AGE-RELATED CHANGES IN WEIGHT AND BODY COMPOSITION:
IMPLICATIONS FOR HEALTH IN THE ELDERLY
by
WENDY CHRISTINA STEPHEN
A thesis submitted to the School of Kinesiology and Health Studies
In conformity with the requirements for
the degree of Master of Science
Queen’s University
Kingston, Ontario, Canada
May, 2008
Copyright © Wendy Christina Stephen, 2008
ABSTRACT
The objective of this thesis was to examine age-related changes in weight and body
composition as they relate to health in older adults. This thesis was completed in manuscript
format and consists of two studies, both of which are based on the Cardiovascular Health Study
(CHS) cohort. The CHS is a prospective cohort study of community-dwelling older (≥65 years)
men and women who were followed for 8 years.
The first manuscript examined whether physical activity (PA) attenuates age-related
weight loss in the elderly. Mixed modeling procedures were employed to create body weight
trajectory curves for the 8 year follow-up period according to physical activity level in a sample
of 4512 CHS participants. Body weight declined over the follow-up period in all physical activity
groups, with an accelerated weight loss occurring in the final years of follow-up. Over the 8 year
follow-up, body weight was reduced by 2.72 kg in the least active PA quartile. Compared to the
least active quartile, weight loss was attenuated by 0.55 kg (20%), 0.80 kg (29%), and 0.69 kg
(25%) within the second through fourth physical activity quartiles. Therefore, participation in
modest amounts of PA attenuated age-related weight loss by approximately 25%.
The second manuscript examined whether sarcopenic-obesity is a stronger predictor of
cardiovascular disease (CVD) than either sarcopenia or obesity alone, and whether muscle mass
or strength is a stronger marker of CVD risk. CHS participants who were free of CVD at baseline
(n=3400) were classified as normal, sarcopenic, obese, or sarcopenic-obese based on measures of
waist circumference and either muscle mass or strength. Participants were followed for CVD
development over 8 years and proportional hazard regression models were used to compare risk
estimates for CVD after adjustment for covariates. When based on measures of waist
circumference and muscle mass, CVD risk was not increased in sarcopenic, obese, or sarcopenicobese groups in comparison to the group with a normal body composition. When categorized
based on waist circumference and muscle strength, CVD risk was significantly increased (by
i
38%) in the sarcopenic-obese group but not in either the sarcopenic or obese groups. Thus,
sarcopenic-obesity, based on muscle strength, was associated with increased CVD risk implying
that strength is more important than muscle mass for cardiovascular health in old age.
In summary, the findings of this thesis support the continuation of public health efforts to
promote regular PA and balanced nutrition to assist with maintenance of optimal body
composition and weight through adulthood and into old age.
Key words: body weight, physical activity, sarcopenic-obesity, cardiovascular disease, body
composition
ii
CO-AUTHORSHIP
This thesis presents the work of Wendy Stephen in collaboration with her advisor, Dr. Ian
Janssen.
Manuscript 1: Influence of Physical Activity on Age-Related Weight Loss in the Elderly.
This manuscript has been submitted to Journal of the American Geriatrics Society and is
presented as requested by the journal. Wendy Stephen helped develop the study concept and
design and was primarily responsible for performing the data analysis, interpreting the study
results, and drafting the original manuscript. Dr. Ian Janssen secured the necessary funding for the
project and assisted with the development of the study concept and design. He also provided
guidance regarding data interpretation, statistical analysis, and editorial feedback during
preparation of the manuscript. Mr. Eric Bacon also assisted with the statistical analysis for this
project.
Manuscript 2: Sarcopenic-Obesity and Cardiovascular Disease Risk in the Elderly.
This manuscript has been submitted to Obesity and is presented as requested by the journal.
Wendy Stephen helped develop the study concept and design and was responsible for performing
the data analysis, interpreting the study results, and drafting the original manuscript. Dr. Ian
Janssen secured the funding for the project and assisted with the development of the study
concept and design. He also provided guidance regarding data interpretation, statistical analysis,
and editorial feedback during preparation of the manuscript.
iii
ACKNOWLEDGEMENTS
This thesis would never have come together if it weren’t for the guidance provided by Dr.
Ian Janssen. Ian took me on despite my glaring lack of epidemiology training and I dare say he
has almost turned me into a competent researcher. He has been a role model to me in many ways,
but in particular he has impressed me time and again with his efficiency. Ian has to be one of the
most capable and well-organized people I have ever met, which is only one of his qualities that
made this whole journey an enjoyable one. I am very thankful that I had the chance to work with
a world class researcher who is also a wonderful person.
Someone else who certainly deserves mention here is my husband, Geoff. Despite his
various connections to south western Ontario, he moved to Kingston with me so I could pursue
my Master’s degree. Throughout this process Geoff offered encouragement and a friendly ear
while I rattled on about a statistical test that was giving me grief or something “interesting” (to
me, at least) that I had read about that day. His love and support continually amaze me and I feel
truly fortunate to have such a remarkable person in my life.
My lab mates also made this experience far more enjoyable than it would have been on
my own. The folks in the Epidemiology Lab are wonderfully unique people who I will certainly
miss seeing every day. One person in particular, Caitlin, provided me with guidance and
reassurance every step of the way. Both as a colleague and as a friend, our time together is
something that I will never forget. Another important co-worker and friend, Travis, was always
ready and willing to debate or discuss the wonders of science and research. He constantly
challenged and entertained me and I can’t imagine having gone through this without him.
My old and newly-acquired family, Queen’s friends, Western friends, and high school
friends have all played an instrumental part in this process as well. Staying balanced in grad
school, and life, doesn’t just happen—it has to be consciously worked at. All of these important
people served as a reminder of this and for that I am eternally grateful.
iv
TABLE OF CONTENTS
ABSTRACT……………………………………………………………………..…..…………… .i
CO-AUTHORSHIP……………………………………………...…..…………………………....iii
ACKNOWLEDGEMENTS………………………………………………………………..……...iv
TABLE OF CONTENTS……………………………………………………………….…………v
LIST OF TABLES…………………………………………………..…………………….……..viii
LIST OF FIGURES……………………………………………………………………………….ix
CHAPTER 1: GENERAL INTRODUCTION……………………………………….……………1
Objectives and hypotheses ……………………………………………………………….2
Thesis organization……………………………………………….………………………3
References……………………………………………………………..………………….4
CHAPTER 2: LITERATURE REVIEW……………………………………….…………………6
2.1 The aging population…………………………………………………………………6
2.2 Aging and weight loss………………………………………..………………………6
2.2.1 Prevalence of age-related weight loss……………………………..………7
2.2.2 Health consequences of age-related weight loss…………….………….…8
2.2.3 Causes of age-related weight loss…………………………….……..…….8
2.2.4 Physical activity as a preventive measure…………….………………...…9
2.3 Aging and body composition………………………………………………..………10
2.3.1 Measurement of human body composition……………………………….10
2.3.2 Sarcopenia………………………………………………………….……..11
Classification and prevalence of sarcopenia…………………...………11
Health consequences of sarcopenia……………….…………...………12
2.3.3 Obesity……………………………………………………………………13
Classification and prevalence of obesity………………………..……..14
v
Health consequences of obesity………………………………………15
2.3.4 Sarcopenic-obesity………………………………………………………15
Classification and prevalence of sarcopenic-obesity………………....15
Health consequences of sarcopenic-obesity…………………….……16
2.4 Summary ………………………………………………………….………………18
References……………………………………………………………………………..20
CHAPTER 3: MANUSCRIPT 1………………………………….……………………………28
Abstract……………………………………………………….……………………….29
Introduction………………………………………...………………………………….30
Methods………………………………………………………………………………..31
Results…………………………………………………………………………………35
Discussion………………………………………………..……………………………39
References………………………………………………..……………………………43
CHAPTER 4: MANUSCRIPT 2……………………………………………………………….54
Abstract………………………………………….…………………………………….55
Introduction……………………………………………...…………………………….56
Methods………………………………………………………………………………..58
Results…………………………………………………………………………………63
Discussion………………………………………..……………………………………65
References…………………………………………………….………….……………68
CHAPTER 5: GENERAL DISCUSSION……………………………..………………………77
5.1 Summary of key findings……………………………………….…………………77
5.2 Limitations of the thesis………………………………………….………………..77
5.3 Strengths of the thesis………………………………………….………………….78
5.4 Future research directions………………………………………..………………..79
5.5 Public health implications…………………………………………………………80
vi
References…………………………………………………………..………….……81
SUMMARY AND CONCLUSIONS………………………………………………….…….82
References………………………………………………………….………….……84
APPENDIX A: Glossary of terms and abbreviations………………………………….……85
APPENDIX B: Ethics approval……………………………………………….……….……87
APPENDIX C: Physical activity questionnaire………………………………………..……91
APPENDIX D: Sample Chi Square analysis output………………………………...………94
APPENDIX E: Sample PROC MIXED analysis output……………………….……………96
APPENDIX F: Sample Cox proportional hazards regression analysis output…………..…108
APPENDIX G: Sample Excel output for an adjusted weight trajectory curve……….…….111
APPENDIX H: Determination of body composition categories based on abdominal
obesity x muscle mass……………………………………………………..114
APPENDIX I: Sample multiple comparisons analysis of variance with Bonferroni
post-hoc tests………………………………………………………………116
APPENDIX J: Sample partial correlations output………………………………………….119
vii
LIST OF TABLES
CHAPTER 3: Manuscript 1
Table 1.
Prevalence of baseline covariates among Cardiovascular Health Study
participants…………………………………………………………..…..……47
Table 2.
Projected absolute (kg) and relative (%) weight loss between the first and last
visits in physically inactive (quartile 1) and active (quartiles 2-4) groups…....49
Table 3.
Hazard ratios for clinically significant weight loss according to sex, age,
number of comorbid conditions, and survival status………………….………50
CHAPTER 4: Manuscript 2
Table 1.
Characteristics of study participants at baseline………………………..…….72
Table 2.
Baseline characteristics based on groups according to sarcopenia and obesity
classification……………………………………………………………….….73
Table 3.
Partial correlations between anthropometric and strength measures…….……74
Table 4.
Cardiovascular disease risk according to sarcopenia and obesity status….…..75
viii
LIST OF FIGURES
CHAPTER 2: Literature Review
Figure 1.
Cross-sectional estimated means for body weight by 2-year age groups from 12
to 20 years and 10-year age groups from 20 to 80 years for non-Hispanic white,
non-Hispanic black and Mexican-American females……………………………7
Figure 2.
Odds ratios for 3 or more physical disabilities in the past year according to
cross-sectional analysis of body composition in the New Mexico Aging Process
Study………………………………………………………………….…………17
CHAPTER 3: Manuscript 1
Figure 1.
Weight loss over 8 years of follow-up based on sex-specific physical activity
quartiles. The plots represent regression lines that were generated using the
coefficients obtained from an unstructured mixed model that included the
following independent variables: age, sex, race, height, household income,
smoking, alcohol, number of comorbid conditions at baseline, incident diseases,
the interaction of each of the previous variables with visit number, visit number,
and visit number2 ……………………………………………….………………52
Figure 2.
Weight loss over 8 years of follow-up with physically Inactive (quartile 1) and
Active (quartiles 2-4) groups according to sex (Panel A), baseline age (Panel B),
number of comorbid conditions at baseline (Panel C), and survival status during
follow-up (Panel D). The plots represent regression lines that were generated
using the coefficients obtained from an unstructured mixed model that included
the following independent variables: age, sex, race, height, annual household
income, smoking, alcohol, number of comorbid conditions at baseline, incident
ix
diseases, the interaction of each of the previous variables with visit number, visit
number, and visit number2………………………………………………………53
CHAPTER 4: Manuscript 2
Figure 1.
Number of cardiovascular disease events per 10,000 person-years according to
abdominal obesity and sarcopenia classified using either obesity X muscle mass
(Panel A) or obesity X muscle strength (Panel B)………………………………76
x
CHAPTER 1
GENERAL INTRODUCTION
The proportion of Canadian adults over the age of 65 years is on the rise. Current
projections estimate that by 2041 elderly adults will comprise nearly one quarter of the Canadian
population (1). Life expectancy estimates are also increasing (2) signifying that we are on the
verge of a dramatic demographic shift in Canada, which will inevitably have major implications
for population health.
Of the numerous physiological changes that accompany advancing age, changes in body
composition may be among the most significant in terms of health. Often a simple measure, such
as body weight, can be a warning sign of more complex problems. For instance, the occurrence of
age-related weight loss in the elderly is a phenomenon that often reflects poor health (3) and loss
of lean body mass (4). Weight loss in this group is typically unintentional and is associated with
increased risk of functional impairment (5) and mortality (6). Outside of disease and under
nutrition, the predictors of weight loss in the elderly are unclear. There is a need to better
understand the various causes of weight loss in the elderly population as well as practical methods
to slow its progression. Physical activity is associated with several health benefits in people of all
ages and may also be a predictor of weight loss in old age. Only one study (7) has examined the
ability of physical activity to attenuate age-related weight loss. Although the results of this study
appear promising, the researchers utilized crude measures of both the exposure (e.g., a physical
activity index based on only four activities) and outcome (e.g., self-reported weight). Therefore,
further research is required to confirm the initial findings of this investigation.
An additional health consequence of advancing age is the development of cardiovascular
disease (CVD), which is the leading cause of death in Canada (8). Among the body composition
changes that accompany aging, obesity has been pushed to the forefront as a risk factor for CVD.
Of particular importance is the increase in abdominal fat with aging as it is adversely linked to
1
numerous CVD risk factors including insulin resistance, hypertension, and dyslipidemia (9).
More recently, the role of sarcopenia, operationally defined as a low skeletal muscle mass and
strength, has been explored in terms of cardiometabolic health. Within elderly people low muscle
mass and strength have been associated with arterial stiffness (10), poor glucose tolerance (11),
and the metabolic syndrome (12).
Although the age-related increase in fat and decline in muscle often occur simultaneously (4),
these two phenomena have typically been studied separately. Recently, the idea of sarcopenicobesity has emerged whereby elderly persons experience a low muscle mass coupled with a high
body fat (13). Sarcopenic-obesity has been shown to be a particularly strong predictor of poor
physical function (13, 14). However, the cardiometabolic implications of sarcopenic-obesity have
been poorly studied and are unclear. While one study reported a negative metabolic effect of
sarcopenic-obesity (15) a second study reported a more desirable cardiometabolic risk factor
profile in obese women with sarcopenia versus obese women with a normal muscle mass (16).
These conflicting results may reflect the different approaches for assessing sarcopenia in these
two studies (muscle strength vs. muscle mass), raising the question as to whether muscle mass or
strength is more important for cardiovascular health. Further research is required to elucidate the
true effect of sarcopenic-obesity on cardiovascular health.
Objectives and Hypotheses
1) To determine whether participation in physical activity attenuates age-related weight loss
in community dwelling elderly men and women. It was hypothesized that weight would
decline over time and that physical activity would minimize, but not prevent, age-related
weight loss.
2) To determine if sarcopenic-obesity is a stronger predictor of CVD development in the
elderly than either sarcopenia or obesity alone, and whether low muscle mass or low
muscular strength is a stronger marker of CVD risk. It was hypothesized that: i) elderly
2
persons who were sarcopenic-obese would have a higher risk of CVD than those with
either a healthy body composition, obesity alone, or sarcopenia alone; and ii) muscle
mass is a stronger predictor of CVD risk than muscle strength.
Thesis Organization
This thesis conforms to the regulations outlined in the Queen’s School of Graduate
Studies and Research “General Forms of Theses.” The second chapter summarizes previous
studies in the areas of health research in elderly persons with a specific focus on age-related
weight loss and sarcopenic-obesity. The third chapter of the thesis contains the first manuscript
which is a study exploring the modifying effects of physical activity on age-related weight loss.
This manuscript has been formatted for submission to the Journal of the American Geriatrics
Society. Chapter 4 of the thesis is the second manuscript, prepared for submission to Obesity. It
examines the risk of CVD according to various combinations of sarcopenia and obesity while
concentrating on their convergence (e.g., sarcopenic-obesity). Chapter 5 contains a general
discussion, which is followed by the overall summary and conclusions.
3
References
1.
Health Canada and the Interdepartmental Committee on Aging and Seniors Issues.
Canada's Aging Population. Cat. H39-608/2002E. Minister of Public Works and
Government Services of Canada: Ottawa, 2002.
2.
Statistics Canada. The Daily. Communications Division, Statistics Canada: Ottawa,
2004.
3.
Wannamethee SG, Shaper AG, Lennon L. Reasons for intentional weight loss,
unintentional weight loss, and mortality in older men. Arch Intern Med 2005;165:10351040.
4.
Gallagher D, Ruts E, Visser M et al. Weight stability masks sarcopenia in elderly men
and women. Am J Physiol Endocrinol Metab 2000;279:E366-375.
5.
Launer LJ, Harris T, Rumpel C, Madans J. Body mass index, weight change, and risk of
mobility disability in middle-aged and older women. The epidemiologic follow-up study
of NHANES I. JAMA 1994;271:1093-1098.
6.
Newman AB, Yanez D, Harris T, Duxbury A, Enright PL, Fried LP. Weight change in
old age and its association with mortality. J Am Geriatr Soc 2001;49:1309-1318.
7.
