Studies were included if: (i) the population comprised people bereaved by suicide; (ii) they utilised quantitative, qualitative, or mixed-methods; (iii) they reported data on the psychosocial characteristics of men bereaved by suicide, including outcomes related to grief, mental health, social functioning, or suicidal behaviour; (iv) they were published in English in a peer-reviewed journal; (v) they compared men bereaved by suicide to non-bereaved men, men bereaved by other means, suicide-bereaved women, or did not use a comparator.

Studies were excluded if: (i) the population did not include people bereaved by suicide; (ii) they utilised other methods such as systematic review or case study; (iii) they did not report suicide bereavement-related data on men; or (iv) they were not available as full-text, in English, published in a peer-reviewed journal.

Studies rarely differentiate between cisgender men, transgender men, and other individuals categorised as male in large datasets. Therefore, the term “men” in this review encompasses all groups, with an acknowledgment that samples likely consist predominantly of cisgender men.

Researcher NL conducted searches on 21 April 2023 in MEDLINE, Embase, Emcare, PsycINFO and Scopus. The PRISMA diagram for this search is depicted in Figure 1. The search strings comprised concepts of suicide, bereavement, and men. MEDLINE was searched with the following combination of MeSH and keywords: (suicid*.mp OR Suicide/) AND (Grief/ OR grie*.mp OR Bereavement/ OR bereave*.mp OR suicide bereave*.mp OR loss by suicide.mp OR bereave* by suicide.mp OR suicide expos*.mp OR suicide loss survivor.mp) AND (Male/ OR male*.mp OR men.mp OR Men/ OR man.mp OR boy*.mp OR Fathers/ OR father*.mp OR Men’s Health/ OR men* health.mp OR Masculinity/ OR masculin*.mp).

The other Ovid-associated databases (Embase, Emcare, PsycINFO) were searched using the same search string, and similar searches were conducted in Scopus. Searches were limited to English-language studies published in peer-reviewed journals, but not by location or date of publication. Eligible articles were subjected to a forward citation search and their references lists were searched to identify additional articles. The database searches were re-run on 16 November 2023, with two additional terms, “suicide survivor*.mp” and “bereave* by *al suicide.mp” included. The PRISMA diagram for the updated search is depicted in Figure 2.

Researcher NL imported all records into Covidence, a collaborative, web-based systematic review programme (34), and removed duplicate records. Reviewer NL screened the titles and abstracts of the studies to assess their potential eligibility. Researchers NL and KA independently screened full-text articles of potentially relevant studies to confirm eligibility. Disagreement regarding the inclusion of three articles was resolved through consultation with a third researcher, KK. Figures 1, 2 present the search and selection process.

The research team created a data extraction form including the following variables: study design, eligibility criteria, sample size, age of participants, time since bereavement, relationship to the deceased, comparator or control group, outcome and outcome measure, and main findings. Researchers NL and KK independently extracted data from the included studies, and they resolved any discrepancy through discussion.

Evaluation of the quality and potential bias in included studies was conducted using the Mixed Methods Appraisal Tool (MMAT) (35). This tool provides distinct criteria for different types of studies: qualitative studies, quantitative randomised controlled trials, quantitative non-randomised studies, and quantitative descriptive studies. For mixed-methods studies, the MMAT requires evaluation based on the criteria for each specific method used within the study and additional criteria exclusive to mixed-methods designs. Overall scores were summarised for each method based on the suggestions of Hong with a star rating based on the number of quality criteria met (36). Researchers NL and KA independently assessed each study, using MMAT criteria to evaluate the clarity of research questions, appropriateness of methodology and study design, sampling and data collection methods, and the integrity of interpretations made based on the data. Any disagreement was resolved through discussion and consultation with researcher KK.

