BMI, Body Composition, and Physical Functioning in Older Adults

Authors


  • The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Department of Medicine and Therapeutics, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, N.T. Hong Kong. E-mail: jeanwoowong@cuhk.edu.hk

Abstract

Objective: Recent studies have emphasized the importance of muscle and fat mass in relation to age-related decline in physical function. Our objective was to determine whether BMI, as a surrogate measurement of fat mass, may be used as a measure of risk factor for physical functioning in older adults and whether body composition measurements confer any advantage over BMI.

Research Methods and Procedures: Four thousand men and women ≥65 years of age living in the community, stratified by age and sex, underwent the following measurements: body composition by DXA; grip strength; and timed 6-m walk. Subjects were grouped into five categories of BMI using Asian criteria for health-related risks, and between-group differences in physical performance measures and body composition were analyzed using analysis of covariance adjusting for age, physical activity level, and presence of chronic disease.

Results: Subjects in the two obese categories had a significantly greater number of instrumental activities of daily living (IADL) impairments compared with the underweight and normal-weight groups. Those with BMI ≥30 kg/m2 had the worst walking performance, and the groups with BMI in the normal and overweight range had optimal performance. Fat mass, but not appendicular muscle mass, was associated with walking speed after adjusting for BMI.

Discussion: Fat mass seems to be a more important factor than appendicular muscle mass in determining walking speed in community-living older adults, even after adjusting for BMI.

Introduction

The etiology of age-related decline in physical function has been studied in terms of changing body composition, in particular the role of sarcopenia (1, 2, 3, 4, 5) and, in recent years, the condition of sarcopenic obesity (6, 7, 8, 9, 10, 11, 12). Increasing emphasis has been placed on the definitions and prevalence of sarcopenia and sarcopenic obesity (9, 13). In parallel, suitable methods of body composition measurement for this purpose have been examined, such as calculations derived from measurement of skinfold thickness (10), bioimpedance (8), and DXA (14). Measurements of body composition, in terms of muscle mass and fat mass, are thought to give a more accurate reflection of physical function compared with BMI, because, for any given value of BMI, the ratio of fat to muscle mass may vary. Such measurements would be important both from an individual health assessment perspective and also from a population health perspective. BMI has been widely used as a health indicator for both perspectives, having been shown to be related to mortality and morbidity in populations of diverse cultures and ethnicity, with established health-related cut-off points (15, 16, 17). It is easily measured and does not require costly equipment, unlike the measurement of body composition, so that it is particularly suitable for population surveys to assess health risk and changing time trends in health risks as a monitor of health promotion efforts. In this study, we examined whether the current cut-off points for health-related risks for BMI may also be applied to physical functioning, comparing BMI with body composition measurements using DXA, to determine whether body composition measurements confer any advantage over BMI in the association with physical functioning.

Research Methods and Procedures

Through recruitment notices placed in community centers for the elderly and housing estates, 2000 men and 2000 women ≥65 years of age living in the community were invited to attend a health check carried out in the School of Public Health of the Chinese University of Hong Kong. Several talks were also given at these centers explaining the purpose, procedures, and studies to be carried out. Subjects who were ambulant and living at home were included in the study. Those with terminal illness or dependent on oxygen were excluded. The sample was stratified by age group and sex so that ∼33% were in each of these age groups: 65 to 69, 70 to 74, and 75+ years for both men and women. The study was approved by the Clinical Research Ethics Committee of the Chinese University of Hong Kong.

Information regarding the presence or absence of disease based on subjects’ report of diagnosis by their doctors, physical activity level, and instrumental activities of daily living was obtained as part of a questionnaire administered by an interviewer. The list of diseases included were stroke, coronary heart disease/myocardial infarction, angina, Parkinson's disease, osteoporosis, arthritis, chronic obstructive pulmonary disease, cataract/glaucoma, cancer, thyroid disease, diabetes, and hypertension. Physical activity level was assessed using the Physical Activity Scale of the Elderly. This is a 12-item scale measuring the average number of hours per day spent in leisure, household, and occupational physical activities over the previous 7-day period. Activity weights for each item were determined based on the amount of energy expended, and each item score was calculated by multiplying the activity weight by activity daily frequency. A summary score of all of the items reflect the daily physical activity level (18). Activity of daily living questions covered difficulty in performing the following activities: walking two to three blocks outside on level ground, climbing up 10 steps without resting, preparing own meals, doing heavy housework such as scrubbing floors or washing windows, and doing own shopping for groceries or clothes.

