Abstract
Men are an established high-risk group for suicide, and one of the strongest risk factors for suicide is depression imbued with complex connections to men’s mental health help-seeking. 1 Men with depression are known to have significant challenges with service engagement 2 and previous research has found that men experience a range of barriers to accessing mental health care. 3 Men who are seemingly in treatment have also been reported to experience suboptimal outcomes, evidenced by fragmented pathways in and out of care, including being lost to follow-up. 2 Understanding the factors that differentiate primary mental health help-seeking (both in terms of accessing care and treatment compliance) in men who experience moderate-severe depression is essential to ongoing efforts at improving detection rates and treatment engagement.
Over the past decade, population-based initiatives for improving men’s access to mental health care have focused on addressing attitudinal and structural barriers. Government-funded initiatives aiming to reduce the financial costs associated with accessing mental health care appear to have increased men’s uptake with services. 4 Additionally, population health campaigns have sought to address men’s attitudinal factors in relation to norming the uptake of mental health care, particularly for depression. 5 While the empirical outcomes and attribution of these campaigns are difficult to gauge, such transformative efforts affirming men’s help-seeking as strength-based are espoused as precursors to men’s behavior change. 6 At a population level, stoicism, and idealized masculine traits–reflecting the belief that men must solve their problems independently–are significant attitudinal predictors of men’s reticence for help-seeking and heightened suicide risk. 7 In essence, men are typically socialized in a manner that dissuades them from acknowledging or displaying vulnerability.1,3,8 Hence, men may conceal emotions for fear of being perceived as–or perceiving themselves to be–weak or nonmasculine. 9
Men’s reluctance to access mental healthcare has also been attributed to structural issues whereby services and interventions are insufficiently sensitive to masculine ideals. 10 Indeed, there is growing interest in how men’s interactions within existing primary care services settings can facilitate adaptive help-seeking.11,12 Men’s low rates of access to mental health care have been attributed to services and interventions not being sufficiently sensitive to masculine identity-related factors, 13 and recent work highlights a lack of attention to men’s health and gender-related constructs (eg, masculinity) among medical and allied health training programs. 2 Primary care consultations, in particular, provide an important setting for the identification of depression and suicide risk among men. 14 However, research has shown that while men may endorse major mental health symptoms (eg, current thoughts of suicide) on a self-report questionnaire prior to a physician consultation, these same symptoms are unlikely to be disclosed as part of the consultation itself. 15
The aim of the present study was to determine the relative predictive contributions of a range of attitudinal and structural barriers that may account for mental health help-seeking among men specifically experiencing moderate severity depression symptoms. Given the importance of masculine norms in shaping men’s health-related behaviors,3,8 we hypothesized that relative to structural barriers, men’s attitudinal factors would have greater predictive effect in determining the likelihood that men with elevated depression symptoms would have a history of mental health help-seeking.
Methods
Participants
Data were collected in September 2017 from an online sample of 530 Canadian men (age range 19-88 years; mean 47.91 years, SD = 14.51), of which those endorsing symptoms of depression (moderate severity; Patient Health Questionnaire–Depression Module ≥10) were selected for the present analysis. Weighted randomization was used to select the full sample. Stratification quotas (age and region) reflected national census data. Respondents were sourced via advertisements placed on social media by a Canadian online survey provider.
Measures
Barriers to mental health services were assessed using a modified version of the Barriers to Mental Health Services Scale–Revised 16 (BMHSS-R), which assesses factors known to prevent individuals from seeking help. Due to the length of the BMHSS-R, the present study used 18 items from BMHSS-R. Items that included “I” or “my” statements were chosen to represent more direct personal statements, rather than selecting items that reflected more general attitudes. For example, selecting the item I would feel embarrassed or ashamed to see a psychotherapist to represent stigma rather than the item Normal people do not go to psychotherapy. Of the 18 BMHSS-R items, 6 were structural (eg, My insurance does not cover mental health care) and 12 attitudinal (eg, I need to solve my own problems); see Table 1.
Group Differences, Inferential Tests, and Adjusted Odds Ratios (AORs) Predicting to Help-Seeking Status.
Abbreviations: SBQ-R, Suicidal Behaviors Questionnaire–Revised; PHQ-9, Patient Health Questionnaire–Depression Module; BMHSS-R, Barriers to Mental Health Services Scale–Revised.
