Mental health is a basic human right for all people which is crucial to personal, community, and socio-economic development (1). However, the discrimination and stigma based on sexual orientation impede the right of men who have sex with men (MSM) to the highest attainable standard of physical and mental health (2). MSM are more vulnerable to poor mental health as they face more stigma and stress in their daily lives (3). Despite significant progress in creating LGBTQ+-friendly societies and mental health promotions, MSM still reported disproportionate mental and psychological conditions like depression, anxiety, distress, negative feelings, and emotions, or low self-identification (4, 5). It is estimated that around 35% of MSM worldwide had experienced depressive symptoms, much higher than the general population (13%) (6). The pooled prevalence of depression among MSM was 37% in Asia, 26% in Europe, 34% in Africa, and 35% in the Americas (6). Another study estimated that 21% of MSM had suicidal ideation, with a pooled prevalence of 20% in Asia and 19% in the Americas (7). These regional differences both reflect widespread contextual issues highlighting regional needs and expectations. Meanwhile, inadequate mental health care for MSM could further exacerbate the mental state of MSM, leading to a poor quality of life (8).

Systemic factors critically shape MSM’s mental health through legal and policy environments. Criminalization of same-sex acts exacerbates stigma, limits health access, and increases psychological distress (9). Conversely, protective laws that reduce structural discrimination by ensuring access to healthcare, in particular mental health services, and legal recourse against discrimination, have shown to improve mental well-being of MSM population (9, 10). The variety of laws and policies across different countries underscores their impact on either exacerbating or alleviating mental health disparities, highlighting the need for context-specific interventions.

Psychosocial interventions refer to non-pharmacological interventions that focus on psychological or social factors, aimed to improve symptoms, functioning, quality of life, and social inclusion when used in people with mental health conditions (11). Psychosocial interventions are an essential component of mental health services and contribute to improving the mental health and well-being of MSM (12). The current interventions for MSM are mainly cognitive behavior therapy (13), supportive groups (14), education (15), online self-care (16), etc. Some studies showed a better intervention effect for MSM on depressive symptoms, distress, self-efficacy, and coping by education and other forms of interventions (12, 15). Previous reviews showed that psychosocial interventions could reduce depressive and anxiety symptoms in sexual and gender minorities and MSM, and decrease substance abuse in MSM (17, 18).

According to World Health Organization, as a state of well-being, mental health is more than the absence of mental disorders, it also includes emotional well-being, good behavioral adjustment, relative freedom from anxiety and disabling symptoms, and a capacity to establish constructive relationships and cope with the ordinary demands and stresses of life (19). From the global perspective, it is necessary to conduct a comprehensive evaluation of the overall effectiveness of psychosocial interventions and their components for MSM to assess the effectiveness and provide evidence for developing quality and affordable mental health care to improve the well-being of MSM.

This study aimed to conduct a systematic review and meta-analysis to evaluate the effectiveness of psychosocial interventions for mental health among MSM, with a particular focus on both overall effects and specific outcomes including depression symptoms, anxiety symptoms, substance abuse, suicidal ideation, stress, coping, emotion, social function and identity. This study would also compare the effects of interventions with different characteristics, such as face-to-face intervention and online intervention, self-care and interventions provided by professionals, etc. The hypothesis was that psychosocial interventions would have superior effects in improving overall mental health and specific outcomes compared to the control. This study would provide evidence-based recommendations for mental health care providers to better address the mental health challenges faced by MSM and improve their overall health and well-being. This study addresses key policy gaps in global mental health by evaluating psychosocial interventions for MSM. It demonstrates how interventions can reduce inequalities, safeguard the human rights of MSM, and inform reforms that align mental health services with global commitments to equality, non-discrimination, and equitable healthcare, thereby providing policymakers with evidence to improve services and reduce disparities.

A total of 11,950 results were searched from four databases and two registers. 1,669 records were duplicates and 10,087 results were excluded by title and abstract screening. Fourteen studies met the inclusion criteria and were included in this study after the full-text screening of 194 studies ( Figure 1 ) (13–16, 28–37).

There were 11 (78·57%) RCT studies and 3 (21·43%) quasi-experimental studies ( Table 1 ). All the studies were published between 2010 and 2024, and most (78·57%) were published in the last five years. Eleven studies (78·57%) were conducted in the USA, one (7·14%) from China, one (7·14%) from Austria and Germany, and one (7·14%) from Canada.

