Impact From Quality of Life to End of Life
Depression wields a sledgehammer impact on the clinical course of people with HIV infection. Dozens of studies confirm links between depression and dangerous risk behavior, poor HIV control, comorbidities including cardiovascular disease, and -- at the end of this trainwreck scenario -- death. Glenn Treisman, a depression expert at Johns Hopkins University, lists 10 life-changing, or life-ending, consequences of depression in people with HIV:18
- Impaired quality of life
- Decreased cognition
- Increased risk behaviors
- More frequent medical visits
- Longer hospital stays
- Higher treatment costs
- Decreased antiretroviral adherence
- Decreased survival
Five studies link depression to higher mortality in people with HIV.8,19-22 Three of these five studies took place in the United States early in the combination antiretroviral era,19-21 when regimens were less effective and many people shunned antiretroviral therapy (ART) at higher CD4 counts for fear of toxicity. And these older studies19-21 had HIV-related death, not all-cause mortality, as an endpoint. But the most recent studies, in Switzerland8 and the United States,22 ran up to the most recent antiretroviral era (from 2004 to 2014) and focused on all-cause mortality.
Compared with HIV-positive women without depressive symptoms, those with chronic symptoms had a doubled risk of HIV-related death (adjusted relative risk [aRR] 2.0, 95% CI 1.0 to 3.8) in a US analysis controlled for clinical and other risk factors.19 This analysis focused on 765 women seen from 1993 through 2000 in the prospective 4-city US HIV Epidemiologic Research Study (HERS) cohort. The HERS team also linked chronic depressive symptoms to greater drops in CD4 count. Only 38% of these women had taken ART for at least 1 year.
Prospective analysis of 1716 HIV-positive women in the US Women's Interagency HIV Study (WIHS) in the same era (1994-2001) found that women with chronic depressive symptoms had a 70% higher risk of AIDS death than women with limited or no depressive symptoms (aRR 1.7, 95% CI 1.1 to 2.7).20 Half of these women had taken ART for 1 year or more. Using mental health services halved the risk of AIDS mortality in these women (aRR 0.5, 95% CI 0.3 to 0.7).
Depressive symptoms boosted the risk of AIDS death by half (adjusted hazard ratio [aHR] 1.49, 95% CI 1.00 to 2.21, P = 0.05) in a prospective study of 338 men and 152 women, two thirds of them nonwhite, seen in five Southeastern US states in the early 2000s.21 There was no association between depressive symptoms and allcause mortality. Four in 5 people were taking ART when follow-up began.
The biggest assessment of depression and mortality used all-cause mortality as the endpoint and ran from January 2010 through July 2013 in the Swiss HIV Cohort Study (SHCS).8 This analysis of 4422 SHCS members without initial depression involved 3294 men (74%), 1934 MSM (44%), 1128 women (26%), and 432 injection drug users (10%). The researchers identified depression through diagnosis by a psychiatrist (63%) or an SHCS infectious diseases specialist. HIV care is free and accessible to all in Switzerland.
During follow-up 193 people died, mostly from non-AIDS deaths (59%), AIDS deaths (11%), or suicide (9%). Overall mortality measured 0.96 per 100 person-years, and mortality proved more than one third higher in people with than without depression (1.17 versus 0.86 per 100 person-years, P = 0.033). Eliminating drug injectors from the analysis rendered this difference nonsignificant. The researchers did not perform an adjusted analysis.
The most recent and most convincing study linking depression to a higher risk of death involved 4001 HIV-positive adults in a prospective US study, the CFAR Network of Integrated Clinical System (CNICS) Cohort.22 Participants entered care in 2004 or later and follow-up continued for up to 10 years. About 15% of cohort members were women, about 30% black, and about 17% Hispanic.
Defining depression as a Patient Health Questionnaire-9 (PHQ-9) score at or above 10, the investigators determined that 1246 people (31%) had depression in their first year of CNICS enrollment.22 At the end of follow-up, 121 people (3%) had died of any cause. A Cox proportional hazards model adjusted for adherence, CD4 count, HIV suppression, and other variables determined that people with depression had almost a two thirds higher risk of death (adjusted hazard ratio 1.64, 95% CI 1.06 to 2.53).
The Swiss and US studies8,22 offer strong contemporary evidence that depression shortens survival, and the intuitive strength of that conclusion speaks for itself. As the CDC observes, depression has strong associations with undeniable correlates of mortality such as smoking, drinking, and a sedentary lifestyle.23
Depression, Heart Failure, and the HIV Care Cascade
It's easy to add other factors that correlate with both depression and mortality, like injecting drugs, abusing other substances, and multiple comorbidities. A recent US general-population trial randomized 20 primary care practices to evidence-based depression care or usual care.24 During 2 years of follow-up involving 1204 older primary care patients, those in usual-care practices with the highest comorbidity and depression levels had a tripled risk of death compared with depressed patients who had minimal comorbidities (HR 3.02, 95% CI 1.32 to 8.72). In contrast, patients in depression-care practices with the highest levels of comorbidity and depression did not run a higher death risk than depressed patients with minimal comorbidity. The bottom line is that active depression management contributes to prolonged survival in older people with a list of comorbid diseases -- like a growing proportion of people with HIV infection.
