These findings make several contributions to our understanding of the patterns of cannabis use prior to non-medical use legalization among a community sample of Canadian men, specifically those living in the eastern province of NB. Notably, the overall prevalence of cannabis use in the MVGHS sample in the past year (46.7%) was almost twice that reported in the 2018 Canadian Cannabis Survey (CCS) for NB men (26.1%) and for Canadian men (26.5%) (Health Canada 2018). Some of this difference may be accounted for by age; men in the 2018 CCS were aged 16 years and older as compared to 19 to 65 years in the MVGHS. Interestingly, the 2017 U.S. National Survey on Drug Use and Health (NSDUH) rate of cannabis use in the past year for men aged 12 and older was also much lower at 17.8% (Center for Behavioral Health Statistics and Quality 2018).

Unlike cannabis research that relies on prevalence rates derived from a dichotomous measure (i.e., yes/no) of cannabis use in a specific period (e.g., past 12 months or past 3 months), we analysed data on a pattern of use categorized as daily, sometimes, or never. Association between health and social indicators and pattern of use is important because the frequency daily or almost daily is applied in epidemiological studies as a proxy for heaviest use, identifying those who may be at greater risk (World Health Organization 2016). Of the 275 men in the MVGHS who had used cannabis in the past year, 45.1% (n = 124) reported using daily, a rate almost twice that of the 27.9% of Canadian men who used cannabis daily or almost daily in the past year reported in the 2018 CCS (Health Canada 2018). Therefore, the proportion of NB men at possible risk for problematic cannabis use may be considerably higher than for Canadian men in general.

Despite violence exposure not being a criterion for taking part in this study, 97.5% of participants reported experiences of violence on the CLVS scale. This high prevalence is consistent with violence being a major public health problem for men (Haegerich and Hall 2011) and may be attributed to the CLVS scale measuring a comprehensive range of violence experiences as target and/or perpetrator across the lifespan (Scott-Storey et al. 2018). Nonetheless, applicability of findings may be limited to men living primarily in widely dispersed medium-sized cities, towns, and rural areas similar to the province of NB. Our findings add to our understanding of patterns of cannabis use and lifetime cumulative violence. In a previous analysis, we dichotomized our measure of cannabis use as daily or almost daily and not daily (i.e., never, once or twice, monthly, weekly) and found that men in the higher CLVS group were significantly more likely to use daily and less likely to use not daily than men with lower CLVS (Scott-Storey et al. 2018). Our current analysis supports our past finding by showing that men with higher CLVS were more likely to use cannabis daily as compared to men with lower CLVS (ES = medium) but expands our knowledge of the relationships among never and sometimes use and lifetime cumulative violence. Specifically, men in the higher CLVS group were significantly less likely to use never but neither more or less likely to use sometimes than men in the lower CLVS group, suggesting that lifetime cumulative violence may not be a factor in occasional cannabis use, but may be for daily use. Although our analysis has expanded our understanding of these relationships, further study of cannabis use patterns and other sociodemographic variables in CLVS groups is needed.

In both the total sample and the higher CLVS group, significantly more men than expected ages 19 to 24 years had used cannabis sometimes in the past year and the number of daily users did not differ from expected (ES = medium). This may signify that more younger men, particularly those with histories of higher CLVS, are choosing to use cannabis occasionally rather than daily, a consumption pattern consistent with guidelines recommended by Fischer et al. (2017) to prevent the health and social consequences of intensive use. In addition, we found that significantly more men than expected in the total sample and the higher CLVS group who were single, divorced or separated used cannabis both sometimes and daily whereas significantly fewer married or cohabiting men used sometimes or daily (ES = large). Although lower prevalence of cannabis use among married men has been reported elsewhere (Han et al. 2017; Merline et al. 2004), dichotomous measurement of cannabis use and non-use in these studies does not permit differentiation between daily and sometimes use and their unique associations with marital status. Our findings highlight the importance of using non-dichotomous measures of cannabis in analysis to enable exploration of meaningful patterns of usage in the context of lifetime cumulative violence over time for the purpose of identifying who may be at lesser or greater risk for cannabis-related problems.

In the MVGHS total sample, daily cannabis use was associated with having a high school education or less, being unemployed, having an annual income of less than $25,000 CAD and having difficulty living on one’s income, with medium ES. With respect to education and difficulty living on income, this pattern was also evident in both CLVS groups, suggesting that irrespective of lifetime cumulative violence exposure, these factors are associated with greater daily cannabis use. In contrast, the significant and medium relationship between daily cannabis use and unemployment in the total sample held for the higher CLVS group but not for the lower CLVS group, suggesting that severity of lifetime cumulative violence influences this association.

