Sarcopenia Characteristics Are Associated with Incident Mobility Limitations in African Caribbean Men: The Tobago Longitudinal Study of Aging

Abstract

Background

Sarcopenia varies by ethnicity, and has a major impact on health in older adults. However, little is known about sarcopenia characteristics in African ancestry populations outside the United States. We examined sarcopenia characteristics in 2,142 African Caribbean men aged 59.0 ± 10.4 years (range: 40–92 years) in Tobago, and their association with incident mobility limitations in those aged 55+ (n = 738).

Methods

Body mass index (BMI), grip strength, dual-x-ray absorptiometry (DXA) appendicular lean mass (ALM), and self-reported mobility limitations were measured at baseline, and 6 years later. Change in sarcopenia characteristics, including grip strength, grip strength/BMI, ALMBMI, and ALM/ht², were determined. Foundations for the National Institutes of Health Sarcopenia Project (FNIH) and European Working Group for Sarcopenia in Older People 2 (EWGSOP2) cut-points were also examined. Odds ratios (OR) and 95% confidence intervals (CI) for mobility limitation were calculated using multivariable linear regression models adjusted for covariates.

Results

Overall, sarcopenia prevalence was quite low using the FNIH (0.3%) and EWGSOP2 (0.6%) operational cut-points, but was higher in those aged 75+ (2.1% [FNIH] and 3.7% [EWGSOP2]). Prevalence was also higher when based on “weakness”, versus “low ALM.” When sarcopenia markers were examined separately, baseline levels, but not changes, were associated with incident mobility limitations. Baseline grip strength/BMI was a particularly strong risk factor for incident mobility limitations (OR per SD: 0.50; 95% CI: 0.37–0.68).

Conclusions

Our findings suggest that grip strength normalized to body mass, measured at one time point, may be a particularly useful phenotype for identifying African Caribbean men at risk for future mobility limitations.

“Sarcopenia,” the age-related loss of muscle mass, strength, and function, varies by age, sex, and ethnicity and has a major impact on health and functional status among older adults (1–3). Sarcopenia and its clinical consequences have an enormous impact on health care costs. Compared with Caucasian Americans, African Americans generally have a lower prevalence of sarcopenia (4,5). However, little is known about sarcopenia and its consequences in African ancestry populations outside the United States, especially in developing, low- and middle-income countries. The population aged ≥60 years and older is estimated to be growing nearly three times faster in low- to middle-income countries compared with higher-income nations (6,7). Thus, a better understanding of the epidemiology, risk factors, and clinical consequences of sarcopenia in low- and middle-income countries is urgently needed.

In the current study, we determined the prevalence of sarcopenia in a large cohort of African Caribbean men using definitions and cut-points proposed by The Foundations for the National Institutes of Health (FNIH) Sarcopenia Project and Second European Working Group on Sarcopenia in Older People (EWGSOP2). We also measured changes in sarcopenia characteristics over a 6-year period and examined their association with mobility limitations.

Methods

Study Population

Originally, 3,170 men aged ≥40 years and older on the Caribbean island of Tobago between 1998 and 2003 for a population-based prostate-specific antigen screening study (8). Approximately 60% of all age-eligible men on the island participated and participation was representative of the island parishes. Eligibility criteria included ambulatory, noninstitutionalized, adults without terminal illness. Written informed consent was obtained from all study participants and the study was approved by the Institutional Review Boards of the University of Pittsburgh and the Tobago Division of Health and Social Services.

Between 2004 and 2007, 70% of survivors (9) returned for a follow-up clinic visit (baseline for the current analyses), during which body composition using dual-x-ray absorptiometry (DXA) and mobility limitations were assessed for the first time (baseline for the current analysis). At the time of this visit, 451 new participants aged ≥40 years and older were recruited using similar criteria as the baseline exam (9). Between 2010 and 2013, all men were invited to attend a follow-up visit, which included a repeat DXA scan and reassessment of mobility limitations (10).

