INTRODUCTION

We aimed to quantify how diseases accumulate and diminish among ageing populations, and examine how modifiable risk factors influence these progressions.

METHODS

In this multicohort study with four cohorts, China Health and Retirement Longitudinal Study, the English Longitudinal Study of Ageing (ELSA), the Health and Retirement Study, and the Survey of Health, Ageing and Retirement in Europe (SHARE), and 75 874 participants, we employed a multistage model that accommodated bidirectional transitions between four health stages (0, 1, 2, ≥3 conditions) from baseline to 8 years, and conducted matching analyses to examine the influence of age, sex, socioeconomic status (SES) and lifestyle factors on these transitions.

RESULTS

Disease accumulated faster than diminished (0.08-0.44 vs 0.00-0.06). Transitions accelerated towards severe multimorbidity (0→1: 0.29 (95% CI 0.28 to 0.29), 1→2: 0.27 (95% CI 0.27 to 0.28) and 2→≥3: 0.44 (95% CI 0.43 to 0.45)). Mortality risk escalated with condition count: 0.08 (95% CI 0.08 to 0.09) for 0 conditions, 0.13 (95% CI 0.12 to 0.13) for 1 condition 0.17 (95% CI 0.16 to 0.18) for 2 conditions, and 0.27 (95% CI 0.26 to 0.27) for ≥3 conditions. Cohorts exhibited broadly similar progression patterns, though ELSA demonstrated slower transitions to ≥3 conditions and SHARE showed elevated mortality from 0 and 1 conditions. Key risk factor effects emerged: disease accumulation peaked at 55-65 years; females had higher disease accumulation but lower transitions to death than males; Low-SES populations had higher probabilities of developing ≥3 conditions than the middle-SES group, while middle-SES populations had higher accumulation probabilities for 0→≥2 and 2→≥3. Lifestyle factors exerted differential impacts: smoking increased 1→3 transitions and drinking increased 0→2 transitions, while physician inactivity increased 0→3 transitions. Sensitivity analyses confirmed robustness across 11 condition-specific models.

CONCLUSIONS

Multimorbidity progression accelerates nonlinearly, with risk factors exerting varying effects, depending on the magnitude of risk factors and initial health states. Precision interventions should target age, sex, SES and lifestyle-specific strategies.