Dziura J, Mendes de Leon C, Kasl S, DiPietro L. Can physical activity attenuate agingrelated weight loss in older people? The Yale Health and Aging Study, 1982-1994. Am J
Epidemiol 2004;159:759-767.
8.
Statistics Canada. Causes of death. Chapter IX: Diseases of the circulatory system. Cat.
84-208-XIE. 2002.
9.
Villareal DT, Apovian CM, Kushner RF, Klein S. Obesity in older adults: technical
review and position statement of the American Society for Nutrition and NAASO, The
Obesity Society. Obes Res 2005;13:1849-1863.
4
10.
Snijder MB, Henry RM, Visser M et al. Regional body composition as a determinant of
arterial stiffness in the elderly: The Hoorn Study. J Hypertens 2004;22:2339-2347.
11.
Snijder MB, Dekker JM, Visser M et al. Larger thigh and hip circumferences are
associated with better glucose tolerance: the Hoorn Study. Obes Res 2003;11:104-111.
12.
Jurca R, Lamonte MJ, Barlow CE, Kampert JB, Church TS, Blair SN. Association of
muscular strength with incidence of metabolic syndrome in men. Med Sci Sports Exerc
2005;37:1849-1855.
13.
Baumgartner RN. Body composition in healthy aging. Ann N Y Acad Sci 2000;904:437448.
14.
Baumgartner RN, Wayne SJ, Waters DL, Janssen I, Gallagher D, Morley JE. Sarcopenic
obesity predicts instrumental activities of daily living disability in the elderly. Obes Res
2004;12:1995-2004.
15.
Schrager MA, Metter EJ, Simonsick E et al. Sarcopenic obesity and inflammation in the
InCHIANTI study. J Appl Physiol 2007;102:919-925.
16.
Aubertin-Leheudre M, Lord C, Goulet ED, Khalil A, Dionne IJ. Effect of sarcopenia on
cardiovascular disease risk factors in obese postmenopausal women. Obesity (Silver
Spring) 2006;14:2277-2283.
5
CHAPTER 2
LITERATURE REVIEW
2.1 The Aging Population
The fastest growing population segment in Canada is comprised of seniors. In 2001,
approximately 12% of the population, almost 4 million Canadians, were 65 years of age or older.
By 2041, the proportion of seniors in the Canadian population is expected to rise to nearly 25%
(1). In addition, life expectancy is increasing for Canadians. According to Statistics Canada, the
life expectancies of 65 year old men and women in 1979 were 14.6 years and 19.0 years,
respectively (2). By 2002, 65 year old men and women were expected to live an additional 17.2
and 20.6 years (3). This growing number of elderly Canadians coupled with their increased life
expectancy has major implications for population health. There are numerous physiological
changes that occur in the elderly that account for their diminished health. This thesis will focus on
changes in body weight, obesity, and sarcopenia.
2.2 Aging and Weight Loss
Several cross-sectional studies illustrate an increase in body weight throughout early and
middle adulthood (4-7) until approximately age 60 at which point the weight trajectory begins to
decline (5, 6, 8). An example of this effect is shown in Figure 1 on page 7. Longitudinal studies
have confirmed these cross-sectional observations and have shown a decline in body weight in
both sexes after 60 (6, 9, 10) or 70 (11-13) years of age. While it is clear that body weight
decreases in elderly populations, the overall severity and consistency of this decline is not fully
understood.
6
90
Weight (kg)
80
70
non-Hispanic white
non-Hispanic black
60
Mexican-American
50
40
20
-2
9.
9
30
-3
9.
9
40
-4
9.
9
50
-5
9.
9
60
-6
9.
9
70
-7
9.
9
12
-1
3.
9
14
-1
5.
9
16
-1
7.
9
18
-1
9.
9
30
Age (years)
Figure 1. Cross-sectional estimated means for body weight by 2-year age groups from 12 to 20
years and 10-year age groups from 20 to 80 years for non-Hispanic white, non-Hispanic black
and Mexican-American females (5).
The weight loss that occurs in older persons is likely not the same as the weight loss that is
typically observed in younger individuals. In young and middle-aged persons, weight loss is often
intentional (e.g., diet or exercise-induced) and represents improved body composition resulting in
better health. Conversely, seniors that experience weight loss often do so unintentionally and this
involuntary weight loss is often a marker of clinical and/or subclinical disease (14). With
advancing age the two major body tissues, skeletal muscle and fat, change in opposite directions.
Skeletal muscle decreases (15) while there is a gradual increase in body fat (16). Typically weight
loss is observed when the loss of skeletal muscle exceeds the concurrent increase in body fat.
2.2.1 Prevalence of Age-Related Weight Loss
The prevalence of weight loss among elderly adults is not well established as the literature on
this topic is limited to a few smaller studies that utilize a variety of definitions for weight loss.
Over the first 3 years of follow-up more than 15% of the Cardiovascular Health Study cohort
7
experienced ≥5% weight loss while an additional 5% had weight loss of ≥10%, with little
difference between sexes (17). In a separate study of male Veterans living in Seattle, one quarter
of the sample experienced ≥4% weight loss over a 2 year period (18). Thus, the weight loss
prevalence ranges from 15% to 25% depending on the population under study and the definition
of weight loss that is employed.
2.2.2 Health Consequences of Age-Related Weight Loss
Weight loss in the elderly is associated with several deleterious consequences. For instance,
older women from the first National Health and Nutrition Examination Survey who experienced a
weight loss of more than 5% over 5 years had a twofold increase in risk of disability compared
with weight-stable women (19). In addition to this increase in disability risk, weight loss in the
elderly has been associated with an increased risk of mortality in several studies (17, 18, 20-22).
For instance, within the Cardiovascular Health Study cohort, a weight loss of 5% or more over 3
years was associated with a 67% higher risk of mortality (17). Interestingly, weight loss in the
elderly is believed to be detrimental regardless of weight loss intention (18). The severe
consequences of age-related weight loss underscore the need to better understand the mechanisms
of weight loss in the elderly and the need to develop strategies that will slow its progression.
2.2.3 Causes of Age-Related Weight Loss
A number of factors have been identified as causes of age-related weight loss in the elderly.
These include the development of illness and disease, psychological issues, and sociodemographic factors (e.g., low socioeconomic status) which may contribute to insufficient caloric
intake (23). Moreover, other factors such as a decline in smell and taste (24), slowed gastric
emptying (25), reduced chewing efficiency (26), and alterations in the neuroendocrine system
(27, 28) are together associated with early satiety and a decline in appetite. Unfortunately, many
of these risk factors may be difficult for an older individual to change. Fortunately, a modifiable
8
cause of age-related weight loss has been identified: muscle disuse (29). Not surprisingly, this
disuse is mirrored by a gradual decline in physical activity with advancing age (30, 31).
2.2.4 Physical Activity as a Preventive Measure
Since age-related weight loss often reflects a reduction in skeletal muscle mass, maintenance
of skeletal muscle should help to attenuate this decline and its associated consequences. One
strategy that has demonstrated effectiveness for minimizing muscle wasting in older age is
resistance exercise (32); however, resistance training is not a preferred exercise modality for most
elderly adults as evidenced by the low prevalence (<15%) of elderly Americans who perform
muscle strengthening activities at least twice per week (33). The broader category of physical
activity, that includes resistance exercise and many other exercises and activities, may offer a
reasonable alternative to resistance training since it provides a greater selection of activities to
suit all tastes. Moreover, it can be incorporated into one’s daily routine through a healthy lifestyle
rather than in regimented training sessions. Several studies in young and middle-aged adults have
demonstrated that physical activity can prevent or attenuate unwanted weight gain (34, 35). This
raises the question of whether physical activity can minimize the magnitude of weight loss
observed in the elderly.
Presently, only one study has examined the impact of physical activity on the weight
trajectory of elderly persons. Over 12 years of follow-up, Dziura et al. (36) found that being
physically active attenuated weight loss amongst 2300 elderly adults such that every unit increase
in the 9 point physical activity score corresponded to weight loss attenuation of 0.04 kg per year.
The mean score of 2 could be achieved by participating in one activity (e.g., walking) “often in
the past month.” Therefore, a person who scored 2 would lose 0.08 kg per year less than a
sedentary person. Although informative, this study was limited in that it employed crude
measures of both the physical activity exposure (e.g., a physical activity index based on only four
activities) and the weight loss outcome (e.g., self-reported weight). Furthermore, this study did
9
not consider incident disease and its potential effect on the weight loss trajectory. Therefore,
while it appears promising that physical activity may attenuate age-related weight loss, more
research is needed to confirm these initial findings.
2.3 Aging and Body Composition
Generally speaking, cross-sectional and longitudinal studies have shown that with
advancing age there is an increase in fat mass and a decrease in muscle mass (11, 37-39).
Gallagher and colleagues (11) demonstrated that after age 60 total skeletal muscle mass declined
by 0.8 kg and 0.4 kg over a 5-year period in men and women, respectively. The corresponding
values for changes in fat mass were a gain of 1.2 kg and a non-significant loss of 0.8 kg. These
observations highlight the concurrent changes in these two major body tissues. Interestingly, in
this sample body weight remained constant illustrating that these changes can manifest
themselves even without changes in body weight.
2.3.1 Measurement of Human Body Composition
Various techniques can be used to measure human body composition. Overweight and
obesity are commonly assessed in the research and clinical setting using anthropometric measures
including the body mass index (BMI) as a measure of overall adiposity. This simple index of
weight-for-height is calculated as weight in kilograms divided by the square of height in meters
(kg/m2). Waist circumference (WC) can also be used as a simple anthropometric index of
abdominal obesity. Unlike total and abdominal obesity, skeletal muscle mass is not typically
measured in the clinical setting and is primarily assessed for research purposes alone. One of the
most straightforward measures for muscle mass that is appropriate for large, epidemiological
studies involves bioelectrical impedance analysis (BIA). Simply put, BIA involves a small
electric current that is sent through the body from hand to foot while the body’s resistance to the
flow of the current is measured (40). Based on basic assumptions involving human tissue
10
hydration, the measured resistance values allow for estimation of skeletal muscle mass or total
lean body mass. BIA output can also be used to estimate body fat (e.g., body fat = body mass –
lean body mass). Other more sophisticated imaging techniques also exist for measurement of
body composition, including dual-energy X-ray absorptiometry (DXA), computed tomography
(CT) and magnetic resonance imaging (MRI). While appropriate and useful for smaller laboratory
based studies, these techniques are often impractical for routine use in the clinical setting and in
large epidemiological studies due to high cost, time considerations, and difficulty accessing such
machines (41, 42). As such, anthropometric indicators and BIA are more commonly employed in
these types of studies.
2.3.2 Sarcopenia
One of the most striking and clinically significant anatomical changes in aging humans is
the loss of skeletal muscle mass. In 1989 Dr. Irwin Rosenberg first published the term
“sarcopenia” to refer to the process of age-related skeletal muscle loss (43).
Classification and Prevalence of Sarcopenia
Sarcopenia has traditionally been measured in terms of the quantity of muscle lost (e.g.,
mass) during the aging process. The loss of muscle quality (e.g., strength) is often considered to
be secondary to loss of muscle mass. As a result, prevalence estimates for sarcopenia are typically
based on the loss of muscle mass with advancing age. Since all humans, even those who remain
healthy and disease-free, lose muscle mass as they age, one could argue that all elderly persons
are sarcopenic. However, skeletal muscle mass varies widely in older adults due to individual
differences in peak muscle mass and the rate at which different people lose muscle with aging.
Some elderly persons have skeletal muscle mass values comparable to young healthy adults
whereas others are unable to perform even simple daily activities. Thus, for clinical purposes and
for statistical comparisons in research studies, an operational definition of sarcopenia is typically
11
used to indicate which elderly persons have skeletal muscle values in the healthy (e.g., normal)
and unhealthy (e.g., sarcopenic) ranges.
In 1998, Baumgartner and colleagues (44) proposed a dichotomous breakdown to
determine which elderly persons have sarcopenia. They defined sarcopenia as a height-adjusted
appendicular (arm + leg) skeletal muscle mass (muscle mass/height2) of two standard deviations
or more below the mean of a young and healthy reference population. Using this approach, the
prevalence of sarcopenia in the New Mexico Elder Health Survey cohort was 14% in those aged
65-69 years and over 50% in those greater than 80 years (44). Since this landmark paper (44),
many other researchers have employed a similar threshold approach for classifying sarcopenia
(45-48). Recently, Janssen and colleagues (49) proposed sarcopenia thresholds based on the
relation between muscle mass and physical disability. Statistical techniques were applied to a
representative dataset of elderly Americans to determine which skeletal muscle thresholds were
associated with a high likelihood of physical disability. Height-adjusted whole-body skeletal
muscle mass thresholds of <6.75 kg/m2 and <5.75 kg/m2 were selected to denote moderate and
severe sarcopenia levels in women. The corresponding values in men were 10.75 kg/m2 and 8.50
kg/m2. Based on these thresholds, 9% of the older American women and 11% of the older
American men were considered to have severe sarcopenia, while 22% of these women and 53%
of the men were considered to have moderate sarcopenia (49). These findings suggest that the
prevalence of sarcopenia, defined here as having a muscle mass that increases physical disability
risk, is extremely high in the elderly population. At present, there is no published data on the
prevalence of sarcopenia in elderly Canadians.
Health Consequences of Sarcopenia
Most of the sarcopenia literature has focused on the functional implications of the loss of
muscle mass with aging (44, 45, 50-52). In general, these cross-sectional and longitudinal
observational studies have found sarcopenia to be associated with functional impairment and
12
physical disability. For example, in a nationally representative cross-sectional sample of adults
aged 60 and older, Janssen and colleagues (45) observed that older men with sarcopenia had a
two-fold greater likelihood of functional impairment and disability compared to older men with a
normal muscle mass. Older women with sarcopenia were three times more likely to have
functional impairment and disability. Additional evidence suggests an important role for muscle
strength beyond that of muscle size. Two published reports on the Health Aging and Body
Composition Study cohort indicate that low muscle is predictive of a loss in physical function
over 2-3 years of follow-up (50, 53). In these studies the effects of muscle mass on loss of
function were attributable to muscle strength, implying that the association between low muscle
mass and functional decline is a function of the underlying loss in muscle strength.
While there is a general consensus in the scientific community regarding the effect of
sarcopenia on functional outcomes, limited attention has been paid to the metabolic implications
of sarcopenia. A cross-sectional analysis of 2484 elderly subjects from the Hoorn Study revealed
that thigh circumference was strongly and negatively associated with markers of glucose
metabolism in women, but not in men, independent of WC, BMI, and age (54). Further
examination of 648 Hoorn Study participants showed that larger leg lean mass was consistently
protective of arterial stiffness (55). In terms of muscle strength, a recent longitudinal study
reported lower incidence of the metabolic syndrome across muscular strength tertiles in men (56).
Taken together, the results from these three studies suggest that low skeletal muscle quantity
(e.g., mass) and quality (e.g., strength) may negatively impact precursors of cardiovascular
disease (CVD) and increase a person’s risk of a cardiovascular event. Due to the limited research
conducted to date, it is currently not clear as to whether muscle mass and strength are equally
important for cardiovascular health.
2.3.3 Obesity
13
In its simplest sense, obesity can be defined as a condition of excessive fat accumulation
to the extent that health may be impaired. Obesity reflects a continual positive energy balance, in
which energy intake consistently exceeds energy output and where undesirable weight gain
occurs. Although a common feature in obese individuals is an increase in body fat, the degree of
excess fat and its distribution within the body varies considerably between obese individuals.
Classification and Prevalence of Obesity
As mentioned previously, overweight and obesity are commonly assessed in the research
and clinical setting using BMI as a measure of overall adiposity. Globally accepted BMI
thresholds of 25 kg/m2 and 30 kg/m2 are used to define overweight and obesity, respectively, in
men and women of all ages (57).
In addition to BMI, WC can be used as a simple anthropometric index of abdominal
obesity. Unlike BMI, there is currently no consensus as to what WC thresholds should be used to
denote increased health risk. The most commonly used cut-points in men are 94 cm, which
denotes a moderately increased risk of obesity-related complications, and 102 cm, which denotes
a substantially increased risk of obesity-related complications. The corresponding WC thresholds
in women are 80 and 88 cm, respectively (58).
Overweight and obesity are at epidemic proportions globally, with more than 1 billion
overweight adults and at least 300 million of these clinically obese (57). Based on BMI, recent
estimates from the United States indicate that 40% of the population aged 60 years or older is
overweight while an additional 31% is obese (59). Likewise, Canadian Community Health
Survey data indicates that about 50% of the Canadian population aged 65 years and older is
overweight while approximately 20% more are considered obese (60). Even more troubling,
measures of waist circumference obtained on a representative sample of older Americans indicate
that approximately 70% of those aged 60 years or older have abdominal obesity (≥102 cm in men,
14
≥88 cm in women) (61). Clearly, a large proportion of elderly persons are at increased health risk
due to their excess total and abdominal fat.
