Type of study: A case control study comparing prevalence, comorbidity and correlates of behavioral health disorders in HIV-infected women with women in the general population

The finding: This large, nationally representative study expanded our limited knowledge of the epidemiology of psychiatric and substance use disorders in women living with HIV. Derived from a cohort of HIV-infected women participating in the Women’s Interagency HIV Study (WIHS), the prevalence, comorbidity and correlates of psychiatric and substance use disorders were found to be significantly elevated. Compared to the general women’s population, HIV-infected women had a higher prevalence of lifetime mood disorders (34.2% vs 23.7%), anxiety disorders (61.6% vs 37.0%) and substance use disorders (58.3% vs 8.8%). For 95.5% of participants, DSM-IV disorder onset preceded HIV diagnosis. The prevalence of 12-month disorders was similarly elevated compared to the general women’s population: mood disorders (22.1% vs 10.8%), anxiety disorders (45.4% vs 23.2%) and substance use disorders (11.1% vs 2.2%). Major depressive disorder was the most common mood disorder (20.0%). Those participants with lifetime and 12-month mood disorders had the highest rates of comorbidity compared to other disorders. There was a high co-occurrence of both lifetime and 12-month anxiety and substance use disorders. Women who were working were less likely to have either a lifetime or 12-month mood or anxiety disorder. Substance use disorders were less likely among Black women and Hispanic/Latina women. There was a strong association between 12-month mood disorders, especially bipolar disorder and 12-month substance use disorders and sexual risk behaviors, after adjusting for age, race, education and study site. Women with any mood disorder and co-occurring substance use disorder were over 15 times as likely to engage in risky sexual behaviors.

Strengths and limitations: Estimates of psychiatric and substance use prevalence among HIV-infected individuals have typically derived from small, non-representative samples using screening instruments. In contrast, this study used a large, population-based, multisite cohort to identify a broad range of behavioral health disorders based on a full diagnostic assessment using the World Health Organization Composite International Diagnostic Interview (CIDI). It confirmed findings of the only other similar nationally representative prevalence study conducted almost 30 years ago, the HIV Cost and Services Utilization Study. Additionally, this study provided insight into associations between specific disorders and HIV risk behaviors as well as between specific psychiatric disorders and co-occurring substance use disorders. It also included a comparison population, the 2003 National Comorbidity Survey Replication, women’s cohort. Because a study cohort was used, findings may not be generalizable to the US population of women living with HIV, to younger women or to women not involved in HIV medical care. The full CIDI diagnostic assessment was not used for the entire cohort which may bias results especially those participants who spoke only Spanish. Lastly, assessment of HIV risk behaviors relied on self-report measures.

Relevance: There is growing optimism about significantly reducing HIV incidence in the US through universal HIV testing, “treatment as prevention” and the use of pre-exposure prophylaxis. The success of these strategies depends on effectively identifying those individuals at highest risk for HIV acquisition as well as those HIV-infected individuals at risk for transmission. The important role that mental health professionals play in this effort is strongly underscored by this study’s findings. The study documented that the prevalence of behavioral health disorders far exceeded that of the general population of women and onset of disorders preceded HIV diagnosis by an average of 19 years. In addition, there were significant associations between having a 12-month psychiatric disorder, with or without a substance use disorder and risky sexual behaviors. Thus, mental health professionals may be especially well positioned to identify those individuals at risk for acquiring HIV who might benefit from pre- and post-exposure prophylaxis. Higher than expected rates of comorbid psychiatric conditions and co-occurring psychiatric and substance use disorders strengthens existing evidence for the need to develop and test integrated models of HIV medical care. For example, the collaborative care model in primary care should be considered and adapted to identify co-existing psychiatric disorders and simultaneously address comorbid substance use in a manner that is culturally sensitive, trauma-informed and overcomes systemic barriers to integrated care.