Reciprocal risks of diabetes and sarcopenia in aging Chinese adults: A prospective cohort study using a semi-Markov multi-state framework

Type 2 diabetes mellitus and sarcopenia are increasingly recognized as interrelated syndromes sharing pathophysiological pathways, yet empirical evidence for their reciprocal, time-dependent relationship in aging populations remains limited. Conventional analytical approaches fail to capture the dynamic, trajectory-based nature of their coprogression because of restrictive assumptions regarding state transitions and sojourn time. We conducted a prospective cohort study using harmonized data from 2 nationally representative Chinese longitudinal surveys: the China Health and Retirement Longitudinal Study and the Chinese Longitudinal Healthy Longevity Survey, with follow-up spanning 2008-2018. A semi-Markov multi-state model was employed to estimate sojourn-time-dependent transition hazards across 5 clinically defined states: diabetes-free/sarcopenia-free, diabetes-only, sarcopenia-only, comorbid diabetes-sarcopenia, and death (absorbing state). All models adjusted for time-varying sociodemographic, lifestyle, and functional covariates and accounted for competing risks. Population attributable fractions were derived via counterfactual simulation. The semi-Markov model demonstrated superior fit over the Markov specification (global ΔAkaike information criterion = 316.4), confirming strong sojourn-time dependence in transition risks. Incident diabetes was associated with an 82% higher hazard of progressing to comorbid diabetes-sarcopenia (hazard ratio = 1.82, 95% confidence interval [CI] = 1.47-2.25), while incident sarcopenia conferred a 65% elevated risk of subsequent diabetes onset (hazard ratio = 1.65, 95% CI = 1.35-2.02). These bidirectional associations remained robust across multiple sensitivity analyses. Counterfactual modeling revealed that 38.2% (95% CI = 31.5%-44.1%) of new comorbid cases could be prevented by blocking diabetes→sarcopenia progression, and 32.6% (26.8%-38.0%) by interrupting sarcopenia→diabetes progression. Strikingly, 58.9% (53.4%-64.1%) of all 5-year mortality was attributable to entry into any chronic disease state. These findings suggest that diabetes and sarcopenia are dynamically and bidirectionally associated in aging Chinese adults. Integrated assessment of glycemic status and muscle health may help identify older adults at elevated risk of comorbidity and mortality. Further interventional studies are needed to determine whether dual-domain prevention strategies can modify these trajectories.