Health Consequences of Obesity
In the elderly, obesity is associated with numerous health problems including impaired
physical function and quality of life, development of type 2 diabetes, hypertension, dyslipidemia,
and CVD (16). Of particular importance for cardiometabolic disease is the increase in abdominal
obesity that occurs with advancing age. The two fat depots represented by abdominal obesity
change in opposition with advancing age: visceral fat increases, while subcutaneous fat in the
abdomen and in other regions of the body (e.g., thighs, calves) decreases (62). The increase in
visceral fat is important to note since it is independently associated with several cardiometabolic
risk factors in elderly persons including insulin resistance, hypertension, and dyslipidemia (16). In
addition to the functional and metabolic implications of abdominal obesity, Kuk and colleagues
(63) demonstrated that for every standard deviation increase in visceral fat mass there was an
81% increase in risk of all-cause mortality in men.
2.3.4 Sarcopenic-Obesity
Most of the body composition research in the elderly has focused on the separate roles of
sarcopenia and obesity (overall or abdominal). However, the concept of sarcopenic-obesity has
been proposed wherein older adults experience a low skeletal muscle mass coupled with a high
fat mass (64).
Classification and Prevalence of Sarcopenic-Obesity
The few published studies conducted thus far have defined sarcopenic-obesity using a
variety of approaches. Sarcopenic-obesity has typically been considered in terms of high body fat
coupled with low muscle mass rather than low muscle strength. In a series of studies,
15
Baumgartner and colleagues defined sarcopenic-obesity as height-adjusted appendicular (arms +
legs) skeletal muscle mass of 2 standard deviations or more below the sex-specific mean for a
young adult reference population, coupled with percentage body fat greater than the sex-specific
median for older adults (64, 65). Using this approach, Baumgartner (64) performed crosssectional analyses on 2 samples of elderly adults. Within the New Mexico Aging Process Study
cohort, 10% of the subjects were sarcopenic-obese, with approximately 15% having sarcopenia
alone, 35% having obesity alone, and 40% having a normal body composition. When similar
methods were applied to the New Mexico Elder Health Survey cohort, only 4% were classified as
sarcopenic-obese, with 21% sarcopenic, 46% obese, and 29% having a normal body composition.
Thus, a relatively small proportion of the elderly population appears to be sarcopenic-obese based
on the criteria proposed by Baumgartner.
Although not called sarcopenic-obesity, other studies have also considered skeletal
muscle and fat measures simultaneously. Approaches such as normalizing muscle mass for total
body weight (45, 66) and adjusting muscle mass for height and fat mass (66, 67) have been
employed. The lack of consensus regarding the definition of this condition means that the true
prevalence of sarcopenic-obesity is unknown in Canada and elsewhere.
Health Consequences of Sarcopenic-Obesity
With some exceptions (66, 68), the limited literature regarding the health consequences of
sarcopenic-obesity have reported that sarcopenic-obese persons are at increased risk of functional
impairment and physical disability (64, 65, 69). In a cross-sectional study of sarcopenic-obesity,
Baumgartner examined participants in the New Mexico Aging Process Study cohort. He reported
that the odds ratios for disability in sarcopenic, obese, and sarcopenic-obese groups relative to the
group of seniors with a normal body composition were 2.07, 2.33, and 4.12, respectively. This is
illustrated in Figure 2 (64).
16
Odds Ratio
5.0
4.0
3.0
2.0
1.0
0.0
Normal
Sarcopenic
Obese
SarcopenicObese
Body Composition Group
Figure 2. Odds ratios for 3 or more physical disabilities in the past year according to crosssectional analysis of body composition in the New Mexico Aging Process Study (64).
These findings suggest that sarcopenia and obesity have additive effects when it comes to
physical disability risk in elderly persons. More recently, Baumgartner and colleagues attempted
to replicate these initial cross-sectional findings in a longitudinal study based on the same cohort
(65). In this study, elderly participants with either sarcopenia or obesity alone at baseline were not
at increased risk of functional decline over 8 years of follow-up compared to elderly individuals
with a normal body composition. However, individuals with sarcopenic-obesity were at 2.5 times
greater risk of functional decline. Thus, although the findings of this longitudinal study did not
support sarcopenia or obesity alone as risk factors for physical disability, these results still
support the idea that the combination of sarcopenia and obesity has a greater impact on physical
function than either body composition condition alone.
What remain unclear are the cardiovascular and metabolic implications of sarcopenicobesity. As highlighted in previous sections of the literature review, obesity, particularly
abdominal and visceral obesity, contributes to numerous cardiometabolic health problems such as
insulin resistance (70), type 2 diabetes (71), dyslipidemia (72), and CVD events (72, 73).
Likewise, low muscle mass and strength are associated with CVD risk factors including arterial
stiffness (55), glucose intolerance (54), and the metabolic syndrome (56). Since both obesity and
17
low muscle mass/strength predict cardiovascular risk factors and outcomes in the elderly, it is
possible that the combination of obesity and sarcopenia would be associated with an even greater
risk, similar to the findings for physical function (64, 65).
In support of the aforementioned notion, cross-sectional research in a sample of 871
sarcopenic-obese elderly persons demonstrated that abdominal obesity and low muscular strength
are characterized by high circulating levels of proinflammatory cytokines (74), which are
recognized risk factors for CVD (75). On the contrary, a recent cross-sectional study of 22
postmenopausal women reported that the cardiometabolic risk factor profile was worse in obese
women with a normal muscle mass compared to sarcopenic-obese women (76). This evidence
suggests that the muscle wasting observed in the sarcopenic-obese group was protective in some
way in these obese women. The inconsistency in the findings of these two studies may reflect the
different approaches for assessing sarcopenia (muscle strength vs. muscle mass), raising the
question as to whether muscle mass or strength is more important for cardiovascular health.
While these two studies report some interesting findings, their small sample sizes, cross-sectional
designs, and contradictory findings indicate that more research is needed to clarify the impact of
sarcopenic-obesity on CVD risk.
2.4 Summary
To date only one study has investigated the utility of physical activity for attenuating agerelated weight loss in the elderly (36). As a result, it is difficult to say conclusively whether
physical activity has a positive effect on attenuating the unhealthy weight loss that is typically
observed in older persons. For this reason, further research is required to confirm these initial
findings.
Recent research in the field of geriatrics has pointed to the additive effects of sarcopenia
and obesity on physical function in the elderly. Limited research has been conducted on the role
of sarcopenic-obesity in terms of cardiovascular health and those studies that do exist report
18
conflicting results (74, 76). This discordance may be attributable to the different measures of
sarcopenia (e.g., low strength vs. low mass). Therefore, additional research is necessary to clarify
the impact of sarcopenic-obesity on cardiovascular health and to determine whether muscle mass
or muscle strength is a stronger indicator of adverse cardiovascular outcomes.
19
References
1.
Health Canada and the Interdepartmental Committee on Aging and Seniors Issues.
Canada's Aging Population. Cat. H39-608/2002E. Minister of Public Works and
Government Services of Canada: Ottawa, 2002.
2.
Statistics Canada. Healthy Canadians: Life expectancy. Cat. 82-401-XIE. 2004.
3.
Statistics Canada. The Daily. Communications Division, Statistics Canada: Ottawa,
September 27 2004.
4.
Guo SS, Zeller C, Chumlea WC, Siervogel RM. Aging, body composition, and lifestyle:
the Fels Longitudinal Study. Am J Clin Nutr 1999;70:405-411.
5.
Chumlea WC, Guo SS, Kuczmarski RJ et al. Body composition estimates from NHANES
III bioelectrical impedance data. Int J Obes Relat Metab Disord 2002;26:1596-1609.
6.
Carmelli D, McElroy MR, Rosenman RH. Longitudinal changes in fat distribution in the
Western Collaborative Group Study: a 23-year follow-up. Int J Obes 1991;15:67-74.
7.
Lewis CE, Smith DE, Wallace DD, Williams OD, Bild DE, Jacobs DR, Jr. Seven-year
trends in body weight and associations with lifestyle and behavioral characteristics in
black and white young adults: the CARDIA study. Am J Public Health 1997;87:635-642.
8.
Kyle UG, Genton L, Hans D et al. Total body mass, fat mass, fat-free mass, and skeletal
muscle in older people: cross-sectional differences in 60-year-old persons. J Am Geriatr
Soc 2001;49:1633-1640.
9.
Visser M, Pahor M, Tylavsky F et al. One- and two-year change in body composition as
measured by DXA in a population-based cohort of older men and women. J Appl Physiol
2003;94:2368-2374.
10.
Shimokata H, Andres R, Coon PJ, Elahi D, Muller DC, Tobin JD. Studies in the
distribution of body fat. II. Longitudinal effects of change in weight. Int J Obes
1989;13:455-464.
20
11.
Gallagher D, Ruts E, Visser M et al. Weight stability masks sarcopenia in elderly men
and women. Am J Physiol Endocrinol Metab 2000;279:E366-375.
12.
Raguso CA, Kyle U, Kossovsky MP et al. A 3-year longitudinal study on body
composition changes in the elderly: role of physical exercise. Clin Nutr 2006;25:573-580.
13.
Hughes VA, Roubenoff R, Wood M, Frontera WR, Evans WJ, Fiatarone Singh MA.
Anthropometric assessment of 10-y changes in body composition in the elderly. Am J
Clin Nutr 2004;80:475-482.
14.
Wannamethee SG, Shaper AG, Lennon L. Reasons for intentional weight loss,
unintentional weight loss, and mortality in older men. Arch Intern Med 2005;165:10351040.
15.
Doherty TJ. Invited review: Aging and sarcopenia. J Appl Physiol 2003;95:1717-1727.
16.
Villareal DT, Apovian CM, Kushner RF, Klein S. Obesity in older adults: technical
review and position statement of the American Society for Nutrition and NAASO, The
Obesity Society. Obes Res 2005;13:1849-1863.
17.
Newman AB, Yanez D, Harris T, Duxbury A, Enright PL, Fried LP. Weight change in
old age and its association with mortality. J Am Geriatr Soc 2001;49:1309-1318.
18.
Wallace JI, Schwartz RS, LaCroix AZ, Uhlmann RF, Pearlman RA. Involuntary weight
loss in older outpatients: incidence and clinical significance. J Am Geriatr Soc
1995;43:329-337.
19.
Launer LJ, Harris T, Rumpel C, Madans J. Body mass index, weight change, and risk of
mobility disability in middle-aged and older women. The epidemiologic follow-up study
of NHANES I. JAMA 1994;271:1093-1098.
20.
Losonczy KG, Harris TB, Cornoni-Huntley J et al. Does weight loss from middle age to
old age explain the inverse weight mortality relation in old age? Am J Epidemiol
1995;141:312-321.
21
21.
Dey DK, Rothenberg E, Sundh V, Bosaeus I, Steen B. Body mass index, weight change
and mortality in the elderly. A 15 y longitudinal population study of 70 y olds. Eur J Clin
Nutr 2001;55:482-492.
22.
Corrada MM, Kawas CH, Mozaffar F, Paganini-Hill A. Association of body mass index
and weight change with all-cause mortality in the elderly. Am J Epidemiol 2006;163:938949.
23.
Alibhai SM, Greenwood C, Payette H. An approach to the management of unintentional
weight loss in elderly people. CMAJ 2005;172:773-780.
24.
Schiffman SS. Taste and smell in disease (first of two parts). N Engl J Med
1983;308:1275-1279.
25.
Clarkston WK, Pantano MM, Morley JE, Horowitz M, Littlefield JM, Burton FR.
Evidence for the anorexia of aging: gastrointestinal transit and hunger in healthy elderly
vs. young adults. Am J Physiol 1997;272:R243-248.
26.
Ship JA, Duffy V, Jones JA, Langmore S. Geriatric oral health and its impact on eating. J
Am Geriatr Soc 1996;44:456-464.
27.
Morley JE. Decreased food intake with aging. J Gerontol A Biol Sci Med Sci 2001;56:8188.
28.
Wilson MM, Morley JE. Invited review: Aging and energy balance. J Appl Physiol
2003;95:1728-1736.
29.
Bortz WM. Disuse and aging. JAMA 1982;248:1203-1208.
30.
Yusuf HR, Croft JB, Giles WH et al. Leisure-time physical activity among older adults.
United States, 1990. Arch Intern Med 1996;156:1321-1326.
31.
Bijnen FC, Feskens EJ, Caspersen CJ, Mosterd WL, Kromhout D. Age, period, and
cohort effects on physical activity among elderly men during 10 years of follow-up: the
Zutphen Elderly Study. J Gerontol A Biol Sci Med Sci 1998;53:M235-241.
22
32.
Hunter GR, McCarthy JP, Bamman MM. Effects of resistance training on older adults.
Sports Med 2004;34:329-348.
33.
Trends in Strength Training--United States, 1998-2004. JAMA 2006;296:1459-1460.
34.
Di Pietro L, Dziura J, Blair SN. Estimated change in physical activity level (PAL) and
prediction of 5-year weight change in men: the Aerobics Center Longitudinal Study. Int J
Obes Relat Metab Disord 2004;28:1541-1547.
35.
Littman AJ, Kristal AR, White E. Effects of physical activity intensity, frequency, and
activity type on 10-y weight change in middle-aged men and women. Int J Obes (Lond)
2005;29:524-533.
36.
Dziura J, Mendes de Leon C, Kasl S, DiPietro L. Can physical activity attenuate agingrelated weight loss in older people? The Yale Health and Aging Study, 1982-1994. Am J
Epidemiol 2004;159:759-767.
37.
Baumgartner RN, Stauber PM, McHugh D, Koehler KM, Garry PJ. Cross-sectional age
differences in body composition in persons 60+ years of age. J Gerontol A Biol Sci Med
Sci 1995;50:M307-316.
38.
Forbes GB. Longitudinal changes in adult fat-free mass: influence of body weight. Am J
Clin Nutr 1999;70:1025-1031.
39.
Hughes VA, Frontera WR, Roubenoff R, Evans WJ, Singh MA. Longitudinal changes in
body composition in older men and women: role of body weight change and physical
activity. Am J Clin Nutr 2002;76:473-481.
40.
Chumlea WC, Sun SS. Chapter 6 Bioelectrical Impedance Analysis. In: Heymsfield SB,
Lohman TG, Wang Z and Going SB. Human body composition. Champaign, IL, 2005,
79-88.
41.
Lohman TG, Zhao C. Chapter 5 Dual-energy X-ray Absorptiometry. In: Heymsfield SB,
Lohman TG, Wang Z and Going SB. Human body composition. Champaign, IL, 2005,
63-78.
23
42.
Ross R, Janssen I. Chapter 7 Computed Tomography and Magnetic Resonance Imaging.
In: Heymsfield SB, Lohman TG, Wang Z and Going SB. Human body composition.
Champaign, IL, 2005, 89-108.
43.
Rosenberg IH. Summary comments. Am J Clin Nutr 1989;50:1231-1233.
44.
Baumgartner RN, Koehler KM, Gallagher D et al. Epidemiology of sarcopenia among
the elderly in New Mexico. Am J Epidemiol 1998;147:755-763.
45.
Janssen I, Heymsfield SB, Ross R. Low relative skeletal muscle mass (sarcopenia) in
older persons is associated with functional impairment and physical disability. J Am
Geriatr Soc 2002;50:889-896.
46.
Melton LJ, 3rd, Khosla S, Crowson CS, O'Connor MK, O'Fallon WM, Riggs BL.
Epidemiology of sarcopenia. J Am Geriatr Soc 2000;48:625-630.
47.
Tanko LB, Movsesyan L, Mouritzen U, Christiansen C, Svendsen OL. Appendicular lean
tissue mass and the prevalence of sarcopenia among healthy women. Metabolism
2002;51:69-74.
48.
Iannuzzi-Sucich M, Prestwood KM, Kenny AM. Prevalence of sarcopenia and predictors
of skeletal muscle mass in healthy, older men and women. J Gerontol A Biol Sci Med Sci
2002;57:M772-777.
49.
Janssen I, Baumgartner RN, Ross R, Rosenberg IH, Roubenoff R. Skeletal muscle
cutpoints associated with elevated physical disability risk in older men and women. Am J
Epidemiol 2004;159:413-421.
50.
Visser M, Goodpaster BH, Kritchevsky SB et al. Muscle mass, muscle strength, and
muscle fat infiltration as predictors of incident mobility limitations in well-functioning
older persons. J Gerontol A Biol Sci Med Sci 2005;60:324-333.
51.
Visser M, Kritchevsky SB, Goodpaster BH et al. Leg muscle mass and composition in
relation to lower extremity performance in men and women aged 70 to 79: the Health,
Aging and Body Composition Study. J Am Geriatr Soc 2002;50:897-904.
24
52.
Sowers MR, Crutchfield M, Richards K et al. Sarcopenia is related to physical
functioning and leg strength in middle-aged women. J Gerontol A Biol Sci Med Sci
2005;60:486-490.
53.
Goodpaster BH, Park SW, Harris TB et al. The loss of skeletal muscle strength, mass,
and quality in older adults: the Health, Aging and Body Composition Study. J Gerontol A
Biol Sci Med Sci 2006;61:1059-1064.
54.
Snijder MB, Dekker JM, Visser M et al. Larger thigh and hip circumferences are
associated with better glucose tolerance: the Hoorn Study. Obes Res 2003;11:104-111.
55.
Snijder MB, Henry RM, Visser M et al. Regional body composition as a determinant of
arterial stiffness in the elderly: The Hoorn Study. J Hypertens 2004;22:2339-2347.
56.