Due to the heterogeneity of the included studies, meta-analysis was not suitable. We adopted narrative synthesis instead, as this method allowed for descriptive analysis of results from qualitative, quantitative, and mixed-method studies, using textual explanations. We adhered to the Enhancing Transparency in Reporting the Synthesis of Qualitative Research (ENTREQ) statement for qualitative synthesis and the Synthesis Without Meta-analysis (SWiM) guidelines for quantitative synthesis (37, 38). We synthesised qualitative and quantitative findings separately. Meta-analysis of effect estimates was not possible due to limited reporting of significance, incompleteness of the estimates and utilisation of different effect measures across studies. In alignment with recommended methods of statistical synthesis without meta-analysis, quantitative findings regarding the same outcomes were synthesised using vote counting based on the direction of effect. Findings were included regardless of reported significance or absence thereof (39). Included full-text qualitative studies were imported into NVivo R1 (40) for thematic synthesis. Adopting an iterative process involving researchers KK and KA, researcher NL coded the themes, supporting quotations, and conclusions in these texts. Research NL then examined the coded information to identify patterns related to men’s experiences of suicide bereavement. These were inductively extracted and organised hierarchically into overarching themes (41).

The searches yielded 35 studies on 31 samples across various countries (Table 1), including 25 quantitative, eight qualitative and two mixed-methods studies published between 1995 and 2023. The data extracted from the quantitative, qualitative and mixed-methods studies are included in Supplementary Tables 1–4.

Sample sizes of the included quantitative studies (n = 25) varied, with individuals bereaved by suicide ranged from 60 (59, 60) to 15,607 (46). Two studies sampled only suicide decedents, with sizes ranging from 9011 (47) to 29,513 (48). Of the studies reporting gender distribution, most (n = 17) had fewer than half male participants (42, 47, 49, 50, 52, 54, 55, 59–66, 70, 71) with only seven having gender parity or being predominantly male (46, 48, 51, 56, 67–69). Participant relationships to the deceased varied (Table 2). Mean age ranged from 20 (67) to 54 years (46), with two studies (67, 68) including those under 18. Mean time since bereavement spanned from 1 month (59) to 15 years (77).

Sample sizes for suicide-bereaved individuals participating in included qualitative studies (n = 8) ranged from 2 (57) to 346 (43). While one study had a notably lower proportion of men (43), most had a gender-balanced sample (44, 45, 74, 75), and three included only men (57, 58, 73). Studies typically examined a single relationship type such as parents of the deceased (45, 74, 75) and domestic partners (57), with one covering both parents and spouses (73) and another including friends, domestic partners, or family members (58). Two studies omitted participants’ relationships to the deceased (43, 44). Of those that reported on age, most focused on individuals aged 40 years and older (45, 57, 74, 75), with others including those at least 30 (73) and two encompassing ages as young as 18 and 20 years (43, 58).

Two mixed-methods studies were included in this review. Time since bereavement ranged between 15 and 38 months in one study (53) and was not reported in the other study (76). Both studies (53, 76) had a sample size of N = 18, with a slightly lower proportion of men compared to women. One sample comprised friends of the deceased (76), while the other included partners, parents, siblings and offspring (53).

All included quantitative studies (n = 25) were non-experimental with half (n = 13) employing a cross-sectional design that either exclusively surveyed individuals bereaved by suicide (42, 52, 54, 59, 60, 62, 64, 77) or included individuals not bereaved by suicide (49, 63, 66, 69, 71). Eight quantitative studies used a cohort study design (46, 50, 51, 55, 65, 68, 70, 72), with an additional two studies using a nested case–control design (47, 48). Two quantitative studies used a case–control design (56, 67).

All included qualitative studies (n = 8) undertook thematic analysis of qualitative data. Four studies collected this data using semi-structured interviews (45, 73–75), two with photovoice techniques (57, 58), and one as part of a realist evaluation (44). An additional study analysed free-text survey responses (43).

