Background: Polypharmacy has been associated with depression in studies of older adults, but evidence on the relationship between polypharmacy and depression in younger adult populations remains limited.
Objectives: To evaluate associations between prevalence of depressive symptoms and polypharmacy in a general adult population overall and according to symptom severity
Methods: NHANES data from the years 2005-2016 were combined. US adults 18 and older were included if they had complete data for depressive symptoms and medication use. Presence and severity of depressive symptoms were assessed using overall score on the Mental Health – Depression Screener and categorized as no depression (0-4 points), mild depression (5-9 points), moderate depression (10-14 points) and moderately severe or severe depression (15-27 points). Polypharmacy was defined as the concurrent use of 5 or more medications and assessed based on interviewer review of prescription containers produced by the participant. Modified Poisson regression models adjusted for age, sex, multimorbidity, and healthcare access were used to estimate prevalence ratios and 95% CIs for polypharmacy overall and stratified by depressive symptom severity. As a sensitivity analysis, we excluded antidepressants from our assessment of polypharmacy and repeated these analyses.
Results: Among the 19,182 adults in our sample, mean participant age was 45 (SD: 19), and 48% were female. Polypharmacy was identified in 17% of participants. Depressive symptoms were reported by 25% of participants, with 64% of these categorized as having mild symptoms, 22% as having moderate symptoms, and 14% as having moderately severe or severe symptoms. Compared to participants who reported no depressive symptoms, polypharmacy was more prevalent among participants with depressive symptoms of any severity (aPR 1.47 (95% CI 1.41, 1.54)), participants with mild symptoms (aPR: 1.33 (95% CI 1.26, 1.40)), participants with moderate symptoms (aPR: 1.69 (95% CI 1.58, 1.83), and participants with moderately severe or severe symptoms (aPR: 1.74 (95% CI 1.60, 1.89)). Excluding antidepressants from the assessment of polypharmacy did not meaningfully change the magnitude of the estimated associations.
Conclusions: Depressive symptoms are associated with a greater prevalence of polypharmacy regardless of symptom severity. As expected, the magnitude of association increased with increasing severity of symptoms. Further research using longitudinal data is needed to investigate causality and should consider subsyndromal depression as well as clinical depression.