Jurca R, Lamonte MJ, Barlow CE, Kampert JB, Church TS, Blair SN. Association of
muscular strength with incidence of metabolic syndrome in men. Med Sci Sports Exerc
2005;37:1849-1855.
57.
Obesity: preventing and managing the global epidemic. Report of a WHO consultation.
World Health Organ Tech Rep Ser 2000;894:i-xii, 1-253.
58.
Lean ME, Han TS, Morrison CE. Waist circumference as a measure for indicating need
for weight management. BMJ 1995;311:158-161.
59.
Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of
overweight and obesity in the United States, 1999-2004. JAMA 2006;295:1549-1555.
60.
Tjepkema M. Nutrition: Findings from the Canadian Community Health Survey.
Measured Obesity. Adult Obesity in Canada: Measured height and weight. 2005;1-32.
61.
Ford ES, Mokdad AH, Giles WH. Trends in waist circumference among U.S. adults.
Obes Res 2003;11:1223-1231.
62.
Zamboni M, Mazzali G, Zoico E et al. Health consequences of obesity in the elderly: a
review of four unresolved questions. Int J Obes (Lond) 2005;29:1011-1029.
25
63.
Kuk JL, Katzmarzyk PT, Nichaman MZ, Church TS, Blair SN, Ross R. Visceral fat is an
independent predictor of all-cause mortality in men. Obesity (Silver Spring) 2006;14:336341.
64.
Baumgartner RN. Body composition in healthy aging. Ann N Y Acad Sci 2000;904:437448.
65.
Baumgartner RN, Wayne SJ, Waters DL, Janssen I, Gallagher D, Morley JE. Sarcopenic
obesity predicts instrumental activities of daily living disability in the elderly. Obes Res
2004;12:1995-2004.
66.
Zoico E, Di Francesco V, Guralnik JM et al. Physical disability and muscular strength in
relation to obesity and different body composition indexes in a sample of healthy elderly
women. Int J Obes Relat Metab Disord 2004;28:234-241.
67.
Newman AB, Kupelian V, Visser M et al. Sarcopenia: alternative definitions and
associations with lower extremity function. J Am Geriatr Soc 2003;51:1602-1609.
68.
Davison KK, Ford ES, Cogswell ME, Dietz WH. Percentage of body fat and body mass
index are associated with mobility limitations in people aged 70 and older from
NHANES III. J Am Geriatr Soc 2002;50:1802-1809.
69.
Morley JE, Baumgartner RN, Roubenoff R, Mayer J, Nair KS. Sarcopenia. J Lab Clin
Med 2001;137:231-243.
70.
Després JP, Lemieux S, Lamarche B et al. The insulin resistance-dyslipidemic syndrome:
contribution of visceral obesity and therapeutic implications. Int J Obes Relat Metab
Disord 1995;19 Suppl 1:S76-86.
71.
Boyko EJ, Fujimoto WY, Leonetti DL, Newell-Morris L. Visceral adiposity and risk of
type 2 diabetes: a prospective study among Japanese Americans. Diabetes Care
2000;23:465-471.
26
72.
Després JP, Moorjani S, Lupien PJ, Tremblay A, Nadeau A, Bouchard C. Regional
distribution of body fat, plasma lipoproteins, and cardiovascular disease. Arteriosclerosis
1990;10:497-511.
73.
Nicklas BJ, Penninx BW, Cesari M et al. Association of visceral adipose tissue with
incident myocardial infarction in older men and women: the Health, Aging and Body
Composition Study. Am J Epidemiol 2004;160:741-749.
74.
Schrager MA, Metter EJ, Simonsick E et al. Sarcopenic obesity and inflammation in the
InCHIANTI study. J Appl Physiol 2007;102:919-925.
75.
Pearson TA, Mensah GA, Alexander RW et al. Markers of inflammation and
cardiovascular disease: application to clinical and public health practice: A statement for
healthcare professionals from the Centers for Disease Control and Prevention and the
American Heart Association. Circulation 2003;107:499-511.
76.
Aubertin-Leheudre M, Lord C, Goulet ED, Khalil A, Dionne IJ. Effect of sarcopenia on
cardiovascular disease risk factors in obese postmenopausal women. Obesity (Silver
Spring) 2006;14:2277-2283.
27
CHAPTER 3
MANUSCRIPT 1
Influence of Physical Activity on Age-Related Weight Loss in the Elderly
This manuscript has been submitted to Journal of the American Geriatrics Society
and is presented as requested by the journal.
Wendy C. Stephen, BSc* and Ian Janssen, PhD*†
*School of Kinesiology and Health Studies, and
†
Department of Community Health and Epidemiology
Queen’s University
Kingston, Ontario, Canada
Corresponding Author:
Ian Janssen, PhD
School of Kinesiology and Health Studies
Queen’s University
Kingston, Ontario, Canada, K7L 3N6
Phone: (613) 533-6000 ext 78631
Fax: (613) 533-2009
Email: [email protected]
Funding Sources: National Heart, Lung, and Blood Institute; Canadian Institutes of Health
Research
Running Head: Physical Activity and Age-Related Weight Loss
28
ABSTRACT
OBJECTIVES: To determine whether physical activity attenuates age-related weight loss in the
elderly.
DESIGN: Observational longitudinal study.
SETTING: Four U.S. communities.
PARTICIPANTS: 4512 community-dwelling older (≥ 65 yr) men and women from the
Cardiovascular Health Study.
MEASUREMENTS: Physical activity (PA) energy expenditure was determined from a
questionnaire at baseline and subjects were divided into sex-specific PA quartiles. Weight was
measured at baseline and annually over the 8 years of follow-up; subjects had a minimum of 2
and a maximum of 9 weight measures. The influence of PA on longitudinal changes in body
weight was examined using mixed models while adjusting for lifestyle variables,
sociodemographic characteristics, and disease status.
RESULTS: Body weight declined in a curvilinear manner over time with accelerated weight loss
occurring in the final years of observation. Over the 8 yr follow-up period, the least active PA
quartile lost 2.72 kg. Weight loss was attenuated by 0.55 kg (20%, P=0.057), 0.80 kg (29%,
P=0.05), and 0.69 kg (25%, P=0.016) within the second through fourth PA quartiles. The effects
of physical activity did not differ by gender, but increased with advancing age.
CONCLUSION: Participation in modest amounts of physical activity attenuated age related
weight loss by approximately 25% with little additional benefit observed with higher levels of
physical activity. The attenuation of age-related weight loss in the elderly adds to an ever
expanding number of health outcomes that are known to be positively impacted by physical
activity.
KEY WORDS: aging; body weight; leisure time activity; longitudinal study
29
INTRODUCTION
Cross-sectional (1, 2) and longitudinal (3-7) studies have consistently illustrated an increase
in body weight throughout early and middle adulthood until around the seventh decade of life, at
which point the weight trajectory begins to decline. This weight loss is typically unintentional and
is often a marker of undetected or undiagnosed disease (8) and not a sign of improved health.
Weight loss in elderly persons is associated with a reduced health-related quality of life (9) as
well as an increased risk of physical disability (10) and mortality (11, 12) regardless of weight
loss intention (13). Thus, it is important to explore potential strategies for minimizing weight loss
in the elderly.
Resistance exercise is an effective method to help maintain muscle mass in the elderly (14)
and therefore may have similar effects on overall body mass. However, resistance exercise is not
commonly engaged in as less than 15% of American elders perform muscle-strengthening
activities at least twice a week (15). The promotion of general physical activity, including
walking and other low intensity aerobic activities, may offer a reasonable alternative to resistance
training. To our knowledge, only one study has examined the impact of physical activity on the
weight trajectory of elderly persons. Over 12 years of follow-up, Dziura et al. (16) reported that
for every unit increase in a 9 point physical activity scale, weight loss was attenuated by 0.04 kg
per year in a cohort of 2300 elderly adults. Although informative, this study was limited in that it
employed crude measures of both the physical activity exposure (e.g., a physical activity index
based on only four activities) and the weight loss outcome (e.g., self-reported weight).
Furthermore, this study did not consider incident disease and its potential effect on the weight
loss trajectory. Therefore, further research is necessary to confirm these initial findings.
The purpose of this study was to determine whether physical activity attenuated age-related
weight loss in a cohort of elderly adults (≥65 yr). It was hypothesized that weight would decline
during the follow-up period and that physical activity would minimize, but not prevent, weight
loss over time.
30
METHODS
Study Sample
Subjects were sampled from the Cardiovascular Health Study (CHS), a population-based
longitudinal study of cardiovascular disease, as described in detail elsewhere (17). Briefly, 5201
men and women were sampled from Medicare eligibility lists in Washington County, MD;
Sacramento County, CA; Forsyth County, NC; and Pittsburgh, PA. Eligible participants were 65
yr and older, noninstitutionalized, expected to remain in the area for the next three years, able to
give informed consent, and did not require a proxy respondent at baseline. Of those eligible, 57%
enrolled in the CHS. The institutional review boards approved the project at each study site, and
written informed consent was obtained from all subjects. Additional ethics approval was granted
by the Research Ethics Board at Queen’s University (Appendix B).
The baseline examination was conducted between June 1989 and June 1990, and the CHS
cohort was followed annually for 8 yr after baseline. All examinations consisted of a home or
telephone interview and standardized clinical examination, as explained in detail elsewhere (17).
Briefly, in the interview, information was obtained on demographics, medical history, and
physical activity while the clinical examination included anthropometric (height, weight, etc.) and
other measures. A total of 613 individuals were sequentially excluded from the sample for not
reporting or not having measures of the study exposures, outcomes, or covariates which consisted
of baseline measures of physical activity (n=5), weight (n=15), height (n=1), alcohol consumption
(n=52), smoking status (n=221), diabetes status (n=30), hypertension status (n=14), cognitive
function (n=5), lung disease (n=74), or at least one follow-up measure of body weight (n=196).
An additional 76 individuals were excluded because they did not grant permission to have their
data used in the public data files. Therefore, the study described here was limited to 4512 (87%)
of the CHS participants.
The CHS was conducted and supported by the National Heart, Lung, and Blood Institute
(NHLBI) in collaboration with the CHS investigators. The NHLBI and CHS investigators have
31
created public access data sets that are available, on appropriate terms and conditions, to qualified
investigators. The present study is based on the public access data.
Physical Activity (exposure variable)
During the baseline examination, physical activity was assessed using a modified version
of established questionnaires (18, 19) (Appendix C). Participants were asked whether they had
engaged in 11 common leisure time activities, excluding chores or work, in the previous two
weeks (20), including walking, hiking, jogging, biking, exercise cycling, dancing, aerobics,
bowling, golfing, calisthenics, and swimming. Subjects also listed their participation in up to two
additional activities not included in this list. Each activity was assigned a per-minute caloric
expenditure value, which was summed over all minutes of activity over the week, resulting in a
weekly energy expenditure value for each subject (kcal·week-1) (18, 20). Initially subjects were
divided into sex-specific quartiles based on their energy expenditure values. In subsequent
analyses the top 3 quartiles were combined as the ‘Active’ group due to similarities in the weight
loss trajectory in these quartiles, while the bottom quartile remained as the ‘Inactive’ referent
group.
Body Weight (outcome variable)
Body weight was measured at baseline and annually over the 8 yr follow-up to the nearest
0.5 pound. As such, subjects could have a minimum of 2 weight measures (baseline + at least one
follow-up) to as many as 9 measures. For the purposes of this study, weight is expressed in
kilograms (1 kg = 2.2 lbs). Clinically significant weight loss was defined as ≥5% weight loss over
any one year interval of the follow-up period or ≥10% loss of baseline body weight during the
cumulative follow-up period (21). Separate regression models were used for each type of weight
loss in which follow-up time for significant weight losers was the number of days from baseline
until the first visit in which they lost ≥5% body weight from the previous year or the number of
days to the first visit at which they lost ≥10% of their baseline body weight. Subjects who did not
32
lose a significant amount of weight were followed for up to 8 years after their baseline
examination or until death if this occurred before 8 years.
Covariates
Several baseline covariates were included in the analyses. These consisted of age, sex,
race, annual household income, smoking, alcohol consumption, and several prevalent comorbid
conditions. A more complete description and categorization of these covariates is provided in
Table 1. Incident cases of diabetes, cancer, myocardial infarction, stroke, and congestive heart
failure that occurred over the follow-up period were also included. These conditions were
considered absent until the first visit in which a subject had a given disease after which point the
subject was coded positive for the condition.
Statistical Analyses
SAS software version 9.1 (SAS Institute, Inc., Cary, NC) was used for all data
management and analyses. Univariate statistics were calculated for the main study variables and
covariates. χ2 analysis was used to determine differences in baseline characteristics between the
Inactive (Quartile 1) and Active (Quartiles 2–4) physical activity groups (Appendix D). To
address the primary study question, we used unstructured multivariable mixed models to model
weight trajectories over 8 years according to baseline physical activity level using the PROC
MIXED procedure (Appendix E). In the subgroup analysis pertaining to survival status, subjects
were followed for up to 7 years since those who died during follow-up did not have weight
measures in the final year of observation.
Appropriate model selection was determined by fitting a variety of covariance structures
and selecting the one that minimized the Bayesian information criteria (BIC) (22). The fullyadjusted mixed models accounted for baseline age, sex, race, height, annual household income,
smoking, alcohol consumption, the number of comorbid conditions at baseline (lung disease,
cognitive impairment, hypertension, diabetes, stroke, congestive heart failure, and myocardial
infarction), and incident disease (cancer, diabetes, myocardial infarction, stroke, and congestive
33
heart failure) to account for their effect on the intercept. Also included in the models were visit
number, the interaction of all covariates with visit number, and visit number2 to account for the
covariates’ effect on the slope of the regression line. Subgroup analyses were performed to
examine the effect of physical activity on weight change over time according to sex, age category,
number of comorbid conditions at baseline, and survival status during follow-up.
In the final set of analyses, Cox proportional hazards regression was used to determine
hazard ratios for clinically significant weight loss according to baseline physical activity level
(Appendix F). These models included all baseline covariates and incident disease variables used
in the mixed models as well as the number of visits attended by each subject.
34
RESULTS
Subject Characteristics
Descriptive information of the 4512 participants at baseline is shown in Table 1. The
mean follow-up time for the final weight measure was 7.0 ± 1.8 years, with an average weight
loss over the follow-up period of 1.39 ± 5.82 kg (range: 32.4 kg weight loss to 29.8 kg weight
gain), representing an average weight loss of -1.8 ± 8.0% of body weight (range: loss of 39% to
gain of 52%). A total of 707 subjects lost ≥10% of their baseline body weight during the followup, while 1771 people lost ≥5% of their body weight over any given year during the follow-up.
These groups were not mutually exclusive. A total of 1050 subjects died during the follow-up
period.
The mean physical activity energy expenditure at baseline was 1258 kcal·week-1. Energy
expenditure values were positively skewed with values of 158 kcal·week-1, 692 kcal·week-1, and
1647 kcal·week-1 corresponding to the 25th, 50th, and 75th percentiles, respectively. Of the 4512
participants, 772 (17.1%) had a physical activity energy expenditure value of 0 kcal·week-1.
When examined by sex, women had a mean energy expenditure value of 903 kcal·week-1, with 60
kcal·week-1, 480 kcal·week-1, and 1215 kcal·week-1 reflecting the 25th, 50th, and 75th percentiles.
The corresponding values in men were 1730 kcal·week-1, 405 kcal·week-1, 1080 kcal·week-1, and
2430 kcal·week-1.
Physical Activity and Weight Trajectories
The results outlined in this section were obtained from projected weight curves that were
created based on the coefficients obtained from the mixed modeling procedure in SAS. The
weight curves were corrected for the average of the participants’ risk factors by taking the sum of
the coefficients for each risk factor category after they were multiplied by the proportion of the
study population who fit within that category (Appendix G). The median number of weights
available per person was 8, with 48.4% of the study sample having weights available for all 9
measures.
35
All Subjects. As shown in Figure 1, subjects in physical activity quartiles 2 – 4 weighed
significantly less than the least active quartile at baseline (P ≤.002). Body weight declined in a
curvilinear manner over the 8 yr follow-up in all physical activity groups with accelerated weight
loss occurring in the final years of observation. Compared to the least active quartile, the rate of
weight loss was not significantly slower in the second quartile (P=.057) but was in the third
(P=.005) and fourth (P=.016) quartiles. Over the 8 yr follow-up period, the least active quartile
lost a total of 2.72 kg (3.7%). Weight loss was attenuated by 0.55 kg (20%), 0.80 kg (29%), and
0.69 kg (25%) within the second, third, and fourth quartiles, respectively, compared with weight
loss in the first quartile. This corresponds to a weight loss attenuation of approximately 0.09 kg
per year in the most active quartile compared with the least active quartile.
Subsequent mixed model analyses were performed in a variety of population subgroups
to consider the modifying effects of sex, age, comorbidity, and survival status on the relation
between physical activity and weight loss. Given the similar weight loss trajectories that were
observed, the upper 3 physical activity quartiles were collapsed to create an ‘Active’ group while
the first quartile remained as the ‘Inactive’ referent group for these analyses.
Sex. As illustrated in Figure 2 (Panel A), men weighed more than women at baseline and
experienced a slightly faster weight loss trajectory within both physical activity groups. Over the
follow-up period the Inactive men lost 3.07 kg, and this weight loss was attenuated by 0.77 kg
(25%) in the Active men (P=.042) (Table 2). Inactive women lost 2.55 kg and weight loss was
attenuated by 0.68 kg (27%) in the Active women (P=.027). The difference in the weight loss
trajectories between the Active and Inactive groups was not modified by sex (P=.156).
