OBJECTIVES To assess whether frailty is a risk factor for skeletal muscle mass decline in community-dwelling elderly people.

DESIGN Prospective observational cohort study.

SETTING Seongnam, Gyeongi Province, Korea.

PARTICIPANTS Community-dwelling Koreans aged 65 and older (n = 341).

MEASUREMENTS Bioimpedance analysis (BIA) was used to measure body composition at baseline and 5 years later. Laboratory examination and comprehensive geriatric assessment were performed at both times. Lean mass index (LMI) was defined as total body lean mass/height(2). A decrease of more than 5% in the LMI was considered to be significant. Frailty status was defined using the Cardiovascular Health Study criteria.

RESULTS LMI decline occurred in 196 (54.1%) subjects during the follow-up period (5.0 ± 0.7 years). Baseline LMI was highest in robust (17.6 ± 1.8 kg/m(2), n = 126), lower prefrail (17.0 ± 1.7 kg/m(2), n = 185), and lowest in frail (16.7 ± 1.3 kg/m(2), n = 30) subjects (P < .001). Frailty status was associated with LMI decline at 5-year follow-up (robust 0.81 ± 0.78 kg/m(2), prefrail 1.00 ± 0.92 kg/m(2), frail 1.35 ± 0.85 kg/m(2), P < .001). This effect of frailty on LMI decline persisted after adjusting for covariables (P = .02). The risk of significant LMI decline was 2.9 times as great in frail elderly adults as in those who were robust even after adjusting for covariates (95% confidence interval = 1.01-8.55).

CONCLUSION Frailty status was found to be independently associated with subsequent LMI decline in community-dwelling older adults.