Relative Weight and Mobility: A Longitudinal Study in a Biracial Population of Older Adults CARLOS F. MENDES DE LEON, PHD, MARIA R. HANSBERRY, MD, JULIA L. BIENIAS, SCD, MARTHA C. MORRIS, SCD, AND DENIS A. EVANS, MD
PURPOSE: We examined the association of relative weight with mobility and changes in mobility over time and whether these associations differed by race. METHODS: Data come from a prospective, population-based, observational study of adults aged 65 years or older. Mobility outcomes were assessed at baseline and two follow-up interviews at 3-year intervals. The study included 4195 participants with a mean age of 73.8 6.3 (SD) years; 61.4% were women, and 60.9% were black. Assessment of mobility included a brief self-report instrument and a performance-based walk test. Body mass index (BMI, kilograms per square meter) was used as a measure of relative weight. We used generalized estimating equation models to examine change in mobility outcomes over time as a function of BMI. RESULTS: Average BMI was 26.6 5.7 kg/m2, with 34.0% overweight and 23.4% obese. BMI showed a significant curvilinear association with mobility outcomes at baseline (p ! 0.001), but was not associated with change in mobility during follow-up. Maximum mobility levels occurred at a significantly higher level of BMI among blacks than whites. CONCLUSIONS: Higher levels of BMI may lead to mobility impairments earlier in life, but there is little evidence that they increase the rate of decline in mobility in older age itself. Ann Epidemiol 2006;16:770e776. Ó 2006 Elsevier Inc. All rights reserved. KEY WORDS:
Mobility, Body Mass Index, Longitudinal Studies, Racial Differences.
INTRODUCTION Overweight and obesity affect an increasingly greater proportion of Americans, including older adults (1, 2). According to the 1999 to 2000 National Health and Nutrition Examination Survey, 35% of all Americans older than 60 years are obese, and 68% are overweight or obese (1, 3). There is abundant evidence that increased relative body weight is associated with increased risk for major adverse health outcomes, such as heart disease, stroke, diabetes, and osteoarthritis, and mortality (4, 5). High levels of relative weight also are associated with increased risk for disability and functional limitations in older adults (6e10). Impairment in basic mobility, such as the ability to walk, forms an important aspect of the progression toward disability in older age (11, 12). Mobility impairment also is associated with decreased survival independent of other disabilities or From the Departments of Internal Medicine (C.F.M.L., M.R.H., J.L.B., M.C.M., D.A.E.), Preventive Medicine (C.F.M.L.), and Neurological Sciences (D.A.E.), Rush University Medical Center, Chicago, IL. Address correspondence to: Carlos F. Mendes de Leon, PhD, Rush Institute for Healthy Aging, 1645 West Jackson Blvd, Suite 675, Chicago, IL 60612. Tel.: (312) 942-3350; fax (312) 942-2861. E-mail: cmendes@ rush.edu. This research was supported by grants from the National Institute on Aging (AG 11101) and the National Institute of Environmental Health Sciences (ES 10902). Received November 4, 2005; accepted May 2, 2006. Ó 2006 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010
comorbidities (13, 14). Although high levels of relative weight were associated with increased risk for mobility impairment (10, 15, 16), important issues regarding the nature of this association have not been fully addressed. First, increased risk for mobility impairment may not be restricted to excessive body weight. Some studies reported that older adults with very low body mass index (BMI) are at increased risk for disability (7, 9, 17), but few studies to date considered the full spectrum of relative weight in relation to mobility impairment in older adults. Second, the effect of either underweight or overweight on progression in mobility impairments in older age itself remains mostly unclear. Previous studies suggested that the adverse effect of overweight or obesity on mobility is strongest in middle age and may weaken considerably at more advanced ages (7, 15, 18, 19). Furthermore, these studies focused primarily on initial mobility impairment, rather than progression along the full spectrum of mobility impairment that occurs in older populations. Finally, there has been little systematic investigation of the relationship between relative weight and mobility across different ethnic or racial groups; in particular, blacks. The prevalence of overweight and obesity is greater among blacks, especially black women (1). Blacks also tend to have greater levels of mobility impairment and disability than whites (20, 21). The purpose of the present study is to: i) describe the association of relative weight with mobility impairment along 1047-2797/06/$esee front matter doi:10.1016/j.annepidem.2006.05.002
