Suicide is a leading cause of death worldwide, with an estimated 700 000 people dying by suicide annually (World Health Organisation, 2019). It is a highly sensitive and complex issue – and one that is disproportionately affecting men. Today on #WorldSuicidePreventionDay we summarise a recent review that has clinical implications and recommendations for future research.

Suicide attempts are more likely to occur among women, while suicide deaths are more common among men. This ‘gender paradox in suicide’ has been found consistently over time and throughout most Western countries (Naghavi, 2019; Canetto, 1998). In the UK specifically, men account for around 75% of all suicides (Samaritans, 2021).

Some evidence suggests that the gender differences in suicidal behaviour may be due to different methods used by men and women (men tend to use more lethal means), stigma (men might be less likely to report suicide attempts), as well as differences in social, emotional, and life experiences (Freeman et al., 2017; Hawton, 2000; Canetto 1998; Chandler 2019). However, there are still major gaps in our understanding of why this difference exists.

To prevent suicide deaths and better understand this gender paradox, it is crucial to know what the specific risks of suicidal behaviour are for men. This timely systematic review by Richardson and colleagues (2021) aimed to determine the nature and extent of risk factors to predict suicidal behaviour in men over time.

Methods

The authors conducted a literature search (in March 2019, repeated in 2020) on six databases: CINAHL, PsycINFO, Web of Science Core Collection, PubMed, Embase, and Psychology and Behavioural Sciences Collection, with no date restriction. Their broad search strategy included the appropriate terms “Men” or “Male” and “Suicid*” and “risk*”.

Inclusion criteria:

• Primary level research employing a retrospective or prospective research design;
• Participants aged 18 years of age;
• Participants who experienced suicidal behaviour (suicide attempts, or death by suicide);
• Male and female results reported separately or male-only results;
• Included association between gender and risk factors, not solely stating male gender as a risk factor;
• English language.

They recorded their search strategy and decisions in a Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) flow chart and extracted data from each included study using a uniform data extraction form. Information extracted included: methodology, results, and authors interpretation of the data.

Each included study was assessed for Risk of Bias (RoB) using a modified 9-item index based on a quality assessment tool used by O’Connor et al. (2016). The score ranges from 0-13, with a lower score representing a higher probability of methodological bias. One researcher assessed the bias, with 10% cross-checked by a second researcher (concordance was 100%).

Lastly, the authors conducted a narrative synthesis of findings, grouped by study design (prospective or retrospective). Emergent risk factors were clustered into seven subheadings: sociodemographic characteristics, physical health/illness, mental health problems/psychiatric illness, psychological factors, negative life events/trauma, characteristics of suicidal behaviour, and other factors.

Results

Searching the databases yielded 26,307 records. After duplicate removal and screening, 105 studies were included in the review (62 prospective studies and 43 retrospective studies). The mean RoB score was 7.19 (+/- 1.19) for the prospective studies and 7.67 (+/-1.64) for the retrospective studies.

Overall, the studies yielded 68 different risk factors for suicidal behaviour (attempts and death) in men. Most studies were conducted in high-income countries, such as Sweden, USA, UK, and Denmark. Participants were from a variety of settings (general population and psychiatric inpatients). In both the prospective and retrospective studies, the most consistent evidence was for socio-demographic factors, mental health/psychiatric illness, physical health/illness, and negative life events/trauma.

Below are some of the key findings (risk factor reported by ≥2 studies)

	Prospective studies (number of studies)	Retrospective studies (number of studies)
Sociodemographic characteristics	– Marital status: unmarried, divorced, widowed, separated, or single relative to those who are married (8) – Same-sex married relationship (2) – Low levels of education (8) – Low household income as a child (2) – Social material deprivation (3) – Living alone (4) – Short stature among men (2)	– Marital status: unmarried, single, divorced, or widowed (14) – Low level of education (5) – Unemployment (5)
Physical health and illness	– Being underweight (6) – Obesity in men aged 40-69, and in a general population cohort of men and women (2) – Smoking among a range of populations (5) – Cancer diagnosis (year after as a significant risk period) (2) and poor cancer prognosis (1) – Diabetes (2)	– Cancer diagnosis (5) – Physical health problems (3) – Current smoking (2)
Mental health problems/psychiatric illness	– Alcohol and/or drug use/dependence (15) – Depression (12) – Any diagnosis of psychiatric disorder (9) -Diagnosis of a personality disorder (5) – Anxiety (5) – Schizophrenia (5) – Bipolar disorder (4) – Neurotic disorder (2)	– Alcohol and/or drug use/dependence (9) – Depression (7) – Psychiatric diagnosis (3) – Mental health comorbidities (2)
Psychological factors – personality and individual differences	– Low IQ (5) – Poor emotional control (3)	– Impulsive aggression and non-impulsive aggression (2)
Negative life events/trauma	– Adverse childhood experiences (5) – Bereavement (2) – Involvement in criminal activity (2)	– Experiencing a recent crisis (3) – Bereavement (2)
Characteristics of suicidal behaviour	– History of previous suicide attempts (6)	– Disclosing intent to harm self (3) – Previous suicide attempt or previous self-harm (2)

Conclusions

68 risk factors were identified for suicidal behaviour in men, over time. Across both prospective and retrospective studies, the most consistent evidence was for socio-demographic factors, mental health/psychiatric illness, physical health/illness, and negative life events/trauma.

Key risk factors:

• Alcohol and/or drug/use dependence;
• Being unmarried, divorced, widowed, or single;
• Diagnosis of depression;
• Low level of education.

However, the authors note:

The paucity of psychological research may be a by-product of the types of studies included in this review. Most studies were large epidemiological designs, and as such, they do not tend to routinely assess psychological factors and characteristics of suicidal behaviour. This major weakness needs to be addressed urgently.

