Associations of Depressive Symptom Severity with High-Sensitivity C-Reactive Protein Among U.S. Adults: NHANES...

Background: Depressive symptoms have been linked to systemic inflammation, yet estimates in population-representative data vary by symptom severity and analytic specifications. We quantified the association between depressive symptom severity and high-sensitivity C-reactive protein (hs-CRP) in U.S. adults using design-based inference. Methods: We analysed pooled NHANES 2015-2018 data for adults aged ≥ 20 years (unweighted n = 9164; complete-case adjusted models n = 8173). Depressive symptom severity was categorised using the Patient Health Questionnaire-9 (PHQ-9) with 0-4 as the reference group and a pre-specified primary contrast of 10-14 versus 0-4. Outcomes were (i) continuous hs-CRP modelled on the log scale, reported as geometric mean ratios (GMR), and (ii) elevated inflammation defined as hs-CRP > 3 mg/L, modelled using a log-link to obtain prevalence ratios (PR). Models incorporated NHANES complex sampling and adjusted for a pre-specified core covariate set (age, sex, race/ethnicity, education, poverty-income ratio, and smoking). Sensitivity analyses excluded hs-CRP > 10 mg/L and added BMI. Results: After adjustment, the geometric mean hs-CRP was 1.43 mg/L (95% CI 1.21-1.70) for PHQ-9 0-4 and 1.63 mg/L (95% CI 1.29-2.08) for PHQ-9 10-14. For the primary contrast (10-14 vs. 0-4), the adjusted GMR was 1.14 (0.96-1.35) and the PR was 1.15 (0.95-1.39). Using a clinically relevant dichotomy (PHQ-9 ≥ 10 vs. <10), depressive symptoms were associated with higher hs-CRP (GMR 1.24 (1.07-1.43)) and a higher prevalence of hs-CRP > 3 mg/L (PR 1.19 (1.01-1.39)). Associations were strongest for PHQ-9 15-19 (GMR 1.62 (1.20-2.19); PR 1.49 (1.15-1.92)). In sensitivity analyses for the primary contrast, GMR estimates ranged from 1.01 to 1.14 and PR estimates ranged from 1.05 to 1.15, with attenuation towards the null after excluding hs-CRP > 10 mg/L and after additional adjustment for BMI. Conclusions: Higher depressive symptom severity was associated with higher hs-CRP and a higher prevalence of low-grade systemic inflammation in U.S. adults, with the clearest elevations observed among those with moderately severe symptoms. For the pre-specified moderate-symptom contrast, point estimates were modest and sensitive to handling of high hs-CRP values and adiposity-related adjustment.