AEP Vol. 16, No. 10 October 2006: 770e776
Selected Abbreviations and Acronyms BMI Z body mass index CHAP Z Chicago Health and Aging Project SES Z socioeconomic status
the full spectrum of relative weight, ii) test whether the relationship between relative weight and mobility impairment is the same among older blacks and whites, and iii) test the association between relative weight and progression of mobility impairment over time. An important strength of this study is the availability of both a self-report and a performance-based measure of basic mobility. METHODS Study Population Data come from the Chicago Health and Aging Project (CHAP). CHAP is an ongoing, longitudinal, populationbased study of adults 65 years and older residing in three contiguous neighborhoods on the south side of Chicago. Details of the study design are described elsewhere (22). Briefly, the study began with a complete census of the community between 1993 and 1996. All residents 65 years and older were identified and asked to participate. Of 7813 eligible residents, 6158 (78.9%) enrolled and participated in an in-home interview at baseline. Participants completed an in-home interview and examination consisting of structured questions about sociodemographic characteristics, health, disability, measured weight, and performance-based tests of physical function. The present analysis is based on data from the baseline interview and two follow-up interviews conducted at 3-year intervals. The study was approved by the Institutional Review Board of Rush University Medical Center (Chicago, IL), and all participants (or legal guardians) provided written consent. Measures Basic Mobility. We used a self-report and a performance-based measure of walking as indicators of basic mobility. The self-report measure included three commonly used questions on basic mobility: the ability, without assistance, to i) walk across a small room, ii) walk up and down stairs, and iii) walk half a mile (23, 24). Positive responses were given a score of 1 and summed across all 3 items to yield a score for self-reported mobility ranging from 0 (lowest mobility) to 3 (highest mobility). Mobility impairment is defined as reporting a negative response on one or more of the three questions (mobility score < 2). The performance test was an 8-foot walk timed in seconds (3). Performance times were categorized in quintiles
Mendes de Leon et al. RELATIVE WEIGHT AND CHANGE IN MOBILITY
771
and assigned scores from 1 to 5, with higher codes indicating faster walking times. A score of 0 was assigned to those unable to complete the walk test. Higher scores represent better mobility. Relative Weight. BMI was used as measure of relative body weight. Weight was measured in pounds by using a digital freestanding scale placed on a hard flat surface, with the participant’s shoes removed. Height was self-reported in feet and inches. In general, self-reports of weight and health are reasonably reliable and accurate among adults, although there is a tendency for more obese persons to underreport their weight (25). Pounds and inches were converted to kilograms and meters to compute BMI (kilograms/square meters), respectively. World Health Organization criteria were used to define underweight (BMI ! 18.5 kg/m2), overweight (BMI > 25, but !30 kg/m2), and obesity (BMI > 30 kg/m2). The continuous measure of BMI was used for analysis, centered at 25 kg/m2. Control Variables. Other variables in the analysis included age (in years, centered at 75), sex, self-reported race, lifetime socioeconomic status (SES), and number of chronic conditions. Race was coded as black or nonblack (white, Asian American, or American Indian). Less than 1% reported being of Hispanic origin (asked separately from race) or being either Asian American or American Indian. We used four components to construct a measure of lifetime SES: childhood SES, education (years of regular schooling), occupational status most of life, and current income. Childhood SES was based on parental education, father’s occupational prestige (discussed next), and financial status during childhood (26). Occupational status was classified according the 1990 US Census indices of industries and occupations and assigned prestige scores based on occupational earnings and educational requirement data derived from the 1990 US Census (27). Total household income was measured by using a color-coded card with 10 income categories ranging from less than $5000/y to greater than $75,000/y. The measure of lifetime SES was obtained after z scoring each of the four component measures and averaging across components. Higher scores indicate higher SES. Number of chronic conditions was computed by summing positive responses to nine self-reported physiciandiagnosed medical conditions, including myocardial infarction, stroke, cancer, diabetes, hypertension, thyroid disease, shingles, Parkinson disease, and hip fracture. Joint-related problems were measured by using questions about joint pain (present/absent) and joint stiffness (present/absent). Statistical Analysis Racial differences in baseline characteristics were tested by using t-tests for continuous variables and chi-square tests