Our socio-demographic findings are in line with data from the 2018 CCS which showed that men with the highest prevalence of cannabis use in the previous 12 months were those who had a high school diploma (34.9%) or less (29.0%), were unemployed (34.4%), and had an income under 10,000 CAD (35.2%) or from 10,000 to 24,999 CAD (37.5%) (Health Canada 2018). Because overall unemployment rates in NB were 8.1% in 2017, higher than the national Canadian rate of 6.5% (NBjobs 2018), NB men may be at greater risk for daily cannabis use than their national counterparts. Other research suggests that unemployment in early adulthood is associated with concurrent and ongoing cannabis use through mid-life, and that early adulthood use of cannabis is associated with decreased subsequent participation in the workforce (Hara et al. 2013). Thus, the 13.5% unemployment rate for NB youth aged 15 to 24 years (NBjobs 2018) could have implications not only for the prevalence of daily cannabis use but also for the participation of NB men in the workforce into middle adulthood. A preventive strategy for reducing daily cannabis use among NB men might be strengthening employment opportunities for youth. Overall, our findings lend support to health promotion strategies targeted to socio-demographic factors associated with daily cannabis use.

Our efforts to address the knowledge gap regarding how gender, that is ideas about what it means to be a man, influences patterns of cannabis use are important because, in Canada, men have been found to be significantly more likely than women to both try cannabis and to exhibit patterns of cannabis use deemed problematic (Bonner et al. 2017). Although total scores of multidimensional gender measures reflecting greater adherence to dominant masculine norms have, for the most part, been associated with detrimental health outcomes such as substance use, associations between unidimensional subscale scores and health outcomes are often mixed, with some found to be protective (Gerdes and Levant 2018). Such findings highlight the complex relationships between gender and health outcomes that may be culturally, situationally or contextually dependent (Gerdes and Levant 2018). Our analysis of relationships between CMNI-46 subscales and patterns of cannabis use among NB men ages 19 to 65 years was additionally nuanced because our measure of cannabis use was not dichotomous (i.e., yes/no). Using a dichotomous cannabis use measure, Liu and Iwamoto (2007) in a sample of Asian-American university students, found cannabis use was associated with the norm of Playboy, a result consistent with our finding for the total sample that daily and sometimes users were more likely than never users to endorse having sexual activity with casual partners as a masculine norm. However, our findings for the norm of Heterosexual Self-presentation for the total sample suggests a more complex relationship than their finding of cannabis use being associated with this norm (Liu and Iwamoto 2007). We found that sometimes users had fewer concerns about how others perceive their sexual orientation than never or daily users, suggesting that never and daily users, despite their distinctly different usage patterns, adhere more strongly to the dominant masculine norm of Heterosexual Self-presentation than occasional users. The cultural dissimilarity of the two samples may account for this difference and also for our finding that daily use was associated with the norm of Self-reliance in the total sample, an association not found in the Asian American sample. Overall, although the relationships for each gender indicator and patterns of cannabis use were significant, because the ES was small our findings for the total sample must be interpreted with caution.

Our findings further suggest that lifetime cumulative violence may be a contextual factor that influences how particular masculine gender norms relate to patterns of cannabis use. In the higher CLVS group, significant relationships were found between the pattern of cannabis use and the same gender norms as in the total sample. Although the patterns of use for Heterosexual Self-presentation were the same in the higher CLVS group as those for the total sample, they differed for Self-reliance and Playboy. Both daily and sometimes users had significantly higher scores on self-reliance than never users, suggesting that men with higher CLVS scores may value the belief that men should solve their own problems and avoid seeking help. It may be that men with higher CLVS had historically been unable to count on others. With respect to Playboy, only sometimes users had significantly higher endorsement scores compared to never users, an outcome that is difficult to interpret. For the lower CLVS group, the finding that never users had significantly higher scores on the norm of having control over women than sometimes users is also puzzling. These discoveries, particularly those for the higher CLVS group, highlight the need for further research examining the contextual intersections between gender, CLVS and patterns of cannabis use.

One limitation in the MVGHS is our failure to collect data about whether men’s intent in using cannabis was for recreational and/or therapeutic purposes. Prior to legalization of cannabis for non-medical use, Canadians engaged in therapeutic use either as authorized users of approved medical cannabis or as unauthorized therapeutic users of recreational sources; however, few differences were found between these groups with respect to medical conditions and patterns of cannabis use (Walsh et al. 2013). In interpreting our finding of patterns of cannabis use by health problem, we recognize that cannabis use may not be intentionally therapeutic. Nonetheless, our finding for the full sample that men who used cannabis daily had significantly more chronic illnesses (ES = small) than men who used sometimes or never suggests it is possible that they may be using for therapeutic purposes. Further, the fact that no significant differences were found by pattern of cannabis use and in either the lower or higher CLVS group, suggests that lifetime cumulative violence severity may not influence usage related to number of chronic health problems.

