Age Differences in the Associations of Dietary and Urinary Caffeine and Its Metabolites with Depressive Symptoms

We examined associations between caffeine intake, urinary caffeine biomarkers, and depressive symptoms across age groups. We analyzed the National Health and Nutrition Examination Survey from 2005 to 2018 data from 30,817 adults aged≥18 years, categorized as younger (18–39 y), middle-aged (40–59 y), and older (≥ 60 y). Dietary caffeine, assessed using two 24-hour recalls, was modeled continuously and categorically (< 100, 100 to<200, and≥200 mg/d). Urinary caffeine and metabolites were measured in a 2009–2014 subsample (n=4,378). A nine-item Patient Health Questionnaire score of≥10 was considered to indicate depressive symptoms. Survey-weighted multivariable models estimated associations and caffeine-by-age interactions. The adjusted dietary caffeine and age interaction was significant for a nine-item Patient Health Questionnaire score of≥10 (F[4,77]=2.912, p=0.027), but not for the total nine-item Patient Health Questionnaire score (p=0.522). Among older adults, high versus low intake was associated with greater odds of a nine-item Patient Health Questionnaire score of≥10 (adjusted odds ratio=1.64, 95% confidence interval: 1.23–2.20). The spline model similarly showed an age interaction (F[2,80]=3.782, p=0.027), with predicted probabilities increasing mostly among older adults. Higher log urinary caffeine was associated with greater odds of a nine-item Patient Health Questionnaire score of≥10 among older adults (adjusted odds ratio=1.24, 95% confidence interval: 1.07–1.42; false discovery rate–adjusted p=0.012), although the global age interaction was not significant (false discovery rate–adjusted p=0.470). No adjusted urinary biomarker association with the total nine-item Patient Health Questionnaire score was significant. Associations were outcome-specific: age modified the association between dietary caffeine and the presence of depressive symptoms, but not mean symptom severity. Among older adults, high intake was associated with greater odds of depressive symptoms.