772
Mendes de Leon et al. RELATIVE WEIGHT AND CHANGE IN MOBILITY
for categorical variables. Generalized estimating equation models were used to analyze the two outcome variables while accounting for the correlated structure of data across repeated measurement (28). Because of the nonnormal distribution of the self-reported mobility measure, these scores were considered the proportion of mobility tasks a person was able to do of the total number of tasks (three). The resulting scores were analyzed by using a logistic link function and binomial error structure. The performance-based mobility variable was analyzed by using an identity link and normal error structure. We first modeled each outcome as a function of time since baseline, age at baseline, sex, and race. We included terms for interactions for age with sex and age with race to allow for the differential association of age with each mobility outcome by sex and race. We also included two-way interaction terms for age, sex, and race with time since baseline and the three-way interaction between age, race, and time since baseline to account for significant differences in the rate of change in one or both mobility outcomes as a function of age, sex, race, and the age-by-race interaction. We then tested the effect of BMI on mobility outcomes as follows: we added to the previous model terms for BMI and BMI squared to see whether there was a significant curvilinear association between BMI and mobility. We included an interaction term for BMI by race to see whether the association between BMI and mobility differed by race. We also tested the interaction between BMI squared and race, but these test results were negative and the corresponding terms were omitted from the final models. Finally, we included a term for BMI by time since baseline to see whether the rate of change in mobility differed by level of BMI. Interaction terms between BMI squared and time since baseline were tested as well, but were not significant and were not included in the final models. Model assumptions were validated analytically and graphically. We conducted two sets of secondary analyses. In the first analysis, we reran the primary models, excluding participants with a very low relative weight, defined as BMI less than 18 kg/m2 (N Z 152) and a very high relative weight, defined as BMI greater than 40 kg/m2 (N Z 87), to make sure that our results were not dependent on the inclusion of extreme values of BMI. In the second analysis, we added additional covariates (lifetime SES, number of chronic conditions, joint pain, and joint stiffness), as well as their interactions with time since baseline, to see whether the association between BMI and mobility outcomes was independent of other characteristics that might account for their association. All longitudinal analyses were performed using the GENMOD procedure of SAS, version 8 (SAS Institute, Cary, NC) (29).
AEP Vol. 16, No. 10 October 2006: 770e776
RESULTS Of 6158 participants in the CHAP Study, we excluded participants who died (n Z 1205) or who refused or were lost to follow-up (n Z 554) before the first follow-up interview because they had fewer than two waves of outcome data. Of the remaining 4399 participants, 204 (4.6%) had missing data on one or more of the primary variables, leaving 4195 participants for analysis. Average age at baseline was 73.8 6.3 (SD) years, 61.4% were women, and 2554 (60.9%) participants were black (Table 1). Average BMI was 26.6 5.5 kg/m2. More than a third (34.0%) of participants were overweight, and 23.4% were obese. On average, blacks were younger and had lower SES, higher average BMI, and a greater prevalence of obesity (Table 1). Compared with participants included in the analysis, those omitted from analysis for reasons other than death (refusal, lost to follow, or missing variables; n Z 758) were significantly older at baseline and had lower SES and lower self-reported and performance-based mobility scores (all p ! 