A total of 2,676 participants were recruited including 1,209 participants in intervention groups (median: 60, IQR ¹ ‡: 33-96) and 1,467 participants in comparison groups (median: 56, IQR: 32·25-126·75). The participants were primarily adolescents or young adults, with a mean age between 15 and 34 years old. Four studies (28·57%) only recruited MSM population or gay and bisexual men, and the rest recruited both MSM and other sexual minorities like lesbian and bisexual women with a proportion of MSM ranging from 12·1% to 59·1% ( Table 1 ). The included studies measured various mental health and psychosocial function outcomes, mostly self-reported. The most frequently measured outcomes were symptoms of depression, which were evaluated in 11 of the studies and symptoms of anxiety, which were evaluated in nine studies. Substance abuse was assessed in six of them. Seven studies measured coping-related outcomes; eight on emotion-related outcomes; five on self-identity-related and social function-related outcomes, and five on stress-related outcomes ( Table 1 ).

Among the 14 studies ( Table 1 ), nine interventions (64·28%) were delivered online or internet-based (15, 16, 29, 31, 32, 34–37), and five (35.17%) were delivered face-to-face (13, 14, 28, 30, 33). Six interventions (42·85%) were mainly self-care or without a trained specialist (15, 16, 29, 31, 34, 35), while the other eight (57·14%) were mainly provided by trained specialists (13, 14, 28, 30, 32, 33, 36, 37). The median number of sessions was 8 with an interquartile range from 3 to 10. The longest intervention duration was 18 weeks with a weekly session (33) and the shortest duration was 10 minutes with only one online video session (35). Based on the duration and session, there were five low-frequency interventions (35·17%) (16, 29, 31, 34, 35) with a duration of less than four weeks or a maximum of three sessions and eight high-frequency interventions (57·14%) (13–15, 30, 32, 33, 36, 37) with a duration of more than four weeks or a maximum of 18 sessions. All interventions have a basic standardized process. At the end of the intervention, the median adherence rate was 90% with IQR from 74% to 100%. Most studies had good adherence of more than 80%. Six studies (14, 15, 30–32, 37) had recorded adverse events during the intervention and four studies (14, 15, 31, 37) reported no adverse events. One study reported five active suicidalities (four in the intervention group and one in the comparison group) (30). Another study reported two suicidal ideation and/or attempts (one in the intervention group and one in the comparison group) (32). The rest had no adverse event information. Few studies reported adverse events, but adverse events were generally rare.

Cognitive behavior therapy (CBT) was one of the most widely adopted therapeutic interventions for the treatment of various mental health disorders. Seven interventions included in this review were developed based on the principles of CBT (13, 16, 28, 30, 32, 36, 37). Another four studies adopted psychoeducation as a main approach to improve the mental health of participants (14, 15, 34, 35). Two further studies adopted expressive writing to improve mental health (29, 31). Meanwhile, most interventions adopted more than one therapy or technique. Some interventions based on CBT also contain the contents of psychoeducation (37), mindfulness (13, 30), or dialectical behavior therapy (28). Psychoeducation might involve cognitive adjustment (34), or dialectical discussion as well (14). Most interventions were adapted for sexual minorities and mainly focused on minority stress (13, 15, 16, 30–32, 36, 37). The minority stress theory was commonly adopted in the intervention design, development, and implementation which believed that sexual minorities usually faced extensive distal and proximal stress leading to poor mental health (3). It provided a comprehensive framework to understand the stressors from different levels and aspects. Researchers can design and develop their interventions based on that making their interventions more targeted and specific. Several studies adopted this theory into CBT interventions to help participants identify stressors and develop strategies to enhance stress coping (13, 16, 30, 32, 36, 37). Some studies adopted this model in psychoeducation materials to strengthen the self-adjustment (15).

The psychosocial intervention had a significant effect size on improving overall mental health status (Hedges’ g = 0·14, 95%CI: 0·08-0·21, p<0·001, n=14) ( Figure 2 ). There was no significant heterogeneity between studies (I² = 28·23%, Q = 18·11, p = 0·15). Sensitivity analysis showed that effect size remained significant after excluding four studies considered to be at high risk-of-bias (Hedges’ g = 0·15, 95%CI: 0·07-0·22, p<0·001, I² = 9·51%, n=10). The overall effect size at the follow-up was 0·14 (95%CI: -0·05-0·33, p = 0·16, I² = 81·95%, n=8).