Age. The impact of age on the relation between physical activity and weight loss is
summarized in Table 2 and depicted in Figure 2 (Panel B). Older subjects weighed less than
younger subjects at baseline and experienced a more rapid weight decline independent of physical
activity status. Compared to the Inactive group, the rate of weight loss was significantly slower in
the Active group for both the youngest (P=.020) and oldest (P=.005) age groups, but this was not
36
the case in the 73-80 year old group (P=.160). The differences in the weight loss trajectories
between the Active and Inactive groups were modified by age (P=.002) such that the differences
in weight loss between the Inactive and Active groups were 0.67 kg (43%), 0.56 kg (15%), and
2.39 kg (38%) in the youngest, middle, and oldest age groups, respectively.
Baseline Comorbid Conditions. Subjects with ≥2 comorbid conditions at baseline
weighed the most at baseline while those with none weighed the least (Figure 2, Panel C).
Subjects with no comorbid conditions maintained a fairly stable weight over the follow-up, while
the group with ≥2 experienced the most rapid decline, resulting in the greatest weight loss (Table
2). Active subjects weighed less than Inactive subjects at baseline irrespective of the number of
comorbid conditions (P ≤.003). However, the rate of weight loss between physical activity groups
did not differ with the exception of those with 1 comorbid condition in which the Inactive group
lost 2.99 kg compared to 1.51 kg (49% attenuation) within the Active group (P<.001).
Survival Status. For the survival status analysis, weight trajectories were only modeled up
to visit 8 since the subjects who died during the follow-up period did not have weight measures
for visit 9. Figure 2 (Panel D) illustrates that those who died during follow-up weighed less at
baseline and experienced a more dramatic weight loss trajectory than those who survived,
irrespective of physical activity level. Compared to their Inactive counterparts, Active survivors
weighed significantly less than the Inactive group at baseline (P<.001) and experienced a 35%
slower rate of weight loss (P=.016, Table 2). Within non-survivors, baseline weight and the rate
of weight loss during the follow-up period were not statistically different between the physical
activity groups (P>.1, Table 2). However, inspection of Table 2 shows that, although not
statistically significant, the magnitude of the difference in weight loss between the Active and
Inactive non-survivors was larger than that observed in the survivors (1.08 vs. 0.54 kg).
Physical Activity and Risk of Clinically Significant Weight Loss
All Subjects. Compared to the lowest physical activity quartile, the hazard ratios (95%
confidence intervals) for having a weight loss ≥5% between any 2 consecutive annual visits were
37
1.04 (0.91–1.19), 0.98 (0.86–1.12), and 0.96 (0.84–1.10) in the second, third, and fourth quartiles,
respectively. The corresponding values for a total weight loss of ≥10% of baseline body weight
were 0.91 (0.74–1.11), 0.85 (0.69–1.05), and 0.86 (0.70–1.07). Thus, physical activity was not a
significant predictor of clinically significant weight loss. Hazard ratios were adjusted for all
baseline covariates listed in Table 1, as well as height, incident cancer, diabetes, myocardial
infarction, stroke, and congestive heart failure, and the number of weight measures taken over the
8 year follow-up period.
Moderating Effects of Sex, Age, Comorbidity, and Survival Status. As indicated in Table
3, within the different population subgroups the risks of having ≥5% weight loss in any given
year or ≥10% weight loss from baseline were not different in the Active and Inactive groups
(P>.1) with two exceptions. First, the Active subjects with 1 comorbid condition at baseline
significantly reduced the risk of having a ≥10% weight loss by 32% compared to their Inactive
counterparts (Table 3). Second, the Active subjects aged ≥81 years reduced the risk of having a
≥5% weight loss in 1 year by 26%, although this finding was not statistically significant (P=.071).
38
DISCUSSION
While an important goal of physical activity participation in young and middle-aged
adults is often to prevent weight gain, the goal of physical activity within older adults may be
better directed towards minimizing weight loss. As confirmed in this large prospective study of
adults aged 65 yr and older, body weight tends to decline over time in most elderly persons. We
were able to demonstrate that participation in modest amounts of physical activity attenuated this
age-related weight loss by approximately 25%, with little additional benefit observed with higher
levels of physical activity. The attenuation of weight loss in the elderly adds to an ever expanding
number of health outcomes that are known to be positively impacted by physical activity, further
demonstrating the importance of this lifestyle behaviour.
Previous research in middle-aged adults has highlighted the effectiveness of physical
activity in preventing weight gain over time (23, 24). For example, Di Pietro et al. (24) observed
an inverse association between daily physical activity level and weight gain over 5 years in men
aged 20 – 55 yr from the Aerobics Center Longitudinal Study. However, the findings of the
present study suggest that the relationship between physical activity and body weight may be
different in the elderly, a population in which weight loss is not usually considered a positive
outcome.
In the CHS cohort as a whole, the least active quartile lost 0.34 kg per year while the
second, third, and fourth quartiles experienced weight loss that was 0.07, 0.10, and 0.09 kg per
year less than the least active quartile (21-29% attenuation of the total weight loss). This finding
is consistent with, but less substantial than, that reported by Dziura and colleagues (16). In a
similar repeated measures study conducted over 12 years, these authors found that weight loss
was attenuated by 0.16 kg per year in subjects who scored 4 on a 9 point physical activity scale
compared to subjects who were completely inactive. In that study, a score of 4 could be obtained
by participating in 2 activities such as walking and gardening/housework “often in the past
month” (16). Due to several methodological differences between the present study and that of
39
Dziura et al., it is difficult to postulate as to why the beneficial effects of physical activity were
less pronounced here.
Because Inactive subjects in the present study lost more weight than their Active
counterparts, the differences in the weight loss trajectories in these two groups was clearly not a
function of energy expenditure. Thus, it seems logical that the greater weight loss in Inactive
subjects was driven by insufficient energy intake. There are a variety of age-associated
physiologic changes that can affect energy intake and predispose the elderly to weight loss
including a decline in smell and taste (25), slowed gastric emptying (26), reduced chewing
efficiency (27), and alterations in the neuroendocrine system (28, 29), which are together
associated with early satiety and a decline in appetite. Because the relation between physical
activity and these energy intake determinants has been poorly studied in the elderly, the exact
physiological mechanism(s) that explains our findings is unknown. However, a study conducted
in the Netherlands found that a modest 17 week physical activity intervention in elderly persons
(e.g., walking, stooping, and chair stands for 45 minutes performed twice a week) resulted in a
small increase in energy intake (120 kcal, P =.05) compared with non-exercisers (30). The effect
of physical activity on energy intake in that study was independent of a change in appetite or in
sensory perception (30). Thus, limited evidence suggests that physical activity may somehow
assist with maintenance of energy intake at a level that helps to prevent weight loss, although the
mechanism behind this effect is unknown.
Several potential effect modifiers were examined in the present study to determine if the
effects of physical activity on age-related weight loss differed according to gender, age, disease
status, and survival status. Apart from gender, all of these variables impacted the effect of
physical activity. Regarding age, the beneficial effects of physical activity were most pronounced
in the oldest subjects, perhaps because that is where the greatest weight loss occurred. Similarly,
and consistent with earlier findings (16), physical activity was most beneficial in subjects with at
least 1 comorbid condition; comorbidity was also associated with accelerated weight loss.
40
Likewise, weight loss was greater in non-survivors than survivors as previously reported (8, 1113, 31-33), and physical activity had an approximately twofold greater effect on attenuating
absolute weight loss in non-survivors compared to those who survived (1.08 kg vs. 0.54 kg).
Thus, the effects of physical activity on weight loss attenuation were generally more pronounced
in the subgroups of subjects who experienced an accelerated weight loss.
In this study, physical activity attenuated weight loss to the greatest extent within the
older subjects. Within the older age group, Active individuals lost 2.39 kg less than the Inactive
individuals over the 8 year follow-up. This equates to preservation of 3.3% of total body weight,
or ~0.4% of body weight per year. To put these values into perspective, within elderly persons a
weight loss of >5% over 5 years has been associated with an increased risk of physical disability
(10), a weight loss of >2.25 kg over 4 years as been associated with an impaired quality of life in
normal weight and underweight elders (9), and a weight loss of 5% over 3 years has been
associated with an increased mortality risk (11). Therefore, although the general effect of
attenuating age-related weight loss appears promising and reached statistical significance in the
present study, the magnitude of effect induced by physical activity may not have been sufficient
enough to achieve clinical relevance. However, given that weight loss tends to continue
progressively throughout old age, and that we only followed subjects over 8 years, it could be
argued that any slowing of weight loss in the elderly is a positive outcome. Furthermore, it is
possible that more specific and vigorous types of physical activity, such as resistance training,
would have a greater effectiveness than the lower intensity activities of an aerobic nature that
tended to be performed by the participants of this study.
Notable strengths of the present study include the longitudinal study design, measured
body weights, and the mixed models used to predict weight trajectories over time. Specifically,
mixed modeling makes use of repeated measures while allowing for the inclusion of subjects with
some missing body weight measures and time-varying covariates (e.g., incident disease).
However, this study is not without its limitations. Energy expenditure was estimated based on
41
self-reported physical activity at baseline only; therefore we were unable to detect changes in
exposure status over the follow-up period. Although physical activity questionnaires are a
repeatable and valid tool for elderly persons (34), some degree of imprecision was surely
introduced with this measure. Furthermore, we only considered energy expenditure, and the
components of physical activity (e.g., intensity and duration) were not assessed individually.
Finally, we do not have a direct measure of weight loss intention; although some studies have
found that weight loss is detrimental in the elderly regardless of intention (11, 13, 35).
Physical activity modestly attenuates age-related weight loss in the elderly. At this time,
the physiological mechanism that explains these findings and its potential health implications is
unclear. Future studies in this area should consider employing objective measures of physical
activity (e.g., accelerometers) to determine if a stronger measure of the exposure affects the
strength of the relationship. In addition, the impact of different types and intensities of physical
activity on its relation with body weight changes in the elderly warrants investigation.
42
REFERENCES
1.
Chumlea WC, Guo SS, Kuczmarski RJ et al. Body composition estimates from NHANES
III bioelectrical impedance data. Int J Obes Relat Metab Disord 2002;26:1596-1609.
2.
Kyle UG, Genton L, Hans D et al. Total body mass, fat mass, fat-free mass, and skeletal
muscle in older people: cross-sectional differences in 60-year-old persons. J Am Geriatr
Soc 2001;49:1633-1640.
3.
Guo SS, Zeller C, Chumlea WC, Siervogel RM. Aging, body composition, and lifestyle:
the Fels Longitudinal Study. Am J Clin Nutr 1999;70:405-411.
4.
Lewis CE, Smith DE, Wallace DD, Williams OD, Bild DE, Jacobs DR, Jr. Seven-year
trends in body weight and associations with lifestyle and behavioral characteristics in
black and white young adults: the CARDIA study. Am J Public Health 1997;87:635-642.
5.
Visser M, Pahor M, Tylavsky F et al. One- and two-year change in body composition as
measured by DXA in a population-based cohort of older men and women. J Appl Physiol
2003;94:2368-2374.
6.
Carmelli D, McElroy MR, Rosenman RH. Longitudinal changes in fat distribution in the
Western Collaborative Group Study: a 23-year follow-up. Int J Obes 1991;15:67-74.
7.
Shimokata H, Andres R, Coon PJ, Elahi D, Muller DC, Tobin JD. Studies in the
distribution of body fat. II. Longitudinal effects of change in weight. Int J Obes
1989;13:455-464.
8.
Wannamethee SG, Shaper AG, Lennon L. Reasons for intentional weight loss,
unintentional weight loss, and mortality in older men. Arch Intern Med 2005;165:10351040.
9.
Fine JT, Colditz GA, Coakley EH et al. A prospective study of weight change and healthrelated quality of life in women. JAMA 1999;282:2136-2142.
43
10.
Launer LJ, Harris T, Rumpel C, Madans J. Body mass index, weight change, and risk of
mobility disability in middle-aged and older women. The epidemiologic follow-up study
of NHANES I. JAMA 1994;271:1093-1098.
11.
Newman AB, Yanez D, Harris T, Duxbury A, Enright PL, Fried LP. Weight change in
old age and its association with mortality. J Am Geriatr Soc 2001;49:1309-1318.
12.
Losonczy KG, Harris TB, Cornoni-Huntley J et al. Does weight loss from middle age to
old age explain the inverse weight mortality relation in old age? Am J Epidemiol
1995;141:312-321.
13.
Wallace JI, Schwartz RS, LaCroix AZ, Uhlmann RF, Pearlman RA. Involuntary weight
loss in older outpatients: incidence and clinical significance. J Am Geriatr Soc
1995;43:329-337.
14.
Hunter GR, McCarthy JP, Bamman MM. Effects of resistance training on older adults.
Sports Med 2004;34:329-348.
15.
Trends in Strength Training--United States, 1998-2004. JAMA 2006;296:1459-1460.
16.
Dziura J, Mendes de Leon C, Kasl S, DiPietro L. Can physical activity attenuate agingrelated weight loss in older people? The Yale Health and Aging Study, 1982-1994. Am J
Epidemiol 2004;159:759-767.
17.
Fried LP, Borhani NO, Enright P et al. The Cardiovascular Health Study: design and
rationale. Ann Epidemiol 1991;1:263-276.
18.
Taylor HL, Jacobs DR, Jr., Schucker B, Knudsen J, Leon AS, Debacker G. A
questionnaire for the assessment of leisure time physical activities. J Chronic Dis
1978;31:741-755.
19.
Paffenbarger RS, Jr., Wing AL, Hyde RT. Physical activity as an index of heart attack
risk in college alumni. Am J Epidemiol 1978;108:161-175.
44
20.
Siscovick DS, Fried L, Mittelmark M, Rutan G, Bild D, O'Leary DH. Exercise intensity
and subclinical cardiovascular disease in the elderly. The Cardiovascular Health Study.
Am J Epidemiol 1997;145:977-986.
21.
Alibhai SM, Greenwood C, Payette H. An approach to the management of unintentional
weight loss in elderly people. CMAJ 2005;172:773-780.
22.
Littell RC, Milliken GA, Stroup WW, Wolfinger RD, Schabenberger O. SAS for mixed
models. SAS Institute, Inc.: Cary, N.C., 2006.
23.
Littman AJ, Kristal AR, White E. Effects of physical activity intensity, frequency, and
activity type on 10-y weight change in middle-aged men and women. Int J Obes (Lond)
2005;29:524-533.
24.
Di Pietro L, Dziura J, Blair SN. Estimated change in physical activity level (PAL) and
prediction of 5-year weight change in men: the Aerobics Center Longitudinal Study. Int J
Obes Relat Metab Disord 2004;28:1541-1547.
25.
Schiffman SS. Taste and smell in disease (first of two parts). N Engl J Med
1983;308:1275-1279.
26.
Clarkston WK, Pantano MM, Morley JE, Horowitz M, Littlefield JM, Burton FR.
Evidence for the anorexia of aging: gastrointestinal transit and hunger in healthy elderly
vs. young adults. Am J Physiol 1997;272:R243-248.
27.
Ship JA, Duffy V, Jones JA, Langmore S. Geriatric oral health and its impact on eating. J
Am Geriatr Soc 1996;44:456-464.
28.
Morley JE. Decreased food intake with aging. J Gerontol A Biol Sci Med Sci 2001;56:8188.
29.
Wilson MM, Morley JE. Invited review: Aging and energy balance. J Appl Physiol
2003;95:1728-1736.
45
30.
de Jong N, Chin APMJ, de Graaf C, van Staveren WA. Effect of dietary supplements and
physical exercise on sensory perception, appetite, dietary intake and body weight in frail
elderly subjects. Br J Nutr 2000;83:605-613.
31.
Dey DK, Rothenberg E, Sundh V, Bosaeus I, Steen B. Body mass index, weight change
and mortality in the elderly. A 15 y longitudinal population study of 70 y olds. Eur J Clin
Nutr 2001;55:482-492.
32.
Corrada MM, Kawas CH, Mozaffar F, Paganini-Hill A. Association of body mass index
and weight change with all-cause mortality in the elderly. Am J Epidemiol 2006;163:938949.
33.
Reynolds MW, Fredman L, Langenberg P, Magaziner J. Weight, weight change,
mortality in a random sample of older community-dwelling women. J Am Geriatr Soc
1999;47:1409-1414.
34.
Harada ND, Chiu V, King AC, Stewart AL. An evaluation of three self-report physical
activity instruments for older adults. Med Sci Sports Exerc 2001;33:962-970.
35.
Yaari S, Goldbourt U. Voluntary and involuntary weight loss: associations with long term
mortality in 9,228 middle-aged and elderly men. Am J Epidemiol 1998;148:546-555.
36.
Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for
grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-198.
37.
Chobanian AV, Bakris GL, Black HR et al. The Seventh Report of the Joint National
Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure:
the JNC 7 report. JAMA 2003;289:2560-2572.
38.
American Diabetes Association. Screening for type 2 diabetes. Diabetes Care
2003;26:21-24.