AOR values at step 3 (final step) of the logistic regression model.
PHQ-9 severity category data not entered into logistic model as PHQ-9 total score used; boldfaced values significant at P < .05.
Depression was assessed by the Patient Health Questionnaire–Depression Module 17 (PHQ-9). The PHQ-9 is a widely used self-report measure of the 9 DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, 5th edition) symptoms of depression (eg, Feeling down, depressed or hopeless) with ratings made over the preceding 2-week period where 0 = “not at all” to 3 = “almost every day.”
Previous mental health care engagement was assessed using a single item designed for the present study: Have you ever received counselling or psychotherapy for mental health concerns? Response options for this item were coded as either “never,” “currently,” or “previously.”
Suicide risk was assessed by the Suicidal Behaviors Questionnaire–Revised 18 (SBQ-R). The SBQ-R is a validated 4-item self-report measure of suicide risk that assesses past suicide planning, ideation, history of verbalization of suicidality, and future likelihood of attempt. The general population SBQ-R cutoff score of 7 was used to identify those at risk of suicide.
Procedure
Ethical approval was granted by the University of British Columbia and all participants consented to be involved. Respondents received reimbursement for their time via proprietary panel points, which could be exchanged for various rewards. All participants were directed to the HeadsUpGuys website (www.headsupguys.org) following provision of data, providing male-specific psychoeducation and information for depression and suicide risk. 19
Data Analysis
Descriptive statistics characterized the sample. Inferential tests evaluated group differences for those with and without previous mental health help seeking, reporting either Cohen’s d or Cramer’s V effect sizes. Hierarchical logistic regression was undertaken in order to identify factors predictive of mental health help-seeking status, coded as no = 0, yes = 1. Step 1 accounted for current depression severity (PHQ-9 total score) and suicide risk (SBQ-R ≥ 7). In step 2, age and cultural group affiliation were entered into the model, and in step 3, the 18 items from the BMHSS-R were entered. Model interpretation was guided by Nagelkerke R2 values, the proportion of cases correctly identified and adjusted odds ratios (AORs) with 95% confidence intervals (95% CIs). The relative proportions of significant predictor items comprising each barrier domain were examined to evaluate the study hypothesis.
Results
The mean age of participants was 42.36 years (SD = 13.31), most identified as heterosexual (81.4%, n = 92), and were working full-time (54.7%, n = 64), with a minority currently studying (12.8%, n = 15), unable to work due to disability (11.1%, n = 13), retired (6.8%, n = 8), or identifying with cultural groups other than Canadian (20.2%, n = 22). Of the 117 cases included in the study sample, 51.3% (n = 60) indicated previous mental health help-seeking, 8.5% (n = 10) of whom indicated that they were currently receiving counseling or psychotherapy for a mental health problem. The remaining men (48.8%; n = 57) indicated no previous help-seeking. With the exception of cultural identity (see Table 1), there were no differences on demographic variables according to mental health help-seeking status. Each of the measures reported satisfactory internal consistency in the present sample (eg, BMHSS α = .85; PHQ-9 α = .70; SBQ-R α = .67.
Group comparisons (see Table 1) indicated small effects, whereby men who endorsed previous mental health help-seeking reported higher mean PHQ-9 scores (d = 0.43). Endorsement of past 2-week suicide and self-harm ideation was relatively frequent across the sample, with only 36.8% (n = 43) reporting not at all, 34.2% (n = 40) reporting several days, 17.1% (n = 20) reporting more than half the days, and 12.0% (n = 14) reporting nearly every day, with no association between the help-seeking groups (P = .365).