Body weight was measured, with subjects wearing a light gown, with the Physician Balance Beam Scale (Healthometer, Bridgeview, IL). Height was measured with the Holtain Harpenden stadiometer (Holtain, Crosswell, UK). Body composition was measured using DXA (Hologic 4500 W, software version 11.2; Hologic, Inc., Waltham, MA). Total appendicular skeletal muscle mass was calculated as the sum of lean body mass minus bone mineral content of both arms and legs.

Grip strength was measured using a dynamometer (JAMAR hand dynamometer 5030 J1; Sammons Preston, Bolingbrook, IL). Two readings were taken from each side, and the average value of the right and left was used for analysis. Subjects were asked to walk 6 m twice by following a straight line painted on the floor of a long corridor. The number of steps and the time taken (in seconds) were recorded. The values for the faster of the two trials were used.

The χ2 test or ANOVA was used to examine differences in categorical or continuous variables by age group. Subjects were grouped into different BMI categories, separately for men and women, to represent underweight (<18.5 kg/m2), normal-weight (18.5 to <23 kg/m2), overweight (23 to 24.9 kg/m2), and obesity I (25 to 29.9 kg/m2) and obesity II (≥30 kg/m2), using Asian criteria (19). The differences in body composition, instrumental activities of daily living (IADL)1 impairments, and physical performance measures between BMI groups were examined using analysis of covariance or logistic regression, adjusting for age, physical activity, and the number of chronic diseases. The analysis was repeated by quintiles of fat mass adjusted for BMI and by quintiles of appendicular mass adjusted for height and fat mass. Linear and quadratic trends (U-shape relationship) were examined by analysis of covariance for both continuous and categorical variables. All statistical analyses were performed using SAS, V.9.1 (SAS Institute, Inc., Cary, NC). An α level of 5% was used as the level of significance.

Results

BMI, body composition, physical performance measures, and other characteristics are shown by age groups separately for men (Table 1) and women (Table 2). The percentage of subjects with an underweight BMI (<18.5 kg/m2) increased with age, whereas that for obesity decreased with age in both men and women. Fat mass was similar for the three age groups in men but was lower in older women compared with younger women. Grip strength and walking speed also declined with age in both men and women. The number of IADL impairments and chronic diseases increased with age in both men and women, whereas the level of physical activity as measured by the Physical Activity Scale of the Elderly score decreased. Differences in sex, body composition, IADL, grip strength, and walking performance were examined by BMI groups adjusting for age, physical activity, and presence of chronic diseases (Table 3). The percentage of women in each BMI category was greater with increasing BMI. Subjects with BMI <18.5 kg/m2 had the lowest appendicular muscle mass/fat mass ratio and fat mass/body weight ratio. These values increased with successively higher BMI categories. Subjects in the two obese categories had a significantly greater number of IADL impairments compared with the underweight and normal-weight groups, although the underweight group had the lowest grip strength. The group with BMI ≥30 kg/m2 had the worst walking performance (longest time to walk 6 m and slowest walking speed). The underweight (BMI <18.5 kg/m2) and obese (BMI 25 to 29.9 kg/m2) groups had slower walking speed compared with the groups with normal (BMI 18.5 to <23 kg/m2) and overweight (BMI 23 to 24.9 kg/m2) categories, suggesting a U-shaped relationship between physical performance and BMI (p < 0.0001 testing for quadratic trend).

Table 1. . Characteristics by age (men)
VariablesAge group (yrs)p*
 65 to 6970 to 7475+ 
  • IADL, instrumental activities of living; SD, standard deviation. Values are mean (SD).

