Intervening for HIV prevention and mental health: a review of global literature

Abstract Introduction Numerous effective HIV prevention options exist, including behaviour change interventions, condom promotion and biomedical interventions, like voluntary medical male circumcision and pre‐exposure prophylaxis. However, populations at risk of HIV also face overlapping vulnerabilities to common mental disorders and severe mental illness. Mental health status can affect engagement in HIV risk behaviours and HIV prevention programmes. We conducted a narrative review of the literature on HIV prevention among key populations and other groups vulnerable to HIV infection to understand the relationship between mental health conditions and HIV prevention outcomes and summarize existing evidence on integrated approaches to HIV prevention and mental healthcare. Methods We searched five databases for studies published from January 2015 to August 2020, focused on HIV prevention and mental health conditions among key populations and individuals with serious mental illness. Studies were included if they evaluated an HIV prevention intervention or assessed correlates of HIV risk reduction and included assessment of mental health conditions or a mental health intervention. Results and discussion We identified 50 studies meeting our inclusion criteria, of which 26 were randomized controlled trials or other experimental designs of an HIV prevention intervention with or without a mental health component. Behaviour change interventions were the most common HIV prevention approach. A majority of studies recruited men who have sex with men and adolescents. Two studies provided distinct approaches to integrated HIV prevention and mental health service delivery. Overall, a majority of included studies showed that symptoms of mental disorder or distress are associated with HIV prevention outcomes (e.g. increased risky sexual behaviour, poor engagement in HIV prevention behaviours). In addition, several studies conducted among groups at high risk of poor mental health found that integrating a mental health component into a behaviour change intervention or linking mental health services to combination prevention activities significantly reduced risk behaviour and mental distress and improved access to mental healthcare. Conclusions Evidence suggests that mental health conditions are associated with poorer HIV prevention outcomes, and tailored integrated approaches are urgently needed to address overlapping vulnerabilities among key populations and other individuals at risk.

oral pre-exposure prophylaxis (PrEP) [7], the dapivirine ring and other new, longer-acting PrEP modalities [8] and combination prevention packages incorporating behavioural, biological and structural interventions [9,10]. However, the over-representation of key populations and other vulnerable groups among new HIV infections suggests that it is critical to improve their engagement with these effective HIV prevention strategies.
Among key populations and other vulnerable groups, the same inequities that drive HIV infection also increase the risk for mental health conditions [11]. Social exclusion and marginalization, poverty, violence and discrimination create cycles of vulnerability to both HIV and worsening mental health status [12]. A review of the mental health of sexual minorities reported elevated rates of depression, bipolar disorder, suicide attempts and drug use disorders across sexual orientation and genders [13]. Mental health conditions are disproportionately prevalent among incarcerated people; people in prisons are at increased risk of all-cause mortality, suicide, self-harm, violence and victimization [14].
Vulnerability to HIV and mental health conditions also intersect in adolescence, a sensitive period of neuropsychological and social development during which adolescents seek greater autonomy, takes risks and initiate sexual activity, with genderspecific consequences [15,16]. Globally, 1.7 million adolescents (ages 10 to 19) live with HIV, and nearly 60% of adolescents living with HIV are girls [17]. Most mental disorders that persist in adulthood begin in adolescence [18]. Young key populations have elevated rates of depression, suicidal ideation and intent and traumatic stress [19,20]; meanwhile, adolescents with mental health conditions may be vulnerable to deficits in emotion regulation when making decisions about sex [21].
Lastly, people with SMI experience specific challenges related to living with symptoms that can be disabling, in addition to social, economic and gender-specific vulnerabilities [22][23][24][25][26][27]. These increase their risk of coercive sexual encounters, transactional sex, and sex with partners at high risk of HIV and unsafe drug use [28][29][30]. Notably, these groups are not mutually exclusive: individuals often have multiple marginalized identities leading to overlapping and intersecting vulnerability to both HIV and mental health conditions. Few studies have examined HIV prevention and mental health among HIV-negative key populations and vulnerable groups eligible for combination prevention, condom promotion, PrEP and other HIV prevention services. HIV prevention programmes providing behavioural counselling and PrEP interventions have found that young key populations have a high prevalence of depressive, anxiety and traumatic stress symptoms [19,20,31,32]. Consequently, expert commentaries have called for integrated, culturally appropriate mental health and HIV prevention interventions to address the needs of people living with mental health conditions; these include social interventions, like cash transfer programmes, economic empowerment strategies and gender-based violence services, alongside HIV prevention counselling, or PrEP delivery with psychotherapy and social support interventions to address mental health conditions [19,20,33,34].
From 2014 to 2016, several reviews reported the evidence on the need for and results of HIV prevention interventions focusing on people with SMI [2,[35][36][37]. These reviews found that there was considerable heterogeneity in the outcomes of HIV prevention interventions in this population. Behavioural skills training and HIV risk reduction counselling were effective in reducing the number of sexual partners and condomless sex in some trials, but few studies demonstrated lasting effects post-intervention [36].
A synthesis of more recent literature on the range of mental health conditions (from stress to SMI) is needed to understand the latest evidence on HIV prevention interventions and their outcomes for people managing or at risk for mental health conditions. We conducted a narrative review of global research on mental health and HIV prevention among adolescents and young women, key populations and people with SMI. Our primary aim was to understand how symptoms of mental health conditions influence HIV prevention intervention engagement and outcomes among these groups. We also reviewed evidence of integrated approaches to HIV prevention and the management of mental health conditions. We organize our findings around two questions: 1) what do we know about the relationship between mental health and HIV prevention or risk behaviours? 2) what do we know about interventions that address mental health in the context of HIV prevention?

