
individuals among the Tsimane, topcoding of age among the Taiwanese, and differential
use of nursing homes and thus inclusion in the studies among the very old.
Table 2 gives the characteristics of each dataset; the year of the survey, eligible
sample universe, whether it was nationally-representative, and the method of interview.
Most surveys focused on older adults, but two included individuals of all ages. All surveys
were conducted using in-person interviews, except for the HRS, which used a combination
of in-person and telephone interviews. Finally, most, but not all surveys were nationally
50
representative. Data for Bolivia are not representative of the national population, but
instead come from an ongoing study of an indigenous forager-horticulturist group: the
Tsimane. The UNM-UCSB Tsimane Health & Life History Project (THLHP) began studying
the Tsimane in 2002. The Tsimane Amerindians are comprised of approximately 9,000
individuals who live in the Beni region of Bolivia (Beheim, n.d.). They reside in 80+ small
villages of 50-150 people living in extended family clusters. In this traditional population,
average life expectancy is very low, and levels of infection and inflammation are high. This
population represents the lowest level of development relative to other contexts.
< Table 2 about here >
Table 3 shows the total sample size, age-eligible sample size (55-85 years), and
analytic sample size for each dataset. The analytic sample size excludes individuals missing
data on sex or ADL disability, and those interviewed via proxy. In some cases, sample sizes
for other outcomes may be smaller, particularly for physical performance measures. On
average, respondents were in their mid to upper 60s. The proportion of women was more
variable, ranging from 46.8% among the Tsimane to 55.4% in the United States.
< Table 3 about here >
Measures
Outcomes include measured physical performance, and self-reported functional
limitations and difficulty with activities of daily living (ADLs). ADLs were available for all
countries, but not all of the physical performance or functional limitation measures were
available for all countries. Table 4 summarizes the measure available across each dataset.
< Table 4 about here >
Physical Performance Measures
51
Physical performance measures included tests of grip strength, lower extremity
function, and balance, although not all measures were available for all countries. Grip
strength is a measure of upper body strength and is the average of 2 or 3 trials using the
dominant hand, measured in kilograms. Those who could not perform the test due to
injury, surgery, pain, or other health/safety reasons were coded as missing. This measure
was available for all but two countries (Mexico and the Tsimane).
Lower extremity function was assessed with a measure of gait speed and time to
complete 5 chair stands. Gait speed was measured with a timed walk (U.S.: 98.5 inches,
Tsimane and Taiwan: 3 meters; China: 2.5 meters) that was an average of two trials. Gait
speed was calculated as meters per second. Those who tried but were unable or refused for
health/safety reasons, and those who used a walking aid were coded as missing. Since this
test was only administered to U.S. adults aged 65+, analysis of this measure was restricted
to those 65-85. Lower extremity function was also assessed using performance on the chair
stand test as the number of seconds taken to stand up from a sitting position 5 times, top-
coded at 25 seconds. Those who were unable to complete 5 repetitions, tried but were
unable, or did not try for health reasons were coded as missing.
Balance was assessed with 2 tests: the time a person was able to hold the tandem
position and balance on one leg. The full tandem test was administered to the Tsimane and
older adults in the U.S. and China. Among the Tsimane, the tandem test measured the
seconds an individual could hold the tandem position, for a maximum of 10 seconds. U.S.
respondents were first asked to complete a semi-tandem test. If individuals younger and
older than 65 could hold the semi-tandem position for 60 seconds and 30 seconds,
respectively, then they were asked to hold the tandem position. Respondents in China were
52
administered the full tandem test if they could hold the semi-tandem position for ten
seconds. Performance on the tandem stand was dichotomized and poor balance was
defined as being unable to hold the tandem position for ten seconds. Individuals who did
not try because of safety reasons or who tried and were unable to do the test were coded as
missing, as were U.S. and Chinese respondents who could not complete the semi-tandem
stand. Tsimane and Mexican older adults were asked to stand on one leg for a maximum of
ten seconds. Those who were unable to hold the position for ten seconds were considered
to have poor balance, as were those who did not try for safety reasons or who tried and
were unable.
Self-Reported Physical Function & ADL Disability
Self-reported measures of physical function were available for some, but not all
countries, and included ability to squat, climb stairs, and carry a heavy load. Disability was
measured using self-reported difficulty with three activities of daily living (ADLs) that were
available in all seven countries (bathing, dressing, and going to the toilet). Each item was
dichotomized to indicate a person either “had no difficulty” or “had difficulty/could not
perform the task.”
Analysis
The prevalence of disability and poor function and average performance was
examined for three age groups (55-64, 65-74 and 75-85 years) in each country. Logistic
regression was used to assess whether the odds of ADL disability, functional limitations
and poor performance were significantly higher among those aged 65-74 and 75-85
compared with the youngest age group. OLS regression was used to analyze continuous
variables physical performance variables. All models controlled for sex and models of grip
53
strength also included BMI to control for the effect of body size. Data were analyzed
separately for each country and 95% confidence intervals were examined to assess
differences in the effect of sex across countries. Data were analyzed using Stata (v. 11) and
Stata’s survey (SVY) command was used to account for the effect of stratified and/or
clustered sampling. Finally, we displayed countries ranked according to GDP in our results
table to examine whether there was a trend in the effect of age with increasing economic
development. We also calculated correlation coefficients between country-level
characteristics and the prevalence/average level at ages 55-64 as well as the effect of age
(75-85 relative to 55-64) to assess whether there was evidence for an association.
Results
Physical Performance Measures
Average physical performance by age and country are shown in Figure 1 while Table
5 shows the regression coefficients and odds ratios associated with greater age relative to
those aged 55-64. For all measures in all countries for which there were data, those aged
65-74 and 75-85 had significantly worse performance compared with the youngest age
group.
Average grip strength was highest in the U.S. and lowest among Indonesians at all
ages, while it was similar among older adults in Taiwan, Korea and China. Older age was
associated with significantly lower grip strength in each country with data, however, when
examining the 95% confidence intervals, the relative effect of age was significantly greater
in the U.S. compared with Korea and China among those aged 75-85. Average gait speed
was greatest among older adults in the U.S., followed by Taiwan and China, while the
Tsimane had the lowest average gait speed. Gait speed declined between the two age
54
groups in all countries, although the relative effect of age was smaller among the Tsimane
compared with the U.S. and China. For chair stands, Indonesians of all ages had the best
performance, followed by the Taiwanese, Chinese, and the Tsimane. Greater age was
associated with longer average time to complete five chair stands, however, the age-
associated increase for those aged 65-74 was significantly larger in Taiwan compared with
China, Indonesia, and the Tsimane. The average increase in time to complete five chair
stands among those aged 75-85 was also significantly greater in Taiwan compared with the
Tsimane.
Turning to balance, age was associated with greater odds of poor balance on both
tests for countries were data were available. The prevalence of poor balance on the tandem
stand was lowest among the Tsimane and similar in the U.S. and China. The relative effect
of age did not differ, except among Tsimane aged 65-74 whose age-related increase in the
odds of poor balance was significantly greater relative to the U.S. or China. With respect to
the one-leg stand, the prevalence of poor balance was also lower among the Tsimane
compared with older adults in Mexico, however, the effect of age did not differ significantly
between the two countries.
Functional Tasks
Turning to self-reported functional limitations (Figure 2), the prevalence of
difficulty squatting varied greatly, ranging from 3% in Indonesia to 44% in China among
adults aged 55-64. Greater age was associated with higher odds of difficulty in all five
countries with data, but according to the lack of overlap in 95% confidence intervals, the
effect of age differed between countries (Table 6). The effect of age among those aged 65-
74 and 75-85 was significantly greater in Taiwan and Indonesia compared with the U.S.,
55
Mexico and China. For difficulty climbing stairs, the prevalence among the youngest age
group was only 10% in Taiwan compared with 60% in China. Logistic regression results
show that the effect of age was significant for all four countries with data, but was larger in
Taiwan compared with the U.S., Mexico and China. With respect to carrying, the prevalence
of difficulty was relatively similar among those aged 55-64 in the five countries with data,
ranging from 10-17%. Yet the age-related increase in difficulty was significantly greater
among Indonesians compared with those in the U.S., Mexico, and China. The effect of age
was also smaller in the U.S. relative to the effect of age in the other four countries.
ADL Disability
In general, prevalence of difficulty with each ADL task increased with age. The
prevalence of difficulty tended to be high for all tasks at all ages among older Tsimane and
Chinese, although the prevalence of difficulty dressing was also quite high among
Indonesians. Prevalence of difficulty was more similar in the remaining countries. In the
U.S., prevalence of difficulty with each ADL task was slightly higher at ages 55-64 compared
with similarly-aged individuals in Mexico, Indonesia, Taiwan, and Korea. Yet there tended
to be smaller age-associated increases in the odds of difficulty, which in many cases was
statistically significant. For example, in terms of dressing, those aged 65-74 in the U.S.
actually had lower odds of difficulty compared with the youngest age group and there was
no significant difference among those aged 75-85. Yet the odds of difficulty associated with
age were significantly higher in the other six countries. For bathing, odds of difficulty
associated with age were smaller in the U.S. relative to Indonesia, Taiwan and Korea. The
age-related increase in odds of difficulty using the toilet was also smaller in the U.S.
compared with Indonesia and Taiwan. Finally, age-related increases in the odds of difficulty
56
with each task also tended to be smaller in China relative to Indonesia, Taiwan and Korea,
although the differences in the effect of age were not always significant.
Turning to country-level indicators, Table 7 shows the correlation coefficients with
the effect of sex (75-85 versus 55-64) for each measure for a more formal test of
hypotheses. While there was no relationship with self-reported measures, there was
evidence for a negative correlation between infant mortality rates and the effect of age for
gait speed and chair stands and a positive correlation between life expectancy/level of
education and age differences in these two measures.
In terms of correlations with prevalence of difficulty and average performance
among those aged 55-64, there was a positive correlation between GNP, education and life
expectancy and average gait speed, while infant mortality rates were negatively correlated
with gait speed (Table 8). Also, grip strength was positively correlated with BMI.
Correlations for prevalence of self-reported difficulties must be interpreted cautiously, but
results suggest that infant mortality rates were positively associated with the prevalence of
difficulty dressing and bathing, while these measures were negatively correlated with
education and life expectancy. However, all correlations must be interpreted with caution
due to the small number of points.
< Insert Table 7 about here >
< Insert Table 8 about here >
Discussion
Overall, significant age differences were observed for all measures in all countries.
This is consistent with the literature on age differences in both self-reported and objective
measures of physical ability and self-reported ADL disability (Beckett et al., 1996; Liang et
57
al., 2008; Samson et al., 2000). For measured physical performance, age-related decline
tended to be fairly consistent across most countries. At younger ages, those in the U.S.
performed the best on test of grip strength and gait speed, however, age differences were
also more pronounced on these measures. On the other hand, age-associated declines in
gait speed and performance on the chair stand test were significantly smaller among the
Tsimane. This may be because the Tsimane fared poorly on these tests, even at younger
ages. The Tsimane also tended to report greater difficulty with ADL tasks, although age-
related differences were not significantly different compared with other countries.
Relative to the other populations, living condition are particularly harsh among the
Tsimane. For example, more than half of Tsimane adults aged 20-84 were infected with 1+
parasite species (Vasunilashorn et al., 2010) and between ages 5 and 54, levels of C-
reactive protein (a marker of inflammation) were found to be significantly higher among
the Tsimane compared with those in the U.S. (Gurven, Kaplan, Winking, Finch, & Crimmins,
2008). In addition, average life expectancy is only 42.8 years (Gurven, Kaplan, & Supa,
2007). Although age differences were smaller among the Tsimane for gait speed and chair
stands, significant decline may have already occurred before age 55. According to a
previous study by Gurven and colleagues (2006), total strength (including chest, arm, hand,
thigh and leg strength) among the Tsimane peaked at age 29 and declined quite rapidly. In
contrast, the prevalence of poor balance was lower among the Tsimane and although age
differences in balance tended to be somewhat larger among the Tsimane, in most cases, the
difference was not significant. Therefore, it appears that balance was fairly well maintained
among the Tsimane, despite declines in other areas.
58
Compared with objective measures of physical performance, there appeared to be
greater variation in the effect of age on reported functional difficulties. In general, age
differences tended to be more pronounced in Taiwan and Indonesia. This may be because
the prevalence of difficulty with several measures was quite low at younger ages, although
between-country differences in absolute levels of difficulty must be interpreted with
caution due to potential differences stemming from language, cultural norms, etc. as well as
some differences in question wording and response categories. However, results from the
gait speed and chair stand tests showed that older Taiwanese had better performance than
older Chinese adults, and this pattern was also seen for squatting and climbing stairs; tasks
that use many of the same lower-body muscle groups. This suggests that the lower
prevalence of difficulty squatting and climbing is not merely an artifact of reporting
differences. Older adults in Taiwan seem to be faring well in terms of physical ability
relative to other countries. However, they also enjoy much higher standards of living
compared with older adults in many of the other countries, although economic
development occurred rapidly. For instance, Taiwanese have been covered by the National
Health Insurance (NHI) program since 1995 (Chen et al., 2007) and per capita GDP was
$38,300 in 2012 (CIA, 2012). Therefore, although the age-related increase in the
prevalence of functional limitations was more pronounced in Taiwan, older Taiwanese
seem to be doing relatively well in terms of physical function because levels were quite low
among the youngest age group.
The age-related increase in prevalence of difficulty was also larger among
Indonesians for difficulty squatting and carrying, and again, there was concordance with
physical measures. Indonesians had both the best performance on the chair stand test and
59
the lowest prevalence of difficulty squatting. On the other hand, they had both a high
prevalence of difficulty carrying and the lowest average grip strength at all ages. As was
observed for the Tsimane, ability in some physical domains appeared to be higher than in
other countries and remained relatively high (lower-body strength) despite evidence for
greater age-related decline. However, functioning in another domain (upper-body
strength) followed a different trajectory. These results further suggest that there may be
differences across countries in physical activity (recreation, labor/chores, etc.) that may
lead to relatively better or relatively worse performance for a given task than what would
be expected given performance on a different task.
However, relative to both objective and self-reported physical ability, ADL tasks
have a much larger environmental component. Tasks such as dressing, bathing and using
the toilet depend on factors such as types of clothing, availability of indoor plumbing, type
of toilet, availability of assistive devices, etc. Therefore, one might expect age differences in
these outcomes to be more variable. Yet in general, although absolute differences in
prevalence varied across countries, age differences were similar, except in the U.S. and
China, where age differences tended to be smaller.
Even among those aged 55-64 in China, difficulty with ADL tasks was very high and
might help explain why age differences appeared to be small. Yet not only did older Chinese
adults report more difficulty with ADL tasks, but they also reported more functional
limitations and displayed poor performance on the gait speed and chair stands tests. In
other words, although age-related differences were relatively small, older Chinese adults
fared poorly on almost all outcomes. Although China has undergone rapid economic
development in recent years, development has been uneven and many current older adults
60
have experienced difficult living conditions in both childhood and adulthood, both of which
have been linked to poor health (Wen & Gu, 2011). Also, several policies implemented by
the Communist Party make the aging context in China somewhat unique (Streib, 1987). For
instance, the “one-child” policy, implemented in 1979, limited the number of children a
couple could have led to rapid fertility decline and population aging (Zhang & Goza, 2006).
It has also led to concerns about caring for the growing elderly population (Zimmer &
Kwong, 2003). Lower availability of support for the elderly could lead to greater (real or
perceived) difficulty with functional and ADL tasks.
In the U.S., the prevalence of reported difficulty with functional and ADL task was
also somewhat high among those aged 55-64, however, there was much less age-associated
increase in the prevalence of difficulty. It is interesting to note that although the U.S. and
Taiwan are the most similar in terms of development, the prevalence of reported difficulty
was low at younger ages and increased more rapidly with age, while the opposite was
observed in the United States. In the U.S., there are several possible explanations for a
higher prevalence of reported difficulty at younger ages. First, those in the U.S. may report
difficulty at lower levels of impairment if they have higher health expectations (Carr,
Gibson, & Robinson, 2001; Salomon, Tandon, & Murray, 2004). In the present study,
average grip strength and gait speed among older adults in the U.S. was better at all ages
compared with other countries, which suggests that U.S. older adults have higher
expectations about their health and functioning. Second, disabled individuals in the U.S.
may be more likely to survive with a disability, whereas in less developed countries,
selective mortality among those in poor health may lead to better health at the population
level. Third, older adults in the U.S. may have had greater lifetime exposure to harmful
61
lifestyle factors, such as smoking, obesity and inactivity. Until recently these factors were
more common in Western industrialized societies (Popkin, 1999; Wong, Ofstedal, Yount, &
Agree, 2008). This may help explain the different patterns observed in Taiwan and the U.S.
In reality, each of these factors may contribute to the observed findings.
Yet overall, there was little evidence that age differences varied systematically by
level of development (and related indicators), as had been hypothesized. This is likely
because age differences in part depend on the initial prevalence of difficulties or level of
performance and by age 55, some “aging” and selection have already occurred.
Limitations
Although this paper makes an important contribution to what is already known
concerning age differences in physical function and disability, several limitations must be
noted. First, the use of cross-sectional data means that it is only possible to compare age
differences in the prevalence of disability and poor physical function, and average
performance at a single point in time. Therefore, it is impossible to disentangle possible
cohort or period effects. Prevalence of difficulty is a function of both onset and duration
(recovery, survival or mortality), but we cannot say what the relative contribution of each
is. In many cases, longitudinal data are not currently available. However, follow-up waves
now exist for many of these studies. In addition, by age 55, there has already been
significant population selection. Unfortunately, many population-based studies do not
measure performance on young and middle-aged adults.
Other issues arise when comparing findings across countries. For example, it is
possible that individuals with similar levels of disability or physical function respond
differently across countries, due to differences in question wording, cultural differences,
62
and other sources of measurement error. However, this study compares not only self-
reported difficulties, but also objectively-measured physical performance. Physical
performance measures may be less prone to cultural bias and findings are generally
consistent across objective and subjective functioning, with the exception of the U.S.
Although not all surveys collected data on all measures, each survey examined contained at
least one objective test of physical performance. Finally, surveys varied in terms of their
sample size, therefore, care must be taken to examine both the magnitude and significance
of effect sizes. In some cases, lack of statistical significance may be due to low power to
detect differences as a result of small sample size.
Conclusion
Overall, this study identified important age differences in physical function and
disability, using both objective and self-reported measures. While significant age
differences were observed for all measures in all countries, age-related declines varied
across countries. Some of this was due to differences in outcomes among the young-old,
since age differences tended to be smaller in countries where the initial prevalence of
difficulties was high. Taken together, differences in baseline level of performance and
difficulties and age-related decline indicate that some populations are “aging” faster than
others.