For specific outcomes at post-intervention ( Table 2 ), the effect size was 0·25 (95%CI: 0·09-0·41, p = 0·002, I² = 57·77%, n=11) for depressive symptoms, 0·20 (95%CI: 0·12-0·29, p<0·001, I² = 46·78%, n=9) for anxiety symptoms, and 0·19 (95%CI: 0·10-0·28, p<0·001, I² = 11·49%, n=6) for substance abuse. The intervention had a positive effect on stress (Hedges’ g = 0·18, 95%CI: 0·03-0·33, p = 0·02, I² = 0%, n=5), coping (Hedges’ g = 0·21, 95%CI: 0·06-0·36, p = 0·006, I² = 0%, n=5), emotion (Hedges’ g = 0·16, 95%CI: 0·06-0·25, p = 0·001, I² = 45·86%, n=8), and identity (Hedges’ g = 0·19, 95%CI: 0·07-0·30, p = 0·001, I² = 49·60%, n=5). The intervention didn’t have a positive effect on suicidal ideation (Hedges’ g = 0·01, p>0·05) and social function (Hedges’ g = 0·01, p>0·05). At follow-up, the intervention groups had a superior effect on depressive symptoms (Hedges’ g = 0·36, 95%CI: 0·09-0·70, p = 0·044, I² = 86·14%, n=6), anxiety symptoms (Hedges’ g = 0·42, 95%CI: 0·01-0·82, p = 0·044, I² = 89·06%, n=5) and substance abuse (Hedges’ g = 0·23, 95%CI: 0·04-0·42, p = 0·015, I² = 50·07%, n=5).

Online interventions had positive effects on the overall mental health status, coping, emotion, and identity, while face-to-face interventions had positive effects on depressive symptoms, anxiety symptoms, and substance abuse ( Table 3 ). Interventions only for MSM and for LGBTQ+ both were effective in overall mental health status. Intervention for MSM had positive effects on depressive symptoms, anxiety symptoms, substance abuse, and identity, while intervention for LGBTQ+ had positive effects on stress and emotion. The self-care interventions had better effects on emotion and overall status, and the interventions provided by trained specialists had better effects on depressive symptoms, anxiety symptoms, substance abuse, stress, coping, and overall status. The low-frequency interventions had better effects on emotion and overall status, and the high-frequency interventions had better effects on depressive symptoms, anxiety symptoms, substance abuse, stress, coping, social function, and overall status. However, these variables had no moderating effect except for the effect sizes of low- and high-frequency interventions on social functioning (p = 0·001). We assessed the moderating effects of “age” and “number of sessions” by meta-regression, but there was no significant effect. Based on the above research findings, we can select appropriate intervention methods in practice by considering the specific context. When mental health resources are limited, short-term online self-care interventions may be a more suitable option.

Three RCT studies and one quasi-experimental study were assessed at “high risk-of-bias” or had “serious bias”. Seven studies were assessed as having “some concerns” or having “moderate bias”, and four studies were at “low risk-of-bias” (S8-9). The funnel plot ( Figure 3 ) and Egger’s test (intercept: 0·32, p = 0·72) did not show a significant publication bias. The quality of evidence was estimated by GRADE to be moderate (S10). The certainty was downgraded from high to moderate because of studies with a high risk-of-bias. This downgrade reflects how bias may influence the reliability of the results. However, the overall findings still support the effectiveness of psychosocial interventions. By downgrading the rating, we aim to provide a more cautious interpretation and highlight the need for future research with more rigorously designed studies to strengthen the evidence base.

This systematic review and meta-analysis provide a comprehensive evaluation of the current psychosocial interventions for MSM mental health. An overall benefit of psychosocial intervention for mental health in MSM was observed in this review with a pooled effect size of 0·14 (95%CI: 0·08-0·21), which supported the research hypothesis. The results showed that the psychosocial interventions had more improvements in general mental status, mental health, and psychosocial functions than the comparison groups. No publication bias or substantial heterogeneity was found in this study. Moreover, the subgroup analyses based on the characteristics of interventions would provide additional information for future intervention development. However, the effectiveness of these interventions must be considered in the context of various external factors that drive poor mental health outcomes for MSM. Violence (38, 39), stigma (40), and discrimination (41) are key contributors that exacerbate mental health disparities. These social stressors not only negatively impact psychological well-being but also limit MSM’s access to mental health support and resources. In addition to providing interventions, addressing these root causes of poor mental health is crucial for improving outcomes in this population. Based on our findings, researchers could have a better understanding of psychosocial interventions for MSM and their characteristics and further develop more effective interventions.