| METHODS
We conducted a structured narrative review of the literature in peer-reviewed journals to examine the relationships between HIV prevention, risk reduction and mental health conditions.

| Search strategy
Two authors worked with a university informationist to select search terms and determine filters for five databases. We searched PubMed, Web of Science, CINAHL, PsychINFO and EMBASE for English-language studies published between 1 January 2010 and 25 August 2020. We constructed search terms on HIV prevention; interventions, services or programmes; and mental disorders. We ran these base search terms with search terms for each of the three populations of interest: (1) adolescents (boys and girls) and young women, (2) people with SMI and (3) UNAIDS-defined key populations (i.e. MSM, sex workers, transgender persons, PWID and prisoners and other incarcerated people as key population groups [38]). (See Additional File S1: Search Strategy.)

| Inclusion and exclusion criteria
Studies were eligible if they met the following criteria: (1) assessed correlates of HIV risk reduction activities or evaluated an HIV prevention intervention; (2) included assessment of mental health conditions, cognitive processes or tailored an intervention for people with mental disorders. We included studies related to the use of structural and behavioural HIV risk reduction approaches as well as biomedical prevention interventions. We excluded qualitative studies. Given the scope of the literature on drug use and HIV prevention, which merits its own review, we eliminated studies focused on harm reduction and prevention or treatment of substance use disorders if there was no mental disorder data. We also eliminated studies exclusively focused on HIV treatment as prevention for this review. Although treatment as prevention studies have the ultimate goal of preventing HIV transmission, they tend to focus on populations living with HIV who may have distinct HIV care and mental health needs compared with at-risk, HIVnegative populations. Eligible studies reported the outcomes of HIV prevention trials, cross-sectional and longitudinal analyses or quasi-experimental studies.

| Study selection and data extraction
We restricted our review of papers to January 2015 to August 2020. There were overlapping topical reviews of the literature prior to 2015 (e.g. for people with SMI). Given the scope of prevention research publications and the rapid evolution of HIV prevention science, the criterion of a 5-year period of review targets current research most relevant to practitioners and researchers. However, if studies from this time period were sub-studies, secondary analyses or otherwise related to a primary trial that met our inclusion criteria, we also added the primary trial to the dataset. We identified five such studies, three of which were published between 2010 and 2014. We included six studies published from 2015 to 2020 identified in the PrEP literature (e.g. references of other publications or adjacent studies in a search database) that met our criteria for inclusion. All studies were imported into Endnote (X9) to remove duplicates, book chapters, conference abstracts and theses. The resulting set was imported into Abstrackr (Tufts Evidence-based Practice Center, beta version) for title and abstract review [39].
Authors (TC, JV, LO) conducted an initial screening of 20 abstracts to arrive at consensus on studies for inclusion and exclusion, followed by screening of all abstracts using Abstrackr Beta [39]. Full-text manuscripts were assessed independently by authors (PC, TC, JV, LO, LC, JS). PC resolved disagreement in the process of study selection.
Authors used spreadsheet software to extract the following information: lead author, year of publication, country of study, sample size, study population, study design, study objective, HIV prevention approach, approach category, mental health component in intervention, HIV prevention outcome, mental health assessment used, and key HIV prevention/mental health finding. PC and JV conducted a second-level review of all eligible manuscripts. Due to the heterogeneity of our sample in terms of intervention type, outcome and populations, we conducted a qualitative synthesis of study findings.