63
Table 3.1. Characteristics of study countries (CIA World Factbook, 2005)
U.S. Taiwan Korea Mexico China Indones. Tsimane
1

Development
GDP per cap ($) 40,100 25,300 19,200 9,600 5,600 3,600 200
Infant mort rate 6.5 6.4 7.05 20.91 24.18 35.6 105.0
Labor force in
agricult/forestry
1% 8% 8% 18% 49% 45% ~100%
Life exp at birth
Males 74.9 74.5 73.4 72.4 70.7 67.1 54.3
Females 80.7 80.3 80.6 78.1 74.1 72.1 54.0
1. Gurven, Kaplan & Supa 2007

64

Table 3.2. Survey characteristics
Country Dataset Year Eligible
Population
Nationally
Rep?
Method of
Interview
U.S. Health & Retirement
Study (HRS)
2006 50+ (and
spouses)
Yes Phone & in-
person
Taiwan Social Environment &
Biomarkers of Ageing
Study (SEBAS)
2006 53+ Yes In-person
Korea Korean Longitudinal
Study of Aging
(KLoSA)
2006 45+ Yes In-person
Mexico Mexican Health &
Aging Study (MHAS)
2001 50+ (and
spouses)
Yes In-person
China China Health &
Retirement
Longitudinal Study
(CHARLS)
2011-2012 45+ No In-person
Indonesia Indonesian Family
Life Study (IFLS4)
2007-2008 All ages 83% of
population
In-person
Tsimane Tsimane Health &
Life History Project
Baseline
(ongoing)
40+ No In-person

65

Table 3.3. Survey sample characteristics

U.S. Taiwan Korea Mexico China Indones. Tsimane
Total Sample 18,469 1,284 10,254 15,402 17,708 44,103 ~3,000
Age-Eligible
Sample (55-85)
15,038 1,092 6,747 9,722 11,071 4,220 596
Analytic Sample* 14,125 1,051 6,532 8,846 10,227 4,196 449
% Women 55% 48% 54% 53% 51% 51% 47%
% Age 55-64 33% 50% 42% 53% 56% 54% 53%
% Age 65-74 42% 27% 40% 32% 28% 35% 34%
% Age 75-85 25% 23% 18% 14% 14% 12% 13%
*Excludes those missing data on sex or ADL disability and those interviewed via proxy

66

Table 3.4. Available measures by country

Objective Measures Self-Reported Measures

Physical Performance Functional Tasks ADLs

Grip Gait Chair Tand 1 Leg Squat Stair Carry Dress Bathe Toilet
U.S. X X

X

X X X X X X
Taiwan X X X

X X X X X X
Korea X

X X X
Mexico

X X X X X X X
China X X X X

X X X X X X
Indon. X

X

X

X X X X
Tsim.

X X X X

X X X
67

Figure 3.1. Average physical performance by age and Country.
68

Figure 3.2. Left: Prevalence of reported functioning difficulties by age and Country. Right:
Prevalence of reported ADL difficulties by age and Country.

69
Table 3.5. Odds ratios and regression coefficients of the effect of age, controlling for sex: Physical performance

U.S. Taiwan Korea Mexico China Indonesia Tsimane
B B B B B B B
Grip Strength (kg)

65-74 -3.76 *** -4.18 *** -3.43 *** ---

-3.54 *** -3.18 *** ---

[-4.25, -3.27] [-5.01, -3.36] [-3.77, -3.09]

[-4.04, -3.05] [-3.79, -2.57]

75-85 -8.26 *** -7.12 *** -6.24 *** ---

-7.00 *** -6.98 *** ---

[-8.79, -7.72] [-8.56, -5.68] [-6.69, -5.79]

[-7.69, -6.31] [-7.94, -6.03]

Gait Speed (m/s)

75-85 -0.12 *** -0.09 *** ---

---

-0.10 *** ---

-0.03 *

[-.14, -.10] [-.14, -.04]

[-.12, -.08]

[-.05, -.01]
Chair Stands (sec)

65-74 ---

2.18 *** ---

---

1.00 *** 1.36 *** 0.61 *

[1.62, 2.73]

[.75, 1.25] [1.11, 1.60] [.07, 1.28]
75-85 ---

3.55 *** ---

---

2.68 *** 2.68 *** 1.69 ***

[2.90, 4.19]

[2.27, 3.09] [2.22, 3.13] [.69, 2.68]

OR

OR

OR

OR

OR

OR

OR

Tandem Stand

65-74 1.94 *** ---

---

---

1.68 *** ---

6.20 ***

[1.61, 2.34]

[1.35, 2.08]

[2.69, 14.28]
75-85 6.23 *** ---

---

---

5.03 *** ---

6.49 ***

[5.16, 7.52]

[3.88, 6.52]

[2.40, 17.54]
One-Leg Stand

65-74 ---

---

---

1.84 ** ---

---

3.33 ***

[1.24, 2.72]

[2.01, 5.53]
75-85 ---

---

---

3.36 *** ---

---

9.43 ***
[1.87, 6.04] [4.82, 18.46]
*p<.05 **p<.01 ***p<.001

70

Table 3.6. Odds ratios of the effect of age, controlling for sex: Self-reported measures

U.S. Taiwan Korea Mexico China Indonesia Tsimane
OR OR OR OR OR OR OR
Phys Function
Squatting
65-74 1.32 *** 3.53 *** ---

1.53 *** 1.16 * 3.51 *** ---

[1.19, 1.46] [2.56, 5.53]

[1.27, 1.84] [1.03, 1.30] [2.43, 5.08]
75-85 1.98 *** 5.84 *** ---

2.35 *** 1.43 *** 7.37 *** ---

[1.77, 2.21] [3.79, 8.99]

[1.82, 3.03] [1.20, 1.69] [4.89, 11.11]
Climbing Stairs
65-74 1.44 *** 4.48 *** ---

1.70 *** 1.19 * ---

---

[1.28, 1.62] [3.01, 6.65]

[1.41, 2.06] [1.04, 1.37]
75-85 1.99 *** 6.28 *** ---

2.88 *** 1.53 *** ---

---

[1.78, 2.22] [4.04, 9.76]

[2.20, 3.76] [1.24, 1.90]
Carrying
65-74 1.21 ** 3.26 *** ---

1.90 *** 1.79 *** 2.71 *** ---

[1.08, 1.36] [2.01, 5.29]

[1.48, 2.43] [1.49, 2.16] [2.25, 3.28]
75-85 1.92 *** 5.43 *** ---

3.22 *** 3.67 *** 6.04 *** ---

[1.66, 2.22] [2.93, 10.09]

[2.39, 4.35] [3.06, 4.41] [4.60, 7.93]

ADLs OR

OR

OR

OR

OR

OR

OR
Dressing
65-74 0.66 *** 3.29 * 2.41 ** 1.68 ** 1.29 * 2.85 *** 2.66 **

[.53, .81] [1.18, 9.17] [1.46, 3.97] [1.23, 2.30] [1.04, 1.59] [2.19, 3.69] [1.45, 4.88]
75-85 0.87

6.00 *** 5.70 *** 2.54 *** 2.38 *** 4.99 *** 3.92 ***

[.69, 1.09] [2.21, 16.25] [3.44, 9.44] [1.72, 3.74] [1.79, 3.16] [3.70, 6.75] [1.87, 8.20]
Bathing
65-74 1.04

4.00 * 2.87 *** 2.23 ** 1.46 *** 2.94 *** 2.42 **

[.84, 1.29] [1.03, 15.56] [1.68, 4.90] [1.40, 3.57] [1.21, 1.76] [2.08, 4.14] [1.31, 4.46]
75-85 2.05 *** 8.67 *** 8.35 *** 4.01 *** 2.89 *** 5.00 *** 3.87 ***

[1.59, 2.64] [2.68, 28.10] [5.03, 13.85] [2.45, 6.56] [2.23, 3.74] [3.36, 7.45] [1.86, 8.08]
Toilet
65-74 1.02

4.65 ** 2.70 ** 1.90 ** 1.40 *** 3.12 *** 4.29 **

[.80, 1.30] [1.71, 12.62] [1.33, 5.48] [1.21, 2.98] [1.22, 1.62] [1.66, 5.86] [1.73, 10.66]
75-85 2.05 *** 7.41 *** 4.00 *** 3.65 *** 2.25 *** 6.54 *** 3.73 *
[1.66, 2.53] [2.63, 20.88] [2.01, 7.95] [2.10, 6.33] [1.85, 2.74] [3.33, 12.86] [1.20, 11.60]
*p<.05 **p<.01 ***p<.001

71
Table 3.7. Correlations between country-level indicators (2005) and the effect of age (75-85yrs versus 55-64yrs)

GNP
Infant Mortality
Rate
Education
(men, yrs)
BMI
(men, kg/m3)
Life Expectancy
(both sexes)
OR N Corr p N Corr p N Corr p N Corr p N Corr p
Dress 7
-.23 .616
7
.02 .964
7
-.03 .951
7
-.62 .140
7
-.06 .905
Bathe 7
.09 .849
7
-.29 .529
7
.31 .505
7
-.31 .506
7
.27 .556
Toilet 7
-.12 .805
7
-.06 .896
7
-.02 .971
7
-.43 .335
7
.04 .932
Squat 5
-.21 .729 5 .31 .608 5 -.19 .759 5 -.57 .315 5 -.37 .539
Stairs 4
.19 .809 4 -.55 .446 4 .23 .772 4 -.17 .828 4 .52 .484
Carry 5
-.52 .372 5 .46 .431 5 -.49 .405 5 -.82 .092 5 -.54 .344
Tandem 3
.23 .853 3 .50 .668 3 -.17 .893 3 .42 .721 3 -.45 .702
One-Leg 2
--- --- 2 --- --- 2 --- --- 2 --- --- 2 --- ---
Coeff