  • *

    p value of χ2 for categorical or ANOVA for continuous variables.

N664708628 
No. of IADL impairments   <0.0001
 089.31%85.69%74.84% 
 1 to 29.49%13.17%20.86% 
 3 to 51.20%1.13%4.30% 
BMI   0.0355
 Underweight (<18.5 kg/m2)4.67%4.66%8.12% 
 Normal-weight (18.5 to <23 kg/m2)37.05%37.29%39.81% 
 Overweight (23 to 24.9 kg/m2)26.81%26.84%24.84% 
 Obese I (25 to 29.9 kg/m2)28.61%29.52%25.96% 
 Obese II (≥30 kg/m2)2.86%1.69%1.27% 
 Mean (SD)23.68 (3.13)23.60 (3.00)23.03 (3.23)0.0002
Average of left/right grip strength (kg)33.79 (6.10)31.35 (5.86)28.31 (6.27)<0.0001
Minimum time to walk 6 m (s)5.41 (1.09)5.92 (2.10)6.47 (1.74)<0.0001
Walk speed using best times (m/s)1.15 (0.22)1.07 (0.21)0.98 (0.22)<0.0001
No. of chronic diseases   <0.0001
 024.40%14.69%12.58% 
 1 to 252.41%52.83%50.16% 
 ≥323.19%32.49%37.26% 
Physical Activity Scale of the Elderly105.69 (54.06)99.25 (47.88)86.18 (46.77)<0.0001
Appendicular lean muscle mass (kg)19.82 (2.67)19.22 (2.46)18.44 (2.60)<0.0001
Fat mass (kg)15.40 (4.66)15.46 (4.60)15.00 (4.72)0.1539
Fat mass/body weight ratio0.24 (0.05)0.24 (0.05)0.24 (0.05)0.1157
Table 2. . Characteristics by age (women)
VariablesAge group (yrs)p*
 65 to 6970 to 7475+ 
  • IADL, instrumental activities of living; SD, standard deviation. Values are mean (SD).

  • *

    p value of χ2 for categorical or ANOVA for continuous variables.

N669665666 
No. of IADL impairments   <0.0001
 074.14%66.87%53.98% 
 1 to 223.02%29.37%35.94% 
 3 to 52.84%3.77%10.08% 
BMI   0.0001
 Underweight (<18.5 kg/m2)3.44%3.31%8.26% 
 Normal-weight (18.5 to <23 kg/m2)36.17%33.38%37.09% 
 Overweight (23 to 24.9 kg/m2)23.62%23.61%24.17% 
 Obese I (25 to 29.9 kg/m2)32.74%34.44%27.03% 
 Obese II (≥30 kg/m2)4.04%5.26%3.45% 
 Mean (SD)24.06 (3.28)24.31 (3.51)23.39 (3.48)<0.0001
Average of left/right grip strength (kg)21.68 (4.23)20.42 (4.06)18.64 (3.78)<0.0001
Minimum time to walk 6 m (s)6.05 (1.34)6.51 (1.77)7.48 (2.70)<0.0001
Walk speed using best times (m/s)1.03 (0.19)0.97 (0.20)0.87 (0.22)<0.0001
No. of chronic diseases   <0.0001
 022.57%15.04%11.11% 
 1 to 253.06%54.59%52.56% 
 ≥324.36%30.38%36.34% 
Physical Activity Scale of the Elderly94.98 (35.23)87.90 (29.96)73.17 (30.24)<0.0001
Appendicular lean muscle mass (kg)14.18 (1.89)13.98 (1.92)13.27 (1.79)<0.0001
Fat mass (kg)19.54 (4.95)19.61 (5.10)18.14 (5.46)<0.0001
Fat mass/body weight ratio0.35 (0.05)0.35 (0.05)0.34 (0.06)0.0016
Table 3. . Prevalence or estimated mean (SE) by BMI groups, analysis of covariance, adjusting for age, physical activity, and number of chronic diseases
VariablesBMI group (kg/m2)p difference*p trend*p quadratic*
 Underweight (<18.5)Normal-weight (18.5 to 22.9)Overweight (23 to 24.9)Obese I (25 to 29.9)Obese II (≥30)   
  • SE, standard error; IADL, instrumental activities of living. Values are estimated mean (SE).