Of the two included mixed-methods studies, one used a concurrent design that collected qualitative and quantitative data in the same timeframe (76). The other used an embedded sequential design with qualitative data being collected from a subsample of participants from a large cohort study (53). Both studies undertook thematic analysis of the qualitative data and descriptive analysis of the quantitative data (53, 76).

Outcomes addressed by the included quantitative studies encompassed six topics related to suicide bereavement: suicidality of the bereaved (42, 46–48, 51, 54, 56, 64, 66, 68, 69, 71, 72), mental health and substance use (46, 49, 52, 54, 56, 59, 60, 65–67, 69, 70, 77), grief (59, 60, 62, 64, 66, 77), relationships and quality of life (42, 54, 56, 60, 63, 64, 69), physical health (46, 59, 60, 70), and employment (46, 50). These topics are documented further in Table 3.

While all included qualitative studies explored the experiences of suicide bereavement, some focused on specific aspects such as grief (57, 73, 74), coping mechanisms (43, 45, 57, 74, 75), relationships with others (45, 57, 58, 74), emotional repression and masculinities (44, 58, 73), and substance use (43, 74, 75).

Of the included mixed-methods studies, one examined general health, substance use, coping styles, prolonged grief and posttraumatic growth (76). Both studies examined severity of depression and anxiety (53, 76). Qualitative outcomes for one study explored how participants coped with bereavement (76), while the other focused on the impacts of grief, support needs and wellbeing (53).

The quality and risk of bias assessment of the included studies is summarised in Table 4. The details of the assessment are presented in Supplementary Tables 5, 6. Most qualitative studies (n = 17) were of high quality, achieving a star rating of 4–5 (46–51, 55, 56, 62, 63, 66, 68–72, 77). Those that achieved a star rating of 2–3 largely scored poorly on domains relating to representativeness and accounting for confounders in design and analysis (52, 54, 59, 60, 67). The qualitative studies were also of high quality with 63% achieving a 5-star rating (44, 45, 58, 74, 75). All studies with less than five stars did not fulfil the criterion relating to adequacy of data collection methods (43, 57, 73). This was largely due to the potential for selection bias being unaddressed. Regarding the mixed-methods studies, their qualitative elements of both were of high quality (53, 76). There was risk of selection bias in the quantitative elements of one study (76) and both did not adequately investigate divergences in the qualitative and quantitative components (53, 76).

Suicide-bereaved men are at greater risk of suicide mortality than non-bereaved men (46–48, 51, 72). There is also emerging evidence that they are more at risk of suicide than men bereaved by other means (48). Current impacts of suicide bereavement on suicide mortality risk appear to be similar between men and women (46, 48, 51, 72). However, suicide-bereaved men still died by suicide at higher rates than suicide-bereaved women, reflecting the overall higher suicide risk for men (78). The findings of this review are consistent with other research that suggests that spousal bereavement by any means, is linked to an increased risk of suicide for men that is comparable to that of women (79). Findings from smaller observational studies on gender differences in suicide ideation among suicide-bereaved individuals were mixed. Some reflected the “gender paradox” (80) of men experiencing lower rates of suicide ideation despite having higher rates of suicide mortality (54, 71), while others described no gender differences in risk of suicide ideation among suicide-bereaved individuals (42, 56, 69). However these studies used small samples from a single community (56) or occupation (69), or that were self-selecting (42, 54), which may not reflect patterns outside of these samples.

The qualitative findings, while also susceptible to selection bias, had very few instances of suicide-bereaved men discussing their own suicidality. This was possibly unexpected given the higher risks of suicidality for suicide-bereaved men reflected in quantitative findings. Some men who had lost other males to suicide described how this experience, and in particular seeing the impacts of the loss on other family and friends of the deceased, emboldened them to seek help for their own mental health challenges (58). The only finding to mention suicidality in bereaved men was Ferlatte et al., who described suicide ideation experienced by gay man whose partner had died by suicide during a bereavement process complicated by stigma related to his sexuality and his partner’s HIV positive status. This reflects other findings on the impacts of social marginalisation on bereavement (81) as well as the heightened risk of suicidality for the LGBTIQ+ community (82).