Grip (kg) 5
-.63 .251 5 .17 .782 5 -.28 .653 5 -.75 .145 5 -.24 .694
Gait (m/s) 4
-.75 .249 4 .94 .058 4 -.92 .081 4 -.60 .402 4 -.94 .058
Chair (sec) 4
.89 .114 4 -.95 .053 4 .99 .006 4 .36 .637 4 .96 .040

72
Table 3.8. Correlations between country-level indicators (2005) and prevalence/average level of difficulties
and performance (age 55-64yrs)

GNP
Infant Mortality
Rate
Education
(men, yrs)
BMI
(men, kg/m3)
Life Expectancy
(both sexes)
Prevalence N Corr p N Corr p N Corr p N Corr p N Corr p
Dress 7
-.40 .380
7
.72 .069
7
-.70 .081
7
-.05 .914
7
-.71 .072
Bathe 7
-.44 .326
7
.69 .087
7
-.67 .099
7
-.28 .551
7
-.71 .076
Toilet 7
-.19 .677
7
.00 .992
7
-.16 .731
7
-.15 .744
7
-.04 .939
Squat 5
.25 .690 5 -.35 .566 5 .22 .716 5 .55 .334 5 .39 .517
Stairs 4
-.57 .434 4 .83 .166 4 -.59 .405 4 -.16 .840 4 -.81 .185
Carry 5
.03 .967 5 .29 .641 5 -.01 .983 5 .31 .606 5 -.21 .730
Tandem 3
.70 .503 3 -1.00 .023 3 .93 .249 3 .54 .635 3 .99 .058
One-Leg 2
--- --- 2 --- --- 2 --- --- 2 --- --- 2 --- ---
Level

Grip (kg) 5
.77 .124 5 -.56 .330 5 .61 .274 5 .91 .033 5 .59 .290
Gait (m/s) 4
.92 .078 4 -.95 .053 4 1.00 .001 4 .74 .265 4 .96 .038
Chair (sec) 4
-.48 .520 4 .76 .236 4 -.77 .231 4 .23 .769 4 -.76 .243

73
CHAPTER 4: RATE OF CHANGE IN PHYSICAL FUNCTION AND DISABILITY IN
TRADITIONAL VERSUS MODERN CONTEXTS
Introduction
Age is one of the most consistent predictors of functional limitations and disability.
This is true for both self-reported (Beckett et al., 1996; Liang et al., 2008) and
performance-based measures (Samson et al., 2000; Seeman et al., 1994) . Examination of
change in functioning ability with age, rather than cross-sectional differences in prevalence
of poor functioning, is important in assessing the process of aging. Declines in functioning
and increases in disability become more likely with increasing age and as length of
observational time increases (Liang et al., 2003) . Most of what is known comes from
studies done in developed countries; much less is known about age changes in physical
function and disability in traditional societies. There is a lacuna of historical data and
limited data on contemporary indigenous populations. We hypothesize that the aging
process will be more rapid and physical deterioration occur at earlier ages in societies
where people have been exposed to more demanding physical regimes, more illness, and
fewer medical resources. In order to examine this question, we look at the rate of change in
physical functioning and disability among older persons in two societies, the Tsimane of
Bolivia and the United States.
Background
There are several reasons to hypothesize that physical aging may be more rapid
among traditional populations. For much of human history, average life expectancy at birth
was only about 25 years, compared with about 75-80 years in wealthy countries today
(Wilmoth, 2000) Although life expectancy has increased nearly 3-fold in most populations,
74
it is still quite low among traditional populations such as the Tsimane forager
horticulturalists of Bolivia (Gurven, Kaplan, Winking, Finch, & Crimmins, 2008) . However,
average life expectancy can be greatly influenced by infant mortality. Among those who
reach adulthood, a substantial proportion may survive to their sixth and seventh decade,
even among hunter-gatherers. For example, Gurven and Kaplan found that adult mortality
rates remain low through the 5
th
decade among hunter-gatherers (2007) and among the
Tsimane, 40% of those born survive to age 60 (Gurven, Kaplan, & Supa, 2007) . Still, adult
mortality is relatively high among the Tsimane and this may be due in part to high levels of
infection and inflammation throughout life (Gurven et al., 2008; Vasunilashorn et al., 2010)
since chronic disease mortality at older ages has been linked to infection and inflammation
in early life (Finch & Crimmins, 2004) . Chronic inflammation is also a risk factor for low
muscle strength and poor physical performance (Barbieri et al., 2003; Cesari et al., 2004) .
Physiological wear and tear may also be greater among traditional populations due to their
lifestyles demanding of physical activity and limited access to healthcare. One-fifth of
deaths among hunter-gatherers have been attributed to accidents and violence (Gurven et
al., 2007) , suggesting that injuries may be more common and relatively more debilitating.
Other environmental factors play a role; physical limitations may be relatively more
disabling in traditional contexts.
On the other hand, increases in life expectancy do not necessarily result in better
population health. There is a long history of debate about whether increased life
expectancy leads to an expansion (Gruenberg, 1977) or compression (Fries, 1980) of
morbidity. Empirical evidence has been mixed in industrialized populations and findings
are sensitive to the measures used. Trends in the US suggest that late-life disability has
75
declined in recent years (Schoeni, Freedman, & Wallace, 2001; Schoeni, Freedman, &
Martin, 2008) , although not all findings are as positive (Seeman, Merkin, Crimmins, &
Karlamangla, 2010) . In more developed contexts, chronic diseases have overtaken
infectious disease as the leading causes of death (Omran, 1971). This is important because
chronic diseases are well-known risk factors for poor functioning and disability. Much of
this increase in chronic disease has been attributed to changes in lifestyle, including
physical activity and obesity (Archer & Blair, 2011; Popkin, 2006; Reddy, 2002) . Yet there
is also evidence that chronic conditions have become less disabling over time in the United
States (Costa, 2002).
Among the Tsimane and similar groups, obesity is rare and adults are more
physically active (Gurven, Jaeggi, Kaplan, & Cummings, 2013) . Although this physical
activity tends to be moderate rather than strenuous, recent evidence suggests that
sedentary behavior has negative health consequences above and beyond lack of physical
activity (Owen, Sparling, Healy, Dunstan, & Matthews, 2010) . Therefore, older adults in
such contexts may maintain higher levels of physical function. On the other hand it is also
possible that selective mortality is higher among the Tsimane, which would result in
seemingly better population health as indicated by functioning and disability.
The present study seeks to understand whether declines in physical function and
increases in disability are indeed greater in a traditional population compared with the U.S.
We examine both objective measures of physical function in addition to self-reports and
use longitudinal data to examine change within individuals.
Methods
Data
76
Tsimane Health & Life History Project
The UNM-UCSB Tsimane Health & Life History Project (THLHP) began studying the
Tsimane in 2002. The Tsimane Amerindians are an indigenous forager-horticulturist group
comprised of approximately 9,000 individuals who live in the Beni region of Bolivia
(Beheim, n.d.). They reside in 80+ small villages of 50-150 people living in extended family
clusters. The UNM-UCSB Tsimane Health & Life History Project (THLHP) began studying
the Tsimane in 2002 and the project currently includes data on approximately 3,000
individuals from 83 villages. The project has collected extensive data on demographic,
social, and health conditions of the Tsimane. In addition to self-reported information
collected via in-person interviews, physical performance measures and biomarkers are
also available. Finally, data collection continues as of 2013, thus, longitudinal measures are
available. There were 263 and 477 individuals with data at 2 or more time points for
disability and balance, respectively, and 209 individuals aged 65+ with data on measured
gait speed at 2+ time points. Detailed information on the number of individuals with data
for multiple time points is show in Appendix D.
Health & Retirement Study (HRS)
Data for the United States come from the Health and Retirement Study (HRS). The
HRS is a longitudinal study that interviews approximately 26,000 Americans aged 50 and
older and their spouses every two years since 1992. The sample is nationally
representative of the older population after applying sample weights. The HRS collects
extensive data on demographic, social, economic, and health characteristics.
In addition, in 2006, HRS began collecting data on physical performance and
biomarkers via enhanced face-to-face interviews. Half of households were randomly
77
selected to participate in this additional component in 2006 and eligible respondents
included non-institutionalized individuals who completed an in-person interview
themselves (not by proxy). The enhanced interview components were administered to the
remaining 50% of the HRS sample in 2008 and in 2010 physical performance and
biomarkers were measured again for the initial subset of individuals selected in 2006.
Thus, data from the 2006 and 2010 waves of the HRS are used to provide longitudinal data
at 2 time points. There were 11,953 respondents with data on disability at both time
points, while 4,474 had data on balance and 2,659 individuals aged 65+ had data on gait
speed.
Measures
There are two types of measures of functioning and disability: self-reports and
measured performance. We use both types in this analysis.
Disability
Disability was measured using self-reported difficulty with 5 activities of daily living
(ADLs), including eating, bathing, dressing, going to the bathroom, and walking. Each item
was dichotomized to indicate a person either did not have difficulty or had difficulty/could
not perform the task. Increased disability was then defined as having difficulty with or
being unable to perform more ADL tasks at follow-up.
Measured Physical Performance
Physical performance measures included tests of lower extremity function and
balance. Lower extremity function (gait speed) was measured with a timed walk that was
an average of two trials. Tsimane individuals aged 40 and older were asked to walk 3
meters while US individuals aged 65 or older were asked to complete a timed walk, which
78
was 98.5 inches or approximately 12 feet (or 3.66 meters) in length. Gait speed was
calculated as meters per second. Those who tried but were unable, refused for safety
reasons, and those who used a walking aid at baseline were coded as missing. Decline in
gait speed was defined as having a 0.05 or slower gait speed at follow-up and included
those who were unable to perform the test.
Balance was assessed with a timed tandem stand. Among the Tsimane, the tandem
test measured the seconds an individual could hold the tandem position, for a maximum of
10 seconds. HRS respondents were first asked to complete a semi-tandem test. If
individuals younger and older than 65 could hold the semi-tandem position for 60 seconds
and 30 seconds, respectively, then they were asked to hold the tandem position. To
facilitate comparison, times on the tandem test were top-coded at 10 seconds. Individuals
who did not try because of safety reasons or who tried and were unable to do the test were
coded as missing at baseline, as were US respondents who could not complete the semi-
tandem stand. Decline in balance was defined as being able to hold the tandem position for
fewer seconds at follow-up and included those who were unable to perform the test.
Independent Variables
Independent variables include sex and age (50-59, 60-69 and 70-95). For
longitudinal analyses, time was measured in months, and was obtained by subtracting the
date of follow-up from the baseline date.
Analysis
Parametric survival analysis with a Weibull model specified was used to analyze
increases in disability and declines in physical performance between baseline and follow-
up. Age and sex were included as covariates and analyses were performed separately by
79
country. Among the Tsimane, change was calculated between baseline measurement and
the longest follow-up measurement; for the United States, change was calculated between
2006 and 2010. Analyses were performed with Stata (v. 11)
Results
Descriptive Results
Descriptive characteristics of the samples are shown in Table 1. Characteristics are
shown for each outcome because only 50% of HRS respondents were eligible for the
physical measures in 2006 and 2010 and because only those aged 65 and older were
eligible for the gait speed test in the U.S.
< Table 1 about here >
In general, the percentage of women was lower among the Tsimane than in the U.S.
The Tsimane also tended to be younger than the U.S. sample, with a lower proportion of
individuals age 70+. In terms of ADL disability, the average number of difficulties was
higher at baseline among the Tsimane, as was the proportion who experienced an increase
in ADL difficulties. However, there was a decrease in the average number of ADL difficulties
among the Tsimane while the opposite was found among U.S. older adults. For balance, the
Tsimane were actually able to hold the tandem position for longer at baseline and fewer
individuals experienced decline in balance. Finally, baseline gait speed was lower among
the Tsimane, yet they also experienced less decline in gait speed. Detailed baseline level
and change by age group are provided in Appendix G.
Due to differences in the study designs, the time span between measures varied
somewhat between the U.S. and the Tsimane. In general, the average number of months
was greater in the U.S., but the time intervals were more variable among the Tsimane.
80