The result of this study indicated that the interventions had a small effect size on the overall mental health status of MSM, which was in line with another study conducted by Pantalone et al. They reported a similar effect size (0·20) of intervention for mental health and substance abuse in MSM (18). For the depression symptoms, our study found a small-to moderate effect (0·26) of intervention, similar to a study for MSM living with HIV (0·28) (12). We found a small effect of intervention for anxiety symptoms (0·21) and stress symptoms (0·18), consistent with another study for sexual and gender minorities (17). From these studies, it was indicated that psychosocial interventions tend to have a small-to-moderate impact on the mental health of MSM (42). Apart from these symptoms, we also assessed the effectiveness in coping, identity, social function, and emotion aspects providing a more comprehensive assessment in this study and finding positive effects in intervention groups. Even if the effect is small, when applied to a large population, it may lead to meaningful improvements at the population level. Furthermore, if interventions are affordable and accessible, their widespread implementation could yield significant public health benefits, especially for communities with limited access to mental health resources.

No significant moderating effect was found for subgroup analyses. While no significant differences were found between online interventions and face-to-face interventions, the online intervention appeared to be more consistent in improving psychosocial aspects (e.g., coping, emotion, and identity), whereas the face-to-face intervention appeared to be more consistent in improving clinical symptoms (e.g., depression, anxiety, and substance abuse). MSM-only interventions were effective on overall status, mental health aspects, and identity, while LGBTQ+ interventions were effective on overall status and psychosocial aspects. However, there were no significant differences between the two interventions. Both self-care interventions and interventions provided by specialists had the same effect on overall mental health. The same results were also shown in the low and high-frequency interventions. This similarity may be attributed to the high degree of overlap between interventions. Most self-care interventions tend to be brief and low-frequency, whereas most high-frequency interventions tend to be provided by specialists. We only found significant moderating effects on social function between low- and high-frequency interventions. It might be that there were fewer studies measuring social function and therefore our results are susceptible to the extremes of particular studies. Although overall statistical significance was not achieved, the trends and consistency observed in the data showed that interventions with certain characteristics may have potential advantages in specific areas, which may still be important and have a positive impact on the development of future interventions. In high-income countries, psychosocial interventions benefited from strong health systems, sufficient trained providers, and supportive policies, making care more accessible for MSM. In contrast, in low- and middle-income countries (LMICs), limited resources made community-based or online approaches more scalable and acceptable. Self-care and digital delivery overcame barriers of distance and workforce shortages, while in urban settings, more intensive face-to-face interventions by professionals were feasible due to greater service availability.

In terms of intervention therapies, more than half of the studies adopted the principle of CBT. CBT is generally considered to be more standardized, structured, and focused as compared to other forms of psychotherapy, which makes it easier for trained intervention providers to implement (43). Moreover, CBT is usually a short-term intervention that can be delivered in various formats such as online, individual or group-based sessions (44). The core approach was to identify negative thought patterns and replace them with more positive, realistic ones (45). This process helped individuals to develop practical skills for coping with difficult emotions and situations (46). The CBT interventions included in this study have shown a positive impact on improving mental health by enhancing emotion regulation, cognitive reconstruction, strengthening coping mechanisms for minority stigma and stress, and facilitating participants to develop motivation and skills for behavioral change.

Other commonly adopted interventions were psychoeducation and expressive writing. Psychoeducation is a cost-effective and efficient way to improve mental health and can be conducted quickly with large sample sizes (47, 48). These interventions were usually passive education by providing information on emotion regulation, sexual minority communities, guidance on stress coping, or dealing with the stress of coming out to participants (15). Information is usually provided through web pages or online videos so that participants can access these resources at a flexible time (14, 15). Psychoeducation for MSM or LGBTQ+ was usually in the form of brief, short-term, or single-session intervention (35). The adoption of social media and Internet technology facelifted psychoeducation, making it easier to disseminate, more effective, and more attractive. Expressive writing prompts participants to write about their sexual minority-related stressful events, which can facilitate cognitive processing of unresolved psychological stressors (28, 31). For MSM, writing about their stressful experiences may enhance exposure to stress-related cues and eventually habituate them. Participants can benefit from exposure and cognitive adjustment during writing tasks.

This systematic review and meta-analysis demonstrate the overall benefits of psychosocial interventions for MSM mental health, with a small-to-moderate effect size. These interventions could reduce clinical symptoms like depression, anxiety, and substance abuse while improving psychosocial functions such as self-identity, coping, and social function. Future research should explore the effectiveness of interventions across different socio-cultural contexts and evaluate the impact of policy measures, while also focusing on enhancing mental health accessibility, reducing minority stress, and promoting equity in MSM populations.