There were three analytic samples included in this paper. First, the present cross-sectional (baseline) analyses included the n = 2,142 men who had complete DXA, grip strength and mobility data from the 2004–2007 visit. Second, change in sarcopenia characteristics was based on the n = 1,536 men with complete baseline and follow-up DXA and grip strength data (76.3% of survivors). A comparison of those with and without valid longitudinal change data (but were included in cross-sectional analyses) is presented in Supplementary Table 1. Third, due to the low incidence of mobility limitations in those aged less than 55 at baseline (2.8%), all incident analyses were limited to those aged 55 years and older, and those without mobility limitations at baseline (analytic sample for incident analyses: n = 738).

Anthropometry, Body Composition, and Grip Strength Measures

Height (cm) was measured using a wall-mounted stadiometer and body weight (kg) was measured without shoes on a balance beam scale using identical methods at the baseline and follow-up. Height and weight were used to calculate body mass index (BMI, kg/m²).

Whole body DXA measurements were made at baseline and follow-up using the same QDR 4500W densitometer (Hologic, Inc.). Scans were analyzed for total fat mass as well as total and appendicular lean mass using QDR software version 8.26a.

Grip strength was measured using a Jamar handheld dynamometer using identical methods at baseline and follow-up. Participants stood with the elbow of the hand being tested tucked to their side with the forearm extended and were asked to squeeze the dynamometer as hard as they could. Each hand was tested two times. For baseline, the maximum grip strength from among the four tests was used in analyses. For follow-up, the maximum grip strength of the hand that was used for baseline was used in analyses.

Sarcopenia Characteristics

Sarcopenia characteristics were defined using recommendations from the Foundations for the National Institutes of Health Sarcopenia Project (FNIH) including: grip strength, grip strength normalized to BMI (gripBMI), and appendicular lean mass normalized to BMI (ALMBMI) (11–13). We also calculated an additional sarcopenia characteristic recommended by the European Working Group on Sarcopenia in Older People (EWGSOP2) (14): ALM normalized to height (m) squared (ALM/ht²). Sarcopenia cut-points were defined as follows: FNIH: weakness = grip strength <26 kg, low gripBMI = gripBMI <1.0, and low ALMBMI = ALMBMI <0.789; and EWGSOP2: weakness = grip strength <27 kg, and low ALM/ht² = ALM/ht² <7.0 (11,14).

Mobility Limitations

As described previously (15), mobility questions were asked in-person by trained interviewers. We asked participants if they had any difficulty walking 2 or 3 blocks outside on level ground or climbing up 10 steps without resting. Participants reporting any difficulty or not performing either of these tasks because of a health or physical problem were categorized as having a mobility limitation. Men with prevalent mobility limitation at the baseline exam (N = 83) were excluded from incident mobility analyses.

Other Measures

Information on lifestyle habits, demographic factors, medical conditions, and medication use was also collected using interviewer-administered questionnaires. Hypertension was defined as having systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mm or self-reported or taking hypertension medication or physician diagnoses. Diabetes was defined as self-reported treatment or fasting blood glucose ≥126 mmol/L. Cardiovascular disease was defined as self-reported history of heart attack, stroke, or congestive heart failure. Smoking was recorded as never (<100 cigarettes or five packs lifetime) or ever smoked (≥100 or five packs cigarettes lifetime). Alcohol consumption was categorized as: >1 drink/day versus ≤1 drink/day. Self-reported frequency of walking outside the home during the past week was recorded and the questions were phrased to assess walking for exercise or leisure. Participants were dichotomized into those who walked “often” (5–7 days per week) to those who walked “sometimes” or less (<5 days per week).