Parametric Survival Analysis
Results from parametric survival models show that older age was associated with
higher hazards of increased ADL disability among U.S. and Tsimane older adults, except
among those aged 60-69 in the U.S. (Table 2). Women also had a significantly greater
hazard of disability increase, particularly Tsimane women. Although separate models
precluded the possibility of testing whether the hazard ratios associated with age differed
significantly between populations, there was no overlap in the 95% confidence intervals
for the hazard ratios for the age groups 60-69 and 70+, suggesting that the hazard of
disability increase associated with age was greater among the Tsimane.
< Table 2 about here >
With respect to decline in measured balance, age was associated with a greater
hazard of decline in both populations, although the effect was not significant among those
aged 60-69 in either population. Controlling for sex and compared with those aged 50-59,
those aged 70 and older had 9.2 times and 2.5 times greater hazard of decline in balance
among the Tsimane and in the U.S., respectively. As was seen for disability increase, lack of
overlap in the 95% confidence intervals of the hazard ratio associated with being age 70+
suggests that the age-related hazard of decline in balance was greater among the Tsimane.
Sex was not associated with decline in balance in either population.
Finally, results from parametric survival models indicate that age was only
significantly associated with a greater hazard gait speed decline among U.S. adults aged
75+ relative to those aged 65-74 and women did not have a greater hazard of decline in
either population. Yet since average gait speed was considerably lower among the Tsimane,
81
we examined an alternative definition of gait speed decline: a half standard deviation or
more decrease in gait speed at follow-up (Appendix H). Although this changed the
proportion of individuals who decline and the hazard ratios associated with age were
somewhat larger, the relative difference between populations was similar.
Discussion
Age was found to be a consistent predictor of increased disability in self-reported
tasks of daily living as well as decline in measured balance and gait speed among both the
Tsimane and U.S. older adults. Although the age coefficient was not statistically significant
for decline in gait speed among the Tsimane, the coefficient was in the expected direction
and the lack of significance was likely a due to the small sample size of Tsimane age 65+
since gait speed was only measured among U.S. older adults aged 65 and older. However,
there was also evidence that the hazard of decline in balance and ability to perform self
care tasks associated with increased age was relatively larger among the Tsimane
compared to U.S. older adults.
Although the hazard of decline in balance associated with age was larger among the
Tsimane, at baseline the Tsimane were able to hold the tandem position for longer, on
average, and fewer individuals displayed a decline in balance at follow-up. It is possible
that Tsimane older adults maintain better balance at older ages as a result of their fairly
high levels of physical activity throughout life as well as the types of tasks they engage in
such as farming, hunting and fishing (Gurven et al., 2013) . In terms of disability, the
Tsimane had more ADL difficulties on average at baseline and a higher proportion showed
decline over time. The hazard associated with age was also larger among the Tsimane,
painting a worrisome picture in terms of self-care.
82
Aging, as indicated by physical performance and disability, may be more rapid
among the Tsimane for several reasons. Although the Tsimane are more active than those
in the U.S., they also have a higher burden of infection and inflammation (Gurven et al.,
2008) . Injuries are more commonplace and are likely to be more disabling due to the lack
of access to adequate health care and other social policies for the aged. In addition, more
rapid age-related increases in disability among the Tsimane may be partly attributable to
differences in the environment, since disability can be conceptualized as a lack of fit
between a person and their environment (Verbrugge & Jette, 1994) . For example, most
Tsimane lack indoor plumbing and electricity, which make it more difficult to perform
basic activities of daily living, such as bathing or going to the bathroom. On the other hand
informal care provided by family may be more readily available given differences in living
arrangements and social structures.
Limitations
This study has some limitations that must be acknowledged. First, the time interval
was somewhat different between the U.S. and Tsimane datasets. In the U.S., change was
examined between 2006 and 2010 with a mean time interval of about 51 months. Among
the Tsimane, change was measured between baseline and the longest follow-up, for a mean
interval of 31-38 months and greater variation in the time interval. To determine whether
this affected results, sensitivity analyses were performed. Several alternative time spans
were examined for the Tsimane (Appendix I). For time intervals closest to 40, 50 or 60
months, substantive results were not changed greatly. When the time interval was
restricted to between 40-60 months, the hazard ratios associated with age were smaller for
decline in balance, although this may be due to the loss of approximately 2/3 of the sample.
83
For the U.S., we compared ADL change between 2006-2008 and between 2006-2008
(Appendix J). Although the percentage with increased disability was greater in the larger
time span, differences in the hazard ratios associated with age did not differ in the model
that included people who were not missing in 2010. Change in physical performance
measures for 2006-2008 could not be compared because these are obtained every other
wave. However, the HRS collected a limited number of performance measures 2004 (pilot
module), including gait speed, but not balance. Therefore change in gait speed (and ADLs)
was compared between 2004-2006 and 2004-2008 (Appendix K). In general, the hazard
associated with age was somewhat greater for both measures for the longer time span,
although the differences were not significant. Yet it must be noted that the gait speed
samples for 2004-2006 and 2004-2008 are mutually exclusive due to the data collection
schedule for physical measures and biomarkers.
Another limitation was the small size of the Tsimane sample. In the model of gait
speed decline, the age-related hazard was not significant despite a substantial effect size,
which could be due to the small sample size. Unfortunately, this limitation cannot
realistically be resolved. Indigenous populations tend to be quite small and are
predominantly young because of their low average life expectancy. This study actually
includes one of the largest samples of indigenous older adults available. We did perform
analyses on a random subsample of the U.S. data and find that despite similar effect sizes,
hazard ratios associated with greater age were not significant (Appendix L). However,
relative differences between the two populations were similar.
Conclusion
84
Overall, the findings of the present study lend support for the hypothesis that
physical aging, as indicated by physical functioning and disability is more rapid among the
Tsimane, a traditional, indigenous population characterized by low average life expectancy
and high burdens of infection and inflammation relative to a more modern population (the
United States), despite higher levels of obesity and cardiovascular disease among older
adults in the U.S. These findings also highlight the importance of the environment in
shaping trajectories of physical performance and disability in later life.
85
Table 4.1. Baseline characteristics & average change (50-95years)

Tsimane U.S.
ADL Disability
Female 48.3% 55.4%
Age
50-59 47.2% 41.0%
60-69 32.3% 32.3%
70+ 20.5% 26.7%
Baseline difficulties (SD) .49 (1.22) .17 (.62)
Change difficulties (per yr) -.14 (-.05) .07 (.02)
% Increased (per yr) 9.1% (3.6%) 5.5% (1.3%)
Months (avg, SD, min, max) 30.7 (10.6) 51.2 (4.3)

[10.1, 45.9] [37.5, 62.9]
N 263 11,953
Balance
Female 47.4% 52.9%
Age
50-59 54.5% 44.3%
60-69 28.3% 31.6%
70+ 17.2% 24.1%
Baseline, sec (SD) 9.91 (.73) 9.76 (1.34)
Change, sec (per yr) -.72 (-.23) -1.11 (-.26)
% Declined (per yr) 9.9% (3.1%) 16.5% (3.9%)
Months 38.4 (19.1) 51.0 (4.1)

[5.7, 80.6] [38.5, 61.9]
N 477 4,474
Gait Speed
Female 46.9% 56.9%
Age
65-75 74.2% 60.1%
75+ 25.8% 39.9%
Baseline, m/s (SD) .34 (.07) .82 (.24)
Change, m/s (per yr) -.02 (-.01) -.10 (-.02)
% Declined (per yr) 31.1% (11.0%) 56.0% (13.0%)
Months 34.1 (19.3) 51.8 (3.8)

[9.9, 79.3] [40.5, 61.9]
N 209 2,659
86
Table 4.2. Hazard ratios from parametric survival models of ADL increase or
balance/gait speed decline

Tsimane (N=263)

U.S. (N=11,953)
ADLs HR p L95 U95

HR p L95 U95
Female 3.24 0.010 1.33 7.92

1.34 0.001 1.13 1.60
Age (50-59)
60-69 4.78 0.022 1.26 18.15

0.78 0.081 0.58 1.03
70+ 10.77 <.001 3.07 37.82

1.77 <.001 1.40 2.24

Tsimane (N=477)

U.S. (N=4,474)
Balance HR p L95 U95

HR p L95 U95
Female 1.62 0.104 0.91 2.91

1.09 0.287 0.93 1.28
Age (50-59)
60-69 1.66 0.262 0.68 4.03

1.04 0.735 0.81 1.34
70+ 9.22 <.001 4.57 18.59

2.47 <.001 1.99 3.06

Tsimane (N=209)

U.S. (N=2,659)
Gait Speed HR p L95 U95

HR p L95 U95
Female 1.29 0.308 0.79 2.10

0.99 0.793 0.88 1.10
Age (65-74)
75+ 1.34 0.269 0.80 2.56 1.31 <.001 1.17 1.47
87
CHAPTER 5: CONCLUSION
This dissertation began by asking the following questions: Do women always report
greater disability and display poorer performance on physical tasks? Are age-related
declines in physical ability and increases in disability consistent across countries or do
individuals in some populations “age” more quickly than others? Are longitudinal (intra-
individual) changes in functioning and disability consistent with cross-sectional age
differences? Are findings consistent with respect to different stages of the disablement
process (functional impairments versus disability) and methods of measurement (objective
measures versus self-reports)? Do age and sex differences vary systematically with respect
to level of development?
With respect to gender differences, women nearly always showed poorer physical
performance and reported more difficulty with functional tasks in the 7 countries
examined. Differences were quite consistent across these more “situation-free” measures,
both objective and self-reported (Figures 1 & 2). Differences were also remarkably
consistent across contexts that varied widely in terms of level of economic development,
gender roles, etc. This may be because women generally have lower muscle mass and
strength and tend to be less physically active. Women also tended to have greater odds of
difficulty with basic self-care tasks, although sex differences were smaller than for physical
functioning and were not always significant. Yet findings were also relatively consistent
across countries and no clear pattern was observed with respect to level of economic
development, as indicated by per capita GDP (Figures 1 & 2). Also, there seemed to be little
pattern with respect to gender equality. In part, this may be because “gender equality” can
be either protective or detrimental, depending on the domain. For example, in the U.S.,
88
gender differences in labor force participation and education were relatively small and
both income and education are associated with lower risk of disability. Yet gender
differences in smoking rates were smaller, which could offset the effect of greater gender
equality in work and education. Gender differences did appear to be smaller in China,
which may be due to their unique history under communism. Gender differences also
tended to be small in Korea.
Turning to age differences, in nearly all measures in all populations, age was
associated with significantly worse physical function and disability. For measured
performance, age differences were quite similar across most countries (Figures 3 & 4). The
Tsimane were an exception; age differences in gait speed and chair stands were
considerably smaller among the Tsimane, which may be because average performance on
these tests was poor at younger ages. This implies that considerable “aging” had already
occurred in these domains. Yet the Tsimane displayed relatively good balance at younger
ages.
Much more variation was observed in the effect of age on reported difficulty with
functional tasks. Age differences tended to be more pronounced in populations with a low
prevalence of difficulty at younger ages. Although differences in the prevalence of self-
reported difficulty must be interpreted with caution, there was evidence that self-reported
functioning and objective performance measures were in concordance. For example, the
Taiwanese performed relatively well in terms of gait speed and chair stands, and also had a
lower prevalence of difficulty squatting and climbing stairs. It also appeared that
populations could do well vis-à-vis the others in one domain, but rank poorly in others.
This could be due to differences across populations in normative physical activity
89
(recreation, labor/chores, transportation etc.) that could promote the maintenance of
abilities in certain tasks.
With respect to ADL tasks, age differences were less variable. Again, age differences
tended to be more pronounced in populations where the prevalence of difficulty with self-
care tasks was relatively low at younger ages. Overall, there was no evidence for a clear
pattern of age differences with respect to level of development (Figures 3 & 4). However, it
is necessary to consider both the level at younger ages as well as age differences since high
levels of difficulty at younger ages indicates that “aging” has already occurred. When
considering both, it appeared that the Tsimane and Chinese populations were “aging” more
quickly and these populations are at the lower end of the development spectrum.
Indonesians appear to be “aging” more rapidly in some domains (e.g. upper body strength),
but show higher levels of lower body function and less age-related decline.
Finally, results from longitudinal parametric survival models were consistent with
findings from cross-sectional models of age differences, although data were only available
for two populations: the U.S. and the Tsimane. The hazard of decline in balance and hazard
of increase in disability associated with age were greater among the Tsimane.
In conclusion, this dissertation found that female disadvantage female is universal
for physical functioning and is frequently observed, but less pronounced for self-care tasks.
Countries at very low levels of development do appear to be “aging” more quickly than
those in middle and high-income countries and this suggests that demand for formal and
informal care will increase. However, future work should examine functioning and
disability at earlier ages, since considerable “aging” and population selection has already
occurred by age 55. Future work should also seek to understand factors that promote
90
higher levels/preservation of functioning in certain domains. For instance, better
understanding of the factors leading to maintenance of lower body strength among
Indonesians could benefit older adults in other populations.
91

Figure 5.1. Per capita GNP (2005) versus the effect of being female (OLS regression
coefficients)
92

Figure 5.2. Per capita GNP (2005) versus the effect of being female (odds ratios from
logistic regression)

93

Figure 5.3. Per capita GNP (2005) versus the effect of age 75-85 relative to 55-64 years
(OLS regression coefficients)
94

Figure 5.4. Per capita GNP (2005) versus the effect of age 75-85 relative to 55-64 years
(odds ratios from logistic regression)

