Mental Health Diagnoses Linked to Poorer HIV Treatment Outcomes in Rural Zimbabwe

Adults living with HIV who also had a documented mental health disorder showed poorer retention in care and lower rates of viral suppression than matched patients without such diagnoses, according to a retrospective cohort study conducted at a rural district hospital in central Zimbabwe. The research, published in BioPsychoSocial Medicine, analyzed routinely collected clinical records from 128 adults receiving antiretroviral therapy (ART) and found that psychosis and suicidality were particularly associated with patients falling out of care. The authors caution that the modest sample size and reliance on routine documentation mean the findings should be interpreted carefully, but they argue the results underline an urgent need to weave mental health services into HIV care in resource-limited settings.

The study team, led by Laston Gonah and Jacqueline Nyamadzawo, set out to examine the association between documented mental health disorders and two key treatment outcomes: retention in care and viral suppression. Rather than deploying structured mental health screening instruments, the researchers deliberately used diagnoses as they were recorded in everyday clinical practice. This choice, they explain, offers a window into real-world programmatic conditions in rural, under-resourced health systems, where mental health detection is constrained by scarce specialist capacity, stigma, and fragmented data systems.

To assemble their cohort, the investigators linked two separate registries: the electronic Patient Monitoring System, which registers people living with HIV on ART, and a chronic diseases register that records diagnoses of non-communicable conditions, including mental health disorders. Record linkage using unique patient identifiers, names, and dates of birth initially identified 52 adults with both HIV and a documented mental health disorder among 1,784 registered patients on ART. A subsequent five-month review of patient booklets, from March to July 2023, during which all registered patients were expected to visit the hospital for medication refills, confirmed 12 additional missed or newly diagnosed cases, bringing the total to 64.

Each of these 64 case records was then matched one-to-one by age within one year, sex, and duration on ART within one year to a control patient who had no documented mental health disorder and no other chronic treatment-requiring condition such as tuberculosis, hypertension, or diabetes. The resulting analytic sample comprised 128 participants in 64 matched pairs. Patients with substance use disorders were excluded because documentation of these conditions was inconsistent and underreported, which the authors note could have compromised both prevalence estimation and outcome associations.

The headline prevalence figure was strikingly low: documented mental health disorders accounted for just 3.6 percent of the 1,784 registered patients living with HIV in the district. Psychosis, recorded in 20 patients, and anxiety, recorded in 22, were the most common diagnoses, followed by depression with 13 cases and suicidality with 9. Women and people aged 35 to 49 appeared disproportionately among the documented cases. The authors argue this low figure almost certainly reflects substantial underdetection rather than a genuinely light burden. Previous research in Zimbabwe has estimated mental health condition prevalence among people living with HIV at between 47 and 67 percent, a gap the study attributes to inadequate screening capacity, limited availability of mental health specialists, stigma, and the separation of HIV and mental health information systems.

The diagnostic context matters for interpreting these numbers. In this rural primary care setting, mental health diagnoses were typically made by medical officers holding undergraduate medical degrees and nurses with three-year Registered General Nurse diplomas, cadres who have general clinical training but usually lack formal postgraduate psychiatric specialization. Diagnoses followed national guidelines aligned with ICD-10 categories, but the absence of standardized screening tools, staff shortages, and reliance on clinical judgment likely reduce diagnostic sensitivity, particularly for less overt conditions such as depression and anxiety, while severe and visible conditions like psychosis are more readily identified and recorded.

On the outcomes themselves, the matched comparison showed that 69 percent of cases had viral suppression, defined as plasma HIV-1 RNA below 1,000 copies per milliliter at the last measurement, compared with 75 percent of controls. Retention in care, measured using the United States National HIV/AIDS Strategy metric of two or more HIV clinical visits within 12 months spaced at least 90 days apart, stood at 44 percent among cases versus 54 percent among controls. Both differences were statistically significant at the p < 0.05 level. Fisher’s exact tests found no significant associations between sex, marital status, or employment status and either outcome within either group, although the direction of effects was generally consistent with poorer outcomes among patients with documented mental health disorders.

Drilling down by diagnosis, psychosis and suicidality were significantly associated with non-retention in care, with p values of 0.0089 and 0.022 respectively, while depression showed no significant association with retention. Notably, anxiety was associated with higher retention, reaching significance at p = 0.0006. Ordered logistic regression models, adjusted for age, sex, employment status, and ART adherence, confirmed that psychosis was significantly associated with lower retention among cases. Adherence emerged as a significant predictor of both retention and viral suppression in both groups. An intriguing pattern appeared in the adherence-viral load models: the association coefficient was higher among controls than among cases, which the authors suggest may indicate that mental health disorders attenuate the impact of adherence on viral load control, though they frame this as a hypothesis-generating observation from a small sample.

The study’s limitations are considerable and the authors enumerate them plainly. The sample of 128 participants limits statistical power to detect weaker associations and raises the possibility of Type II error. Residual confounding may persist from the absence of key psychosocial variables, including detailed socioeconomic status, substance use, and psychosocial support from caregivers, the latter unmeasured due to missing data. Reliance on routinely collected clinical data introduces misclassification bias because diagnoses were not based on standardized screening instruments or structured psychiatric assessments, and record linkage across fragmented data systems may have resulted in incomplete case identification. The retrospective matched design also limits causal inference. The authors additionally note that the study did not capture exposure to community-based mental health interventions such as the Friendship Bench, a Zimbabwean program in which lay health workers deliver psychosocial support, which may have introduced further unmeasured confounding.

Despite these caveats, the researchers conclude that documented mental health disorders were associated with poorer retention in care and lower viral suppression, and that the observed low prevalence likely reflects systemic and data-related constraints rather than true epidemiological patterns. They propose a set of programmatic responses: systematic mental health screening within HIV services, targeted interventions for high-risk groups such as patients with psychosis or suicidality, improved integration of HIV and mental health data systems to reduce fragmentation, and task-shifting to trained non-specialist health workers to expand diagnostic capacity. Scaling such evidence-based models, they argue, could improve early identification and management of psychiatric comorbidities and, in turn, support adherence, retention, and long-term viral suppression in similar resource-limited settings. Future research, they add, should examine the role of psychosocial support in improving outcomes for people living with both HIV and mental health conditions.

Subject of Research: The association between documented mental health disorders and retention in care and viral suppression among adults receiving antiretroviral therapy in rural Zimbabwe

Article Title: Association between documented mental health disorders and HIV treatment outcomes among adults receiving antiretroviral therapy in rural Zimbabwe

Article References: Gonah, L., & Nyamadzawo, J. (2026). Association between documented mental health disorders and HIV treatment outcomes among adults receiving antiretroviral therapy in rural Zimbabwe. BioPsychoSocial Medicine, 20(1), Article 12. https://doi.org/10.1186/s13030-026-00361-3

Image Credits: AI Generated

DOI: 10.1186/s13030-026-00361-3

Keywords: HIV, antiretroviral therapy, mental health disorders, viral suppression, retention in care, Zimbabwe, psychosis, suicidality, retrospective cohort, record linkage, task-shifting, resource-limited settings