46
Table 1. Prevalence of Baseline Covariates among Cardiovascular Health Study Participants
Prevalence (%)
Variable
All Subjects Physically Inactive Physically Active
(n = 4512)
(n = 1129)
(n = 3383)
Male
42.9
42.9
43.0
Female
57.1
57.1
57.0
65 – 70
43.0
36.8
45.1††
71 – 76
33.7
33.0
34.0
77 – 82
17.5
21.4
16.1††
≥ 83
5.8
8.8
4.8††
White
95.1
92.3
96.0††
Other
4.9
7.7
4.0††
≤ $7,999
10.4
13.8
9.2††
$8,000-15,999
25.3
26.1
25.0
$16,000-34,999
34.6
32.3
35.3
$35,000-49,999
9.9
10.1
9.9
≥ $50,000
13.4
11.1
14.2**
Not Reported
6.4
6.6
6.4
None
43.8
42.3
44.2
Passive (lived with regular smoker)
4.3
5.0
4.1
Light (1 – 13 pack-years)
12.9
10.6
13.7**
Moderate (14 – 50 pack-years)
26.3
26.8
26.1
Heavy (> 50 pack-years)
12.7
15.3
11.9**
Less than 1 drink/week
67.2
71.7
65.7††
1 – 7 drinks/week
19.6
17.6
20.3
> 7 drinks/week
13.2
10.7
14.0**
Sex
Age (years)
Race
Socioeconomic status (annual income)
Smoking status
Alcohol consumption
Comorbid conditions
47
Lung disease*
24.3
26.0
23.7
Cognitive impairment†
5.0
7.2
4.3††
Hypertension‡
63.7
67.8
62.4**
Diabetes§
14.9
18.8
13.5††
Stroke║
3.4
4.9
2.9**
Congestive heart failure║
3.8
6.0
3.0††
Myocardial infarction║
9.5
11.3
8.9 #
0
23.1
18.6
24.6 ††
1
42.3
39.3
43.3 #
≥2
34.6
42.1
32.1††
Number of comorbid conditions¶
*
Self-report of chronic bronchitis or emphysema.
†
Based on Mini Mental State questionnaire (36).
‡
Based on the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of
High Blood Pressure guidelines (37).
§
Based on American Diabetes Association Guidelines (38).
║
Cardiovascular diseases were diagnosed by self-reports that were validated by ascertaining
medications, reviewing medical records, and/or standardized examinations (17).
¶
Includes baseline lung disease, cognitive impairment, hypertension, diabetes, stroke, congestive
heart failure, and myocardial infarction.
#
Significantly different from Inactive group based on chi-square test (P<.05).
**
Significantly different from Inactive group based on chi-square test (P<.01).
††
Significantly different from Inactive group based on chi-square test (P<.001).
48
Table 2. Projected Absolute (kg) and Relative (%) Weight Loss between the First and Last* Visits in Physically Inactive (Quartile 1) and Active
(Quartiles 2-4) Groups
Absolute Weight Loss (kg)
Group
Relative Weight Loss (%)
Inactive
Active
Difference
(Inactive – Active)
Inactive
Active
Difference
(Inactive – Active)
Males (n=1937)
3.07
2.30
0.77‡
4.16
3.20
0.96‡
Females (n=2575)
2.55
1.87
0.68‡
3.50
2.65
0.85‡
65 – 72 (n=2550 )
1.56
0.83
0.67‡
2.09
1.14
0.95‡
73 – 80 (n=1528)
3.86
3.30
0.56
5.32
4.68
0.64
6.34
3.95
§
2.39
9.14
5.84
3.30§
0.99
1.02
-0.03
1.40
1.49
-0.09
║
Sex
Age (years)
≥ 81 (n=434)
Number of comorbid conditions
0 (n=1043)
1 (n=1907)
2.99
1.51
1.48
4.09
2.11
1.98║
≥ 2 (n=1562)
3.50
3.37
0.13
4.66
4.60
0.06
Lived (n=3462)
1.54
1.00
0.54‡
2.08
1.40
0.68‡
Died (n=1050)
6.56
5.48
1.08
9.17
7.77
1.40
Survival Status†
*
Weight loss calculated between first and ninth visit.
Weight loss calculated between first and eighth visit.
‡
Rate of weight loss significantly attenuated in Active group (P<.05).
§
Rate of weight loss significantly attenuated in Active group (P<.01).
║
Rate of weight loss significantly attenuated in Active group (P<.001).
†
Table 3. Hazard Ratios for Clinically Significant Weight Loss According to Sex, Age, Number of
Comorbid Conditions, and Survival Status
Population Subgroup
≥ 5% Weight Loss
≥ 10% Weight Loss
All subjects (n = 4512)
0.99 (0.89 – 1.11)
0.88 (0.74 – 1.03)
Males (n = 1937)
0.91 (0.77 – 1.08)
0.85 (0.66 – 1.11)
Females (n = 2575)
1.05 (0.91 – 1.21)
0.88 (0.71 – 1.10)
65 – 72 (n = 2550 )
0.99 (0.85 – 1.16)
0.81 (0.63 – 1.04)
73 – 80 (n = 1528)
1.06 (0.89 – 1.28)
0.98 (0.76 – 1.28)
≥ 81 (n = 434)
0.74 (0.53 – 1.03)
0.81 (0.51 – 1.29)
0 (n = 1043)
0.83 (0.64 – 1.07)
0.96 (0.62 – 1.49)
1 (n = 1907)
0.89 (0.75 – 1.05)
0.68 (0.53 – 0.89)*
≥2 (n = 1562)
1.13 (0.95 – 1.34)
0.98 (0.77 – 1.25)
Lived (n = 3462)
0.99 (0.87 – 1.13)
0.87 (0.71 – 1.07)
Died (n = 1050)
1.01 (0.81 – 1.25)
0.92 (0.69 – 1.21)
Sex
Age (years)
Number of comorbid conditions
Survival Status
Data presented as hazard ratio (95% confidence intervals). The Inactive subjects served as the
referent group (hazard ratio=1.00) for all analyses and the risk estimates are provided for the
Active subjects. Where appropriate, hazard ratios were adjusted for sex, age, race, height,
household income, smoking, alcohol, baseline disease, incident diseases, and the number of
weight measures taken over the 8 year follow-up period.
* Significantly reduced risk compared to Inactive (P<.01).
50
FIGURE LEGENDS
Figure 1.
Weight loss over 8 years of follow-up based on sex-specific physical activity
quartiles. The plots represent regression lines that were generated using the
coefficients obtained from an unstructured mixed model that included the
following independent variables: age, sex, race, height, household income,
smoking, alcohol, number of comorbid conditions at baseline, incident diseases,
the interaction of each of the previous variables with visit number, visit number,
and visit number2.
Figure 2.
Weight loss over 8 years of follow-up with physically Inactive (quartile 1) and
Active (quartiles 2-4) groups according to sex (Panel A), baseline age (Panel B),
number of comorbid conditions at baseline (Panel C), and survival status during
follow-up (Panel D). The plots represent regression lines that were generated
using the coefficients obtained from an unstructured mixed model that included
the following independent variables: age, sex, race, height, annual household
income, smoking, alcohol, number of comorbid conditions at baseline, incident
diseases, the interaction of each of the previous variables with visit number, visit
number, and visit number2.
51
FIGURE 1
Quartile 1 (Least Active)
Quartile 2
Quartile 3
Quartile 4 (Most Active)
74
73
Weight (kg)
72
71
70
69
68
67
66
1
2
3
4
5
Visit
52
6
7
8
9
A
C
Weight (kg)
Weight (kg)
64
66
68
70
72
74
76
64
66
68
70
72
74
76
1
1
2
2
3
3
5
4
Visit
5
≥ 2 Diseases
0 Diseases
1 Disease
Visit
6
6
7
Active
Inactive
7
8
8
Active
Women
4
Inactive
Men
9
9
FIGURE 2
B
D
Weight (kg)
Weight (kg)
53
60
62
64
66
68
70
72
74
76
60
62
64
66
68
70
72
74
76
1
1
2
2
3
3
5
4
Visit
6
7
Active
7
Inactive
6
Died
5
Visit
Active
Inactive
Lived
4
≥ 81 yr
73 – 80 yr
65 – 72 yr
8
8
9
9
CHAPTER 4
MANUSCRIPT 2
Sarcopenic-Obesity and Cardiovascular Disease Risk in the Elderly
This manuscript has been submitted to Obesity and is presented as requested by the journal.
Wendy C. Stephen1 and Ian Janssen1, 2
1
2
School of Kinesiology and Health Studies, and
Department of Community Health and Epidemiology
Queen’s University
Kingston, Ontario, Canada
Running Head: Sarcopenic-obesity and cardiovascular disease
Corresponding Author:
Ian Janssen
School of Kinesiology and Health Studies
Queen’s University
Kingston, Ontario, Canada, K7L 3N6
Phone: (613)533-6000 ext 78631
Fax: (613)533-2009
Email: [email protected]
54
ABSTRACT
The study objectives were to determine: 1) whether sarcopenic-obesity is a stronger
predictor of cardiovascular disease (CVD) than either sarcopenia or obesity alone in the elderly,
and 2) whether muscle mass or muscular strength is a stronger marker of CVD risk. Participants
included 3400 community-dwelling older (≥65 years) men and women who were free of CVD at
baseline. Waist circumference (WC), bioimpedance analysis, and grip strength were used to
measure abdominal obesity, whole-body muscle mass and muscular strength, respectively.
Subjects were classified as normal, sarcopenic, obese, or sarcopenic-obese based on measures of
WC and either muscle mass or strength. Participants were followed for 8 years for CVD
development and proportional hazard regression models were used to compare risk estimates for
CVD in the four groups after adjusting for age, sex, race, income, smoking, alcohol, and cognitive
status. Compared with the normal group, CVD risk was not significantly elevated within the
obese, sarcopenic, or sarcopenic-obese groups as determined from measures of WC and muscle
mass. When determined from measures of WC and muscle strength, CVD risk was not
significantly increased in the sarcopenic or obese groups, but was increased by 38% (95%
confidence interval: 11-72%) within the sarcopenic-obese group. In summary, sarcopenia and
obesity alone were not sufficient to increase CVD risk. Sarcopenic-obesity, based on muscle
strength but not muscle mass, was associated with increased CVD risk. These findings imply that
muscle strength is more important than muscle mass for CVD protection in old age.
KEY WORDS: waist circumference; skeletal muscle; aged; longitudinal study
55
INTRODUCTION
Sarcopenic-obesity represents a reduced skeletal muscle mass coupled with increased
adiposity within the same elderly person (1). The limited research on the health consequences of
sarcopenic-obesity has focused on functional outcomes. Some (1, 2) but not all (3, 4) studies
indicate that sarcopenic-obesity, but not obesity or sarcopenia alone, is a risk factor for functional
impairment.
Within the elderly, obesity, particularly abdominal and visceral obesity, contributes to
numerous cardiometabolic health problems such as insulin resistance (5), type 2 diabetes (6),
dyslipidemia (7), and cardiovascular disease (CVD) events (7, 8). Likewise, low muscle mass and
strength are associated with CVD risk factors including arterial stiffness (9), glucose intolerance
(10), and the metabolic syndrome (11). As both obesity and low muscle mass/strength predict
cardiovascular risk factors and outcomes in the elderly, it is possible that the combination of
obesity and sarcopenia would be associated with an even greater risk.
In support of the aforementioned notion, cross-sectional research in a sample of 871
sarcopenic-obese elderly demonstrated that abdominal obesity and low muscular strength are
characterized by high circulating levels of proinflammatory cytokines (12), which are recognized
risk factors for CVD (13). Conversely, in a cross-sectional study of 22 obese postmenopausal
women, the CVD risk factor profile was more favourable in sarcopenic-obese women than in
obese women with a normal muscle mass (14). The discrepant findings in these studies may
reflect the different approaches for assessing sarcopenia (muscle strength vs. muscle mass),
raising the question as to whether muscle mass or strength is more important for cardiovascular
health. While these two studies provide some interesting findings, their small sample sizes, crosssectional designs, and contradictory findings indicate that more research is needed to clarify the
impact of sarcopenic-obesity on CVD risk.
The primary purpose of this study was to determine if sarcopenic-obesity is a stronger
predictor of CVD risk in the elderly than either sarcopenia or obesity alone. A secondary
56
objective was to determine whether low muscle mass or low muscular strength was a stronger
marker of CVD risk.
57
METHODS
Study Sample
The study sample included elderly men and women from the Cardiovascular Health
Study (CHS), a population-based longitudinal study of CVD in adults aged 65 years and older, as
described in detail elsewhere (15). Briefly, subjects were sampled from Medicare eligibility lists
in Washington County, MD; Sacramento County, CA; Forsyth County, NC; and Pittsburgh, PA.
Eligible participants were 65 years and older, noninstitutionalized, able to give informed consent,
and did not require a proxy respondent. Of those eligible, 5201 (57%) enrolled. The institutional
review boards approved the project at each study site and written informed consent was obtained
from all subjects.
The baseline examination was conducted between June 1989 and June 1990, and the CHS
cohort was followed annually for 10 years. The baseline and follow-up examinations consisted of
a home or telephone interview and clinical examination (15). Briefly, in the home/telephone
interview, information was obtained on demographics and medical history. The clinical
examination included anthropometric measurements and a standardized clinical examination. A
total of 76 individuals were excluded from the study sample because they did not grant
permission to have their data included in the public access data set, 1241 were excluded due to
prevalent CVD at baseline, and an additional 484 were excluded because of missing data on the
study variables. Therefore, this study was limited to 3400 participants.
The CHS was conducted and supported by the National Heart, Lung, and Blood Institute
(NHLBI) in collaboration with the CHS investigators. The NHLBI and CHS investigators created
public access data sets that are stripped of all personal identifiers and are available, on appropriate
terms and conditions, to qualified investigators. The analyses presented in this paper are based on
the public access data files.
58
Body Composition
Abdominal Obesity. Measures of abdominal obesity, as determined by waist
circumference (WC), were chosen over measures of total adiposity (e.g., BMI, percent fat) as the
obesity indicator given that abdominal adiposity is a stronger predictor of cardiovascular risk
factors and disease (7, 8). WC was measured at baseline to the nearest 0.5 cm at the level of the
umbilicus using a flexible tape. Previous research has demonstrated that waist circumference is
highly correlated to total (R2=0.68 in men, 0.87 in women), abdominal (R2=0.68 in men, 0.73 in
women), and visceral (R2=0.55 in men, 0.76 in women) fat as determined by magnetic resonance
imaging (16).
Sarcopenia. Sarcopenia was classified using two different approaches based on either
skeletal muscle mass or skeletal muscle strength. Whole-body skeletal muscle mass was
estimated using bioelectrical impedance analysis (BIA). BIA resistance was obtained using a
TVI-10 Body Composition Analyzer (Danninger Medical, Columbus, OH) (15). BIA
measurements were taken between the right wrist and ankle with the subject in a supine position
after completion of an overnight fast (17). Muscle mass was calculated using the equation
developed by Janssen and colleagues (18): Skeletal muscle (kg) = [Height2/R x 0.401) + (sex x
3.825) + (age x -0.071)] + 5.102, where height is in centimetres; R is BIA resistance in ohms; for
sex, women=0 and men=1; and age is in years. The r2 and standard error of this regression
equation are 0.86 and 2.7 kg (9%), respectively, when compared to whole-body measures
obtained by magnetic resonance imaging. Muscle mass was normalized for height [muscle mass
(kg) / height (m2)] and termed the skeletal muscle index (SMI). As a measure of muscular
strength, maximal dominant hand grip strength was measured 3 times using a Jamar
dynamometer (Asimow Engineering Co., Los Angeles, CA) and averaged to the nearest kg (15).
Grip strength is a commonly employed measure of muscular strength in epidemiological studies
and is well correlated with other maximal isometric strength measures using a strain-gauge
59
system, including elbow flexion (r=0.64), knee extension (r=0.52), trunk flexion (r=0.43), and
trunk extension (r=0.52) (19).
Determination of Obesity and Sarcopenia Categories. Initially, subjects were divided
into sex-specific tertiles (low, moderate, and high) based on their: 1) WC, and 2) SMI. Subjects in
the low or moderate WC tertiles and the moderate or high SMI tertiles were classified as having a
‘normal’ body composition. Subjects in the high WC tertile and either the moderate or high SMI
tertiles were considered ‘obese’. Subjects in the low SMI tertile and either the low or moderate
WC tertile were considered ‘sarcopenic’. Finally, subjects in the high WC tertile and low SMI
tertile were classified as ‘sarcopenic-obese’. A comparable classification approach to that
described above (obesity X muscle mass) was used to classify subjects into four groups based on
tertiles of WC and skeletal muscle strength (obesity X muscle strength).
Cardiovascular Disease
The study outcomes consisted of incident: 1) coronary heart disease (CHD) (first
occurrence of myocardial infarction, silent myocardial infarction, angioplasty, or coronary artery
bypass graft) (20), 2) congestive heart failure (CHF), 3) stroke, and 4) overall CVD (first
occurrence of CHD, CHF, or stroke). Incident CVD events were ascertained over 8 years by selfreport and from the Health Care Financing Administration hospitalized patient database of
International Classification of Diseases, Ninth Revision diagnostic codes and are reported to the
exact day (21). Confirmation of CVD-related deaths was conducted through reviews of
obituaries, medical records, death certificates, and the U.S. Health Care Financing Administration
healthcare utilization database for stays in hospital.