95

Figure 5.5. Per capita GNP (2005) versus prevalence of difficulty among those aged 55-64
96
REFERENCES
Alvarado, B. E., Guerra, R. O., & Zunzunegui, M. V. (2007). Gender differences in lower
extremity function in Latin American elders seeking explanations from a life-course
perspective. Journal of Aging and Health, 19(6), 1004-1024.
Archer, E., & Blair, S. N. (2011). Physical activity and the prevention of cardiovascular
disease: From evolution to epidemiology. Progress in Cardiovascular Diseases, 53(6),
387-396.
Barbieri, M., Ferrucci, L., Ragno, E., Corsi, A., Bandinelli, S., Bonafè, M., et al. (2003). Chronic
inflammation and the effect of IGF-I on muscle strength and power in older persons.
American Journal of Physiology-Endocrinology and Metabolism, 284(3), E481-E487.
Barford, A., Dorling, D., Smith, G. D., & Shaw, M. (2006). Life expectancy: women now on top
everywhere: During 2006, even in the poorest countries, women can expect to outlive
men. BMJ: British Medical Journal, 332(7545), 808.
Beckett, L. A., Brock, D. B., Lemke, J. H., de Leon, C. F. M., Guralnik, J. M., Fillenbaum, G. G., et
al. (1996). Analysis of change in self-reported physical function among older persons in
four population studies. American Journal of Epidemiology, 143(8), 766-778.
Beheim, B. (n.d.) The Tsimane Health and Life History Project. Retrieved from
http://www.unm.edu/~tsimane/index.html.
Beltran, B., Cuadrado, C., Martin, M. L., Carbajal, A., & Moreiras, O. (2001). Activities of daily
living in the Spanish elderly: Association with mortality. The Journal of Nutrition,
Health & Aging, 5(4), 259-260.
Boutayeb, A., & Boutayeb, S. (2005). The burden of non communicable diseases in
developing countries. International Journal for Equity in Health, 4(1), 2-10.
97
Boutayeb, A. (2006). The double burden of communicable and non-communicable diseases
in developing countries. Transactions of the Royal Society of Tropical Medicine and
Hygiene, 100(3), 191-199.
Carr, A. J., Gibson, B., & Robinson, P. G. (2001). Measuring quality of life: Is quality of life
determined by expectations or experience? BMJ: British Medical Journal, 322(7296),
1240.
Caspersen, C. J., Pereira, M. A., & Curran, K. M. (2000). Changes in physical activity patterns
in the United States, by sex and cross-sectional age. Medicine and Science in Sports and
Exercise, 32(9), 1601-1609.
Central Intelligence Agency (CIA). (2012). The World Factbook. Retrieved from
https://www.cia.gov/library/publications/the-world-factbook/.
Cesari, M., Penninx, B. W. J. H., Pahor, M., Lauretani, F., Corsi, A. M., Williams, G. R., et al.
(2004). Inflammatory markers and physical performance in older persons: The
InCHIANTI study. The Journals of Gerontology Series A: Biological Sciences and Medical
Sciences, 59(3), 242-248.
Chan, L., Beaver, S., MacLehose, R. F., Jha, A., Maciejewski, M., & Doctor, J. N. (2002).
Disability and health care costs in the Medicare population. Archives of Physical
Medicine and Rehabilitation, 83(9), 1196-1201.
Chen, L., Yip, W., Chang, M. C., Lin, H. S., Lee, S. D., Chiu, Y. L., et al. (2007). The effects of
Taiwan's national health insurance on access and health status of the elderly. Health
Economics, 16(3), 223-242.
Chuanyi, Z., & Qin, X. (1992). Family care of the elderly in China: Changes and problems.
Sage Focus Editions, 145, 67-67.
98
Chun, H., Cho, S. I., Khang, Y. H., Kang, M., & Kim, I. H. (2012). Trends in gender-based health
inequality in a transitional society: A historical analysis of South Korea. Journal of
Preventive Medicine and Public Health, 45(2), 113-121.
Colbert, L. H., Visser, M., Simonsick, E. M., Tracy, R. P., Newman, A. B., Kritchevsky, S. B., et al.
(2004). Physical activity, exercise, and inflammatory markers in older adults: Findings
from the Health, Aging and Body Composition Study. Journal of the American Geriatrics
Society, 52(7), 1098-1104.
Cooper, R., Kuh, D., & Hardy, R. (2010). Objectively measured physical capability levels and
mortality: Systematic review and meta-analysis. BMJ: British Medical Journal, 341
Costa, D. L. (2002). Changing chronic disease rates and longterm declines in functional
limitation among older men. Demography, 39(1), 119-137.
Costa, D. L. (2005). Causes of improving health and longevity at older ages: a review of the
explanations. Genus, 21-38.
Crimmins, E. M., Kim, J. K., & Seeman, T. E. (2009). Poverty and biological risk: The earlier
“aging” of the poor. The Journals of Gerontology Series A: Biological Sciences and Medical
Sciences, 64(2), 286-292.
Crimmins, E. M., Kim, J. K., & Solé-Auró, A. (2011). Gender differences in health: Results
from SHARE, ELSA and HRS. The European Journal of Public Health, 21(1), 81-91.
Denton, M., Prus, S., & Walters, V. (2004). Gender differences in health: A Canadian study of
the psychosocial, structural and behavioural determinants of health. Social Science &
Medicine, 58(12), 2585-2600.
99
Drageset, J. (2004). The importance of activities of daily living and social contact for
loneliness: A survey among residents in nursing homes. Scandinavian Journal of Caring
Sciences, 18(1), 65-71.
Ferrer, M., Lamarca, R., Orfila, F., & Alonso, J. (1999). Comparison of performance-based
and self-rated functional capacity in Spanish elderly. American Journal of Epidemiology,
149(3), 228-235.
Ferrucci, L., Harris, T. B., Guralnik, J. M., Tracy, R. P., Corti, M. C., Cohen, H. J., et al. (1999).
Serum IL-6 level and the development of disability in older persons. Journal of the
American Geriatrics Society, 47(6), 639.
Finch, C. E., & Crimmins, E. M. (2004). Inflammatory exposure and historical changes in
human life-spans. Science, 305(5691), 1736-1739.
Fried, L. P., Kronmal, R. A., Newman, A. B., Bild, D. E., Mittelmark, M. B., Polak, J. F., et al.
(1998). Risk factors for 5-year mortality in older adults. JAMA: The Journal of the
American Medical Association, 279(8), 585-592.
Fries, J. F. (1980). Aging, natural death, and the compression of morbidity. New England
Journal of Medicine, 303(3), 130-135.
Gakidou, E., Cowling, K., Lozano, R., & Murray, C. J. (2010). Increased educational
attainment and its effect on child mortality in 175 countries between 1970 and 2009: a
systematic analysis. The Lancet, 376(9745), 959-974.
Gapminder. (2014). Institute For Health Metrics and Evaluation Data. Retrieved from
http://www.gapminder.org/data/
Gaugler, J., Duval, S., Anderson, K., & Kane, R. (2007). Predicting nursing home admission in
the US: A meta-analysis. BMC Geriatrics, 7(1), 13.
100
Griffith, L., Raina, P., Wu, H., Zhu, B., & Stathokostas, L. (2010). Population attributable risk
for functional disability associated with chronic conditions in Canadian older adults.
Age and Ageing, 39(6), 738-745.
Gruenberg, E. M. (1977). The failures of success. The Milbank Memorial Fund
Quarterly.Health and Society, , 3-24.
Guralnik, J. M., Branch, L. G., Cummings, S. R., & Curb, J. D. (1989). Physical performance
measures in aging research. Journal of Gerontology, 44(5), M141-M146.
Guralnik, J. M., Fried, L. P., & Salive, M. E. (1996). Disability as a public health outcome in the
aging population. Annual Review of Public Health, 17(1), 25-46.
Gurven, M., & Kaplan, H. (2007). Longevity among hunter‐gatherers: A cross‐cultural
examination. Population and Development Review, 33(2), 321-365.
Gurven, M., Kaplan, H., & Supa, A. Z. (2007). Mortality experience of Tsimane Amerindians
of Bolivia: Regional variation and temporal trends. American Journal of Human Biology,
19(3), 376-398.
Gurven, M., Kaplan, H., Winking, J., Finch, C., & Crimmins, E. M. (2008). Aging and
inflammation in two epidemiological worlds. The Journals of Gerontology Series A:
Biological Sciences and Medical Sciences, 63(2), 196.
Gurven, M., Jaeggi, A. V., Kaplan, H., & Cummings, D. (2013). Physical activity and
modernization among Bolivian Amerindians. PloS One, 8(1), e55679.
He, W., Muenchrath, M.N. & Kowal, P. U.S. Census Bureau. 2012. Shades of gray: A cross-
country study of health and well-being of the older populations in SAGE countries,
2007-2010. (P95/12-01). U.S. Government Printing Office, Washington, DC.
101
Houston, D. K., Ding, J., Nicklas, B. J., Harris, T. B., Lee, J. S., Nevitt, M. C., et al. (2009).
Overweight and Obesity Over the Adult Life Course and Incident Mobility Limitation in
Older Adults The Health, Aging and Body Composition Study. American journal of
Epidemiology, 169(8), 927-936.
Imperial College London. (2014). Global Burden of Metabolic Risk Factors of Chronic
Diseases. Retrieved from
http://www1.imperial.ac.uk/publichealth/departments/ebs/projects/eresh/majidezz
ati/healthmetrics/metabolicriskfactors/.
International Labour Organization (ILO) (2014). Key Indicators of the Labour Market
(KILM). Retrieved from http://www.ilo.org/empelm/what/WCMS_114240/lang--
en/index.htm
Jagger, C., Matthews, R., Melzer, D., Matthews, F., & Brayne, C. (2007). Educational
differences in the dynamics of disability incidence, recovery and mortality: Findings
from the MRC Cognitive Function and Ageing Study (MRC CFAS). International Journal
of Epidemiology, 36(2), 358-365.
Kaneda, T., & Zimmer, Z. (2007). Education, gender, and functional transitions among
Indonesian elderly. Journal of Cross-Cultural Gerontology, 22(3), 303-322.
Kaneda, T., Zimmer, Z., Fang, X., & Tang, Z. (2009). Gender differences in functional health
and mortality among the Chinese elderly. Research on Aging, 31(3), 361-388.
Keeler, E., Guralnik, J. M., Tian, H., Wallace, R. B., & Reuben, D. B. (2010). The impact of
functional status on life expectancy in older persons. The Journals of Gerontology Series
A: Biological Sciences and Medical Sciences, 65(7), 727-733.
102
Kempen, G., Van Heuvelen, M., Van den Brink, R., Kooijman, A., Klein, M., Houx, P., et al.
(1996). Factors affecting contrasting results between self-reported and performance-
based levels of physical limitations. Age and Ageing, 25(6), 458-464.
Kinsella, K. G., & Velkoff, V. A. (2001). An aging world: 2001 Government Printing Office.
Lee, E. O., & Lee, J. (2013). Education, functional limitations, and life satisfaction among
older adults in South Korea. Educational Gerontology, 39(7), 514-526.
Leung, J. C. (1997). Family support for the elderly in China: Issues and challenges. Journal of
Aging & Social Policy, 9(3), 87-101.
Leveille, S. G., Penninx, B., Melzer, D., Izmirlian, G., & Guralnik, J. M. (2000). Sex differences
in the prevalence of mobility disability in old age: The dynamics of incidence, recovery,
and mortality. Journals of Gerontology Series B, 55(1), 41-50.
Liang, J., Bennett, J. M., Shaw, B. A., Quiñones, A. R., Ye, W., Xu, X., et al. (2008). Gender
differences in functional status in middle and older age: Are there any age variations?
The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 63(5),
S282.
Liang, J., Shaw, B. A., Krause, N. M., Bennett, J. M., Blaum, C., Kobayashi, E., et al. (2003).
Changes in functional status among older adults in Japan: Successful and usual aging.
Psychology and Aging, 18(4), 684.
Liu, K., Wall, S., & Wissoker, D. (1997). Disability and Medicare costs of elderly persons.
Milbank Quarterly, 75(4), 461-493.
Luppa, M., Luck, T., Weyerer, S., König, H. H., Brähler, E., & Riedel-Heller, S. G. (2010).
Prediction of institutionalization in the elderly. A systematic review. Age and Ageing,
39(1), 31-38.
103
Macintyre, S., Hunt, K., & Sweeting, H. (1996). Gender differences in health: Are things
really as simple as they seem? Social Science & Medicine, 42(4), 617-624.
Merrill, S. S., Seeman, T. E., Kasl, S. V., & Berkman, L. F. (1997). Gender differences in the
comparison of self-reported disability and performance measures. The Journals of
Gerontology Series A: Biological Sciences and Medical Sciences, 52(1), M19.
Minicuci, N., Noale, M., Pluijm, S., Zunzunegui, M., Blumstein, T., Deeg, D., et al. (2004).
Disability-free life expectancy: A cross-national comparison of six longitudinal studies
on aging: The CLESA project. European Journal of Ageing, 1(1), 37-44.
Monteverde, M., Noronha, K., & Palloni, A. (2009). Effect of early conditions on disability
among the elderly in Latin America and the Caribbean. Population Studies, 63(1), 21-35.
Mor, V., Wilcox, V., Rakowski, W., & Hiris, J. (1994). Functional transitions among the
elderly: Patterns, predictors, and related hospital use. American Journal of Public
Health, 84(8), 1274-1280.
Myers, G. C., Lamb, V. L., & Agree, E. M. (2003). Patterns of disability change associated with
the epidemiologic transition. Determining Health Expectancies, , 59-74.
Ofstedal, M. B., Zimmer, Z., Hermalin, A. I., Chan, A., Chuang, Y. L., Natividad, J., et al. (2007).
Short-term trends in functional limitation and disability among older Asians: A
comparison of five Asian settings. Journal of Cross-Cultural Gerontology, 22(3), 243-261.
Oksuzyan, A., Crimmins, E., Saito, Y., O’Rand, A., Vaupel, J. W., & Christensen, K. (2010).
Cross-national comparison of sex differences in health and mortality in Denmark, Japan
and the US. European Journal of Epidemiology, , 1-10.
Omran, A. R. (1971). The epidemiological transition. Milbank Memorial Fund Quarterly,
49(4), 509-538.
104
Owen, N., Sparling, P. B., Healy, G. N., Dunstan, D. W., & Matthews, C. E. (2010). Sedentary
behavior: Emerging evidence for a new health risk. Paper presented at the Mayo Clinic
Proceedings, , 85. (12) pp. 1138.
Park, B., Jung, M., & Lee, T. (2009). Associations of income and wealth with health status in
the Korean elderly. Journal of Preventive Medicine and Public Health, 42(5), 275-282.
Peek, M. K., & Coward, R. T. (1999). Gender differences in the risk of developing disability
among older adults with arthritis. Journal of Aging and Health, 11(2), 131-150.
Popkin, B. M. (1999). Urbanization, lifestyle changes and the nutrition transition. World
Development, 27(11), 1905-16.
Popkin, B. M. (2002). Part II. what is unique about the experience in lower-and middle-
income less-industrialised countries compared with the very-high income
industrialised countries? Public Health Nutrition, 5(1a), 205-214.
Popkin, B. M. (2006). Global nutrition dynamics: The world is shifting rapidly toward a diet
linked with noncommunicable diseases. American Journal of Clinical Nutrition, 84(2),
289.
Rahman, O., Strauss, J., Gertler, P., Ashley, D., & Fox, K. (1994). Gender differences in adult
health: An international comparison. The Gerontologist, 34(4), 463-469.
Reddy, K. S. (2002). Cardiovascular diseases in the developing countries: Dimensions,
determinants, dynamics and directions for public health action. Public Health Nutrition,
5(1a), 231-237.
Reuben, D. B., Seeman, T. E., Keeler, E., Hayes, R. P., Bowman, L., Sewall, A., et al. (2004). The
effect of self-reported and performance-based functional impairment on future
hospital costs of community-dwelling older persons. The Gerontologist, 44(3), 401-407.
105
Roy, K., & Chaudhuri, A. (2008). Influence of socioeconomic status, wealth and financial
empowerment on gender differences in health and healthcare utilization in later life:
Evidence from India. Social Science & Medicine, 66(9), 1951-1962.
Rueda, S., & Artazcoz, L. (2009). Gender inequality in health among elderly people in a
combined framework of socioeconomic position, family characteristics and social
support. Ageing and Society, 29(4), 625.
Salomon, J. A., Tandon, A., & Murray, C. J. (2004). Comparability of self rated health: Cross
sectional multi-country survey using anchoring vignettes. Bmj, 328(7434), 258.
Samson, M. M., Meeuwsen, I., Crowe, A., Dessens, J., Duursma, S. A., & Verhaar, H. (2000).
Relationships between physical performance measures, age, height and body weight in
healthy adults. Age and Ageing, 29(3), 235-242.
Schoeni, R. F., Freedman, V. A., & Martin, L. G. (2008). Why is Late‐Life disability declining?
Milbank Quarterly, 86(1), 47-89.
Schoeni, R. F., Freedman, V. A., & Wallace, R. B. (2001). Persistent, consistent, widespread,
and robust? Another look at recent trends in old-age disability. The Journals of
Gerontology Series B: Psychological Sciences and Social Sciences, 56(4), S206-S218.
Scott, W. K., Macera, C. A., Cornman, C. B., & Sharpe, P. A. (1997). Functional health status as
a predictor of mortality in men and women over 65. Journal of Clinical Epidemiology,
50(3), 291-296.
Seeman, T. E., Charpentier, P. A., Berkman, L. F., Tinetti, M. E., Guralnik, J. M., Albert, M., et al.
(1994). Predicting changes in physical performance in a high-functioning elderly
cohort: MacArthur studies of successful aging. Journal of Gerontology, 49(3), M97-
M108.
106
Seeman, T. E., Merkin, S. S., Crimmins, E. M., & Karlamangla, A. S. (2010). Disability trends
among older Americans: National health and nutrition examination surveys, 1988–
1994 and 1999–2004. Journal Information, 100(1)
Shumway-Cook, A., Ciol, M. A., Hoffman, J., Dudgeon, B. J., Yorkston, K., & Chan, L. (2009).
Falls in the Medicare population: Incidence, associated factors, and impact on health
care. Physical Therapy, 89(4), 324-332.
Steffen, T. M., Hacker, T. A., & Mollinger, L. (2002). Age-and gender-related test
performance in community-dwelling elderly people: Six-minute walk test, berg balance
scale, timed up & go test, and gait speeds. Physical Therapy, 82(2), 128-137.
Streib, G. F. (1987). Old age in sociocultural context: China and the United States. Journal of
Aging Studies, 1(2), 95-112.
Tromp, A., Pluijm, S., Smit, J., Deeg, D., Bouter, L., & Lips, P. (2001). Fall-risk screening test: A
prospective study on predictors for falls in community-dwelling elderly. Journal of
Clinical Epidemiology, 54(8), 837-844.
United Nations (UN). (2002). World population ageing: 1950-2050. United Nations
Population Division.
Vasunilashorn, S., Crimmins, E. M., Kim, J. K., Winking, J., Gurven, M., Kaplan, H., et al.
(2010). Blood lipids, infection, and inflammatory markers in the Tsimane of Bolivia.
American Journal of Human Biology, 22(6), 731-740.
Verbrugge, L. M., & Jette, A. M. (1994). The disablement process. Social Science & Medicine,
38(1), 1-14.
Verbrugge, L. M., Wingard, D. L. (1987). Sex differentials in health and mortality. Women &
Health, 12(2), 103-145.
107
Wang, D., Zheng, J., Kurosawa, M., & Inaba, Y. (2009). Relationships between age and gender
differentials in health among older people in China. Ageing and Society, 29(7), 1141.
Wen, M., & Gu, D. (2011). The effects of childhood, adult, and community socioeconomic
conditions on health and mortality among older adults in China. Demography, 48(1),
153-181.
Wilmoth, J. R. (2000). Demography of longevity: Past, present, and future trends.
Experimental Gerontology, 35(9), 1111-1129.
Wong, R., Ofstedal, M. B., Yount, K., & Agree, E. M. (2008). Unhealthy lifestyles among older
adults: Exploring transitions in Mexico and the US. European Journal of Ageing, 5(4),
311-326.
Wong, R., Gerst, K., Michaels-Obregon, A., & Palloni, A. (2010). Burden of aging in
developing countries: Disability transitions in Mexico compared to the United States.
RAND Working Paper.
World Bank. (2014). World Development Indicators Database. Retrieved from
http://data.worldbank.org/data-catalog/world-development-indicators
World Health Organization (WHO). (2004). The Global Burden of Disease: 2004 Update.
Department of Health Statistics and Informatics.
World Health Organization (WHO). (2013). Global Health Observatory Data Repository.
Retrieved from http://apps.who.int/gho/data/node.main?lang=en.
Wray, L. A., & Blaum, C. S. (2001). Explaining the role of sex on disability. The Gerontologist,
41(4), 499-510.
Yount, K. M., & Agree, E. M. (2005). Differences in disability among older women and men
in Egypt and Tunisia. Demography, 42(1), 169-187.
108
Yount, K. M., Hoddinott, J., & Stein, A. D. (2010). Disability and self-rated health among older
women and men in rural Guatemala: The role of obesity and chronic conditions. Social
Science & Medicine, 71(8), 1418-1427.
Yu, M., & Sarri, R. (1997). Women's health status and gender inequality in China. Social
Science & Medicine, 45(12), 1885-1898.
Zhang, Y., & Goza, F. W. (2006). Who will care for the elderly in china?: A review of the
problems caused by China's one-child policy and their potential solutions. Journal of
Aging Studies, 20(2), 151-164.
Zimmer, Z. (2005). Health and living arrangement transitions among China’s oldest-old.
Research on Aging, 27(5), 526-555.
Zimmer, Z., Linda, G. M., & Chang, M. C. (2002). Changes in functional limitation and survival
among older Taiwanese, 1993, 1996, and 1999. Population Studies, 56(3), 265-276.
Zimmer, Z., & Kwong, J. (2003). Family size and support of older adults in urban and rural
China: Current effects and future implications. Demography, 40(1), 23-44.