0.001). However, these two subsets did not differ by sex, race, number of chronic conditions, or baseline BMI (all p O 0.10). We first consider results for self-reported mobility. Mobility tended to be lower in older persons at baseline and decline during follow-up, with the rate of decline greater at older ages (Table 2). Blacks and women reported lower mobility levels at baseline than nonblacks and men. However, these groups also tended to have a slower rate of decline in mobility during follow-up. There was a significant curvilinear relationship between ^ Z 0.005; BMI and self-reported mobility at baseline (b p ! 0.001; Table 2). The negative coefficient is indicative of a concave-downward relationship between BMI and mobility, with lower mobility at both the lower and higher ends of BMI. The baseline association of BMI with mobility was ^ Z 0.035; p Z 0.004), indicating that the modified by race (b predicted maximum score of self-reported mobility occurs at a higher level of BMI for blacks than nonblacks. BMI was ^ Z not associated with decrease in mobility over time (b 0.002; p Z 0.08), although the coefficient was in the direction of greater decrease with increasing levels of BMI. Figure 1 illustrates the curvilinear relationship between BMI and self-reported mobility, adjusted for other variables in the multivariate regression model. Predicted mobility scores are lower at both ends of the spectrum of BMI among blacks and nonblacks. As shown in Fig. 1, maximum predicted mobility levels occur at a BMI of 28.5 kg/m2 for a 75-year-old black man and 25 kg/m2 for a similar nonblack man. We next consider results for performance-based mobility. Consistent with results for self-reported mobility, performance-based mobility scores tended to be lower in older persons at baseline and decrease during follow-up, with a greater rate of decrease in older persons (Table 2). Men
AEP Vol. 16, No. 10 October 2006: 770e776
Mendes de Leon et al. RELATIVE WEIGHT AND CHANGE IN MOBILITY
773
TABLE 1. Baseline characteristics of participants in the Chicago Health and Aging Project Variable
Total (N Z 4195)
Black (N Z 2554)
Nonblack (N Z 1641)
p
73.8 6.3
72.9 5.7
75.7 6.4
!0.001 0.2781
1618 (38.6) 2577 (61.4) 12.1 3.7
1005 (39.4) 1549 (60.6) 11.0 3.5
613 (37.4) 1028 (62.6) 13.8 3.3
1603 (39.3) 1417 (34.7) 1061 (26.0) 33.5 14.3 0.1 0.8 1.2 1.0 26.6 5.7
1251 (49.8) 888 (35.4) 372 (14.8) 29.4 12.4 0.5 0.7 1.3 1.0 27.3 5.9
352 (22.4) 529 (33.7) 689 (43.9) 40.8 14.5 0.2 0.6 1.2 1.1 25.6 5.0
210 (5.0) 1576 (37.6) 1426 (34.0) 983 (23.4) 729 (17.4) 3.0 (1.5)
107 (4.1) 858 (33.6) 882 (34.5) 707 (27.7) 477 (18.7) 2.8 (1.5)
103 (6.3) 718 (43.8) 544 (33.2) 276 (16.8) 252 (15.4) 3.4 (1.5)
Age (years) Sex Men Women Education (years) Income ($) !15,000 15,000e30,000 O30,000 Occupational prestige Lifetime socioeconomic status No. of chronic medical conditions Body mass index (kg/m2) Body mass index (kg/m2) !18.5 18.6e24.9 25e29.9 O30 Self-reported mobility impairment Performance-based mobility
!0.001 !0.001
!0.001 0.001 0.34 !0.001 !0.001
.006 !.001
Values expressed as mean SD or number (percent).
tended to have higher mobility scores, with sex differences at baseline greater in older persons, but decreasing during follow-up. Blacks had lower performance-based mobility ^ Z 0.683 for a 75-year-old; p ! scores at baseline (b 0.001), but blackenonblack differences were modified by age and time, such that these differences in mobility scores tended to decrease with age and during follow-up. BMI showed essentially the same type of curvilinear rela^ Z 0.003; p ! 0.001) with performance-based tionship (b mobility scores as it did for self-reported mobility. The black ^ Z 0.028; p ! 0.001) indirace BMI interaction term (b cates that the maximum predicted performance-based mobility score at baseline occurs at a higher level of BMI in blacks than nonblacks. BMI was not associated with change ^ Z 0.000; p Z 0.74). in mobility scores during follow-up (b Figure 2 shows the curvilinear relationship between BMI and predicted mobility at baseline, adjusted for other variables in the multivariate regression model. For a 75-yearold black man, the maximum predicted mobility score was associated with a BMI of 28.5 kg/m2, whereas this point occurred at a BMI of 23.5 kg/m2 for a nonblack man. In secondary analysis (data not shown), basic associations among BMI, race, and mobility were unchanged after excluding persons with extremely low (!18 kg/m2) and extremely high (O40 kg/m2) BMIs. Additional adjustment for SES, medical conditions, and joint-related problems similarly did not change relationships among BMI, race, and either mobility outcome. Results of reanalysis using World Health Organizatione defined BMI categories instead of the continuous BMI