The quantitative studies included in this review suggest that suicide-bereaved men exhibit lower levels of emotional distress and prolonged grief compared to women, but low response rates may limit these findings’ broader applicability (52, 66). Furthermore, their interpretation must consider the timing of data collection. Examining experiences of prolonged grief among spoused bereaved by a death by any means, Lundorff and colleagues indicated that similar proportion of men and women experience prolonged grief, but with markedly different grief trajectories (22). Men generally experienced a greater intensity of prolonged grief symptoms immediately following bereavement while women showed an increase over time. Thus, the results of this review could be influenced by the predominance of studies that gathered data from participants sometime after bereavement, thereby omitting the initial post-loss phase, a period when men’s grief difficulties might have been most pronounced.

Gender differences in coping strategies for suicide bereavement identified by this review, such as men’s lower utilisation of emotional regulation strategies (60, 76) reflect patterns described by other literature (23). However, such literature also highlights how men may not recognise or report their coping behaviours as forms of emotional regulation. Indeed, qualitative findings describe how suicide-bereaved men regulate or even suppress their emotional responses, influenced by cultural norms of grief expression (73), or a desire to shield others from their own mourning (58). Such avoidance or seeking of distraction from grief emotions may be seen as a maladaptive reaction at first, but can also be part of the natural oscillations of more adaptive coping processes depicted by Stroebe and Schut’s Dual Process Model of coping with bereavement (12). Longitudinal studies with suicide-bereaved men are needed to examine their coping styles and grief outcomes with consideration to the way that these experiences may shift over time.

While studies often describe alcohol use as a common coping strategy for bereaved men (25), findings from this review were more mixed. Some studies reported that suicide-bereaved men utilised typical masculine coping strategies involving alcohol, drugs and gambling more commonly than suicide-bereaved women (42), while others reported that suicide-bereaved women had higher rate ratios of alcohol-related disorders than men (46). Qualitative studies noted that some men used alcohol to cope with the death of a child by suicide (74) but did not expressly note that this was different to women and may be a reflection of the authors’ awareness of men’s alcohol use to cope being a common theme in bereavement-related literature. Other studies found no differences in alcohol use by gender (43, 53, 76).

Further research is needed to examine other aspects of suicide-bereaved men’s grief experiences such as continuing bonds (57, 74, 75) and posttraumatic growth (57, 76) as this review did not yield sufficient and consistent findings to draw any clear conclusions.

While only a single study compared mental health outcomes for suicide-bereaved and non-suicide-bereaved men, it found that men bereaved by spousal suicide were at a greater risk of common mental health conditions such as depression, anxiety, and PTSD than non-suicide-bereaved men (46). They were also at greater risk of drug and alcohol use disorders and were more likely to use mental health therapies and services, including psychiatric hospitalisation (46, 70). Studies that only compared risks of depression, anxiety and PTSD in suicide-bereaved men and women largely found the men were at comparatively lower risk of these outcomes (49, 65, 66, 77, 83). This may be driven by gender differences in the risks of these conditions in the broader population where women generally have higher rates of depression, anxiety and PTSD than men (84–86).

Conversely, gender comparisons of rate ratios in Erlangsen et al. found that the rate ratio for PTSD was higher in suicide-bereaved men compared to non-bereaved men, indicating a stronger association between suicide bereavement and PTSD in men than in women, who showed a lower rate ratio when comparing bereaved to non-bereaved individuals. This disparity may also be attributed to typically lower baseline rates of mental health conditions among men, suggesting that the impact of suicide bereavement is more pronounced against a backdrop of lower overall mental health issues among men. However research on spousal bereavement, not specific to suicide, also suggests that bereaved men experience poorer mental health outcomes compared to bereaved women (87).