Covariates
Sex. Males and females were included.
Age. Age was subdivided into 4 subgroups (65-70 years, 71-76 years, 77-82 years, ≥83 years).
Race. Subjects were classified as white or other race.
60
Income. Self-reported family income was used as a proxy for socioeconomic status. Annual
income was categorized as very low (≤$7,999), low ($8,000-$15,999), moderate ($16,000$34,999), high ($35,000-$49,999), or very high (≥$50,000). Participants who did not report their
income (6.3%) were coded in a separate non-response category.
Smoking. Lifetime smoking was categorized as none, passive (lived with regular smoker), light
(1-13 pack-years), moderate (14-50 pack-years), or heavy (>50 pack-years) (15).
Alcohol. Weekly alcohol consumption was categorized as low (<1 drink·week-1), moderate (1–7
drinks·week-1), or high (>7 drinks·week-1).
Cognitive Function. Cognitive function was measured using the 30-point Mini-Mental State
Examination (22). Subjects were categorized as having normal (>24 points) or impaired (≤23
points) cognitive function.
Statistical Analyses
SAS software version 9.1 (SAS Institute, Cary, NC) was used for data management and
analyses. Initially, differences in the descriptive characteristics between the groups were
determined using χ2 analysis for categorical variables (Appendix D) and general linear models
with Bonferroni post-hoc tests for continuous variables (Appendix H). Because 6 pair-wise
comparisons were necessary to compare all four groups to each other, a P-value of <0.008
denoted statistical significance (e.g., 0.05 / 6 = 0.008). Next we calculated age and sex-adjusted
partial correlations for the anthropometric and strength measures (Appendix I). Finally, we
compared the risk of developing the different CVD outcomes using Cox proportional hazards
regression models (Appendix F). The normal body composition group served as the referent
group in the Cox models and covariates included in the models consisted of age, sex, race, annual
household income, smoking, alcohol, and cognitive status. Sex X group and age X group
interaction terms were explored in these models, and without exception these were nonsignificant (P>0.1).
61
Determination of Follow-up Length for Cox Models. Participants were followed for 8
years after their baseline examination. For those who experienced a CVD event, follow-up length
was the number of days between their baseline exam and the initial event. For those subjects who
did not develop CVD but died during the follow-up period, the length of time between their
baseline examination and death was used as their follow-up length.
62
RESULTS
Subject Characteristics
Descriptive information for the 3400 subjects is shown in Table 1. When the sample was
divided into the four groups using WC and SMI (obesity X muscle mass), 39.5% of subjects were
classified as having a normal body composition, 27.6% were sarcopenic, 27.2% were obese, and
5.7% were sarcopenic-obese. The corresponding numbers in the groups classified according to
WC and grip strength (obesity X muscle strength) were 44.4%, 22.6%, 22.4%, and 10.6%. Of
those subjects who were classified as normal based on obesity X muscle mass, 70.5% were also
classified as normal based on obesity X muscle strength, while the remaining 29.5% were
classified as sarcopenic. Only 39.9% of the subjects who were sarcopenic based on obesity X
muscle mass were also sarcopenic based on obesity X muscle strength; the remaining 60.1% were
classified as normal. The majority (70.5%) of subjects who were obese based on obesity X muscle
mass were obese based on obesity X muscle strength; the remaining 29.5% were sarcopenicobese. Finally, of those who were sarcopenic-obese based on obesity X muscle mass, 44.3% were
sarcopenic-obese and 55.7% were obese based on obesity X muscle strength.
Baseline body composition and strength details for the four obesity X muscle mass
groups are outlined in Table 2. Body weight, BMI, WC, and SMI values were lowest in the
sarcopenic groups and highest in the obese groups. This pattern persisted for grip strength,
although the differences in the mean grip strength values of the four groups were <8%. Table 2
also contains the baseline body composition and strength details for the four obesity X muscle
strength groups. Body weight, BMI, and grip strength values were lowest in the sarcopenic and
highest in the obese subjects. Interestingly, sarcopenic-obese subjects had a higher SMI than
subjects in the normal group.
Relation Between Anthropometric and Strength Measures
The partial correlations between the anthropometric and grip strength measures, adjusted
for age and sex, are presented in Table 3. WC was highly correlated with BMI (r=0.81) and SMI
63
(r=0.41). Grip strength was modestly correlated with absolute muscle mass (r=0.22) and weakly
correlated with SMI (r=0.10). Due to the large sample size these correlations were all highly
significant (P<0.001).
Cardiovascular Disease Risk
Table 4 provides a summary of the person-years of follow-up, event rates, and adjusted
risk estimates for the different CVD outcomes. Because the results detailed in this section were
not modified by age or sex, the results for the entire cohort are presented. Inspection of Figure 1
reveals that, irrespective of how sarcopenic-obesity status was classified, the crude event rates for
CVD were higher in the sarcopenic and obese groups relative to the normal group, with the
sarcopenic-obese group displaying the highest CVD event rates. Compared with the normal
group, CVD event rates in the sarcopenic-obese group were elevated by 41% when based on
obesity X muscle mass and 72% when based on obesity X muscle strength.
After adjustment for the covariates, the effect of sarcopenic-obesity on CVD risk was less
pronounced. When based on measures of obesity X muscle mass, the risks of CVD, CHD, CHF,
and stroke were not significantly elevated within the obese, sarcopenic, or sarcopenic-obese
groups compared with the normal group, with one exception. The risk of CHD was elevated by
29% in the obese group compared to the group with a normal body composition (Table 4).
When based on measures of obesity X muscle strength, the risks of CVD, CHD, CHF,
and stroke were not significantly elevated in either the obese or sarcopenic groups compared to
the normal group. However, in the sarcopenic-obese group the risks of CVD, CHD, and CHF
were increased by 38%, 42%, and 53%, respectively (Table 4, P<0.05). The risk for stroke within
the sarcopenic-obese group was elevated by 33%, although this did not reach statistical
significance (P=0.12). The risk estimates for overall CVD and its subtypes within the sarcopenicobese group were not significantly different from those in either the sarcopenic or obese groups as
evident by the overlapping confidence intervals for the hazard ratios (Table 4).
64
DISCUSSION
The impact of sarcopenic-obesity on physical function has been given considerable
attention in the gerontology literature; however, the impact of this condition on cardiovascular
health has been poorly studied. To our knowledge, this is the first prospective study to examine
the relation between sarcopenic-obesity and CVD risk. The findings indicate that muscle strength
is more important for CVD risk than is muscle mass. Furthermore, although obesity and
sarcopenia alone did not significantly predict CVD, when they occurred simultaneously (e.g.,
sarcopenic-obesity) the risk of CVD increased by nearly 40%.
Previous research has defined sarcopenic-obesity using low height-adjusted skeletal
muscle mass coupled with high percent body fat. Using these criteria, literature in this area has
identified an increased risk of physical disability and functional impairment in sarcopenic-obese
persons (1, 2). For example, using a prospective cohort study design Baumgartner and colleagues
(2) found that sarcopenic-obese elderly were 2.5 times more likely to have a decline in physical
function compared to elderly with a normal body composition. Purely sarcopenic and obese
persons were not at increased risk. In the present study, waist circumference was used instead of
percent body fat. When determined by a high waist coupled with low hand grip strength, the
results for CVD risk in this study were similar to those reported by Baumgartner and colleagues
for physical disability (2), whereby only the sarcopenic-obese group experienced significantly
increased risk. However, the magnitude of effect for sarcopenic-obesity was not as strong in the
present study as that previously observed for physical function.
Our finding that sarcopenic-obesity increased the risk of CVD was opposite to the results
of Aubertin-Leheudre and colleagues (14). These authors reported that the cardiometabolic risk
factor profile was, surprisingly, more favourable in sarcopenic-obese postmenopausal women
than in obese postmenopausal women who were not sarcopenic. In that small cross-sectional
study, the purely obese women had 41% more visceral fat than the sarcopenic-obese women,
which may explain why the cardiovascular risk factors were different in these two groups. While
65
the abdominal fat content was not well matched in the two obese groups in the Aubertin-Leheudre
study, in the present study the purely obese and sarcopenic-obese groups had comparable waist
circumference values, suggesting that their abdominal fat content was similar.
The conflicting results between our different definitions of sarcopenic-obesity imply that
muscle quality (e.g., strength) is more important than muscle quantity (e.g., mass) in aging
humans. While muscle strength and muscle mass are related, they are not one and the same as
evidenced by the weak correlations between these measures. Previous work demonstrated that
only leg strength was independently associated with lower extremity functional performance
when leg strength and leg muscle were considered in the same regression model (23). Muscular
strength also appears to be important for cardiometabolic health as evidenced by the present
findings and those of Jurca and colleagues (11), who reported lower incidence of the metabolic
syndrome across muscular strength tertiles in men. Together, these results suggest an important
function for muscular strength that extends beyond the role of muscle mass.
Muscle strength may have a positive influence on cardiovascular health through
improvement of CVD risk factors. For instance, increases in muscular strength have been
associated with improvements in high-density lipoprotein cholesterol (24), blood pressure (25),
and insulin sensitivity (26); although a literature review in this area (27) indicated that the effects
of muscle strengthening activities on CVD risk factors is inconsistent. Thus, the specific
physiologic mechanism through which strength may affect CVD risk is uncertain.
A recent study reported that abdominal obesity was associated with an upregulation of
proinflammatory cytokine production (12). These cytokines in turn were inversely related to
muscle strength, thereby providing a link by which obesity may lead to sarcopenic-obesity over
time. However, due to the cross-sectional design of that study (12), the cause-and-effect nature of
the relation between abdominal fat and muscle is uncertain. Other research has reported that these
same proinflammatory cytokines are risk factors for CVD (28, 29). Therefore, although
66
speculative at this time, abdominal obesity may represent the starting point of sarcopenic-obesity
and its associated CVD risk.
In light of the current findings, public health efforts should continue to promote regular
physical activity and balanced nutrition to assist with maintenance of optimal body composition
through adulthood and into old age. For elderly persons who are sarcopenic-obese, treatment
would ideally focus on decreasing abdominal fat while simultaneously improving muscle
strength. Recently released public health guidelines for physical activity recommend that older
adults participate in moderate-to-vigorous intensity aerobic activities on at least 5 days of the
week and muscle strengthening activities on at least 2 days of the week (30). These guidelines
likely represent a useful intervention approach since the muscle strength observed in elderly
persons is likely a reflection of a physically active lifestyle (30).
There are limitations to the present study that warrant consideration. Although reasonably
accurate for use in large studies, BIA and WC are not criterion measures of body composition.
The imprecision of these measures may have weakened the relationship between sarcopenia and
obesity with CVD risk. Furthermore, although related to more definitive measures of strength
(19), hand grip strength is only a proxy of overall muscular strength. Another limitation of the
present study is its generalizability to non-white racial groups given that the cohort was 95%
white.
In summary, sarcopenic-obesity, as determined by a high WC and low hand grip strength,
was associated with a 38% increased risk of CVD in a large sample of community-dwelling
elderly adults. This relationship was not apparent when body composition was classified using
muscle mass, suggesting that strength may be more important than muscle mass for CVD
prevention. On their own, sarcopenia and obesity were not sufficient to increase CVD risk,
suggesting that there may be a synergistic effect of abdominal fat and muscle strength on CVD
risk.
67
REFERENCES
1.
Baumgartner RN. Body composition in healthy aging. Ann N Y Acad Sci 2000;904:437448.
2.
Baumgartner RN, Wayne SJ, Waters DL, Janssen I, Gallagher D, Morley JE. Sarcopenic
obesity predicts instrumental activities of daily living disability in the elderly. Obes Res
2004;12:1995-2004.
3.
Davison KK, Ford ES, Cogswell ME, Dietz WH. Percentage of body fat and body mass
index are associated with mobility limitations in people aged 70 and older from
NHANES III. J Am Geriatr Soc 2002;50:1802-1809.
4.
Zoico E, Di Francesco V, Guralnik JM et al. Physical disability and muscular strength in
relation to obesity and different body composition indexes in a sample of healthy elderly
women. Int J Obes Relat Metab Disord 2004;28:234-241.
5.
Despres JP, Lemieux S, Lamarche B et al. The insulin resistance-dyslipidemic syndrome:
Contribution of visceral obesity and therapeutic implications. Int J Obes Relat Metab
Disord 1995;19 Suppl 1:S76-86.
6.
Boyko EJ, Fujimoto WY, Leonetti DL, Newell-Morris L. Visceral adiposity and risk of
type 2 diabetes: A prospective study among Japanese Americans. Diabetes Care
2000;23:465-471.
7.
Despres JP, Moorjani S, Lupien PJ, Tremblay A, Nadeau A, Bouchard C. Regional
distribution of body fat, plasma lipoproteins, and cardiovascular disease. Arteriosclerosis
1990;10:497-511.
8.
Nicklas BJ, Penninx BW, Cesari M et al. Association of visceral adipose tissue with
incident myocardial infarction in older men and women: The Health, Aging and Body
Composition Study. Am J Epidemiol 2004;160:741-749.
68
9.
Snijder MB, Henry RM, Visser M et al. Regional body composition as a determinant of
arterial stiffness in the elderly: The Hoorn Study. J Hypertens 2004;22:2339-2347.
10.
Snijder MB, Dekker JM, Visser M et al. Larger thigh and hip circumferences are
associated with better glucose tolerance: The Hoorn Study. Obes Res 2003;11:104-111.
11.
Jurca R, Lamonte MJ, Barlow CE, Kampert JB, Church TS, Blair SN. Association of
muscular strength with incidence of metabolic syndrome in men. Med Sci Sports Exerc
2005;37:1849-1855.
12.
Schrager MA, Metter EJ, Simonsick E et al. Sarcopenic obesity and inflammation in the
InCHIANTI study. J Appl Physiol 2007;102:919-925.
13.
Pearson TA, Mensah GA, Alexander RW et al. Markers of inflammation and
cardiovascular disease: Application to clinical and public health practice: A statement for
healthcare professionals from the Centers for Disease Control and Prevention and the
American Heart Association. Circulation 2003;107:499-511.
14.
Aubertin-Leheudre M, Lord C, Goulet ED, Khalil A, Dionne IJ. Effect of sarcopenia on
cardiovascular disease risk factors in obese postmenopausal women. Obesity (Silver
Spring) 2006;14:2277-2283.
15.
Fried LP, Borhani NO, Enright P et al. The Cardiovascular Health Study: Design and
rationale. Ann Epidemiol 1991;1:263-276.
16.
Janssen I, Heymsfield SB, Allison DB, Kotler DP, Ross R. Body mass index and waist
circumference independently contribute to the prediction of nonabdominal, abdominal
subcutaneous, and visceral fat. Am J Clin Nutr 2002;75:683-688.
17.
Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free mass
using bioelectrical impedance measurements of the human body. Am J Clin Nutr
1985;41:810-817.
18.
Janssen I, Heymsfield SB, Baumgartner RN, Ross R. Estimation of skeletal muscle mass
by bioelectrical impedance analysis. J Appl Physiol 2000;89:465-471.
69
19.
Rantanen T, Era P, Kauppinen M, Heikkinen E. Maximal isometric muscle strength and
socioeconomic status, health, and physical activity in 75-year-old persons. J Aging Phys
Act 1994;2:206-220.
20.
Tracy RP, Arnold AM, Ettinger W, Fried L, Meilahn E, Savage P. The relationship of
fibrinogen and factors VII and VIII to incident cardiovascular disease and death in the
elderly: Results from the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol
1999;19:1776-1783.
21.
Ives DG, Fitzpatrick AL, Bild DE et al. Surveillance and ascertainment of cardiovascular
events. The Cardiovascular Health Study. Ann Epidemiol 1995;5:278-285.
22.
Folstein MF, Folstein SE, McHugh PR. "Mini-Mental State". A practical method for
grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-198.
23.
Visser M, Newman AB, Nevitt MC et al. Reexamining the sarcopenia hypothesis.
Muscle mass versus muscle strength. Health, Aging, and Body Composition Study
Research Group. Ann N Y Acad Sci 2000;904:456-461.
24.
Hurley BF, Hagberg JM, Goldberg AP et al. Resistive training can reduce coronary risk
factors without altering VO2max or percent body fat. Med Sci Sports Exerc 1988;20:150154.
25.
Martel GF, Hurlbut DE, Lott ME et al. Strength training normalizes resting blood
pressure in 65- to 73-year-old men and women with high normal blood pressure. J Am
Geriatr Soc 1999;47:1215-1221.
26.
Ryan AS, Hurlbut DE, Lott ME et al. Insulin action after resistive training in insulin
resistant older men and women. J Am Geriatr Soc 2001;49:247-253.
27.
Hurley BF, Roth SM. Strength training in the elderly: Effects on risk factors for agerelated diseases. Sports Med 2000;30:249-268.
70
28.
Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein and other markers of
inflammation in the prediction of cardiovascular disease in women. N Engl J Med
2000;342:836-843.
29.
Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin,
and the risk of cardiovascular disease in apparently healthy men. N Engl J Med
1997;336:973-979.
30.