variable show that underweight (BMI < 18.5 kg/m2) and obesity were each associated with lower mobility scores at baseline for nonblacks, but overweight (BMI, 25.0 to 29.9 kg/m2) was not (Table 3). The association between obesity and mobility outcomes at baseline was significantly smaller for blacks than nonblacks for both outcomes and also was smaller for overweight blacks relative to overweight nonblacks for performance-based TABLE 2. Body mass index and mobility and changes in mobility over time in the Chicago Health and Aging Project Self-reported mobility Variable Time in study Age Age time in study Male sex Male sex time in study Age male sex Black race Black race age Black race time in study Black race age time in study Body mass index Body mass index body mass index Body mass index time in study Black race body mass index
^ b 0.228 0.116 0.006 0.689 0.028 0.018 0.658 0.013 0.039 0.002
p
Performance-based mobility ^ b
p
!0.001 !0.001 !0.001 !0.001 0.07 0.07 !0.001 0.29 0.01 0.28
0.105 0.094 0.006 0.320 0.017 0.018 0.683 0.022 0.021 0.003
!0.001 !0.001 !0.001 !0.001 0.09 0.002 !0.001 !0.001 0.04 0.04
0.001 0.94 0.005 !0.001
0.008 0.003
0.21 !0.001
0.002
0.08
0.000
0.74
0.035
0.004
0.028
!0.001
774
Mendes de Leon et al. RELATIVE WEIGHT AND CHANGE IN MOBILITY
FIGURE 1. Predicted self-reported mobility at baseline by body mass index (BMI) for typical 75-year-old black and nonblack men.
mobility. No BMI category was associated with change in mobility over time compared with the normal-weight reference group (BMI, 18.6 to 24.9 kg/m2).
DISCUSSION Our findings indicate that in a population of communitydwelling older adults, mobility levels tend to be lower at both the higher and lower spectrum of BMI. Although this curvilinear relationship was seen for blacks and nonblacks, there also were important differences in the relationship between relative weight and mobility impairment by race. BMI was not related to rate of progression in mobility impairment over time. Overall, our findings are consistent with previous findings regarding impaired mobility among
AEP Vol. 16, No. 10 October 2006: 770e776
overweight or obese older adults (6, 15, 19, 30) and extend these findings in showing the extent of mobility impairments across the full spectrum of BMI. In older age, mobility is decreased in both the upper and lower range of BMI (7, 31). A particular strength of this study is the availability of performance-based data for mobility. The similarity of results for the self-report and performance-based measures of mobility is noteworthy and provides further support for the curvilinear relationship between BMI and mobility impairment. Greater mobility impairment at overweight or obese levels of BMI may reflect the effect of age-related changes in lean tissue. Lean tissue, as found in muscle, tends to decrease in older age, leading to a greater imbalance of lean to fat body mass (32). These changes may decrease physical activity levels and increase disability in older adults (10, 16, 33, 34). Overweight and obesity also are associated with increased risk for various chronic conditions, such as cardiovascular disease and diabetes, that commonly lead to increased mobility impairment and disability (4, 5). Higher BMI also may affect the overall condition of the musculoskeletal system, causing wear and tear in joints and tendons that gradually affect mobility over time (15, 35). Age-related decreases in lean tissue also may lead to weight loss, causing BMI to decrease to less-than-normal levels (36). Loss of lean tissue is associated with decreased muscle strength and sarcopenia (32, 37), which may contribute to increased risk for mobility impairment and disability (37e40). Loss of weight also is a key component of the frailty syndrome (41), which is independently predictive
TABLE 3. BMI category by World Health Organization criteria and change in mobility in the Chicago Health and Aging Project Self-reported mobilitya ^ b
Variable b
^ b
p
2
BMI categories (kg/m ) <18.5 25e29.9 >30 BMI categoriesb time in study <18.5 time 25e29.9 time >30 time BMI categoriesb race <18.5 black race 25e29.9 black race >30 black race
FIGURE 2. Predicted performance-based mobility at baseline by body mass index (BMI) for typical 75-year-old black and nonblack men.
p
Performance-based mobility
0.616 !0.01 0.026 0.86 0.712 !0.001
0.510 0.074 0.368
!0.001 0.29 !0.001
0.014 0.017 0.015
0.64 0.33 0.41
0.008 0.017 0.004
0.73 0.14 0.74
0.285 0.234 0.453
0.30 0.14 0.008
0.219 0.184 0.271
0.24 0.02 0.006
BMI Z body mass index. a Models adjusted for time in study, baseline age, age time in study, male sex, male sex time in study, age male sex, black race, black race age, black race time in study, and black race age time in study. b BMI between 18.5 and 24.9 kg/m2 serves as referent group.