  • *

    p value of binary or ordinal logistic regression for categorical variables or analysis of covariance for continuous variables, adjusting for age, physical activity, and number of chronic diseases.

  • p < 0.05 comparing normal-weight, overweight, Obese I, and Obese II with underweight.

  • p < 0.05 comparing overweight, Obese I, and Obese II with normal-weight.

  • §

    p < 0.05 comparing Obese I and Obese II with overweight.

  • p < 0.05 comparing Obese II with Obese I.

N215147110001190124   
Women46.51%48.33%47.60%52.77% ,§68.55% ,,§,<0.0001<0.00010.0111
No. of IADL impairments   bcbc0.00040.00300.0036
 070.09%76.92%76.8%70.23%67.74%   
 1 to 224.3%19.95%18.9%26.16%24.19%   
 3 to 55.61%3.13%4.3%3.62%8.06%   
Average of left/right grip strength (kg)24.38 (0.51)25.68 (0.21)26.71 (0.26) ,26.37 (0.24) ,24.33 (0.67) ,§,<0.00010.7299<0.0001
Minimum time to walk 6 m (s)6.46 (0.13)6.14 (0.05)6.19 (0.07)6.39 (0.06) ,§6.95 (0.17) ,,§,<0.00010.0038<0.0001
Walk speed using best times (m/s)1.02 (0.01)1.04 (0.01)1.03 (0.01)1.00 (0.01) ,§0.9 (0.02) ,,§,<0.0001<0.0001<0.0001
Appendicular lean muscle mass (kg)14.02 (0.23)15.46 (0.09)16.86 (0.11) ,17.82 (0.11) ,,§18.60 (0.29) ,,§,<0.0001<0.00010.0226
Fat mass (kg)7.94 (0.22)13.91 (0.09)17.45 (0.11) ,21.32 (0.11) ,,§28.03 (0.29) ,,§,<0.0001<0.00010.0186
Fat mass/body weight0.19 (0.00)0.27 (0.00)0.30 (0.00) ,0.33 (0.00) ,,§0.38 (0.01) ,,§,<0.0001<0.0001<0.0001

Physical function was examined by quintiles of fat mass and appendicular lean muscle mass after adjusting for BMI (Tables 4 and 5). Increasing fat mass was associated with worsened physical function, whereas no association between appendicular muscle mass and walking speed was observed. The results were essentially the same when the relationship between physical function and fat mass was examined without adjustment for BMI (Table 6). However, both high fat mass and low appendicular lean mass were significantly associated with IADL impairments and low grip strength.

Table 4. . Estimated mean (SE) by quintile of fat mass, analysis of covariance, adjusting for age, physical activity, BMI, and number of chronic diseases
VariablesFat mass (kg)p difference*p trend*p quadratic*
 <12.9 (I)12.9 to 15.7 (II)15.8 to 18.2 (III)18.3 to 21.4 (IV)≥21.5 (V)   
  • SE, standard error; IADL, instrumental activities of living.

  • *

    p value of binary or ordinal logistic regression for categorical variables or analysis of covariance for continuous variables, adjusting for age, physical activity, BMI, and number of chronic diseases.

  • p < 0.05 comparing II, III, IV, and V with I.

  • p < 0.05 comparing III, IV, and V with II.

  • §

    p < 0.05 comparing IV and V with III.

  • p < 0.05 comparing V with IV.

