胃肠药物使用与老年新发抑郁症状

Background: Depressive symptoms contribute substantially to disability and loss of independence in later life, and gastrointestinal illness frequently co-occurs with affective vulnerability. However, evidence linking gastrointestinal medication use to subsequent depressive symptoms remains largely confined to single countries and cross-sectional assessments. We investigated whether baseline gastrointestinal medication use and longitudinal two-wave medication-use patterns are associated with incident depressive symptoms across three prospective ageing cohorts. Methods: We analysed community-dwelling adults aged 50 years or older from the China Health and Retirement Longitudinal Study (CHARLS), the US Health and Retirement Study (HRS), and the Survey of Health, Ageing and Retirement in Europe (SHARE) who scored below cohort-specific depressive-symptom thresholds at baseline, followed for a median of 6-11 years. Baseline medication use was compared with non-use; two-wave patterns comprised continued non-use (reference), initiation, discontinuation, and persistent use at both waves. Cohort-specific Cox proportional hazards models with cluster-robust standard errors adjusted for sociodemographic, behavioural, and chronic health covariates (multiple imputation for missing data); estimates were synthesised using random-effects meta-analysis. Negative- and positive-control medications and Fine-Gray competing-risk regression models evaluated specificity and survival bias. Results: The baseline analysis included 50,467 participants (median age 59-65 years; 44.4%-56.8% female) with 16,896 incident depressive events; the two-wave analysis included 32,686 participants with 8,069 events. Pooled adjusted hazard ratios (HRs) were 1.31 (95% CI 1.21-1.41; I2 = 29%) for baseline use and 1.34 (95% CI 1.22-1.47; I2 = 0%) for persistent use; initiation (1.27, 1.13-1.43) and discontinuation (1.25, 1.11-1.40) showed consistent risk elevations across cohorts. Antihypertensive and lipid-lowering medications (negative controls) showed null associations with incident depression, whereas psychiatric medications (positive control) showed the expected strong relationship (pooled HR 1.99, 95% CI 1.79-2.22). The association remained robust after further adjusting for continuous baseline subthreshold depressive symptoms (pooled HR 1.21, 95% CI 1.14-1.27; I2 = 0%), and medicated digestive disease carried higher risk than untreated digestive disease (pooled HR 1.18, 95% CI 1.05-1.32; I2 = 0%). Accounting for competing mortality did not alter the estimates. Conclusions: Baseline and persistent gastrointestinal medication use were consistently associated with increased risk of incident depressive symptoms across three cross-national ageing cohorts with negligible heterogeneity. These findings support integrating routine affective screening into medication reviews for older adults receiving gastrointestinal pharmacotherapy, while maintaining guideline-directed management of digestive conditions.