AEP Vol. 16, No. 10 October 2006: 770e776
of progression of mobility impairment and disability (41, 42). Consistent with previous reports (20, 21), our data indicate that older blacks have lower mobility levels than whites. Although the general curvilinear shape of the association between BMI and mobility did not differ between blacks and nonblacks, we found that the level of relative weight associated with maximum mobility was substantially higher for blacks than nonblacks. The reason for this difference is unclear and may involve a complex set of factors. Previous studies suggested that blacks may have more skeletal muscle mass for a given BMI, which may favor better mobility at higher levels of BMI (43, 44). Another possibility is that given the greater prevalence of overweight and obesity throughout adulthood (1), BMIs less than 25 kg/m2 in older age may be more likely to be an indication of age-related weight loss for blacks than whites of similar age. This also may explain why among blacks, mobility levels were less in what is normally considered the optimal range of BMI (18.5 to 25 kg/m2) than in the overweight range (BMI, 25 to 30 kg/m2). To some degree, this finding reflects similar findings for mortality because previous reports found that higher levels of BMI appear to have a less deleterious affect on survival in blacks than whites (45, 46). Our data do not provide clear support for the idea that higher levels of BMI are associated with a greater rate of decrease in mobility in older age. It is possible that the adverse effect of overweight on mobility occurs earlier in life, leading to less mobility when reaching older age, but not to an acceleration of mobility losses at more advanced ages. This is consistent with previous observations that high BMI is associated with first onset of mobility disability (15, 31, 34, 47) and changes in mobility over time in middle-aged and younger old adults, but not older old adults (7, 15, 18, 19). Perhaps because of the restricted age range of the population, there was no evidence for such age dependency in the relationship between BMI and mobility levels in our data. The findings parallel the decreased predictive validity of BMI for mortality in older populations (48, 49), suggesting that the magnitude of the detrimental health effects of overweight and obesity diminish in older age. An important strength of this analysis is that data are from a prospective population-based study of community-dwelling older adults, with a large representation of blacks. Performance-based assessment of mobility is another strength. The relatively long intervals between mobility assessments are a potential limitation because we may have missed changes in mobility that occurred between intervals for participants who died before follow-up assessment. In separate analyses (data not shown), we found a significant increase in mortality risk associated with being underweight (BMI ! 18.5 kg/m2), but not with being overweight or obese. Thus,
Mendes de Leon et al. RELATIVE WEIGHT AND CHANGE IN MOBILITY
775
the lack of a more deleterious prospective effect of high levels of BMI on mobility impairment does not seem to be attributable to selective attrition. Overall, the findings indicate that older adults with either high or low levels of BMI have lower mobility levels than those with BMI in the middle range. However, BMI was not associated with rate of decrease in mobility in this population, suggesting that mobility differences as a function of BMI developed earlier in life. The authors thank Michelle Bos, Holly Hadden, Flavio Lamorticella, and Jennifer Tarpey for coordination of the study; George Dombrowski for data management; and Hye-Jin Nicole Kim and Zhaotai Cui for statistical programming.
