Pre-diagnostic low muscle mass and sarcopenia associated with subsequent cancer in Chinese adults: a longitudinal CHARLS study

Loss of skeletal muscle may reflect systemic physiological compromise before cancer is clinically recognized. We examined whether low muscle mass, sarcopenia, and deterioration in these phenotypes were associated with subsequent first self-reported cancer in middle-aged and older Chinese adults. We analyzed China Health and Retirement Longitudinal Study (CHARLS) data from 2011 to 2020. Adults aged 45 years or older who reported no cancer in 2011 were eligible. The primary pre-diagnostic risk set excluded first cancer reports in 2013 or 2015, ensuring that 2015 muscle status preceded cancer reporting in 2018 or 2020. Low muscle mass was defined as height-adjusted appendicular skeletal muscle mass (ASM/height²) < 6.8 kg/m² in men or < 5.69 kg/m² in women. Low muscle strength was defined as grip strength < 28 kg in men or < 18 kg in women. Sarcopenia was identified using the composite c2015_sarcopenia indicator according to the prespecified sarcopenia definition. The analysis variables were c2015_lowmusclemass and c2015_sarcopenia (0 = no, 1 = yes, null=unknown/unavailable). The primary analysis used wave-based Cox regression with censoring at the last observed cancer-status wave. Models adjusted for demographic factors, smoking, drinking, hypertension, diabetes, and dyslipidemia; expanded sensitivity models additionally considered baseline body mass index, estimated glomerular filtration rate, and 2011–2015 weight change. Firth logistic models and a 2020-only longer-lag analysis assessed rare-event robustness. Among 10,407 cancer-free participants at baseline, 254 first reported cancer during follow-up; 212 first reported cancer in 2018 or 2020. Only 22 participants in the pre-diagnostic risk set had cancer status missing at both target waves and were not classified as cancer-free. In the primary Cox models, 2015 low muscle mass (hazard ratio [HR] 2.02, 95% confidence interval [CI] 1.10–3.71; P = 0.023) and sarcopenia (HR 2.33, 95% CI 1.17–4.66; P = 0.016) were associated with later first cancer reporting. Cumulative reporting was 3.39% versus 1.87% for participants with versus without low muscle mass and 4.00% versus 1.90% for those with versus without sarcopenia. In four-level transition models, new/worsened low muscle mass (HR 3.45, 95% CI 1.44–8.25) and new/worsened sarcopenia (HR 4.31, 95% CI 1.80-10.32) showed the strongest associations. The findings remained directionally consistent after expanded confounder adjustment and in longer-lag Firth analyses. Low muscle mass and worsening sarcopenia measured before a first cancer report were associated with subsequent self-reported cancer. Given the low absolute event rates, self-reported outcome, residual confounding, and possibility of occult cancer preceding recognition, these findings support a pre-diagnostic vulnerability phenotype rather than a causal or screening interpretation.