Methods: We tested this hypothesis both in the Veterans Aging Cohort Study (VACS) Virtual Cohort (VC), where baseline major depressive disorder (MDD) was identified by ICD-9 codes (77,232 HIV-/MDD-; 14,606 HIV-/MDD+; 34,150 HIV+/MDD-; 7,030 HIV+/MDD+), and in the subset participating in the VACS Biomarker Cohort (BC), where baseline depression was defined as a Patient Health Questionnaire (PHQ)-9 score ≥10 (572 HIV-/PHQ-9 <10; 263 HIV-/PHQ-9 ≥10; 1175 HIV+/ PHQ-9 <10; 348 HIV+/ PHQ-9 ≥10). Cox proportional hazard models were constructed to examine the risk for all-cause mortality by baseline depression and HIV-infection status while controlling for demographics, BMI, traditional CVD risk factors, substance use, HCV co-infection, reduced renal function, hemoglobin, and use of statin drugs.
Results: Median (IQR) time of follow-up in the VC was 10.0 (5.6, 12.1) years and in the BC was 8.8 (8.0, 9.3) years. In adjusted Cox models, the hazard ratios (95% CI; p-value) for mortality in the VC for the HIV-/MDD+, HIV+/MDD-, and HIV+/MDD+ groups compared to the HIV-/MDD- group were, respectively, 1.07 (1.03, 1.12; p=0.001), 1.49 (1.45, 1.54; p<0.001), and 1.57 (1.50, 1.65; p<0.001). The hazard ratios (95% CI; p-value) for mortality in the BC for the HIV-/PHQ-9 ≥10, HIV+/PHQ-9 <10, and HIV+/PHQ-9 ≥10 groups compared to the HIV-/PHQ-9 <10 group were, respectively, 1.36 (0.99, 1.85; p=0.06), 1.33 (1.04, 1.70; p=0.021), and 1.51 (1.12, 2.03; p=0.007).
Conclusion: In separate analyses using both diagnostic codes and a more sensitive instrument for identifying depression, we found that depression in combination with HIV infection was strongly associated with increased all-cause mortality in U.S. Veterans.
Travel support to scientific conferences
Bristol-Myers Squibb: Investigator , Research support and Travel support to scientific conferences
Janssen Therapeutics: Grant Investigator and Investigator , Research grant and Research support
J. Stewart, None
J. Chang, None
J. L. Goulet, None
M. Freiberg, None