Body Composition and Health-Related Quality of Life in Older Korean Women

Background/Objectives: Ratio-based body composition indices obtained by bioelectrical impedance analysis (BIA) are attractive for repeated, low-burden monitoring of older adults; however, it remains unclear whether their associations with health outcomes reflect muscle mass, adiposity, or both. This exploratory, cross-sectional study examined the associations of three indices—the skeletal muscle index (SMI; appendicular skeletal muscle mass [ASM]/height2), skeletal muscle mass percentage (SMM%; [total SMM/body weight] × 100), and muscle-to-fat ratio (MFR; total SMM/body fat mass)—with health-related quality of life (HRQoL), cognitive function, and depressive symptoms. Methods: We analyzed baseline data from 122 community-dwelling women aged 65–86 years, who enrolled in a community-based exercise service in Korea; the analytic dataset was assembled directly from source records with prespecified linkage and validity checks. Body composition was measured using the InBody 970. HRQoL, cognitive function, and depressive symptoms were assessed using the EQ-5D-3L index (Korean value set), Korean Montreal Cognitive Assessment (MoCA-K), and Short Form of the Korean Geriatric Depression Scale (SGDS-K), respectively. Linear regression models adjusted for age (Model 1) and for age, education, and the number of chronic conditions (Model 2) were estimated with heteroskedasticity-consistent (HC3) standard errors, which were used for all reported p-values and for Benjamini–Hochberg false discovery rate (FDR) correction. Results: In Model 1, higher SMM% was associated with a higher EQ-5D index (β = 0.253, HC3 p = 0.013, FDR q = 0.048), whereas higher SMI was associated with a lower EQ-5D index (β = −0.230, q = 0.048); the MFR association did not survive FDR correction (q = 0.168). None of these associations remained significant in Model 2 (SMM%: β = 0.130, HC3 p = 0.171) or after adjustment for body mass index (BMI). Body weight, BMI, and fat mass were associated with the EQ-5D index at least as strongly as any index (β = −0.333 for BMI), and SMM% was almost perfectly inversely correlated with body fat percentage (r = −0.985). No index was associated with MoCA-K after FDR correction, and the association of SMI with SGDS-K did not persist in Model 2. Conclusions: In community-dwelling older Korean women, associations between BIA-derived muscle indices and HRQoL largely reflected adiposity and overall body size and did not persist after adjustment for education, chronic conditions, or BMI; associations with cognitive function and depressive symptoms were not robust. Given the study’s cross-sectional design, single recruitment setting, incomplete confounder data, and assessments that for some participants followed home-program installation, these indices should not be interpreted as muscle-specific correlates of HRQoL. Longitudinal studies with imaging-based body composition and functional measures are thus needed.