Prevalence and correlates of probable common mental disorders in a population with high prevalence of HIV in Zimbabwe

Common mental disorders (CMD) affect people across the world with a global lifetime prevalence of 29 % [1]. Presenting as a mixture of somatic, anxiety, and depressive symptoms [2], CMD increase the risk of developing both non-communicable and communicable diseases [3]. Amongst people living with HIV (PLWH), CMD are a leading cause of disability [4–7] and are known to hasten HIV disease progression [8], particularly in LMIC [4, 6, 9] where low levels of CMD detection at primary health care level [10] are confounded by a large treatment gap for CMD [11, 12]. In sub-Saharan Africa it is estimated that there is one psychologist for every 2.5 million people, one mental health nurse for every million people and one psychiatrist for every 2 million people [13]. In contrast, in high income countries the ratio of psychiatrists to the population is estimated to be 1:10000 [14].

In recent years the development of interventions that emphasize task-shifting as a way to address the treatment gap for CMD has gained recognition [15–18] with specific structured packages being developed and evaluated [19–21]. The key elements of task-shifting involve the delegation of responsibilities to lower level cadres, often non-professionals, to deliver services while the more qualified professionals provide support and supervision [22]. However, use of targeted task-shifting psychological interventions to address CMD in PLWH requires a thorough understanding of factors associated with CMD in this population [4, 23], due partially to the need for greater monitoring and supervision of work delivered by non-specialists [24–26].

Much is known about factors associated with CMD in non-HIV populations in LMIC [27–31], contrasting with sparse data on CMD among PLWH [6]. In four LMIC, namely India, Zimbabwe, Chile and Brazil, female gender, poverty and recent life events were found to be common correlates of CMD in non-HIV populations [29]. In other parts of the world, correlates of CMD in PLWH include death of a significant other [32], family history of mental illness, negative coping style, alcohol dependency, food insecurity [33] and recent negative life events [34, 35]. Other studies from poor resource settings have found that CMD in PLWH is correlated with trauma, posttraumatic stress disorder (PTSD), stigma and social barriers such as peoples’ attitudes towards disabilities, physical and organizational obstacles, however, these differences appear to be due to a difference in the study setting and study population [36–39].

Early studies of CMD in PLWH in Zimbabwe were based on antiretroviral (ART) -naïve populations using the Shona Symptom Questionnaire (SSQ-14) [40] as the main outcome instrument. Findings were inconsistent, with correlates of CMD including being female, having a male partner who was older by 10 or more years, multi-parity (having borne a number of children), and negative life events [34, 41, 42].

The aim of this study was to determine the prevalence and correlates of probable CMD in a primary health care setting with a high prevalence of HIV, and to compare these correlates of CMD by HIV status. The study was part of the validation of screening tools for CMD in a population with a high prevalence of HIV, which was preparatory work for a randomized controlled trial of a psychological intervention to reduce depression. The trial used the SSQ-14 and the 9-item Patient Health Questionnaire (PHQ-9) as primary and secondary outcome measures respectively. Establishing the prevalence of CMD among primary health care patients helped determine the possible duration of recruitment of participants for the trial, and analysis of correlates of CMD contributed towards the development of appropriate interventions [43].

We found a high prevalence of probable CMD and depression (both over 60 %) among PLWH, underpinning the importance of detecting and addressing common mental disorders in this population. Our results indicate that probable CMD and depression in PLWH, as measured by the SSQ-14 and the PHQ-9, are associated with recent negative life events and female gender. These findings reflect those from non-HIV infected populations [29–31, 49]. In a recent systematic review of screening tools used in LMIC, prevalence of 11 %–55 % was reported, however, none of the studies looked at HIV populations on ART [48]. While the risk of false positive screening results from self-report measures is a concern, such tools can contribute towards identifying those most at risk of CMD in resource-poor settings when validated and used within primary health care facilities [6, 7].