Nelson ME, Rejeski WJ, Blair SN et al. Physical activity and public health in older
adults: Recommendation from the American College of Sports Medicine and the
American Heart Association. Med Sci Sports Exerc 2007;39:1435-1445.
71
Table 1. Characteristics of study participants at baseline
Covariate
Prevalence (%)
Male
39.6
White
94.9
Age
65–70 years
46.0
71–76 years
32.0
77–82 years
16.4
≥83 years
5.6
Annual household income
≤$7,999
10.2
$8,000–15,999
25.2
$16,000–34,999
34.4
$35,000–49,999
10.0
≥$50,000
13.9
Not Reported
6.3
Smoking status
None
45.6
Passive (lived with regular smoker)
4.1
Light (1–13 pack-years)
12.9
Moderate (14–50 pack-years)
25.9
Heavy (>50 pack-years)
11.6
Alcohol consumption
Low (<1 drinks·week-1)
67.3
-1
Moderate (1–7 drinks·week )
19.4
High (>7 drinks·week-1)
13.3
Impaired cognitive function
4.8
72
Table 2. Baseline characteristics based on groups according to sarcopenia and obesity classification
Variable
Normal
Sarcopenic
Obese
Sarcopenic-Obese
n (%)
1342 (39.5)
938 (27.6)
926 (27.2)
194 (5.7)
BMI (kg·m-2)
25.5±2.7b,c,d
23.1±2.3a,c,d
30.8±3.4a,b,d
27.3±2.4a,b,c
WC (cm)
88.2±7.9b,c,d
85.3±8.1a,c,d
106.3±5.9a,b,d
104.0±4.6a,b,c
SM (kg)
22.4±6.2b,c,d
18.9±5.4a,c
24.1±6.5a,b,d
20.0±5.4a,c
SMI (kg·m-2)
8.19±1.60b,c,d
6.85±1.31a,c
8.68±1.62a,b,d
7.08±1.30a,c
28.7±10.5b
26.8±10.0a,c
29.1±10.6b
27.3±10.8
n (%)
1510 (44.4)
770 (22.6)
761 (22.4)
359 (10.6)
BMI (kg·m-2)
24.8±2.8b,c,d
24.1±2.7a,c,d
30.3±3.4a,b
29.9±3.7a,b
WC (cm)
87.0±7.9c,d
86.9±8.3c,d
105.9±5.7a,b
105.9±5.9a,b
SM (kg)
21.3±6.0b,c
20.4±6.5a,c,d
24.0±6.6a,b,d
22.0±6.2b,c
SMI (kg·m-2)
7.68±1.56c,d
7.56±1.76c,d
8.56±1.69a,b,d
8.15±1.69a,b,c
Hand grip strength (kg)
31.4±9.8b,d
21.1±7.6a,c
32.4±9.9b,d
21.0±7.5a,c
Obesity X Muscle Mass
Hand grip strength (kg)
Obesity X Muscle Strength
Significantly different from anormal, bsarcopenic, cobese, and dsarcopenic-obese group (P<0.008).
WC, waist circumference; BMI, body mass index; SM, skeletal muscle; SMI, skeletal muscle index.
Table 3. Partial correlationsa between anthropometric and strength measures
BMI
WC
SM
SMI
BMI
WC
0.81b
SM
0.46b
0.40b
SMI
0.57b
0.41b
0.87b
Grip Strength
0.07b
0.03
0.22b
a
b
0.10b
Correlations are adjusted for age and sex.
P≤0.0001
BMI, body mass index; WC, waist circumference; SM, skeletal muscle; SMI, skeletal muscle
index.
74
Table 4. Cardiovascular disease risk according to sarcopenia and obesity status
Obesity X Muscle Mass
Cardiovascular Disease
Normal
Sarcopenic
Obese
Sarcopenic-obese
Coronary Heart Disease
Normal
Sarcopenic
Obese
Sarcopenic-obese
Congestive Heart Failure
Normal
Sarcopenic
Obese
Sarcopenic-obese
Stroke
Normal
Sarcopenic
Obese
Sarcopenic-obese
Obesity X Muscle Strength
n
Personyears of
follow-up
Events per
10,000
person-yr
1.00 (referent)
1.02 (0.85, 1.22)
1.17 (0.98, 1.39)
1.16 (0.87, 1.56)
1510
770
761
359
10533
4878
5102
2238
290
416
345
500
1.00 (referent)
1.15 (0.96, 1.38)
1.14 (0.94, 1.37)
1.38 (1.11, 1.72) b
146
181
201
174
1.00 (referent)
1.12 (0.87, 1.45)
1.29 (1.01, 1.64) c
1.00 (0.64, 1.57)
1510
770
761
359
10996
5127
5353
2396
149
183
176
242
1.00 (referent)
1.12 (0.86, 1.45)
1.12 (0.87, 1.44)
1.42 (1.04, 1.92) c
9729
6563
6511
1310
137
151
157
237
1.00 (referent)
0.90 (0.69, 1.17)
1.07 (0.83, 1.39)
1.31 (0.88, 1.95)
1510
770
761
359
11099
5193
5411
2410
123
185
135
249
1.00 (referent)
1.11 (0.85, 1.46)
1.05 (0.79, 1.40)
1.53 (1.12, 2.08) b
9832
6566
6552
1356
102
133
133
118
1.00 (referent)
1.12 (0.83, 1.49)
1.23 (0.92, 1.64)
0.91 (0.54, 1.55)
1510
770
761
359
11139
5259
5452
2456
99
146
110
175
1.00 (referent)
1.14 (0.84, 1.54)
1.07 (0.78, 1.47)
1.33 (0.93, 1.91)
n
Personyears of
follow-up
Events per
10,000
person-yr
1342
938
926
194
9225
6186
6099
1242
309
360
384
435
1342
938
926
194
9662
6462
6428
3121
1342
938
926
194
1342
938
926
194
Hazard Ratio
(95% CI)a
a
Hazard ratios (95% confidence intervals) were adjusted for age, sex, race, income, smoking, alcohol, and cognitive status.
b
P<0.01 vs. normal.
c
P<0.05 vs. normal
Hazard Ratio (95%
CI) a
76
Number of CVD events
per 10,000 person-years
0
100
200
300
400
500
A
Normal
Figure 1.
Sarcopenic
Obese
Sarcopenicobese
0
100
200
300
400
500
B
Normal
Sarcopenic
Obese
classified using either obesity X muscle mass (Panel A) or obesity X muscle strength (Panel B).
Sarcopenicobese
Number of cardiovascular disease events per 10,000 person-years according to abdominal obesity and sarcopenia
Number of CVD events
per 10,000 person-years
CHAPTER 5
GENERAL DISCUSSION
5.1 Summary of Key Findings
The purpose of this thesis was to investigate the beneficial effects of physical activity and
the detrimental effects of age-related body composition changes in the elderly. The first
manuscript in this thesis confirms that participation in modest amounts of physical activity is
beneficial for attenuation of age-related weight loss in the elderly. Moderately active subjects lost
approximately 25% less weight than their inactive counterparts, with little additional benefit
observed with higher levels of physical activity. These findings add to an ever expanding number
of health outcomes that are known to be positively impacted by physical activity, further
highlighting the importance of this lifestyle behaviour.
The second manuscript investigated the impact of sarcopenic-obesity on cardiovascular
health, a topic that has been poorly studied. This prospective study examined the relationship
between sarcopenic-obesity and CVD risk and demonstrated that although obesity and sarcopenia
alone did not significantly predict CVD, when they occurred simultaneously (e.g., sarcopenicobesity) the risk of CVD increased by nearly 40%. However, this increased risk was observed
only when sarcopenic-obesity was classified using abdominal obesity and low muscle strength,
implying that muscle quality (e.g., strength) is more important for cardiovascular health than is
muscle quantity (e.g., mass) in the elderly.
5.2 Limitations of the Thesis
There are several limitations common to both studies reported in this thesis that warrant
consideration. First, the CHS data is primarily comprised of white subjects (95%), therefore the
generalizability of the results to non-white ethnic groups is uncertain. Second, the CHS study was
conducted in the United States. While these findings are likely generalizable to Canadians, the
77
differences in the health care systems of the United States and Canada may represent a limitation
if they affected the CHS seniors’ lifelong access to health care (1). This may have in turn
influenced the overall health of the CHS participants causing them to be somewhat different from
Canadian seniors. Unfortunately, there is currently no Canadian data source of a similar nature. A
further limitation of these two studies is the use of non-criterion measures for the exposure
variables (e.g., self-reported physical activity and measured WC, BIA, and grip strength) which
likely introduced some imprecision and may have weakened the strength of the observed
relationships with their respective outcome variables. Additionally, it is possible that physical
activity, body composition, and strength changed over the follow-up period; although we were
unable to detect these changes since these measures were only available at baseline. Basing the
analyses on baseline measures of these variables may have weakened their longitudinal
relationships with both body weight and CVD.
5.3 Strengths of the Thesis
Despite these limitations, both studies contained in this thesis share notable strengths.
The first is the use of a large population-based longitudinal study of elderly persons. The large
sample size provided sufficient statistical power (even within subgroups) to detect small
differences in weight loss trajectories according to physical activity in the first study, and to
detect small increases in CVD risk according to body composition in the second study. The
longitudinal nature of this data also allowed for temporal trends in body weight to be considered.
Furthermore, we were able to demonstrate that body composition abnormalities precede the
development of various CVD outcomes, thereby eliminating the possibility of reverse causality in
this relationship. A further strength of these two studies is the use of precise outcome measures,
which implies that we were able to accurately quantify our outcomes. These include the measured
body weights in Manuscript 1 and the thorough CVD follow-up in Manuscript 2, which was
78
established using International Classification of Diseases, Ninth Revision (ICD-9) diagnostic
codes (2).
The use of complex mixed modeling techniques in the first manuscript is an additional
strength of this thesis. Specifically, this allowed for the longitudinal modeling of the various
weight trajectories over time. As such, we were able to use repeated measures of body weight
while allowing for the inclusion of subjects with some missing body weight measures.
Furthermore, this method allowed for inclusion of time-varying covariates, which enabled the
inclusion of various incident disease variables.
5.4 Future Research Directions
Future investigations in physical activity research as it pertains to weight maintenance in
the elderly should utilize objective measures of physical activity (e.g., accelerometers) to
determine if a stronger measure of the exposure affects the strength of its relationship with weight
loss. Furthermore, examination of different types and intensities of physical activity on its relation
with body weight changes in the elderly would help elucidate the amount and type of activity that
would incur the greatest health benefits.
Manuscript 2 represents the first prospective study to examine the relation between
sarcopenic-obesity and CVD risk. Additional research on this topic would ideally employ more
precise measures of body composition and muscular strength to confirm the present findings.
Moreover, examination of sarcopenic-obesity and its relation with specific CVD risk factors may
provide some insight into the mechanisms involved in this association.
Finally, it would be interesting to see if the relationships observed in this thesis persist
when Canadian data of this nature becomes available. Incidentally, a large, national, long-term
Canadian study is set to launch in the Fall of 2008. The Canadian Longitudinal Study on Aging
will follow approximately 50,000 Canadians between the ages of 45 and 85 for a period of at least
79
20 years. Once this data becomes available there will be many opportunities to explore various
health relationships including those outlined in this thesis.
5.6 Public Health Implications
Although the effects of physical activity on attenuation of age-related weight loss were
modest in this group of elderly persons, this finding adds further value to the benefits of this
important lifestyle behaviour. Additionally, the second manuscript represents the first study of its
kind to demonstrate a longitudinal relation between the combination of abdominal obesity and
low muscle strength and increased CVD risk. Incidentally, physical activity is a key treatment
strategy for both abdominal obesity (3) and low muscle strength (4). In this regard, recently
released public health guidelines for physical activity recommend that older adults participate in
moderate-to-vigorous intensity aerobic activities on at least 5 days of the week and muscle
strengthening activities on at least 2 days of the week (5). Although these guidelines are not based
on weight maintenance or sarcopenic-obesity per se, they are likely a good starting point for
physical activity recommendations for these purposes.
Continued research in the elderly will become increasingly important as the senior
population within Canada and other industrialized countries continues to grow. The more we are
able to understand about this specialized group, the more effectively we will be able to
accommodate their growing health needs in the future.
80
References
1.
Lasser KE, Himmelstein DU, Woolhandler S. Access to care, health status, and health
disparities in the United States and Canada: results of a cross-national population-based
survey. Am J Public Health 2006;96:1300-1307.
2.
Ives DG, Fitzpatrick AL, Bild DE et al. Surveillance and ascertainment of cardiovascular
events. The Cardiovascular Health Study. Ann Epidemiol 1995;5:278-285.
3.
Janiszewski PM, Ross R. Physical activity in the treatment of obesity: beyond body
weight reduction. Appl Physiol Nutr Metab 2007;32:512-522.
4.
Marcell TJ. Sarcopenia: causes, consequences, and preventions. J Gerontol A Biol Sci
Med Sci 2003;58:M911-916.
5.
Nelson ME, Rejeski WJ, Blair SN et al. Physical Activity and Public Health in Older
Adults: Recommendation from the American College of Sports Medicine and the
American Heart Association. Med Sci Sports Exerc 2007;39:1435-1445.
81
SUMMARY AND CONCLUSIONS
The increasing proportion of Canadian adults over the age of 65 years will have major
implications for population health. Among the changes that this group will experience, the
increase in body fat and decline in skeletal muscle mass and strength are of great importance.
Often the loss of lean body mass is reflected by undesirable weight loss in the elderly, while the
coupling of high body fat and low muscle mass or strength (e.g., sarcopenic-obesity) may
negatively affect cardiovascular health.
The results presented in this thesis provide information regarding a further benefit of
physical activity in old age, as well as the increased cardiovascular risk associated with
sarcopenic-obesity. Manuscript 1 entitled Influence of Physical Activity on Age-Related Weight
Loss in the Elderly confirms that although most seniors appear to lose weight with advancing age,
moderate amounts of physical activity can help to attenuate this decline. This has important health
implications since weight loss in the elderly is associated with a variety of deleterious outcomes,
including physical disability (1) and mortality (2).
Manuscript 2, Sarcopenic-Obesity and Cardiovascular Disease Risk in the Elderly, is the
first prospective study to demonstrate an increased risk of CVD in sarcopenic-obese seniors.
Furthermore, this study examined the utility of both muscle mass and muscle strength as
measures of sarcopenia and found that muscle strength (e.g., quality) is in fact more important
than muscle mass (e.g., quantity) for cardiovascular health.
Taken together, the current findings support the continuation of public health efforts to
promote regular physical activity and balanced nutrition to assist with maintenance of optimal
body composition and weight through adulthood and into old age. However, they also highlight
the need for more research in order to discern the optimal dose and intensity of physical activity
necessary to attenuate weight loss as well as the need for research to confirm the findings
pertaining to sarcopenic-obesity and CVD.
82
As the population age distribution continues to shift, continued research in the elderly
will become increasingly important. The greater understanding we have about the health needs
and intricacies of the senior population, the more effectively we will be able to appreciate and
assist with their diverse health needs in the future.
83
REFERENCES
1.
Launer LJ, Harris T, Rumpel C, Madans J. Body mass index, weight change, and risk of
mobility disability in middle-aged and older women. The epidemiologic follow-up study
of NHANES I. JAMA 1994;271:1093-1098.
2.
Newman AB, Yanez D, Harris T, Duxbury A, Enright PL, Fried LP. Weight change in
old age and its association with mortality. J Am Geriatr Soc 2001;49:1309-1318.
84
APPENDIX A
Glossary of Terms and Abbreviations
85
BIA – Bioelectrical Impedance Analysis
BMI – Body Mass Index (kg/m2)
CHD – Coronary Heart Disease
CHF – Congestive Heart Failure
CHS – Cardiovascular Health Study
CI – Confidence Interval
CT – Computed Tomography
CVD – Cardiovascular Disease
DXA – Dual-energy X-ray Absorptiometry
ICD-9 – International Classification of Diseases, Ninth Revision
MRI – Magnetic Resonance Imaging
NHLBI – National Heart, Lung, and Blood Institute
PA – Physical Activity
RR – Relative Risk
SM – Skeletal Muscle (kg)
SMI – Skeletal Muscle Index (kg/m2)
WC – Waist Circumference (cm)
86
APPENDIX B
Ethics Approval
87
88
89
90
APPENDIX C
Physical Activity Questionnaire
91
92
93
APPENDIX D
Sample Chi Square Analysis Output
94
95
APPENDIX E
Sample PROC MIXED Analysis Output
96
97
98
99
100
101
102
103
104
105
106
107
APPENDIX F
Sample Cox Proportional Hazards Regression Analysis Output
108
109
110
APPENDIX G
Sample Excel Output for an Adjusted Weight Trajectory Curve
111
112
113
APPENDIX H
Determination of Body Composition Categories
Based on Abdominal Obesity x Muscle Mass
114
Sex-specific Tertile
Low SMI
Moderate SMI
High SMI
Low WC
Sarcopenic
Normal
Normal
Moderate WC
Sarcopenic
Normal
Normal
High WC
Sarcopenic-obese
Obese
Obese
WC: waist circumference (cm), SMI: skeletal muscle index (kg/m2).
115
APPENDIX I
Sample Multiple Comparisons Analysis of Variance
with Bonferroni Post-hoc Test
116
117
118
APPENDIX J
Sample Partial Correlations Output
119
120