| Study characteristics
A total of 3340 articles were identified through search criteria. After removing duplicates, book chapters, conference abstracts and theses, and restricting to 2015 to 2020 publication years, 1023 articles were eligible for the title/abstract screen. A total of 148 articles were full-text reviewed and 46 were eligible to be included in this review. We added 11 articles after full-text review, yielding a total of 57 articles in this review ( Figure 1). We counted the full texts of multiple articles as one study if the same interventions were administered to the same study population or to a subset of the study population, reducing our total number of studies to 50.

| Adolescents and young women
Nineteen studies enrolled adolescents or young women and reported HIV prevention and mental health-related factors as explanatory or outcome variables (Table 1). Depressive and anxiety symptoms were not associated with retention in a combined HIV risk reduction and alcohol and drug use reduction programme for homeless young adults [40].

| Behavioural intervention trials and quasi experiments with a mental health component
Thirteen studies tested the efficacy of an HIV prevention intervention utilizing theory-based skills building and reported both HIV and mental health outcomes among adolescents and young adults (Table 1). All studies demonstrated statistically significant effects on HIV prevention outcomes through follow-up periods ranging from immediately post-intervention to 12 months [41][42][43][44][45][46][47][48][49][50][51][52].
Several studies clearly articulated behavioural or mental health symptom targets and integrated dialectical behaviour therapy techniques [41,42], emotion regulation [44,49], cognitive behavioural skills training and psychoeducation for self-  harm prevention [43] cognitive processing therapy [47], problem-solving for stress reduction [52] and problem-solving therapy and creative therapies for anxiety and aggressive behaviour [45]. Others identified externalizing/internalizing symptoms [46], emotional and mental health [53] as intervention components, but did not specify the interventional approach. Two studies integrated mental health promotion activities, including psychoeducation and cognitive-behavioural skills-building for parent-youth communication, while introducing behavioural strategies for reducing sexual risk [50,51]. The efficacy of the mental health components was mixed: six studies reported no effect of the mental health intervention on adolescents or young women [41,42,46,[48][49][50]53]. Parental mental health outcomes were indirectly related to adolescent HIV risk behaviours in two behavioural intervention studies [48, 54,55]. One showed that building a closer relationship to support improved communication about sexual health is partially contingent on supporting the mental health of the adolescent and the caregiver [54]. Another produced greater improvements in sexual communication and parental monitoring among parents with more psychiatric symptoms at the three-month follow-up [55]; adolescents reported significant reductions in sexual risk behaviours at three months [48], but not after 12 months [56].

| Structural intervention trials
One randomized controlled trial tested a cash transfer intervention, which significantly reduced depressive symptoms among young men only and delayed sexual debut among young men and women [57].

| The relationship of mental health to HIV prevention behaviours
Five studies assessed mental health as an explanatory variable for HIV risk or preventive behaviours (Table 1). Depressive symptoms were associated with a slightly lower intent to undergo HIV testing and counselling among Kenyan youth [58] and with lower PrEP adherence among South African young women, even after accounting for stigma and for PrEP optimism, i.e. belief in the protective effects of PrEP [59]. Depressive symptoms were associated with a greater likelihood to initiate PrEP, as were low social support and high perceived self-efficacy to take medication daily, among Kenyan adolescent girls and young women [60].
Depressive symptoms were also measured in the context of VMMC service uptake and parental consent among circumcised and uncircumcised adolescents [61]. Circumcision without parental consent was associated with being an orphan, out of school, probable clinical depression (CES-D score >16), and poorer quality of life. In addition, higher proportions of uncircumcised youth were depressed compared to the circumcised, possibly due to social pressure associated with VMMC campaigns or shaming of uncircumcised boys in high uptake communities.
A South African study of HIV risk, adverse childhood experiences (ACEs), adolescent mental health, and free schooling for adolescent girls, found that relationships between ACEs and HIV risk behaviour were mediated by internalizing and externalizing symptoms [62]. Free schooling was associated with fewer externalizing symptoms, suggesting that the mental health-promoting effects of free education confer some protection against HIV risk.

| Key populations
Twenty-six studies enrolled members of key populations and reported HIV prevention outcomes and mental health-related factors as explanatory or outcome variables (Table 2).