Hierarchical logistic regression was undertaken in order to identify factors predictive of mental health help-seeking. At the first step, depression severity and suicide risk were entered, and while the overall model was significant, neither individual predictor was significant with 57.3% of cases predicted correctly, χ2(2) = 7.85, P = .020, Nagelkerke R2 = .087. At the second step, age and cultural group identification were added to the model, with depression severity (P = .046), suicide risk (P = .046), and age (P = .041) identified as significant predictors χ2(4) = 15.08, P = .005, Nagelkerke R2 = .161, 59.8% of cases predicted correctly. At the third step, the 18 items from the BMHSS-R were added to the model. The overall model remained significant, accounting for a substantial R2 change (Δ Nagelkerke R2 = .368) and accurately predicting close to 80% of cases, χ2(22) = 59.09, P < .001, Nagelkerke R2 = .529, 78.6% of cases predicted correctly. At step 3, the significant predictors (see Table 1 for odds ratios, 95% CIs, and P values) were participant age, depression severity, suicide risk, and 4 of the 18 BMHSS items, each of which were attitudinal barriers; nondisclosure of depression symptoms to primary care physician, needing to solve problems independently, not knowing why people go to psychotherapy, and believing that current life circumstances would result in anyone being sad or down. Supporting the study hypothesis, overall, 33% (4 of 12) of attitudinal factors were significant predictors of help seeking history, while 0% (0 of 6) of structural factors were significant.
Discussion
In the present sample of men with moderate probable depression, only 8.5% were currently connected with mental health care. This is of concern given that 63.2% of the sample reported past 2-week suicide or self-harm ideation. Moreover, a substantial proportion of men in the non-help-seeking group had SBQ-R scores that indicated potential suicide risk. As expected, results indicated that attitudinal barriers were more predictive than structural barriers in differentiating men’s mental healthcare seeking. While older age, depression severity, suicide risk and a normalized attitude of context-specific mood symptoms each increased the likelihood of men’s mental healthcare engagement, lower engagement was associated with reluctance to disclose mood-related symptoms to a physician, needing to solve one’s own problems, and uncertainty regarding why people seek psychotherapy. These 3 attitudinal barriers are consistent with previous research highlighting the role of stoicism 7 and concerns related to privacy, or perceptions of others 3 in men’s mental health. These attitudinal barriers can stem from, and further fuel men’s often reported poor mental health literacy regarding the signs and symptoms of distress and suicidality. 20 Despite ongoing investment in health promotion, societal stigma and shame surrounding male mental health continues to limit knowledge transference. 21
While it is somewhat encouraging that men with elevated suicide risk (and depression severity) were more likely to have had previous mental healthcare engagement, the high rate of current self-harm or suicide ideation suggests a substantial gap in care provision. Men’s nondisclosure of depressive symptoms and suicidality in the primary care context is of concern, 15 and clearly attitudinal issues offer important explanatory notes for some of the present findings. Enlisting the use of male-centered assessment approaches 10 and clinician training programs for working with men 22 may help men’s comfort to disclose symptoms. The approaches may also provide physicians with necessary skills to elicit and affirm men’s disclosures. By normalizing men’s emotional responses and offering a reasonable entry point to specialized services, physician “gate-keepers” can limit the impact of shame and stigma and promote the practical utility of mental health treatment. Educational materials could promote the notion that emotional difficulties are normative and expectable, that disclosure can be an act of strength, and that the use of assistance in solving problems is an advantage rather than a sign of weakness. This is a useful leveraging and transformative gendered practice for men’s positive health outcomes. 23 Social marketing approaches can also work to demystify help-seeking processes, and provide information about the legitimacy for men to seek and work with psychotherapy. An avenue for intervention may be advancing men’s depression and suicidality health literacy as a destigmatizing effort and conduit to affirming help-seeking. 24
The current study was unable to determine the temporal relationship between help-seeking and current depression and we call for more research on men’s perceived impediments to accessing mental health services, especially given the possibility of nuanced relationships between potential attitudinal and structural barriers. 25 While reluctance to disclose mood-related symptoms to one’s physician was associated with lower service engagement, seemingly related potential barriers, such as asking for a mental health referral, were not identified as significant predictors. Longitudinal replication studies utilizing longitudinal time frames and more comprehensive assessment of barriers with larger samples are needed.
In summary, the present study highlights men’s modifiable attitudinal help-seeking barriers for depression. Helping men to feel more comfortable to disclose mood-related problems to their physician is a key first step toward demystifying treatments, and reframing self-reliant attitudes, for service engagement to make available critical therapeutic alliances.
Footnotes
Acknowledgements
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding was provided by Movember Canada (grant #11R18455). Movember Canada was not involved in study design, data collection, analysis, or interpretation. SMR was supported by a Career Development Fellowship from the NHMRC.