N800800800800800   
Women26.63%38.75%47.50% ,59.00% ,,§78.13% ,,§,<0.0001<0.00010.0007
No. of IADL impairments ,,,§,<0.0001<0.00010.0507
 079.72%77.10%77.50%73.63%63.28%   
 1 to 216.90%19.27%19.13%22.00%32.21%   
 3 to 53.38%3.63%3.38%4.38%4.51%   
Average of left/right grip strength (kg)31.56 (0.33)28.51 (0.27)26.49 (0.26) ,24.04 (0.28) ,,§19.10 (0.34) ,,§,<0.0001<0.0001<0.0001
Average step length (m)0.56 (0.00)0.56 (0.00)0.54 (0.00) ,0.53 (0.00) ,,§0.51 (0.00) ,,§,<0.0001<0.00010.0041
Minimum time to walk 6 m (s)6.20 (0.09)6.12 (0.07)6.18 (0.07)6.31 (0.07)6.55 (0.09) ,,§,0.00410.00820.0036
Walk speed using best times (m/s)1.05 (0.01)1.05 (0.01)1.03 (0.01)1.01 (0.01) ,0.96 (0.01) ,,§,<0.0001<0.00010.0007
Table 5. . Estimated mean (SE) by quintile of appendicular lean muscle mass, analysis of covariance, adjusting for age, physical activity, height, fat mass, and number of chronic diseases
VariablesAppendicular lean muscle mass (kg)p difference*p trend*p quadratic*
 <13.2 (I)13.2 to 14.9 (II)15.0 to 17.2 (III)17.3 to 19.7 (IV)≥19.8 (V)   
  • SE, standard error; IADL, instrumental activities of living.

  • *

    p value of binary or ordinal logistic regression for categorical variables or analysis of covariance for continuous variables, adjusting for age, physical activity, height, fat mass, and number of chronic diseases.

  • p < 0.05 comparing II, III, IV, and V with I.

  • p < 0.05 comparing III, IV, and V with II.

  • §

    p < 0.05 comparing IV and V with III.

  • p < 0.05 comparing V with IV.

N800800800800800   
Women99.13%87.75%52.63% ,9.38% ,,§1.13% ,,§,<0.0001<0.00010.0004
No. of IADL impairments  ,,,§,<0.00010.00830.0487
 064.13%64.91%73.50%81.23%87.48%   
 1 to 230.25%29.20%22.63%15.52%11.89%   
 3 to 55.63%5.89%3.88%3.25%0.63%   
Average of left/right grip strength (kg)21.30 (0.22)23.03 (0.19)25.49 (0.18) ,28.81 (0.19) ,,§31.44 (0.23) ,,§,<0.0001<0.0001<0.0001
Average step length (m)6.23 (0.09)6.34 (0.07)6.31 (0.07)6.20 (0.08)6.25 (0.09)0.56360.75180.4429
Minimum time to walk 6 m (s)1.02 (0.01)1.01 (0.01)1.01 (0.01)1.03 (0.01)1.04 (0.01)0.18840.12000.1038
Table 6. . Estimated mean (SE) by quintile of fat mass, analysis of covariance, adjusting for age, physical activity, and number of chronic diseases
VariablesFat mass (kg)p difference*p trend*p quadratic*
 <12.9 (I)12.9 to 15.7 (II)15.8 to 18.2 (III)18.3 to 21.4 (IV)≥21.5 (V)   
  • SE, standard error; IADL, instrumental activities of living.

  • *

    p value of binary or ordinal logistic regression for categorical variables or analysis of covariance for continuous variables, adjusting for age, physical activity, and number of chronic diseases.

  • p < 0.05 comparing II, III, IV, and V with I.

  • p < 0.05 comparing III, IV, and V with II.

  • §

    p < 0.05 comparing IV and V with III.

  • p < 0.05 comparing V with IV.

