A descriptive three-domain limitation count is associated with incident cardiovascular disease across three longitudinal ageing cohorts

Ageing populations experience functional and psychological limitations, but their accumulated cardiovascular relevance across settings is uncertain. We analysed 31,535 adults aged ≥ 50 years without baseline cardiovascular disease in three cohorts. A descriptive three-domain analytic count (0–3) comprised daily-function, walking-distance mobility and elevated depressive-symptom limitations. Cohort-specific Cox models used a common adjustment set across cohorts; estimates were pooled using random effects. During follow-up, 8,950 participants developed cardiovascular disease. Each additional domain was associated with incident cardiovascular disease; the pooled hazard ratio was 1.214 (95% confidence interval 1.109–1.330; 95% prediction interval 1.043–1.414). Compared with no domains, pooled hazard ratios were 1.250 for one domain and 1.517 for ≥ 2 domains. After expanded adjustment for available socioeconomic, behavioural and multimorbidity-related factors, the pooled hazard ratio attenuated to 1.163 (95% confidence interval 1.064–1.271). Findings were similar in sensitivity analyses using common comparable ADL/IADL items and leave-one-cohort-out analyses. Competing-risk estimates differed in HRS, and findings at ≥ 75 years were heterogeneous. The primary estimate is not independent of broader measured factors; the descriptive count is not intended for causal inference, clinical prediction or decision-making.