REFERENCES 1. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999-2000. JAMA. 2002;288:1723e1727. 2. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002. JAMA. 2004;291:2847e2850. 3. Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer DG, et al. A short physical performance battery assessing lower extremity function: Association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol. 1994;49:M85eM94. 4. Bray GA. Medical consequences of obesity. J Clin Endocrinol Metab. 2004;89:2583e2589. 5. Field AE, Coakley EH, Must A, Spadano JL, Laird N, Dietz WH, et al. Impact of overweight on the risk of developing common chronic diseases during a 10-year period. Arch Intern Med. 2001;161:1581e1586. 6. 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: 1802e1809. 7. Ferraro KF, Booth TL. Age, body mass index, and functional illness. J Gerontol B Psychol Sci Soc Sci. 1999;54:339e348. 8. Friedmann JM, Elasy T, Jensen GL. The relationship between body mass index and self-reported functional limitation among older adults: A gender difference. J Am Geriatr Soc. 2001;49:398e403. 9. Galanos AN, Pieper CF, Cornoni-Huntley JC, Bales CW, Fillenbaum GG. Nutrition and function: Is there a relationship between body mass index and the functional capabilities of community-dwelling elderly? J Am Geriatr Soc. 1994;42:368e373. 10. Visser M, Harris TB, Langlois J, Hannan MT, Roubenoff R, Felson DT, et al. Body fat and skeletal muscle mass in relation to physical disability in very old men and women of the Framingham Heart Study. J Gerontol A Biol Sci Med Sci. 1998;53:214e221. 11. Ferrucci L, Guralnik JM, Cecchi F, Marchionni N, Salani B, Kasper J, et al. Constant hierarchic patterns of physical functioning across seven populations in five countries. Gerontologist. 1998;38:286e294. 12. Lawrence RH, Jette AM. Disentangling the disablement process. J Geront Series B Psychol Sci Soc Sci. 1996;51:173e182. 13. Brock DB, Lemke JH, Branch LG, Evans DA, Berkman LF. Mortality and physical functioning in epidemiologic studies of three older populations. J Aging Soc Policy. 1994;6:21e37. 14. Gregg EW, Cauley JA, Stone K, Thompson TJ, Bauer DC, Cummings SR, et al. Relationship of changes in physical activity and mortality among older women. JAMA. 2003;289:2379e2386. 15. Launer LJ, Harris T, Rumpel C, Madans J. Body mass index, weight change, and risk of mobility disability in middle-aged and older women.
776
Mendes de Leon et al. RELATIVE WEIGHT AND CHANGE IN MOBILITY
The epidemiologic follow-up study of NHANES I. JAMA. 1994; 271:1093e1098. 16. Sternfeld B, Ngo L, Satariano WA, Tager IB. Associations of body composition with physical performance and self-reported functional limitation in elderly men and women. Am J Epidemiol. 2002;156:110e121. 17. Coakley EH, Rimm EB, Colditz G, Kawachi I, Willett W. Predictors of weight change in men: Results from the Health Professionals Follow-Up Study. Int J Obes Relat Metab Disord. 1998;22:89e96. 18. He XZ, Baker DW. Body mass index, physical activity, and the risk of decline in overall health and physical functioning in late middle age. Am J Public Health. 2004;94:1567e1573. 19. Jenkins KR. Body-weight change and physical functioning among young old adults. J Aging Health. 2004;16:248e266. 20. Mendes de Leon CF, Beckett LA, Fillenbaum GG, Brock DB, Branch LG, Evans DA, et al. Black-white differences in risk of becoming disabled and recovering from disability in old age: A longitudinal analysis of two EPESE populations. Am J Epidemiol. 1997;145:488e497. 21. Mendes de Leon CF, Barnes LL, Bienias JL, Skarupski KA, Evans DA. Racial disparities in disability: Recent evidence from self-reported and performance-based disability measures in a population-based study of older adults. J Gerontol B Psychol Sci Soc Sci. 2005;60(Suppl): S263e271.
AEP Vol. 16, No. 10 October 2006: 770e776
33. Tager IB, Haight T, Sternfeld B, Yu Z, van Der LM. Effects of physical activity and body composition on functional limitation in the elderly: Application of the marginal structural model. Epidemiology. 2004;15:479e493. 34. Visser M, Langlois J, Guralnik JM, Cauley JA, Kronmal RA, Robbins J, et al. High body fatness, but not low fat-free mass, predicts disability in older men and women: The Cardiovascular Health Study. Am J Clin Nutr. 1998;68:584e590. 35. Davis MA, Ettinger WH, Neuhaus JM, Cho SA, Hauck WW. The association of knee injury and obesity with unilateral and bilateral osteoarthritis of the knee. Am J Epidemiol. 1989;130:278e288. 36. Dziura J, Mendes de Leon CF, Kasl S, DiPietro L. Can physical activity attenuate aging-related weight loss in older people? The Yale Health and Aging Study, 1982-1994. Am J Epidemiol. 2004;159:759e767. 37. Baumgartner RN, Koehler KM, Gallagher D, Romero L, Heymsfield SB, Ross RR, et al. Epidemiology of sarcopenia among the elderly in New Mexico. Am J Epidemiol. 1998;147:755e763. 38. Ferrucci L, Guralnik JM, Buchner D, Kasper J, Lamb SE, Simonsick EM, et al. Departures from linearity in the relationship between measures of muscular strength and physical performance of the lower extremities: The Women’s Health and Aging Study. J Gerontol A Biol Sci Med Sci. 1997;52:M275eM285.