N800800800800800   
Women26.63%38.75%47.50% ,59.00% ,,§78.13% ,,§,<0.0001<0.00010.0143
No. of IADL impairments  ,,§,<0.0001<0.00010.0562
 079.72%77.10%77.50%73.63%63.28%   
 1 to 216.90%19.27%19.13%22.00%32.21%   
 3 to 53.38%3.63%3.38%4.38%4.51%   
Average of left/right grip strength (kg)27.48 (0.27)26.92 (0.27)26.39 (0.27)25.42 (0.28) ,,§23.30 (0.28) ,,§,<0.0001<0.00010.0002
Average step length (m)0.56 (0.00)0.55 (0.00)0.54 (0.00) ,0.53 (0.00) ,,§0.51 (0.00) ,,§,<0.0001<0.00010.0050
Minimum time to walk 6 m (s)6.18 (0.07)6.11 (0.07)6.18 (0.07)6.32 (0.07)6.58 (0.07) ,,§,<0.0001<0.00010.0034
N1.05 (0.01)1.04 (0.01)1.03 (0.01)1.01 (0.01) ,0.97 (0.01) ,,§,<0.0001<0.00010.0008

Discussion

Recent studies have attempted to define sarcopenia and sarcopenic obesity and to identify cut-off points relating to increased risk of physical disability (5, 9). There is a suggestion from longitudinal studies that the effect of sarcopenia in the development of disability may be smaller than that observed in cross-sectional studies and that it may not be a risk factor for disability in the absence of obesity (20). This survey showed that fat mass and BMI are main factors contributing to physical performance as measured by walking speed. This finding is compatible with the study in U.S. subjects 55+ years of age using bioimpedance (8) and in those ≥70 years of age using predictive equations from skinfold measurements (10). In the latter study, where the age group of subjects is similar to that in this study, a quadratic relationship between BMI and body composition and functional limitations was also observed. The at-risk BMI values adopted in this study are lower than those for whites, among whom the cut-off values for overweight and obesity are 25 and 30 kg/m2, respectively. Asians are considered to have higher cardiovascular risk at a given BMI, as a result of a higher percentage of body fat (19). The BMI values associated with optimum physical function (18.5 to 24.9 kg/m2) coincide with values associated with lowest risk of morbidity, as shown in previous studies in Chinese populations relating BMI to health outcomes (21, 22) and mobility decline (23). This observation is not unexpected, given the close inverse relationship between walking speed and health-related outcomes in well-functioning older people (24, 25). The findings of this study emphasize yet another adverse effect of obesity in the elderly apart from increased risk of various diseases—that of functional limitation. The declining grip strength with increasing fat mass suggests that muscle function may be adversely affected and may partly account for functional limitations. The association between grip strength and appendicular muscle mass emphasizes this point. In obese elderly people, an exercise component must be included to maintain or increase lean muscle mass and bone mineral density. Such regimens have been shown to result in a reduction in fat mass without changes in fat-free mass, increase physical performance, and improve quality of life (26).

There are limitations in this study because of its cross-sectional design. It is possible that a longitudinal study may have different outcomes. For example, in the U.S. Cardiovascular Health Study of 5036 subjects >65 years of age followed for 8 years, it was observed that the effect of sarcopenia was smaller in longitudinal compared with cross-sectional studies. Nevertheless, the conclusion drawn from this cross-sectional study is similar, in that sarcopenia may not be a risk factor for the development of disability in the absence of obesity. Another limitation is that we relied on subjects’ report of the presence of diseases rather than on detailed physical examination and study. Therefore, some asymptomatic diseases such as diabetes may have been missed. The strength of the study lies in the large number of subjects with body composition measurements by DXA and the detailed assessment of physical activity using the Physical Activity Scale of the Elderly score. These measurements are major potential confounders in examining physical performance measure as an outcome. Despite the limitations, the study confirmed the decrease in muscle and increase in fat mass, with age and female sex as underlying factors in age and sex differences in frailty. Fat mass seems to be a stronger determinant of minimum walk time or walk speed than appendicular muscle mass and remains a determinant even after adjustment for BMI, thus supporting the idea that BMI may not fully reflect fat mass.

Acknowledgments

This study was supported in part by the Research Grants Council of Hong Kong (CUHK4101/02M).

Footnotes

  • 1

    Nonstandard abbreviation: IADL, instrumental activities of living.

Ancillary