| Behavioural intervention trials and quasiexperiments
Eight studies reported the efficacy of a theory-based intervention for HIV risk reduction [63][64][65][66][67][68][69][70]. Most reported favourable changes in risk reduction. One pilot intervention's results did not achieve significance [69], and a community-level intervention yielded positive and negative outcomes [63] ( Table 2). Of these studies, six integrated mental health components including psychoeducation for depression and post-traumatic stress disorder (PTSD) and affect management to reduce dissociation [64], behavioural activation for methamphetamine-related anhedonia [65,66], psychoeducation and cognitive-behavioural and acceptance-based coping with stressors [67], cognitive behavioural therapy (CBT) for trauma [68], and CBT for depression [70]. With the exception of one study that did not describe the mental health component [63], interventions with significant reductions in mental disorder symptoms (depression, anhedonia, PTSD) utilized CBT, behavioural activation and affect management with psychoeducation. All interventions were delivered in group formats.

| Biomedical intervention trials
Analysis of a PrEP safety trial among MSM and TGW revealed a high prevalence of moderate depression across participants in both study arms, but no difference in depression-related adverse events or reports of suicide attempts and self-harm between participants in the PrEP and placebo arms [71]. A sub-group analysis showed possible severe depression for MSM and possible moderate depression for TGW were associated with reduced PrEP adherence, but were infrequent [72]. The authors emphasized the importance of ensuring access to PrEP for people with depressive symptoms [71][72][73].

| The relationship of mental health to HIV prevention behaviours
Eighteen studies examined mental health as a correlate of HIV risk reduction in studies of HIV or STI testing and counselling, condom use, and awareness of, adherence to or engagement with PrEP (Table 2).
Of these, seven examined the relationship of mental health variables to HIV or STI testing in cohorts and community samples (Table 2). In four studies, fewer symptoms of a mental disorder were associated with HIV or STI testing, and greater severity of symptoms was associated with less testing [74][75][76]. HIV testing among TGW in Malaysia was associated with higher current scores of mental health functioning as well as having a previous diagnosis of depression [77]. In contrast, ever having a depressed mood for more than two weeks was independently associated with having an STI test in the past    month among MSM in Burkina Faso [78]. Notably, learning one's HIV status, whether positive or negative, was associated with significant reductions in the severity of depressive symptoms among Ugandan and Zambian FSW populations and was not associated with suicidal ideation [79]. Several studies demonstrated that poorer mental health (especially severe depression, depression and PTSD, or anxiety) was associated with less consistent condom use or less perceived condom self-efficacy among MSM [80], young women who sold sex [81], and homeless men who also used drugs and traded sex [82]. Condom use self-efficacy and depression were partial mediators of the relationship between sexual stigma and inconsistent condom use among MSM [83]. MSM on PrEP who reported poly-drug use and depression were significantly more likely to report receptive condomless anal intercourse than those who only reported poly-drug use or only reported depression at baseline assessment [84]. Conversely, among gay and bisexual MSM receiving PrEP in the United Kingdom, despite increased depressive symptoms over time, neither depressive symptoms nor interpersonal violence were associated with sexual risk behaviours [85,86].
Two studies showed that distinct categories of mental disorder symptoms were associated with lower adherence to PrEP: anxiety symptoms and a history of childhood trauma [87], baseline depression and substance use among MSM in a PrEP adherence trial [88]. A third study reported higher depression scores were significantly related to greater acceptability of peer navigation to assist with PrEP engagement among Black and Latinx MSM [89]. A fourth study found no relationship of depressive symptoms to PrEP adherence, but harmful alcohol use and moderate/high-risk cocaine use predicted nonadherence [90].

| People with severe mental illness
Three studies enrolled people with SMI in a clinical trial or cross-sectional surveys.

| HIV prevention interventions tailored for adults with severe mental disorders
A randomized controlled trial tested a sexual health promotion intervention among adults with SMI [91] (Table 3). Participants attending a three-session theory-based group intervention reported fewer episodes of unprotected sex acts through six months [91]. This study examined sexual health, broadly, as an area of attention for people with SMI.

| The relationship of mental health to HIV prevention behaviours
Two Brazilian studies examined people with SMI and HIV risk correlates (Table 3). One study showed that psychiatric symptom clusters had differential effects on condom self-efficacy: people with more severe negative symptoms (e.g. blunted affect, emotional withdrawal) were less likely to perceive themselves as capable of using condoms, condom negotiation, and/or condom acquisition [92]. In a second study, three-fourths of patients in eight public psychiatry clinics reported unprotected sex, but only 9% had participated in the clinics' risk-reduction programmes [93]. Participation was significantly associated with ethnicity, higher HIV knowledge, and receiving HIV testing in the past three months.

