Depression Not Tied to Mortality Risk in Current HIV Treatment Era
Experiencing symptoms of depression had no impact on risk of all-cause mortality in a 2005-2014 U.S. study of people with HIV. More than one-quarter of the study group participants reported depressive symptoms.
Some research links depression to worse antiretroviral adherence and viral suppression, and in the early combination antiretroviral therapy era, to higher mortality. A U.S. team that conducted the new study suggested two pathways through which depression may affect mortality: first, by causing poor engagement in care and antiretroviral adherence, and second, by adversely affecting immunity. The study examined whether improving antiretroviral therapy and greater access to mental health treatment has changed the impact of depression on mortality.
The analysis involved the Centers for AIDS Research (CFAR) Network of Integrated Clinical Systems (CNICS) cohort, which includes adults in routine care at eight sites across the United States. This study focused on patients at seven sites who completed at least one depression questionnaire between 2005 and 2014. Researchers regularly ascertain vital status of all CNICS participants, regardless of whether they remain in care. They used weighted Kaplan-Meier analysis and marginal structural Cox models to assess the impact of moderate to severe depressive symptoms (Patient Health Questionnaire-9 score >10) on all-cause mortality.
The study group included about 11,000 HIV patients with a median age of 44 years; 85% of participants were men, 51% white, 29% black, and 15% Hispanic. While 63% acquired HIV during sex between men, 21% became infected during sex between men and women, and 14% while injecting drugs. Almost two-thirds of participants (64%) were adherent to antiretroviral therapy, 12% were nonadherent, and 24% were not taking antiretrovirals. Only 55% had an undetectable viral load.
At baseline, 28% of participants reported moderate to severe depressive symptoms and 24% were taking antidepressants. Depressive symptoms were reported at 39% of follow-up visits, and over the eight study years 59% of person-time was spent ever reporting depressive symptoms. Through a median follow-up of 3.1 years, 491 people died to yield an incidence of 13.8 per 1,000 person-years. Mortality was 17.75 per 1,000 person-years in people who had reported depressive symptoms and 11.49 per 1,000 person-years in people who never reported depressive symptoms.
In an unadjusted analysis, reporting depressive symptoms was associated with a one-third higher risk of death (hazard ratio [HR] 1.33, 95% confidence interval [CI] 1.11 to 1.59). But after adjustment for potential confounders and unobserved PHQ-9 measures, depressive symptoms had no impact on all-cause mortality (HR 0.82, 95% CI 0.55 to 1.24). The impact of depressive symptoms on mortality did not differ by sex or baseline psychiatric comorbidity. Excluding alcohol use, drug use, and panic disorder as potential confounders did not affect results.
The CNICS investigators suggest several reasons why their analysis did not confirm prior studies that found an association between depression and mortality:
- Previous studies did not fully account for antiretroviral adherence.
- Better engagement in care, and more efficacious current treatment regimens more robust to moderate adherence, could attenuate the impact of depression on mortality.
- Mortality was low in the CNICS cohort.
- Access to depression treatment may have improved in recent years.
"Given the high burden of depression among HIV-infected adults," the authors stress, "effectively treating depression among people living with HIV continues to be a public health priority."