Strengths and limitations

This is the first systematic review of risk factors for suicidal behaviour in men, over time.

A major strength of this review is that the authors pre-registered the protocol on PROSPERO (international prospective register of systematic reviews) and followed PRISMA guidance, alongside best practice methods for research synthesis (Johnson and Hennessy, 2019). The authors addressed a clearly focussed review question and used a robust search strategy across six different databases. Each included study was quality assessed using a modified and validated RoB tool.

The authors provided a comprehensive synthesis of results, categorised by prospective and retrospective studies, with detailed supplementary material available. The exclusion of cross-sectional studies and limit to prospective and retrospective studies is highly appropriate: to identify something as a risk factor it must have occurred before the outcome, something that is only possible to ascertain in studies where you can establish temporality.

However, there are some limitations.

Although the authors followed PRISMA guidelines, in the supplementary material they use the 2009 PRISMA checklist. There is an updated version available (this was published as a preprint in 2020 and published in March 2021). Updated PRISMA guidelines state ‘Specify how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process’. It is not completely clear from the paper or supplementary material how the initial screening was conducted, or if part of the screening process was cross-checked by a second reviewer. Given the large number of records, second screening might not have been feasible or pragmatic, but it would add extra clarity to acknowledge this in the methods. It is not completely clear if forward or backward citation searching was completed, which is often a useful step when conducting a systematic review to ensure no relevant papers are missed. However, given the authors are experts in the field, this may not have been necessary.

The authors also note that the studies used many measures and definitions of risk factors, which made it difficult to compare studies. Although not a limitation unique to this review, this is something in the field of suicide and self-harm research that can make it difficult to synthesise results. However, in the supplementary material, the authors go to great effort in their data extraction form to record how risk factors were measured, and the instruments used, to make this information easily accessible.

The review was limited to English language papers, and it’s likely that non-English language research exists in this field, so including that may have altered the findings of the review.

Lastly, although outside the scope of this review, I also wonder what the ‘protective’ factors are. I think it is useful to discuss both risk and protective factors in context with each other.

Implications for practice

Suicide in men is a complicated issue. This systematic review is a timely step in helping to better understand the risk factors for suicide in men, over time.

Given the complexity of suicide, it is unsurprising that a wealth of risk factors (68) have been identified in this systematic review. Many of the risk factors identified are not static (e.g., sociodemographic, life events). As the authors note, “these risk factors can change in relevance through an individual’s life”. This makes it apparent that context is so important when identifying who is at risk for suicide.

The limit of focussing on prospective/retrospective epidemiological studies was highly appropriate. However, most studies were large epidemiological designs, and studies that look at psychological risk factors were limited, which is a “major weakness that needs to be addressed”. The authors ask the important question: “Is it time to explore other risk factors?”. I agree with this question. Research into novel risk factors, and particularly psychological ones, is an important next step for research into suicidal behaviour in men. Further qualitative research would be helpful in ascertaining what psychological factors are important to focus on.

Similarly, given that relationship status was a prominent risk factor for suicidal behaviour and marriage is a known protective factor for men: when looking at gender differences, it is key to better understand how social support might differ across gender. Lastly, it is also important to look at protective factors.

To me, the key messages are:

• there are a wealth of risk factors associated with suicidal behaviour in men;
• there is still difficulty in characterising who it at risk for suicide but exploring psychological risk factors might be key;
• it is still not certain ‘why’ this gender paradox in suicide exists and;
• context matters for both suicide and other mental health research.

If you need help and support now and you live in the UK or the Republic of Ireland, please call the Samaritans on 116 123.

Statement of interests

No conflicts of interest.

Links

Primary paper

Richardson, C., Robb, K. A., & O’Connor, R. C. (2021). A systematic review of suicidal behaviour in men: A narrative synthesis of risk factors. Social Science & Medicine, 113831. (Link)

Other references

Canetto, S. S., & Sakinofsky, I. (1998). The gender paradox in suicide. Suicide and Life‐Threatening Behavior, 28(1), 1-23. (Link)

Chandler, A. (2019). Boys don’t cry? Critical phenomenology, self-harm and suicide. The Sociological Review, 67(6), 1350-1366. (Link)

Freeman, A., Mergl, R., Kohls, E., Székely, A., Gusmao, R., Arensman, E., … & Rummel-Kluge, C. (2017). A cross-national study on gender differences in suicide intent. BMC psychiatry, 17(1), 1-11. (Link)

Hawton, K. (2000). Sex and suicide: Gender differences in suicidal behaviour. The British Journal of Psychiatry, 177(6), 484-485. (Link)

Johnson, B. T., & Hennessy, E. A. (2019). Systematic reviews and meta-analyses in the health sciences: Best practice methods for research syntheses. Social Science & Medicine, 233, 237-251. (Link)

Naghavi, M. (2019). Global, regional, and national burden of suicide mortality 1990 to 2016: systematic analysis for the Global Burden of Disease Study 2016. bmj, 364 (Link)

O’Connor, D. B., Ferguson, E., Green, J. A., O’Carroll, R. E., & O’Connor, R. C. (2016). Cortisol levels and suicidal behavior: A meta-analysis. Psychoneuroendocrinology, 63, 370-379. (Link)

World Health Organization. Suicide worldwide in 2019. Retrieved from https://www.who.int/teams/mental-health-and-substance-use/suicide-data

Samaritans. Research Briefing Gender and Suicide. 2021. Retrieved from: https://media.samaritans.org/documents/ResearchBriefingGenderSuicide_2021_v7.pdf

Photo credits

• Photo by Tim Marshall on Unsplash
• Photo by Taylor on Unsplash