| HIV prevention and mental health service integration
Two studies in our sample described distinct approaches to HIV prevention and mental health service integration (Table 4). One assessed the outcomes of the Pehchan programme, a community-level intervention, which linked transgender persons to comprehensive community-based services providing combination prevention to transgender persons in India [94]. The programme yielded significant increases in testing referrals, HIV education reach, and access to mental health support through referrals to psychological services.
A survey of mental health programme directors in New York State showed that a majority of programmes treated people known to have HIV, assessed HIV risk, and provided HIV educational materials, and just over half referred people for HIV testing [95]. Between 20% and 32% of programmes offered services related to End the Epidemic activities in the state (e.g. HIV testing, PrEP education and PrEP prescriptions). Compared to past surveys, fewer mental health programme directors reported integration of HIV services and psychiatric services, and fewer identified themselves as fully integrated in 2017 compared to 2004, despite more programmes reporting larger caseloads of people with HIV or AIDS.

| Summary: Interventions that address mental health in the context of HIV prevention
Thirteen studies in our sample describe HIV prevention interventions (11 at the individual level, one social intervention, and one community systems intervention) with an embedded or linked mental health component(s) that reduced HIV risk behaviours and produced a more favourable mental health outcome ( Table 5). The majority of these interventions or services occurred in the community, and one occurred in a prison.
To our knowledge, this is the first review of global research on mental health and HIV prevention among adolescents and young women, key populations and people with SMI that includes multiple prevention modalities. We aimed to understand how mental disorder symptoms influence risk for HIV and preventive intervention outcomes among three populations vulnerable to HIV infection. Overall, this selection of studies, dominated by behavioural interventions to reduce sexual risk, suggests that poorer mental health is associated with HIV risk behaviour. Study findings help to answer two questions: 1) What do we know about the relationship between mental health and HIV prevention behaviours and 2) What do we know about the interventions that address mental health in the context of HIV prevention? 3.7 | The relationship between mental health and

HIV prevention behaviours
Having symptoms of a mental health condition was more often associated with fewer HIV prevention behaviours. In all but one study, depression impaired PrEP adherence. In the majority of HIV testing and counselling studies, having fewer symptoms of a mental disorder increased HIV testing. Depressive, anxiety and trauma symptoms usually reduced the likelihood of condom use or condom self-efficacy.
In a subset of studies, depressive symptoms (current or past) were associated with a greater likelihood of getting an HIV test, and with PrEP initiation [60]. In these cases, the negative effect of poor mental health may be mitigated by other factors that facilitate taking action (e.g. high perceived self-efficacy [60]), or poor mental health may enhance recognition of vulnerability and a need for support [77,89]. Evidence also showed that learning one's HIV status, whether negative or positive, did not worsen depressive symptoms or increase suicidality [79].
The relationship between HIV risk and mental health is sometimes indirect. For adolescents, parent-child communication skills, parenting styles and parental mental health status influenced successful sexual communication, which in turn reduced HIV risk [55]. Consistent with this finding, a recent review highlighted the benefits of family strengthening interventions for the mental health of youth affected by or living with HIV [96]. Gender and social adversity add to the complexity of understanding these associations. Cash transfers led to better HIV risk outcomes for boys and girls, but better mental health outcomes for boys [57]. Nevertheless, for vulnerable girls, access to social resources such as free school or cash transfers reduced behaviours directly and indirectly related HIV risk [57,62]. HIV programme implementers must also consider how the social vulnerabilities of adolescents may increase the risk of coercive participation in HIV prevention or may indirectly create barriers to interventions (e.g. VMMC), and consequently, greater emotional stressors for young people [61].