Statistical Analyses

Means and standard deviations or frequencies and percentages were calculated for baseline characteristics. Comparisons of baseline characteristics between men with and without mobility limitations, and those include versus not included in longitudinal analyses, were made using t tests, nonparametric, and chi-squared tests as appropriate. Unadjusted baseline, and annualized percentage rates of change in sarcopenia characteristics were compared between men with and without mobility limitations using paired t- or nonparametric tests. Trends for changes in sarcopenia characteristics by 5-year age groups were tested for using analysis of variance. Further, to determine if there were points across the adult life span where changes accelerated/decelerated, pairwise t tests were used to test the difference between individual 5-year age groups, using Tukey’s test to account for multiple comparisons (α <0.05). To determine if those with baseline mobility limitations had accelerated declines in sarcopenia characteristics, these data were also stratified by baseline mobility status. We also calculated the spearman correlation coefficients among sarcopenia measures, and between sarcopenia measures and BMI.

Unadjusted baseline levels and change in sarcopenia characteristics were compared between those who did and not develop incident mobility limitations using t tests. The adjusted likelihood of developing incident mobility limitations associated with sarcopenia characteristics was assessed using separate multivariable logistic regression models for each sarcopenia characteristic or cut-point. Odds ratios (OR) and 95% confidence intervals (CI) were expressed per standard deviation difference in each sarcopenia characteristic. The following covariates were evaluated and were included in final models if they were different between men with and without mobility limitation at p < .15: age, education, marital status, drinks/week, smoking, self-reported cardiovascular disease, arthritis/gout, type 2 diabetes, hypertension, and walking habits. An α level of 0.05 was otherwise used for statistical tests and all analyses were performed using SAS v.9.4 (Cary, NC).

Results

Characteristics at Baseline

The men included in the cross-sectional analyses (n = 2,142) were aged 59.0 ± 10.4 years (range: 40–92 years) with a BMI of 27.4 ± 4.4 (Table 1). Those with mobility limitations at baseline were significantly older, less likely to have completed high school, walked less, and had a higher prevalence of hypertension, cardiovascular disease, diabetes and arthritis/gout (all p < .05).

Sarcopenia and Mobility Limitations at Baseline

Overall, those with mobility limitations also had significantly lower levels of all sarcopenia characteristics and a higher prevalence of “sarcopenia” defined by FNIH and EWGSOP2 cut-points (Table 1). In all men, the prevalence of weakness/BMI was over twice that of low ALMBMI (4.4% vs 1.8%). Sarcopenia characteristics and the prevalence of sarcopenia and mobility limitations by 5-year age group can be found in Supplementary Table 2. As expected, grip strength, ALMBMI, and ALM/ht² were all significantly lower, and the prevalence of both sarcopenia and mobility limitations were significantly higher at older ages (p < .01) The prevalence of sarcopenia and mobility limitations were significantly higher at older ages (p < .01). Prior to age 65, the prevalence of sarcopenia was similar using cut-points for ALM compared with grip strength; however, after age 65, the prevalence was much higher for cut-points using grip strength compared with ALM.

Changes in Sarcopenia Characteristics

A comparison of those with and without valid follow-up data (but were included in cross-sectional analyses) is presented in Supplementary Table 1. Compared with ALMBMI, grip/BMI declined 2–2.5 times faster per year across all ages—until age 75+, where grip/BMI declined 4 times faster than ALMBMI. Men aged 55–59 years had faster rates of decline in ALMBMI compared with those aged 50–54 years, indicating that ALMBMI decline accelerates after age 55. Men aged 55–59 years also appeared to have faster rates of decline in grip strength/BMI than those aged 50–54 years (p = .02), but this difference was no longer significant after considering multiple comparisons.