109
APPENDIX A: Question wording for self-reported measures and grip strength equipment/protocol by dataset

Table A1. Wording of ADL items
HRS SEBAS KLoSA MHAS CHARLS IFLS Tsimane
Intro Here are a few
more everyday
activities. Please
tell me if you have
any difficulty with
these because of a
physical, mental,
emotional, or
memory problem.
Again exclude any
difficulties you
expect to last less
than 3 months.
Next I will
mention some
common daily
activities. Please
tell me if you
have trouble
doing these
activities by
yourself? If YES,
continue asking:
Do you have
some difficulty,
great difficulty,
or are you
unable to do
them at all?
We need to
understand
difficulties people
may have with
various activities
because of a health
or physical problem
and help from others
they need in doing
those activities.
Please tell me
whether you have
any difficulty doing
each of the everyday
activities that I read
to you during the last
week. Exclude any
difficulties you expect
to last less than 3
months
Please tell me if
you have any
difficulty with each
of the activities I
mention. If you do
not do any of the
following activities,
simply tell me. Do
not include
difficulties that you
believe will last
less than 3
months.
We need to
understand
difficulties people
may have with
various activities
because of a health
or physical problem.
Please tell me
whether you have
difficulty performing
any of the following
tasks on a regular
basis. Please tell me
if you have any
difficulties with
these because of a
physical, mental,
emotional or
memory problem.
Exclude any
difficulties that you
expect to last less
than 3 months.
Now we
would like
to know
your
physical
ability in
daily
activity. If
you had
[…], Could
you do it?

Response 1=Yes; 5=No;
6=Can't do;
7=Don't do; 8=DK;
9=Refused
0=No difficulty;
1=Some
difficulty;
2=Great
difficulty;
3=Unable to do
it
1=No I don't need
any help; 3=Yes, I
need help to some
extent; 5=Yes, I
need help in every
respect
1=Yes; 2=No;
6=Can't do;
7=Doesn't do
1=No, I don't have
any difficulty;
2=Yes, I have
difficulty; 3=I
cannot do it
1=Easily;
3=With
difficulty;
5=Unable
to do it
0=Without
difficulty;
1=With slight
difficulty;
2=With
moderate
difficulty;
3=With a lot of
difficulty;
4=Can't do it
alone
110
Dressing Because of a health
or memory
problem do you
have any difficulty
with dressing,
including putting
on shoes and
socks?
Any difficulty
dressing/undres
sing?
Because of health
and memory
problems, do you
have any difficulty
with dressing?
Dressing includes
taking clothes out
from a closet, putting
them on, buttoning
up, and fastening the
belt.
Because of a health
problem, do you
have any difficulty
with dressing
including putting
on shoes and
socks?
Because of health
and memory
problems, do you
have any difficulty
with dressing?
Dressing includes
taking clothes out
from a closet,
putting them on,
buttoning up, and
fastening a belt.
To dress
without
help.

Bathing Because of a health
or memory
problem do you
have any difficulty
with bathing or
showering?
Any difficulty
bathing?
Do you have any
difficulty with bathing
or showering?
Because of a health
problem, do you
have any difficulty
with bathing or
showering?
Because of health
and memory
problems, do you
have any difficulty
with bathing or
showering?
To bathe.
Toilet Because of a health
or memory
problem do you
have any difficulty
with using the
toilet, including
getting up and
down?
Any difficulty
going to the
toilet?
Do you have any
difficulties with using
the toilet, getting up
and down?
Because of a health
problem, do you
have any difficulty
with using the
toilet, including
getting on and off
the toilet or
squatting?
Because of health
and memory
problems, do you
have difficulty with
using the toilet,
including getting up
and down?
To go to the
bathroom
(BM)
without
help.

Eating Because of a health
or memory
problem do you
have any difficulty
with eating, such
as cutting up your
food?
Any difficulty
eating?
Do you have any
difficulty with eating,
such as cutting up
your food? By eating,
we mean eating food
by oneself when it is
ready.
Because of a health
problem, do you
have any difficulty
with eating, such
as cutting your
food?
Because of health or
memory problems,
do you have any
difficulty with
eating, such as
cutting up your
food? By eating, we
mean eating food by
oneself when it is
ready.

Mobility Because of a health
or memory
problem do you
have any difficulty
with walking across
a room?
Any difficulty
moving around
the house?
Because of a health
problem, do you
have any difficulty
with walking across
a room?
To walk
across the
room.

111
Table A2. Wording of functional task items
HRS SEBAS MHAS CHARLS IFLS
Intro We need to
understand
difficulties people
may have with
various activities
because of a health
or physical problem.
Please tell me
whether you have
any difficulty doing
each of the everyday
activities that I read
to you. Exclude any
difficulties that you
expect to last less
than 3 months.
If no one helps
you and you do
not use aids,
would you have
difficulty doing
the following
activities by
yourself? If YES,
continue asking:
Would you have
some difficulty,
great difficulty,
or would you be
unable to do
them at all? If a
respondent has
never done a
certain activity,
ask: If you had
to do it, could
you?
Please tell me if
you have any
difficulty in doing
each of the daily
activities that I
am going to read.
Don't include
difficulties that
you believe will
last less than 3
months
We need to
understand
difficulties people
may have with
various activities
because of a health
or physical
problem. Please
tell me whether
you have difficulty
performing any of
the following tasks
on a regular basis.
Please tell me if
you have any
difficulties with
these because of a
physical, mental,
emotional or
memory problem.
Exclude any
difficulties that you
expect to last less
than 3 months.
Now we
would like to
know your
physical
ability in
daily
activities. If
you had […],
could you do
it?
Response 1=Yes; 5=No;
6=Can't do; 7=Don't
do; 8=Don't know;
9=Refused
0=No difficulty;
1=Some
difficulty;
2=Great
difficulty;
3=Unable to do
it; 7=Don't
know
1=yes, 2=no,
6=can't do,
7=doesn't do,
8=refused,
9=don't know
1=No, I don't have
any difficulty,
2=Yes, I have
difficulty, 3=I
cannot do it
1=Easily,
3=With
difficulty,
5=Unable to
do it
Squat Because of a health
problem do you
have any difficulty
with stooping,
kneeling, or
crouching?
Any difficulty
squatting?
Because of a
health problem,
do you have
difficulty with
stooping,
kneeling, or
crouching?
Do you have
difficulty stooping,
kneeling, or
crouching?
To bow,
squat, kneel
Stairs Because of a health
problem do you
have any difficulty
with climbing several
flights of stairs
without resting?
Any difficulty
climbing 2-3
flights of stairs?
Because of a
health problem,
do you have
difficulty with
climbing several
flights of stairs
without rest?
Do you have
difficulty climbing
several flights of
stairs without
resting?

Carry Because of a health
problem do you
have any difficulty
with lifting or
carrying weights
over 10 pounds, like
a heavy bag of
groceries?
Any difficulty
lifting or
carrying
something
weighing 11-12
kgs (like 2 bags
of rice)?
Because of a
health problem,
do you have
difficulty with
lifting or carrying
objects weighing
over 5kg, like a
heavy bag of
groceries?
Do you have
difficulty lifting or
carrying weights of
10 jin, like a heavy
bag of groceries?
To carry a
heavy load
(like a pail
of water) for
20 meters

112
Table A3. Grip strength protocol & equipment
HRS SEBAS KLoSA CHARLS IFLS
Dynamometer Smedley
Spring-Type
Hand
Dynamometer
North Coast
Hydraulic
Hand
Dynamometer
(NC70142)
Tanata
6103
Model
Yuejian WL-
1000
Dynamometer
Not
specified
Times 2x 3x 2x 2x 3x
Position Mostly
standing
Mostly seated Mostly
seated
Mostly
standing
Mostly
standing

113
APPENDIX B: Characteristics of included versus missing respondents

Table B1. Characteristics of missing versus non-missing of age-eligible sample 55-85 years
Datasets & N Age (avg yrs) Sex (% female)
Measures Inc Miss Inc Miss t-test Inc Miss X2
ADLs

HRS 14125 913 68.5 70.9 -8.7 (p<.001) 59% 37% 167.4 (p<.001)
SEBAS 1051 41 66.0 74.0 -5.9 (p<.001) 48% 46% 0.0 (p=.877)
KLoSA 6532 215 66.8 73.6 -13.1 (p<.001) 56% 55% 0.0 (p=837)
MHAS 8846 876 65.0 66.3 -4.5 (p<.001) 54% 49% 8.4 (p=.004)
CHARLS 10227 836 64.5 65.1 -2.5 (p=.014) 51% 43% 18.0 (p<.001)
IFLS 4196 24 64.7 65.5 -.5 (p=.599) 53% 42% 1.3 (p=.260)
Tsimane 449 147 65.2 63.7 1.3 (p=.208) 47% 56% 1.6 (p=.211)
Func Tasks

HRS 14146 892 68.5 70.9 -8.5 (p<.001) 59% 36% 167.4 (p<.001)
SEBAS 1049 43 66.0 73.7 -5.7 (p<.001) 47% 49% 0.0 (p=.861)
MHAS 7769 1953 64.8 66.4 -7.8 (p<.001) 53% 54% 0.0 (p=.927)
CHARLS 10150 913 64.5 65.2 -2.8 (p=.005) 50% 45% 9.4 (p=.002)
IFLS 4196 24 64.7 65.5 -.5 (p=.598) 53% 42% 1.3 (p=.261)
Grip Strength