| What do we know about interventions that
address mental health in the context of HIV prevention?
Prevention scientists emphasize the importance of combination prevention and comprehensive, layered approaches that address contextual and individual risk factors for HIV prevention [34,97,98]. Relatively few examples of such comprehensive approaches emerged in our search. Although some interventions integrated elements to address intrapersonal, community and structural stressors (e.g. self-esteem, discrimination, minority stress, negative sexual identity, community connectedness, access to care) with HIV prevention [53,63], these may not be sufficient to reduce symptoms of depression, trauma or severe anxiety. When they yield positive effects on mental health, more research is needed to understand which components and delivery modes facilitate these outcomes. Assessment also influences outcomes: most studies in the sample utilized screening tools to assess mental health status, but did not always distinguish between diagnoses and symptoms. Some studies assessed mental health even if the HIV preventive intervention had no mental health component.
The thirteen studies in Table 5 that reported HIV risk reduction and improved mental health outcomes used three broad intervention approaches. At the systems level, effective linkage of key populations to HIV testing and mental health services occurred through robust referrals within a community setting managed by a trusted community organization and peer network [94]. The individual-level behavioural interventions, conducted in North America and Africa, applied elements of evidence-based psychological therapies or psychoeducation within a structured HIV risk reduction intervention for adolescents and young women and for key populations. Across these studies, the sample size, duration and strength of effects, and specification of the mental health component varied considerably. However, several interventions integrating CBT approaches for the treatment of trauma, depression, or selfharm reduction showed enduring mental health effects nine months to one-year post-intervention [43,68,70]. Though this latter group of interventions were often delivered by trained mental health specialists in our study sample, current global mental health literature demonstrates that less specialized providers can be trained and supervised to deliver effective, evidence-based psychological interventions like CBT in community settings [99][100][101].
Notably, few new intervention studies for people with SMI have been published since 2012, and studies from Africa, Asia, or Latin America are scarce [2]. The absence in the literature is mirrored by diminishing attention to HIV prevention, access to HIV services in public mental health settings in the United States [95]. Although social functioning-including establishing intimacy and expressing sexuality-is an essential part of the recovery process for people living with SMI, sexual health is largely unaddressed in typical mental health services [24,102].

| Implications for integrating HIV prevention and mental health and achieving global HIV targets
The Global AIDS Targets for 2025 call for «people-centred and context-specific integrated approaches» so that 90% of people at high risk of HIV are linked to services for mental health, sexual and gender-based violence and other relevant care [103]. The results of our review suggest that mental health and HIV prevention could be integrated by (1) identifying community partners and leaders for co-design, delivery and linkage of HIV and mental health resources and services; (2) using task shifting to train adherence counsellors, peers and nurses to administer evidence-based psychological therapies (e.g. problem-solving therapy, CBT, cognitive processing therapy) embedded in theory-based risk reduction for populations experiencing trauma and depression [104]; (3) making use of available mental health capacity-building resources from the World Health Organization and other sources [105-107]; (4) supporting gender-sensitive structural interventions for young people, like access to free schooling for girls and vulnerable youth; (5) expanding access to family-based interventions that enhance parenting and communication skills and (6) introducing a mental health component to adherence support for PrEP. True integration requires shared human resources, budgeting, and planning across HIV and mental health services in partnership with diverse community stakeholders [106, [108][109][110]. Prevention service providers can learn from task-sharing interventions that integrate mental health and HIV care [111][112][113].

| Limitations
Our study has several limitations. The included studies were heterogeneous and precluded the use of a meta-analysis of the results and effect sizes. Our review was systematized, though not systematic, and we may not have captured all representative studies. We did not rate the quality of the studies or conduct a bias assessment, but reported findings qualitatively. Although we captured studies from a diverse set of countries, the majority of studies are from the United States and interventions reflect the contextual specificities of the study populations and settings. Self-report of HIV risk behaviours, non-randomized study designs and small sample sizes may bias some study findings. Despite these weaknesses, the included study results reflect a broad range of countries and support the assertion that poorer mental health is linked to fewer HIV prevention behaviours and activities globally and that integrated interventions can reduce risk.

| CONCLUSIONS
Consistent with previous studies, current evidence suggests that mental health conditions are more often associated with poor HIV prevention outcomes, and integrated approaches are urgently needed to address overlapping vulnerabilities among key populations, vulnerable groups and individuals with SMI. Our review contributes a new synthesis of global literature on mental health and HIV prevention, spanning a broad range of prevention modalities; studies from high-, middle-and low-income countries; and diverse samples of key populations, high-risk groups, and people with SMI. We highlight the components of interventions that address symptoms of mental illness or psychological processes that influence mental health status. Importantly, these findings, in concert with the broader global mental health literature [11,114,115], suggest that integrating structural, social and individual-level HIV prevention and mental health interventions is feasible in diverse community settings. A renewed focus on implementing these integrated interventions and services could contribute to ending the AIDS epidemic, and specifically, to achieving the 2025 Global AIDS targets.

A C K N O W L E D G E M E N T
We are grateful for the assistance of Lynly Beard in the first stages of the project.

F U N D I N G
BHW was supported by grant number K01MH110599 from the National Institute of Mental Health. JV was supported by grant number K99MH123369 from the National Institute of Mental Health. The content of this paper is solely the responsibility of the authors and does not represent the official views of the National Institutes of Health.