One US study tied major depressive disorder to heart failure in people with HIV.25 This Veterans Aging Cohort Study (VACS) involved 26,908 veterans with HIV and 54,519 without HIV. After 5.8 years of follow-up, HIV-positive vets with major depressive disorder had a two thirds higher risk of heart failure than HIV-negative vets without depression (aHR 1.68, 95% CI 1.45 to 1.95). A separate fully adjusted analysis limited to veterans with HIV determined that major depressive disorder independently boosted heart failure risk (aHR 1.29, 95% CI 1.11 to 1.51). Among veterans with major depressive disorder, those taking antidepressants when follow-up began had a 24% lower risk of heart failure (aHR 0.76, 95% CI 0.58 to 0.99).
And for people with HIV, comorbidities represent only one set of hurdles to healthy longevity. To control HIV and comorbidities, they have to get into care, stay in care, start and adhere to antiretroviral therapy, gain CD4 cells, and make their viral load undetectable. Research from the past decade offers evidence that depression can narrow passage through each of these gateways in the HIV care continuum -- starting with linkage to care12 and retention in care,26 and proceeding through antiretroviral adherence,7,27-32 CD4 gains,19,32,33 and viral control.32-35
Retrospective analysis of 3359 HIV patients in the Kaiser Permanente healthcare system addressed the last three intertwined outcomes in people starting their first antiretroviral regimen from January 2000 through December 2003.32 This 8-state analysis included 1961 people (58%) starting ART without depression and 1398 (42%) starting ART with depression. Among people diagnosed with depression, only 508 (36%) got a prescription for a selective serotonin reuptake inhibitor (SSRI) antidepressant. Most study participants, 83%, were men, and median age stood at 40 when followup began. The Kaiser team calculated adherence by pharmacy refills.
An analysis adjusted for age, gender, antiretroviral regimen, and temporal trend determined that people with depression but not taking an SSRI had about a 20% lower chance of at least 90% antiretroviral adherence than the control group of people starting ART without depression (adjusted odds ratio [aOR] 0.81, 95% CI 0.70 to 0.98, P = 0.03) (Figure 2). But antiretroviral adherence did not differ significantly between the control group and people with depression taking an SSRI.
Impact of Depression and SSRIs on ART Outcomes
Figure 2. Retrospective analysis of 3359 people starting their first antiretroviral regimen showed that depression without SSRI therapy independently predicted (1) a lower chance of at least 90% adherence to antiretroviral therapy (ART) and (2) a lower chance of reaching a viral load below 500 copies/mL after 12 months of ART. Also after 12 months of ART (3) people with depression and better than 80% SSRI adherence had a 38-cell CD4-count advantage over people with depression but not taking an SSRI.32
Credit: Teresa B. Southwell/The Center for AIDS Information & Advocacy
In an analysis adjusting for the same variables plus baseline CD4 count, odds of reaching a viral load below 500 copies/mL 12 months after starting ART proved 23% lower in people with depression and not taking an SSRI than in the control group of HIV-positive people without depression (aOR 0.77, 95% CI 0.62 to 0.95, P = 0.02). Chances of viral control in 12 months did not differ significantly between SSRI takers with depression and the control group. Twelve-month CD4 responses were similar in people with and without depression. But among people with depression, those with better than 80% SSRI adherence had significantly greater 12-month CD4 gains than people not taking an SSRI (+19 versus -19 cells/mm3, P = 0.01).
CDC analysis of a nationally representative sample also linked depression to lower chances of reaching an undetectable viral load.35 Among 18,095 HIV-positive people in the Medical Monitoring Project in 2009-2012, 25% had a depression diagnosis, 91% took antiretroviral therapy, and 69% had a viral load below 200 copies/mL on all measures in the past 12 months. An analysis adjusted for antiretroviral adherence and race determined that a depression diagnosis independently conferred a 7% lower chance of attaining sustained viral suppression (adjusted prevalence ratio 0.93, 95% CI 0.91 to 0.96).
The Worse the Depression, the Worse the Adherence
The link between depression and poor antiretroviral adherence appears to grow stronger as depression deepens. That conclusion emerged from a cross-sectional study of 624 HIV-positive adults at the Washington University HIV Clinic in 2009.33 Participants completed the Patient Health Questionnaire-9 (PHQ-9), which focuses on nine diagnostic criteria for DSM-IV depressive disorders, and researchers used responses to rate patients as having no depression, minimal depression, mild depression, moderate depression, moderately severe depression, or severe depression.
Ninety-six people (15%) had symptoms of major depressive disorder. Statistical analysis adjusted for age, race, tobacco use, and treatment with a protease inhibitor versus a nonnucleoside determined that more severe depression predicted a greater chance of worse than 95% antiretroviral adherence. For example, probability of poor adherence lay around 20% or lower in people with minimal or mild depression, above 20% in those with moderate or moderately severe depression, and as high as 40% in people with severe depression (P < 0.05).
If depression fosters poor adherence, one might assume treating depression improves adherence. Meta-analysis of 29 studies involving 12,243 people with HIV confirms that assumption.27 Overall odds of antiretroviral adherence proved 83% greater in people treated for depression (standardized odds ratio 1.83, 95% CI 1.27 to 2.55). In contrast, people not treated for depression ran a 35% higher risk of nonadherence (standardized relative risk 1.35, 95% CI 1.13 to 1.60). Compared with people not treated for depression, treated people had a doubled chance of improvement in depressive symptoms (standardized odds ratio 2.07, 95% CI 1.38 to 3.30). Adherence definitions varied from study to study.
In a review of major depression and other psychiatric disorders, Andrew Angelino and Glenn Treisman from Johns Hopkins call these conditions "a vector for infection with HIV and a barrier to its successful treatment."36 Evidence like that reviewed in this article led them "to conclude that treatment of these disorders greatly improves patient adherence and outcomes of HIV infection."