Abstract / Summary
Sarcopenia is a common geriatric syndrome associated with adverse health outcomes in older adults. Body mass index (BMI), a simple anthropometric measure, has been associated with sarcopenia; however, the dose–response relationship between BMI and sarcopenia remains incompletely understood. To investigate the dose–response relationship between BMI and the prevalence of sarcopenia in older adults. PubMed, Web of Science, Embase, CNKI, Wanfang, Sinomed, and VIP were systematically searched for eligible studies through March 10, 2026. Studies examining the association between BMI and prevalent sarcopenia in older adults using established diagnostic criteria were included. Two reviewers independently performed study selection, data extraction, and study-level risk-of-bias assessment. A one-stage random-effects dose-response meta-analysis using restricted cubic splines was conducted, with a BMI of 22.5 kg/m² as the reference. A total of 14 studies involving 6,449 participants were included. Relative to BMI of 22.5 kg/m², the model-predicted ORs for prevalent sarcopenia were 3.30 (95% CI , 2.92–3.72) at 18.5 kg/m², 0.66 (95% CI , 0.62–0.70) at 24 kg/m², 0.23 (95% CI , 0.16–0.34) at 28 kg/m², and 0.14 (95% CI , 0.08–0.25) at 30 kg/m². The restricted cubic spline model showed a significant inverse linear component (coefficient = − 0.311, P < 0.001), with no significant evidence of departure from linearity ( P for non-linearity = 0.499). Across the included cross-sectional studies, higher BMI was associated with lower odds of prevalent sarcopenia. The analysis did not detect clear evidence of departure from a linear association, although uncertainty was substantial and data support was limited at the extremes of the BMI distribution. These findings do not establish a causal protective effect. BMI may serve as an accessible initial indicator; however, because it cannot distinguish between fat and skeletal muscle mass, it must be interpreted alongside direct measures of body composition and physical function. Prospective studies with direct measures of adiposity and lean mass are required. Not applicable.