HRS 6029 417 68.4 69.4 -2.3 (p=.021) 57% 71% 33.1 (p<.001)
SEBAS 1000 92 65.9 70.9 -5.3 (p<.001) 47% 57% 3.3 (p=.071)
KLoSA 6001 745 66.5 70.7 -14.3 (p<.001) 55% 66% 32.2 (p<.001)
CHARLS 8522 2541 64.4 65.0 -3.4 (p<.001) 50% 49% 0.8 (p=.357)
IFLS 4108 112 64.6 70.1 -7.9 (p<.001) 52% 77% 25.9 (p<.001)
Gait (65-85)

HRS 3655 627 72.9 73.8 -3.5 (p<.001) 55% 61% 8.0 (p=.005)
SEBAS 527 35 73.5 76.9 -3.7 (p<.001) 43% 43% 0.0 (p=.967)
CHARLS 3304 1320 71.4 73.2 -11.1 (p<.001) 48% 52% 4.9 (p=.026)
Tsimane 182 29 72.0 73.5 -1.5 (.132) 45% 55% 1.1 (p=.284)
Chair Stands

SEBAS 931 161 65.4 71.5 -8.4 (p<.001) 45% 61% 13.5 (p<.001)
CHARLS 7966 3097 63.9 66.1 -13.6 (p<.001) 50% 51% 1.8 (p=.180)
IFLS 3940 280 64.3 70.5 -13.9 (p<.001) 52% 66% 21.4 (p<.001)
Tsimane 390 59 64.6 66.8 -2.4 (p=.016) 46% 53% 1.7 (p=.192)
Tandem

HRS 6302 144 68.5 67.8 1.1 (p=.270) 58% 56% 0.1 (p=.757)
CHARLS 8554 2509 64.3 65.3 -6.1 (p<.001) 50% 48% 3.4 (p=.067)
Tsimane 422 27 65.0 65.3 -0.3 (p=.795) 47% 52% 0.7 (p=.417)
One Leg

MHAS 1533 131 64.9 66.9 -2.8 (p=.005) 54% 53% 0.0 (p=.846)
Tsimane 424 25 65.1 65.0 0.0 (p=.957) 47% 52% 0.7 (p=.418)

114
APPENDIX C: OLS models of grips strength controlling for height and weight versus BMI

Table C1. OLS regression models of grip strength controlling for BMI versus height and weight
U.S. Taiwan Korea China Indonesia
B B B B B
Female -15.93 *** -14.49 *** -12.29 *** -11.11 *** -10.30 ***
Age 65-74 -3.76 *** -4.18 *** -3.43 *** -3.54 *** -3.18 ***
Age 75-85 -8.26 *** -7.12 *** -6.24 *** -7.00 *** -6.98 ***
BMI 0.05 * 0.17 * 0.02

0.27 *** 0.32 ***

Female -11.50 *** -10.49 *** -10.54 *** -8.07 *** -7.48 ***
Age 65-74 -3.46 *** -3.60 *** -3.12 *** -2.98 *** -2.63 ***
Age 75-85 -7.34 *** -6.10 *** -5.51 *** -5.83 *** -5.90 ***
Height (m) 28.24 *** 28.03 *** 7.93 *** 20.50 *** 16.09 ***
Weight (kg) 0.03 *** 0.08 ** 0.11 *** 0.08 *** 0.14 ***
*p<.05 **p<.01 ***p<.001

115
APPENDIX D: Comparison of OLS and Heckman selection models of sex differences for continuous physical performance
measures

Table D1. Regression estimates of the effect of being female: Physical performance measures

U.S. Taiwan Korea Mexico China Indonesia Tsimane
B B B B B B B
Grip Strength
SVY: OLS Regression -15.9 *** -14.5 *** -12.3 *** ---

-11.1 *** -10.3 *** ---

[-16.4,-15.5] [-15.6,-13.4] [-12.5,-11.9]

[-11.6,-10.7] [-10.9,-9.7]

SVY: Heckman -16.1 *** -14.3 *** -12.1 *** ---

-11.2 *** -10.6 *** ---

[-16.6,-15.6] [-15.4,-13.2] [-12.4,-11.8]

[-11.6,-10.7] [-11.2,-10.0]

Gait Speed
SVY: OLS Regression -0.07 *** -0.14 *** ---

---

-0.04 *** ---

-0.03 **

[-.09, -.06] [-.18, -.09]

[-.06, -.02]

[-.05,-.01]
SVY: Heckman -0.08 *** -0.13 *** ---

---

-0.05 *** ---

-0.04 ***

[-.10, -.07] [-.17, -.09]

[-.07, -.03]

[-.06, -.02]
Chair Stands
SVY: OLS Regression ---

1.30 *** ---

---

0.87 *** 1.44 *** 1.36 ***

[.74, 1.86]

[.67, 1.07] [1.21, 1.67] [.74, 1.99]
SVY: Heckman ---

0.66 * ---

---

0.88 *** 1.59 *** 1.26 ***
[.01, 1.32] [.68, 1.08] [1.25, 1.92] [.61, 1.90]
*p<.05 **p<.01 ***p<.001

116
APPENDIX E: Comparison of models of sex differences with data pooled across countries (including a sex*country interaction
term) versus models stratified by country (with and without Stata’s SVY command)

Table E1. Regression coefficients of the effect of being female, controlling for age: Physical performance measures

U.S. Taiwan Korea Mexico China Indonesia Tsimane
B B B B B B B
Grip Strength
Pooled -16.0 *** -14.4 *** -12.2 *** ---

-10.9 *** -9.9 *** ---

[-16.4,-15.5] [-15.4,-13.5] [-12.5,-11.9]

[-11.4,-10.5] [-10.5,-9.4]

Unpooled -15.9 *** -14.5 *** -12.3 *** ---

-11.1 *** -10.2 *** ---

[-16.4,-15.5] [-15.4,-13.5] [-12.6, -12.0]

[-11.6,-10.7] [-10.8,-9.7]

Unpooled (SVY) -15.9 *** -14.5 *** -12.3 *** ---

-11.1 *** -10.3 *** ---

[-16.4,-15.5] [-15.6, -13.4] [-12.5, -11.9]

[-11.6, -10.7] [-10.9,-9.7]

Gait Speed
Pooled -0.07 *** -0.14 *** ---

---

-0.04 *** ---

-0.03 *

[-.09,-.06] [-.17,-.10]

[-.06,-.02]

[-.05,-.01]
Unpooled -0.07 *** -0.14 *** ---

---

-0.04 *** ---

-0.03 **

[-.09,-.06] [-.18,-.09]

[-.06,-.02]

[-.05,-.01]
Unpooled (SVY) -0.07 *** -0.14 *** ---

---

-0.04 *** ---

[-.09, -.06] [-.18, -.09]

[-.06, -.02]

Chair Stands
Pooled ---

1.19 *** ---

---

0.88 *** 1.47 *** 1.37 ***

[.73, 1.65]

[.67, 1.10] [1.27, 1.66] [.75, 1.99]
Unpooled ---

1.30 *** ---

---

0.87 *** 1.46 *** 1.36 ***

[.84, 1.76]

[.66, 1.09] [1.27, 1.66] [.74, 1.99]
Unpooled (SVY) ---

1.30 *** ---

---

0.87 *** 1.44 *** 1.36 ***
[.74, 1.86] [.67, 1.07] [1.21, 1.67] [.74, 1.99]
*p<.05 **p<.01 ***p<.001

117

Table E2. Odds Ratios of the effect of being female, controlling for age: Physical performance measures

U.S. Taiwan Korea Mexico China Indonesia Tsimane

OR

OR

OR

OR

OR

OR

OR
Tandem Stand
Pooled 1.59 *** ---

---

---

2.03 *** ---

2.36 *

[1.37, 1.85]

[1.68, 2.45]

[1.20, 4.64]
Unpooled 1.59 *** ---

---

---

2.01 *** ---

2.45 **

[1.37, 1.86]

[1.67, 2.41]

[1.25, 4.81]
Unpooled (SVY) 1.59 *** ---

---

---

2.01 *** ---

2.45 **

[1.37, 1.86]

[1.70, 2.37]

[1.25, 4.81]
One-Leg Stand
Pooled ---

---

---

1.16

---

---

2.39 ***

[.80, 1.68]

[1.56, 3.68]
Unpooled ---

---

---

1.14

---

---

2.59 ***

[.79, 1.65]

[1.62, 4.14]
Unpooled (SVY) ---

---

---

1.14

---

---

2.59 ***
[.79, 1.65] [1.62, 4.13]
*p<.05 **p<.01 ***p<.001

118

Table E3. Odds ratios of the effect of being female, controlling for age: Self-reported functioning measures

U.S. Taiwan Korea Mexico China Indonesia Tsimane
OR OR OR OR OR OR OR
Squatting
Pooled 1.61 *** 1.78 *** ---

1.83 *** 1.24 *** 1.36 * ---

[1.48, 1.75] [1.34, 2.36]

[1.54, 2.17] [1.10, 1.40] [1.06, 1.76]

Unpooled 1.61 *** 2.04 *** ---

1.84 *** 1.22 ** 1.35 * ---

[1.48, 1.75] [1.49, 2.79]

[1.55, 2.20] [1.09, 1.37] [1.04, 1.76]

Unpooled (SVY) 1.61 *** 2.04 ** ---

1.84 *** 1.22 *** 1.37 * ---

[1.48, 1.75] [1.42,2.93]

[1.55, 2.20] [1.10, 1.36] [1.05, 1.79]

Climbing stairs
Pooled 1.96 *** 2.20 *** ---

1.90 *** 1.33 *** ---

---

[1.80, 2.13] [1.60, 3.02]

[1.61, 2.25] [1.18, 1.51]

Unpooled 1.96 *** 2.63 *** ---

1.97 *** 1.33 *** ---

---

[1.80, 2.13] [1.85, 3.73]

[1.66, 2.34] [1.17, 1.50]

Unpooled (SVY) 1.96 *** 2.63 *** ---

1.97 *** 1.33 *** ---

---

[1.80, 2.13] [1.93, 3.57]

[1.66, 2.34] [1.17, 1.50]

Carrying
Pooled 2.67 *** 4.55 *** ---

2.60 *** 2.25 *** 2.58 *** ---

[2.38, 3.00] [3.07, 6.75]

[2.04, 3.32] [1.93, 2.62] [2.21, 3.01]

Unpooled 2.67 *** 5.13 *** ---

2.62 *** 2.28 *** 2.68 *** ---

[2.38, 3.00] [3.37, 7.83]

[2.03, 3.38] [1.95, 2.66] [2.28, 3.16]

Unpooled (SVY) 2.66 *** 5.13 *** ---

2.62 *** 2.28 *** 2.66 *** ---
[2.37, 2.97] [3.09, 8.54] [2.03, 3.38] [1.97, 2.63] [2.23, 3.16]
*p<.05 **p<.01 ***p<.001

119
Table E4. Odds ratios of the effect of being female, controlling for age: Self-reported ADL measures
Dressing
Pooled 1.53 *** 1.97 * 0.82

1.41 * 1.03

2.04 *** 2.33 **

[1.25, 1.87] [1.01, 3.86] [.56, 1.20] [1.06,1.87] [.84, 1.26] [1.62, 2.56] [1.37, 3.99]

Unpooled 1.53 *** 2.16 * 0.74

1.43 * 1.04

2.05 *** 2.42 **

[1.25, 1.87] [1.10, 4.25] [.51, 1.08] [1.07,1.90] [.85, 1.26] [1.62, 2.58] [1.40, 4.18]

Unpooled 1.58 *** 2.16 * 0.74

1.43 * 1.04

2.00 *** 2.42 **
(SVY) [1.36, 1.84] [1.12,4.19] [.52, 1.06] [1.07,1.90] [.85, 1.26] [1.58, 2.53] [1.40, 4.19]

Bathing
Pooled 1.31 ** 1.96 * 1.07

1.27

1.09

2.00 *** 2.05 **

[1.08, 1.60] [1.00, 3.90] [.76, 1.51] [.86, 1.86] [.91, 1.31] [1.46, 2.75] [1.19, 3.53]

Unpooled 1.33 ** 2.15 * 0.96

1.29

1.09

1.99 *** 2.09 **

[1.09, 1.62] [1.07, 4.32] [.69, 1.35] [.88, 1.90] [.91, 1.21] [1.44, 2.74] [1.20, 3.64]

Unpooled 1.33 ** 2.15 *** 0.96

1.29

1.09

1.96 *** 2.09 **
(SVY) [1.10, 1.62] [1.67, 2.78] [.68, 1.37] [.88, 1.90] [.92, 1.30] [1.42, 2.70] [1.20, 3.64]

Toilet
Pooled 2.08 *** 2.09 * 0.56 * 1.47

1.22 ** 1.32

2.08

[1.68, 2.59] [1.01, 4.30] [.33, .98] [.96, 2.25] [1.06, 1.40] [.80, 2.21] [.97, 4.48]

Unpooled 2.09 *** 2.29 * 0.53 * 1.50

1.22 ** 1.30

2.15

[1.69, 2.59] [1.11, 4.72] [.31, .93] [.98, 2.30] [1.06, 1.40] [.78, 2.18] [.99, 4.68]

Unpooled 2.09 *** 2.29 ** 0.53 * 1.50

1.22 ** 1.26

2.15

(SVY) [1.64, 2.66] [1.30, 4.02] [.31, .92] [.98, 2.30] [1.06, 1.40] [.74, 2.16] [.99, 4.67]
*p<.05 **p<.01 ***p<.001

120
APPENDIX F: Tsimane longitudinal data

Table F1. Number of time points by measure (Tsimane)
Time Points Tandem Tandem Gait (>65)
1x 322 231 161
2x 104 224 118
3x 112 171 66
4x 47 69 23
5x 0 43 16
6x 0 7 1
Total 585 745 385

121
APPENDIX G: Detailed descriptive data on change by age group

Table G1. Level and change in measures by age group
Tsimane U.S.
ADLs
Age 50-59
Baseline (# difficulties) 0.37 0.16
Avg Change (# difficulties) -0.30 -0.02
% Increased 2.4% 3.1%
Age 60-69
Baseline (# difficulties) 0.48 0.15
Avg Change (# difficulties) -0.07 0.03
% Increased 9.4% 3.8%
Age 70+
Baseline (# difficulties) 0.76 0.21
Avg Change (# difficulties) 0.13 0.26
% Increased 24.1% 11.3%
Balance (Tandem) Tsimane U.S.
Age 50-59
Baseline (sec) 9.91 9.92
Avg Change (sec) -0.24 -0.66
% Declined 4.2% 9.1%
Age 60-69
Baseline (sec) 9.95 9.79
Avg Change (sec) -0.43 -0.83
% Declined 6.7% 12.8%
Age 70+
Baseline (sec) 9.82 9.43
Avg Change (sec) -2.68 -2.31
% Declined 32.9% 34.9%
Gait Speed Tsimane U.S.
Age 65-75
Baseline (m/s) 0.35 0.86
Avg Change (m/s) -0.01 -0.06
% Declined 28.4% 53.0%
Age 75+
Baseline (m/s) 0.33 0.75
Avg Change (m/s) -0.04 -0.15
% Declined 38.9% 62.0%

