Abstract Background Age-related muscle decline and obesity are associated with impaired bone health and substantially increased fragility and fracture risk; however, evidence in frail populations remains inconsistent. We investigated differences in bone microarchitecture across body composition phenotypes in frail older women. Methods This cross-sectional study included 280 older women, of whom 109 met the Fried frailty criteria. Body composition was assessed by dual-energy X-ray absorptiometry and categorized into four phenotypes: low appendicular lean mass (LALM; 20th percentile of residuals, −1.45), obesity (fat mass index 13 kg/m2), obesity with LALM, or neither condition. Bone microarchitecture was evaluated using high-resolution peripheral quantitative computed tomography (HR-pQCT). Statistical analyses included generalized estimating equations (GEE) with Bonferroni correction, χ2 tests, and Fisher’s exact tests. Results Compared with the obesity group, the LALM group exhibited smaller cortical area (Ct.Ar 7123 vs. 9018 mm2; age-adjusted p=.008) and thinner cortex (Ct.Th 0.8580.202 vs. 1.0310.217 mm; p=.043) at the tibia. Similar differences were observed at the radius (Ct.Ar 3312 vs. 4310 mm2; p=.002; Ct.Th 0.6420.175 vs. 0.7790.157 mm; p=.005). Additionally, compared with the obesity plus LALM group, the LALM group showed lower Ct.Ar, Ct.Th, total volumetric bone density (Tt.vBMD), and trabecular number (Tb.N), along with greater trabecular area (Tb.Ar) and trabecular separation (Tb.Sp) at the radius. Bone strength was significantly lower in the LALM group than in the obesity and obesity plus LALM groups. Conclusions Among pre-frail and frail older women, obesity was associated with favorable cortical and trabecular bone microarchitecture than those with low muscle mass without obesity, suggesting that excess adiposity may partially mitigate skeletal deterioration. Longitudinal studies are warranted to clarify the impact of body composition on fracture risk in frail adults.
Fernandes et al. (Mon,) studied this question.