Incident Mobility Limitations

Baseline characteristics of the 738 men who were free of mobility limitations, and aged 55 years and older at baseline can be found in Supplementary Table 4. Of these men, 11.7% developed mobility limitations after 6 years. Unadjusted baseline, and change in sarcopenia characteristics by incident mobility status are displayed in Table 2. Table 3 shows the adjusted odds of developing incident mobility limitations associated with baseline, and 6-year change in the sarcopenia characteristics. Except for ALM/ht2, concurrent change in sarcopenia characteristics were not associated with incident mobility limitations (Table 3). Conversely, baseline levels of sarcopenia characteristics were strongly associated with mobility limitations. Baseline grip Strength/BMI was a particularly strong predictor of developing incident mobility limitations. Higher ALM/ht² at baseline was associated with higher odds of developing incident mobility limitations; but, losing ALM/ht² was associated with greater odds of incident mobility limitations. However, ALM/ht² was strongly associated with BMI (r = .78, Supplementary Table 5); and when included in the same model—only higher BMI was associated with incident mobility limitations (odds ratio [OR]: 1.17, 95% confidence interval [CI]: 1.06–1.28), while the association with ALM/ht2 was attenuated to nonsignificance and flipped to below 1.0 (0.80, 0.53–1.21). The Variance Inflation Factor for both ALM/ht² and BMI was 2.7, indicating estimates were not inflated due to collinearity. Finally, due to the low prevalence of sarcopenia based on FNIH and EWGSOP2 cut-points (Table 3), these definitions corresponded with large confidence intervals, and only weakness/BMI (FNIH) and weakness (EWGSOP2) were associated higher odds of incident mobility limitations.

Discussion

The current analyses yielded several unique findings. First, we found that baseline sarcopenia characteristics, but not change in these measures, such as grip strength, and appendicular lean mass normalized to BMI (ALMBMI) were strongly associated with the development of incident mobility limitations in this cohort of African Caribbean men. Counterintuitively, while higher baseline ALM/ht² was associated with incident mobility limitations, losing ALM/ht² was associated with greater odds of incident mobility limitations. However, this discrepancy was explained by the strong relationship between ALM/ht² and BMI, and it appears this association was reflective of the well-known association between obesity and mobility limitations (16). Baseline grip strength/BMI was a particularly strong predictor of incident mobility limitations; in fact, compared with measures based on ALM, measures including grip strength were stronger, or equally as strong, predictors of future mobility limitations. This is an important finding, as grip strength and BMI are quickly, easily, and cheaply measured in clinical settings (17)—especially compared with DXA measured ALM. Thus, grip strength/BMI may be a particularly useful clinical phenotype in this population. The fact that concurrent changes in grip strength were not associated with mobility limitations may be due to grip strength being reflective of upper body and not lower extremity physical function. Indeed, the only other observational study to examine concurrent change in grip strength and mobility yielded similar findings. Specifically, the Invecchiare in Chianti study examined baseline and 3-year change in grip strength in relation to 3- and 6-year change in mobility disability and found that baseline, but not 3-year change in grip strength predicted 3- and 6-year declines in mobility (18). Thus, grip strength/BMI measured at one point in time, and not concurrent change, appears to be predictive of change in mobility status.

In this cohort, the prevalence of “sarcopenia” was quite low. For example, in the men aged 65 years and older included in the original FNIH analyses, the prevalence of sarcopenia was 1.3% (19) compared with 0.9% of men aged 65 years and older in our study. In a study of West African Gambian Men aged 40–75 years, the prevalence of sarcopenia using the FNIH ALM cut-point was 20% (20), compared with only 1.8% with low ALM in these African Caribbean men. Similarly, men age 60+ from the Health and Retirement Study (87% white, 10% black) had a sarcopenia prevalence of 8% and 15% based on low grip strength and low grip strength/BMI, respectively (21)—compared with only 6% and 9% of men aged 60+ in the current study. The low prevalence of sarcopenia in this population may be due to environmental (eg, high sunshine exposure increasing vitamin D (22)), socioeconomic status, cultural differences, access to medical care (eg, universal health system in Tobago), genetic admixture, and/or lifestyle and dietary factors (higher occupational physical activity and higher fish intake among the African Caribbeans (23)) that warrant further investigation. Additional research is needed to investigate these possible mechanisms.