122
APPENDIX H: Alternative definition of gait speed decline

Table H1. Comparison of definitions of gait speed decline
Tsimane U.S.
>.05m/s decline
% Declined 29.0% 53.9%
>.5 SD decline >.170 m/s >.405 m/s
% Declined 10.5% 12.3%

Table H2. Hazard ratios from parametric survival models of gait speed decline
Tsimane (N=209) U.S. (N=2,659)
HR p L95 U95 HR p L95 U95
Definition 1
Female 1.29 0.308 0.79 2.10

0.99 0.793 0.88 1.10
Age (65-74)
75+ 1.34 0.269 0.80 2.56

1.31 <.001 1.17 1.47
Definition 2 HR p L95 U95

HR p L95 U95
Female 1.12 0.788 0.48 2.60

0.85 0.172 0.67 1.07
Age (65-74)
75+ 2.29 0.054 0.99 5.31

1.72 <.001 1.36 2.18
1. Gait speed decline defined as .05m/s slower gait speed at time 2
2. Gait speed decline defined as >.5 SD slower gait speed at time 2

123
APPENDIX I: Comparison of different time intervals: Tsimane

Tandem Stand

Time interval (US)

(A) Original time variable (between longest 2 time points)

(B) Time interval closest to 60 months
124

(C) Time interval closest to 50 months

(D) Time interval closest to 40 months

(E) Time interval 40-60 months

Gait Speed

125
Time interval (US)

(A) Original time variable (between longest 2 time points)

(B) Time interval closest to 60 months

(C) Time interval closest to 50 months

(D) Time interval closest to 40 months
126

(E) Time interval 40-60 months

Table I1. Different specifications of time
N Mean (SD) Min Max % Worse
Tsimane
A. Original Time Variable
Tandem 477 38.4 (19.1) 5.7 80.6 9.9
Gait (>65) 209 34.1 (19.3) 9.9 79.3 31.1
B. Time Closest To 60 mo
Tandem 477 37.4 (17.5) 5.7 78.9 8.8
Gait (>65) 209 33.2 (17.7) 9.9 72.9 30.6
C. Time Closest To 50 mo
Tandem 477 34.7 (14.6) 5.7 78.9 8.2
Gait (>65) 209 31.2 (15.0) 9.9 72.9 28.7
D. Time Closest To 40 mo
Tandem 477 31.8 (12.1) 5.7 78.9 6.9
Gait (>65) 209 28.3 (12.1) 9.9 64.4 26.8
E. Time 40-60 mo
Tandem 162 49.4 (5.3) 40.9 59.7 8.0
Gait (>65) 57 49.9 (5.4) 40.3 59.7 19.3
U.S.
Tandem 4474 51.0 (4.1) 38.5 61.9 16.5
Gait (>65) 2659 51.8 (3.8) 40.5 61.9 56.0

127
Table I2. Survival analysis results with different time specifications

Time A Time B Time C Time D Time E
HR p HR p HR p HR p HR p
Tandem
Female 1.62 0.104 1.68 0.100 1.66 0.120 1.90 0.078 1.80 0.312
Age (50-59)
60-69 1.66 0.262 1.61 0.319 1.34 0.552 2.34 0.128 0.59 0.515
70+ 9.22 <.001 9.10 <.001 7.83 <.001 12.56 <.001 2.47 0.152
Gait Speed
Female 1.29 0.308 1.34 0.240 1.20 0.475 1.19 0.505 0.47 0.282
Age (65-74)
75+ 1.34 0.269 1.32 0.297 1.37 0.255 1.48 0.173 2.32 0.241
128
APPENDIX J: Comparison of ADL change: 2006-2008 versus 2006-2010 (United States)

Time interval: 2006 to 2008

Time interval: 2006 to 2010

Table J1. Baseline characteristics & average change (50-95years)

U.S. 2006-
2010
U.S. 2006-
2008
U.S. 2006-
2008*
ADL Disability
Female 55.4% 54.9% 55.3%
Age
50-59 41.0% 40.0% 40.7%
60-69 32.3% 31.1% 31.9%
70+ 26.7% 29.3% 27.3%
Baseline (# difficulties) 0.17 0.19 0.17
Avg Change (# difficulties) 0.07 -0.01 -0.03
% Increased 5.5% 3.2% 2.4%
Months (avg, SD, min, max) 51.2 (4.3) 24.0 (3.2) 23.9 (3.2)

[37.5, 62.9] [13.1, 35.6] [13.1, 35.6]
N 11,953 13,658 12,005
*Excluding those missing in 2010

129
Table J2. Hazard ratios from parametric survival models of ADL increase
U.S. 2006-10 (N=11,953) U.S. 2006-08 (N=13,658) U.S. 2006-08 (N=12,005)*
ADL Increase HR p L95 U95 HR p L95 U95 HR p L95 U95
Female 1.34 0.001 1.13 1.60

1.31 0.003 1.10 1.55

1.36 0.035 1.02 1.83
Age (50-59)
60-69 0.78 0.081 0.58 1.03

1.16 0.31 0.87 1.56

1.02 0.921 0.69 1.51
70+ 1.77 <.001 1.40 2.24

2.87 <.001 2.22 3.72

1.78 <.001 1.27 2.49
*Excluding those missing in 2010

130
APPENDIX K: Comparison of ADL and gait speed change: 2004-2006 versus 2004-2008
(United States)

Time interval (ADLs): 2004-2006 (left), 2004-2008 (right)

Time interval (gait speed): 2004-2006 (left), 2004-2008 (right)

131
Table K1. Baseline characteristics (2004) and average change
ADL Disability (50-95yrs)
All (N=14,028)
Female 55.4%
Age
50-59 44.1%
60-69 29.9%
70+ 26.0%
Difficulties (#)
2004 0.15
2006 0.13
2008 0.18
Change in difficulties (#), 2004-06 -0.02
Change in difficulties (#), 2004-08 0.03
% Increased, 2004-06 2.3%
% Increased, 2004-08 3.8%
Months, 2004-6 (avg, SD, min, max) 24.3(3.2)

[14.1, 35.6]
Months, 2004-8 (avg, SD, min, max) 48.3(3.2)

[38.5, 59.9]

Gait Speed (65-95yrs) 2004-2006 2004-2008
Female 54.8% 59.2%
Age
65-75 39.1% 43.3%
75+ 60.9% 56.7%
Gait speed (m/s)
2004 0.83 0.84
2006 0.70
2008

0.63
Change in gait speed (m/s) -0.13 -0.22
% Declined 60% 71%
Months (avg, SD, min, max) 25.2(2.7) 49.4(2.6)

[16.2, 33.5] [40.5, 59.9]
N 696 635

132
Table K2. Hazard ratios from parametric survival models (Weibull) of ADL increase and gait
speed decline
ADL Increase
(N=14,028)
2004-2006

2004-2008
HR p L95 U95

HR p L95 U95
Female 1.51 0.002 1.17 1.95

1.29 0.007 1.07 1.55
Age (50-59)
60-69 1.07 0.694 0.76 1.51

0.93 0.653 0.71 1.24
70+ 2.34 <.001 1.75 3.15

3.08 <.001 2.42 3.92
Gait Speed Decline* 2004-2006 (N=696)

2004-2008 (N=635)
Female 1.18 0.160 0.94 1.48

1.03 0.828 0.80 1.32
Age (65-74)
75+ 1.07 0.597 0.83 1.38 1.44 0.005 1.12 1.84
*Different individuals

133
APPENDIX L: Random subset of U.S. data to obtain sample sizes equal to the Tsimane data

Table L1. Baseline characteristics & average change (50-95years)

Tsimane U.S. U.S.
(Resampled)
ADL Disability
Female 48.3% 55.4% 53.1%
Age
50-59 47.2% 41.0% 44.7%
60-69 32.3% 32.3% 30.3%
70+ 20.5% 26.7% 24.9%
Baseline (# difficulties) .49 (1.22) 0.17 0.21
Avg Change (# difficulties) -.14 (1.25) 0.07 0.10
% Increased 9.1% 5.5% 7.9%
Months (avg, SD, min, max) 30.7 (10.6) 51.2 51.1

[10.1, 45.9] [37.5, 62.9] [40.5, 60.9]
N 263 11,953 263
Balance (Tandem)
Female 47.4% 52.9% 52.8%
Age
50-59 54.5% 44.3% 41.4%
60-69 28.3% 31.6% 36.2%
70+ 17.2% 24.1% 22.4%
Baseline (sec) 9.91 (.73) 9.76 9.80
Avg Change (sec) -.72 (2.74) -1.11 -0.93
% Declined 9.9% 16.5% 13.7%
Months 38.4 (19.1) 51.0 (4.1) 50.7 (4.0)

[5.7, 80.6] [38.5, 61.9] [41.6, 58.9]
N 477 4,474 477
Gait Speed
Female 46.9% 56.9% 61.7%
Age
65-75 74.2% 60.1% 52.9%
75+ 25.8% 39.9% 47.1%
Baseline (m/s) .34 (.07) 0.82 0.79
Avg Change (m/s) -.02 (.11) -0.10 -0.08
% Declined 31.1% 56.0% 55.4%
Months 34.1 (19.3) 51.8 (3.8) 51.8 (3.9)

[9.9, 79.3] [40.5, 61.9] [40.5, 60.9]
N 209 2,659 209
134
Table L2. Hazard ratios from parametric survival models of ADL increase or balance/gait speed decline

Tsimane (N=263)

U.S. (N=11,953)

U.S. Resampled (N=263)
ADLs HR p L95 U95

HR p L95 U95

HR p L95 U95
Female 3.24 0.010 1.33 7.92

1.34 0.001 1.13 1.60

1.52 0.38 0.59 3.85
Age (50-59)

.
60-69 4.78 0.022 1.26 18.15

0.78 0.081 0.58 1.03

0.41 0.184 0.10 1.54
70+ 10.77 <.001 3.07 37.82

1.77 <.001 1.40 2.24

1.52 0.469 0.49 4.69

Tsimane (N=477)

U.S. (N=4,474)

U.S. Resampled (N=477)
Balance HR p L95 U95

HR p L95 U95

HR p L95 U95
Female 1.62 0.104 0.91 2.91

1.09 0.287 0.93 1.28

0.84 0.51 0.49 1.42
Age (50-59)
60-69 1.66 0.262 0.68 4.03

1.04 0.735 0.81 1.34

1.01 0.991 0.45 2.23
70+ 9.22 <.001 4.57 18.59

2.47 <.001 1.99 3.06

1.67 0.129 0.86 3.25

Tsimane (N=209)

U.S. (N=2,659)

U.S. Resampled (N=209)
Gait Speed HR p L95 U95

HR p L95 U95

HR p L95 U95
Female 1.29 0.308 0.79 2.10

0.99 0.793 0.88 1.10

1.24 0.283 0.84 1.84
Age (65-74)
75+ 1.34 0.269 0.80 2.56

1.31 <.001 1.17 1.47

1.23 0.306 0.83 1.83

Asset Metadata

Creator Wheaton, Felicia Victoria (author)

Core Title International sex and age differences in physical function and disability

Contributor Electronically uploaded by the author (provenance)

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Gerontology

Publication Date 08/04/2014

Defense Date 05/07/2014

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

Tag aging,Disability,Gerontology,international differences,OAI-PMH Harvest,physical function

Format application/pdf (imt)

Language English

Advisor Crimmins, Eileen M. (committee chair), Hagedorn, Aaron (committee member), Lee, Jinkook (committee member)

Creator Email feliciawheaton@gmail.com,fwheaton@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-451454

Unique identifier UC11287590

Identifier etd-WheatonFel-2756.pdf (filename),usctheses-c3-451454 (legacy record id)

Legacy Identifier etd-WheatonFel-2756-0.pdf

Dmrecord 451454

Document Type Dissertation

Format application/pdf (imt)

Rights Wheaton, Felicia Victoria

Type texts

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

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

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

Abstract (if available)

Abstract Worldwide population aging will undoubtedly be accompanied by an increase in the number of disabled older adults. Female gender and increased age are two of the most widely identified risk factors for poor physical functioning and disability. Yet the contexts in which people are aging vary markedly across countries. Countries differ greatly in their level of economic development, both past and present. Economic development is in turn related to improvements in infrastructure, health care, public health, education, etc. that are hypothesized to be related to improved physical function and less disability. ❧ Therefore, this dissertation examined whether sex and age differences/changes in both objectively‐measured physical performance and reported difficulties with functional tasks and activities of daily living (ADLs) were similar or varied across seven countries whose per capita GNP ranged from $200 to $40,100 (United Sates, Taiwan, Korea, Mexico, China, Indonesia, and the Tsimane of Bolivia). It also sought to determine if sex differences and age differences varied systematically in terms of macro‐level indicators including GDP, life expectancy, and measures of gender equality. ❧ Overall, sex differences were remarkably consistent across countries with very different contexts. Sex differences in physical performance and functional limitations were more pronounced than sex differences in difficulty with basic self‐care tasks, but the magnitude of differences did not vary systematically in relation to country‐level measures of development or gender equality. This may be because gender equality can be either protective or detrimental, depending on the domain. ❧ In terms of age differences, it was necessary to consider both the level of performance/prevalence of difficulty at younger ages as well as age differences, since poor performance/high levels of difficulty among the young‐old indicate that “aging” has already occurred. Some populations did appear to be “aging” more rapidly, particularly those at the lowest end of the development spectrum, however, there was no clear evidence for a linear correlation between macro‐level indicators of development and age differences. Interestingly, findings showed that functioning in some domains could be fairly well maintained despite declines in other domains, and these varied across countries. For example, Indonesians appeared to be “aging” more rapidly in terms of upper body strength, but showed relatively high levels of lower body function and less age‐related decline. This may be due to differences across populations in patterns of work, physical activity, the built environment, etc.