Appendicular Lean Mass and Mortality among Prefrail and Frail Older Adults

Abstract

Objective: Prefrail and frail older adults are a heterogeneous population. The measurement of appendicular lean mass (ALM) may distinguish those at higher versus lower risk of poor outcomes. We examined the relationship between ALM and mortality among prefrail and frail older adults.

Design: This was a population-based cohort study.

Setting: The Third National Health and Nutrition Survey (NHANES III; 1988-1994).

Participants: Older adults (age ≥65 years) with pre-frailty or frailty defined using the Fried criteria.

Measurements: ALM was quantified using bioimpedance analysis. Multivariable-adjusted Cox regression analysis examined the relationship between ALM and mortality. Logistic regression analysis was used to determine if ALM added to age and sex improved the predictive discrimination of five-year and ten-year mortality.

Results: At baseline, the average age was 74.9 years, 66.7% were female, 86.3% and 13.7% were prefrail and frail, respectively. The mean ALM was 18.9 kg [standard deviation (SD): 5.5]. During a median 8.9 years of follow-up, 1,307 of 1,487 study participants died (87.9%). Higher ALM was associated with a lower risk of mortality. In a multivariable-adjusted regression model that accounted for demographic, behavioral, clinical, physical function, and frailty characteristics, each SD increase in ALM was associated with an 50% lower risk of mortality [Hazard Ratio: 0.50 (95% CI: 0.27-0.92); P=0.026]. The addition of ALM to age and sex improved the predictive discrimination of five-year (P=0.027) and ten-year (P=0.016) mortality.

Conclusion: ALM distinguishes the risk of mortality among prefrail and frail older adults. Additional research examining ALM as a therapeutic target is warranted.

Keywords: Lean mass; bioimpedance analysis; geriatrics; population-based.

Figure

Risk of mortality associated with appendicular lean mass, and results from multivariable model analyses. NOTE: Solid curve represents estimate of the hazard ratio (HR) for mortality. Dashed curves represent pointwise 95% confidence intervals (95% CI). The reference point (hazard ratio of 1) represents the sample mean appendicular lean mass (18.9 kg) and each standard deviation represents 5.5 kg. ^aModel 1 is adjusted for age, sex, body mass, and height. ^bModel 2 is adjusted for the covariates in model 2, and race, smoking status, cognitive function, hypertension, hyperlipidemia, COPD, cancer, arthritis, myocardial infarction, stroke, heart failure, kidney disease, self-rated health, hospitalization, falls, hemoglobin, c-reactive protein, glycated hemoglobin, insulin, glucose, and creatinine. ^cModel 3 is adjusted for the covariates in model 3, and poor balance and gait speed. ^dModel 4 is adjusted for the covariates in model 3, and stage of frailty (prefrail versus frail).