Interestingly, sarcopenia prevalence was higher in our cohort when definitions were based on grip strength compared with ALM, especially after age 65 years. This is the opposite of what was shown in men included in the original FNIH analyses, where the prevalence of “weakness” was much lower than the prevalence of those with “low ALM” (12,13,24). This suggests that African Caribbean men may have lower muscle strength than their ALM would suggest—compared with other populations—and may have poorer muscle quality (ie, low strength per unit of muscle) than quantity. There are many possible biological reasons for why these men might have poor muscle quality, including but certainly not limited to, muscle fat infiltration (25), capillary density/perfusion (26), neuromuscular junction number and function (27–29), and single muscle fiber properties (30). Likewise, these men also developed mobility limitations at higher levels of muscle strength than expected. For example, African Caribbean men in the current study who developed incident mobility limitations had an average baseline grip strength of 40.2 kg. This value is more than 14 kg, or over 1.5× higher, than the FNIH grip strength cut-point, which was developed using gait speed (12). Thus, population—and/or race-specific sarcopenia cut-points may be needed to better define subgroups at risk of mobility limitation and should be pursued in future research. Additional research is also needed to investigate the mechanisms underlying population differences in muscle strength, lean mass, muscle quality, and mobility limitations. Finally, it should be noted that we previously reported that changes in peripheral computed tomography measures of skeletal muscle cross-sectional area, but not DXA appendicular lean mass, were related to incident mobility limitations in this African Caribbean population (15). Thus, measures other than or in addition to DXA indices of lean mass may need to be included in future sarcopenia definitions (2).

Changes in lean mass and skeletal muscle strength with aging are fairly well characterized in developed countries in North America, Europe, and Asia (2). However, less is known about changes in lean mass and muscle strength with aging in populations from developing countries of the world. In the African Caribbean men in the current study, declines in ALMBMI and grip strength/BMI accelerated after age 55 and again after age 70. Thus, African ancestry individuals may have a later initial onset of accelerated lean mass loss compared with Caucasians. For instance, in Caucasians, muscle mass declines steadily from age 25 to 49 years, with an acceleration at age 50 and again around age 70 (31,32). Compared with ALMBMI, grip/BMI declined ~2.5 times faster per year across all ages—until age 75+, where grip/BMI declined by 4 times faster than. This suggests that muscle quality decline accelerated at age 75 years in these men. As stated above, more research is needed to determine specific underlying mechanisms.

The current study has several potential limitations. Body composition and mobility limitations were measured at only two time points; thus, we may have missed intermediate body composition changes that could positively or negatively impact mobility. Our primary outcome was self-reported mobility limitations, as opposed to an objective walking test; however, these questions are commonly used to define mobility limitation and disability. Further, the sarcopenia prevalence was quite low according to FNIH and EWGSOP2 definitions, resulting in high confidence intervals in incident analyses, limiting statistical power, as well as the precision, and robustness of odds ratio estimates. Finally, this study included only men, and results may not be generalizable to women. The current study also has several notable strengths. First, this study was conducted in a rapidly aging (6,7), but under-studied segment of the population. Second, we investigated widely recommended sarcopenia characteristics using DXA measured lean mass and grip strength. Third, the study included a relatively large sample of men with a wide age range, which allowed us to examine the prevalence and changes in sarcopenia throughout adulthood.

In conclusion, we found that grip strength/BMI measured at one time point may be a particularly useful phenotype for identifying those at higher risk of developing future mobility limitations in African Caribbean men. Additionally, the loss of grip strength/BMI far exceeded the loss of appendicular lean mass/BMI, especially in those aged 75+. Finally, these men appeared to have lower muscle strength than their ALM would suggest, and developed mobility limitations at higher levels of grip strength than other populations. Additional research is needed to investigate the lifestyle, environmental and biological factors underlying age-related changes and differences in muscle mass, strength, and quality in African-ancestry populations.

Funding

The research was supported by funding or in-kind services from the Division of Health and Social Services and Tobago House of Assembly, by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (grants AR050107, AR049747) and National Institute of Diabetes and Digestive and Kidney Diseases (grant R01-DK097084). A.J.S. was supported by a career development award from the National Institute on Aging (K01 AG057726).

Acknowledgments

The authors would like to thank the participants of the study and all supporting staff.

Conflict of Interest

None reported.

References