Across a range of quality of life and social outcomes, including life satisfaction and family functioning, suicide bereavement appeared to have a negative effect on these aspects of men’s lives to a lesser extent than they did women’s (42, 46, 74). In other outcomes, such as religious participation, suicide bereavement did not appear to have a differential effect by gender (63). Loneliness is often experienced by those grieving a suicide loss (88), and for men, particularly those at greater risk of social isolation such as gay men, it can be more pronounced (57, 58). Some suicide-bereaved men can find purpose in their social networks by focusing on their role as protector of their family and supporting others in their grief (54, 58). However, others used this coping style to avoid addressing their own grief and emotional expression. While these men did not describe this suppression as harmful, it aligns with typical masculine experiences of subduing the emotions of grief both to avoid the stigma of inappropriate emotions and to comfort others (89).

Men bereaved by the suicide of a spouse or partner, or sibling, were at greater risk of death by any cause than non-bereaved men (46, 51). Men bereaved by suicide also may be at higher risk of hospitalisation for physical health conditions, however these findings were mixed based on how often they were hospitalised and the conditions for which they were hospitalised (46, 70). Findings were also mixed regarding the effect of gender on the association between suicide bereavement and a range of physical health conditions. While population-level studies observed similar patterns for suicide-bereaved men and women compared to non-suicide bereaved men and women (46), other studies found men to experience fewer adverse physical health effects following suicide bereavement (42). Research on bereavement by any means suggests that this experience can be associated with adverse changes to physical health (90) and the findings of this review generally reflect this.

This review is the first to explore the impacts of suicide bereavement on men. It includes a broad range of outcomes, including suicide mortality and suicidal behaviours, mental health conditions and service utilisation, relationships with others and quality of life, experiences in the workplace, and physical health. The review systematically investigates and synthesises results specific to men across suicide bereavement literature. While studies have used a variety of measures and methods, this review investigates and reports both their agreement and heterogeneity while acknowledging the extent to which its conclusions can be considered reliable. It provides a comparative analysis by contrasting findings from suicide-bereaved men with non-bereaved men and suicide-bereaved women. The included qualitative findings offer in-depth understanding of the emotional and psychological landscapes of suicide-bereaved men, as well as the ways that these aspects may differ from statistical representations. The review also takes an intersectional approach, exploring the different ways that culture and sexuality can impact men’s experiences of suicide bereavement. This review adheres to SWiM guidelines for reporting and discussing its findings (38). It describes and justifies its grouping of studies by comparator, uses quantitative synthesis methods appropriate to heterogenous effect estimates (such as vote counting), investigates heterogeneity of reported effects, uses appropriate data presentation methods and reports the limitations of the synthesis methods and included studies.

The review was limited to peer-reviewed studies, identified by a systematic search of the literature. Nonetheless, it is possible that some relevant studies may not have been captured. The review only included findings related to men bereaved by the suicide of someone close to them, and did not include other relationships, such as clinicians bereaved by a patient’s suicide, or that of first responders. Most studies were based in Europe and other Western countries, with only a few studies from other countries. As such, the findings may not be generalisable to men in non-Western countries, though some are specific to men from East Asian nations. Similarly, while this review suggests that some aspects of suicide bereavement to be specific for gay men, the experiences of other men from the LGBTIQA+ community were not elucidated. Identification of the effects of bereavement by suicide compared to bereavement by other means is limited as most studies compared the experiences of suicide-bereaved individuals to non-bereaved individuals. Similarly, findings regarding differences in the experiences of suicide bereavement for men compared to women may be confounded by baseline gender differences in these outcomes. While this study includes some population-level findings that include all suicide-bereaved men in a certain country and time period, many of the smaller studies had markedly lower proportions of male participants and may be subject to selection bias that often arises from the self-selection of study participants (91, 92). Finally, as meta-analysis was not feasible in this review, qualitative synthesis relied on vote counting methods to summarise the direction of effects and associations, which may oversimplify findings and fail to account for the magnitude or statistical significance of the reported outcomes.