Although we did not measure adherence to HAART in our study population, there is growing evidence of an association of CMD and depression with poor adherence to HAART [8, 50, 51]. There is also evidence suggesting that treatment of depression increases adherence to HAART [50, 52]. Despite being on HAART, PLWH face a magnitude of problems and are twice as likely to suffer from depression when compared to HIV negative matched controls [34, 53, 54]. The findings of our study have implications for some of the newer WHO initiatives such as the Option B+ [55] which advocate for immediate commencement of HAART for HIV infected pregnant women. Over 30 % of women attending the Prevention of Mother to Child Transmission of HIV (PMTCT) program in Zimbabwe have signs of post-natal depression [34], therefore the inclusion of CMD packages of care within these initiatives is critical as this will contribute towards improving outcomes for PLWH.

In LMIC the improvement of HAART access will need to be matched with an equally focused integration of evidence-based CMD and depression care packages [6] because there is a large body of data showing that treating CMD among PLWH improves adherence to HAART and quality of life [56–58]. Both HIV positive and HIV negative women are more likely to be affected by CMD and depression in our setting [29, 34]. Furthermore they are more likely to utilize primary health care facilities than their male counterparts [59]. Therefore focusing on women who are living with HIV at primary health care level, particularly those with negative life events who screen positive on the SSQ-14 and PHQ-9 will strengthen existing HIV related programs such as the PMTCT and the newer Option B+. There is evidence showing the feasibility of using screening tools at primary health care level in Zimbabwe [34, 41, 44, 60] but these tools are not integrated into HIV care programs.

There are several psychological interventions that have been shown to work in LMIC, which could be adapted for use among PLWH [61–64]. However, most of the studies evaluating the effectiveness of these interventions for PLWH have been of poor quality [25]. Therefore there is a need to conduct rigorous formative research to design interventions and evaluate them before they are introduced or integrated into health care facilities in LMIC. Part of this process should involve establishing factors that are associated with CMD and depression, and using this information to determine the content of interventions for specific populations. Our findings support the use of a problem solving therapy approach because of the strong association between negative life events and CMD [65]. Problem solving therapy has shown promising results in earlier pilot studies in Zimbabwe, particularly among participants who have experienced negative life events [44, 61]. It guides participants through a process to list the problems (negative events) that they face, select a problem to tackle based on the client’s priorities, develop a specific and measurable way of addressing the problem and execute an action plan aimed at addressing the problem while motivating and encouraging the participant to be pro-active in the process [44].

There is evidence supporting the use of lay health workers to deliver such interventions for CMD in LMIC [24] with a number of well designed randomized controlled trials suggesting that this is a cost-effective way of addressing the treatment gap for MNS [21, 66–68].

Our study’s main limitation is the lack of data on HIV staging and viral load markers. These factors can be associated with increased risk of developing CMD and depression [69–71]. Although similar high rates of CMD and depression have been reported from other LMIC [6, 72], the use of a clinic-based sample of people predominately visiting the clinic for physical ailments reduces the generalizability of our findings. Furthermore, the use of screening tools without a gold standard to confirm diagnosis may result in the inclusion of participants who are not actually clinically depressed [73]. Nevertheless, our cut-off score was based on a validation study conducted in Zimbabwe and provide an indication of the likely margin of error on these measures. Studies utilizing clinical interviews have found lower rates but these are still higher than those found in non-HIV infected people [7]. In addition, data on the type of HAART being taken by study participants was not available. A number of HAART drugs can increase the risk of psychiatric disorders including depression [74–76]. Although education and employment have been shown in previous studies to be associated with CMD and depression, our study indicates that this was not the case. This may have been due to the homogenous nature of the study population recruited from the same geographical location with similar socio-economic status. The low uptake of male participants is consistent with previous studies carried out in primary health care facilities in the country and other parts of the world [44, 77–79].

This study highlights, as in previous earlier studies in non-HIV infected populations, the association of female gender and negative life events with CMD and depression. Of note is the finding of high prevalence of these disorders in PLWH on HAART. This highlights the need to introduce interventions within HIV care clinics at primary health care level aimed at addressing this burden, which if left unchecked could negatively impact the gains made in the fight against HIV/AIDS in the last 2 decades. We further recommend the use of a two-stage screening process particularly where resources are scarce to avoid the inclusion of false positives, with the second stage consisting of a diagnostic screen. However, in resource-poor settings with large numbers of LHWs a greater sensitivity and a lower specificity would be recommended to ensure that all possible cases are included. We further recommend that the Option B+ program be administered with a psychological intervention. Further studies are needed to look into the development of multiple stage screening delivered by LHWs.