22. Bienias JL, Beckett LA, Bennett DA, Wilson RS, Evans DA. Design of the Chicago Health and Aging Project (CHAP). J Alzheimers Dis. 2003;5:349e355.
39. Landers KA, Hunter GR, Wetzstein CJ, Bamman MM, Weinsier RL. The interrelationship among muscle mass, strength, and the ability to perform physical tasks of daily living in younger and older women. J Gerontol A Biol Sci Med Sci. 2001;56:B443eB448.
23. Branch LG, Katz S, Kniepmann K, Papsidero JA. A prospective study of functional status among community elders. Am J Public Health. 1984;74:266e268.
40. Rantanen T, Guralnik JM, Foley D, Masaki K, Leveille S, Curb JD, et al. Midlife hand grip strength as a predictor of old age disability. JAMA. 1999;281:558e560.
24. Rosow I, Breslau N. A Guttman health scale for the aged. J Gerontol. 1966;21:556e559.
41. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, et al. Frailty in older adults: Evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56:146e156.
25. Rowland ML. Self-reported weight and height. Am J Clin Nutr. 1990;52:1125e1133. 26. Everson-Rose SA, Mendes de Leon CF, Bienias JL, Wilson RS, Evans DA. Early life conditions and cognitive functioning in later life. Am J Epidemiol. 2003;158:1083e1089. 27. Hauser RM, Warren JR. Socioeconomic Indexes for Occupations: A Review, Update, and Critique. CDE Working Paper no. 96e01. Madison, WI: Center for Demography and Ecology, University of WisconsinMadison; 1996. 28. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: A generalized estimating equation approach. Biometrics. 1988;44:1049e1060. 29. SAS Institute. SAS/STAT Software Changes and Enhancements, release 8.1. Cary, NC: SAS Institute; 2000. 30. Coakley EH, Kawachi I, Manson JE, Speizer FE, Willet WC, Colditz GA. Lower levels of physical functioning are associated with higher body weight among middle-aged and older women. Int J Obes Relat Metab Disord. 1998;22:958e965. 31. Ferraro KF, Su YP, Gretebeck RJ, Black DR, Badylak SF. Body mass index and disability in adulthood: A 20-year panel study. Am J Public Health. 2002;92:834e840. 32. Roubenoff R, Hughes VA. Sarcopenia: Current concepts. J Gerontol A Biol Sci Med Sci. 2000;55:M716eM724.
42. Gill TM, Allore H, Holford TR, Guo Z. The development of insidious disability in activities of daily living among community-living older persons. Am J Med. 2004;117:484e491. 43. Gallagher D, Visser M, De Meersman RE, Sepulveda D, Baumgartner RN, Pierson RN, et al. Appendicular skeletal muscle mass: Effects of age, gender, and ethnicity. J Appl Physiol. 1997;83:229e239. 44. Ortiz O, Russell M, Daley TL, Baumgartner RN, Waki M, Lichtman S, et al. Differences in skeletal muscle and bone mineral mass between black and white females and their relevance to estimates of body composition. Am J Clin Nutr. 1992;55:8e13. 45. Durazo-Arvizu RA, McGee DL, Cooper RS, Liao Y, Luke A. Mortality and optimal body mass index in a sample of the US population. Am J Epidemiol. 1998;147:739e749. 46. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA. 2003;289:187e193. 47. Jenkins KR. Obesity’s effects on the onset of functional impairment among older adults. Gerontologist. 2004;44:206e216. 48. Bender R, Jockel KH, Trautner C, Spraul M, Berger M. Effect of age on excess mortality in obesity. JAMA. 1999;281:1498e1504. 49. Stevens J. Impact of age on associations between weight and mortality. Nutr Rev. 2000;58:129e137.