Chronic pain has been found to be one of the main reasons for therapeutic use of cannabis with regular users of all ages attempting to manage their chronic pain with cannabis (Walsh et al. 2013; Fales et al. 2019). In the total sample, 16.8% (n = 99) of men were found to have high disability chronic pain, a rate similar to the 18.9% overall rate for Canadians over 18 years of age who were assessed with persistent, repeated, recent, intense chronic pain (Schopflocher et al. 2011). Our finding (ES = small) that significantly more of the MVGHS men with high disability chronic pain than expected used cannabis used daily (36.4%, n = 36) and neither more nor less than expected used sometimes (22.2%, n = 22), suggests that more than a third might be using cannabis therapeutically. Given the established relationship between violence and chronic pain (Hart-Johnson and Green 2012), we cannot explain why there were no significant differences in patterns of cannabis use by pain disability in the higher CLVS group whereas in the lower CLVS group significantly more men than expected with higher disability pain used cannabis daily (ES = medium). Notably, 72.7% (n = 72) of the men with high disability chronic pain were in the higher CLVS group.

The recent legalization of cannabis for non-medical use in Canada has kindled increased interest in understanding patterns of cannabis use and mental health. Our finding that men in the total sample with possible depression were significantly more likely to use cannabis daily and less likely to never use than men without possible depression (ES = medium) is consistent with outcomes of many cross-sectional studies (Volkow et al. 2014). Crucially, the credibility of these findings is limited by failure to consider other factors that when controlled often eliminate the significant relationship between cannabis use and depression (WHO 2016; Volkow et al. 2014). For example, in a 3-year Swedish study of men and women ages 20 to 64 years, cannabis use no longer predicted possible depression when gender and serious family tensions were controlled, suggesting that these social determinants, not cannabis, might account for the depressive symptoms (Danielsson et al. 2016). Our finding that significantly more men than expected in both the lower and higher CLVS groups (both with medium ES) who had possible depression used cannabis daily must be interpreted cautiously. Although it implies that lifetime cumulative violence severity may not influence the cannabis-depression association, further study of this relationship is needed because the frequency of depression in the higher CLVS group of 50.2% (n = 148) of which 39.2% (n = 58) used cannabis daily is much higher than in the lower CLVS group at 21.1% (n = 62) of which 24.2% (n = 15) used daily.

In our total sample, significantly more men than expected who met the criteria for possible PTSD used cannabis daily and fewer used cannabis never (ES = medium). Of the 210 men with possible PTSD, 61.4% (n = 129) reported using cannabis in the past year, 70 daily and 59 sometimes, suggesting that some men may use cannabis to cope with symptoms of PTSD, a strategy found to be potentially useful in a recent Canadian systematic review of medical cannabis use and mental health (Walsh et al. 2017). Similar to depression, fewer men in the lower CLVS group (15.6%; n = 46) had possible PTSD than those in the higher CLVS group (51.7%, n = 152), a finding consistent with violence being an important factor in the development of PTSD (American Psychiatric Association 2013). Although our findings were significant for possible PTSD in the in the higher CLVS group (ES = medium), cell counts were only different than expected for never users. Although counts for daily (n = 56) and sometimes (n = 45) use did not differ from expected, it is notable that 66.4% (n = 101) of men with higher CLVS and possible PTSD had been using cannabis in the past year. Similar to PTSD, possible moderate to severe anxiety in the total sample was significantly associated with more men than expected using cannabis daily, and fewer using never (ES medium), a finding that is not surprising given that relief of anxiety is one of the most widely reported reasons for therapeutic cannabis use (Walsh et al. 2017). Despite such reports, a recent review indicated that robust evidence supporting the therapeutic effects of cannabis on anxiety is lacking and that longitudinal prospective studies and clinical trials are needed to determine its usefulness (Lowe et al. 2019). It is difficult to interpret our findings that a) in the lower CLVS group significantly more men than expected with anxiety used cannabis daily (ES medium), and b) in the higher CLVS group, despite a small ES and a statistically significant association between anxiety and patterns of cannabis use, no significant differences in cell counts were found. Further study is required to unravel the relationships among lifetime cumulative violence, patterns of cannabis use and possible anxiety.

Our findings of significant relationships for use of other substances and patterns of cannabis use in the total sample and both CLVS groups contribute to the developing knowledge base. Significantly more daily users (52.4%) than expected in the total sample were also current cigarette smokers, a prevalence rate similar to the 55.5% of men in a US NSDUH survey (Pacek et al. 2018). Additionally, 30.6% of daily users in the MVGHS had quit smoking, a finding that may be explained by a general decline in NB in cigarette smoking from 1998 to 2017 (Reid et al. 2019). Similar patterns of cigarette use were found in the lower and higher CLVS groups, although a smaller proportion of daily cannabis users were current smokers in the lower CLVS group as compared to the higher CLVS group. Cigarette smoking has been found to be a coping mechanism more likely to be employed by those experiencing greater violence severity, particularly intimate partner violence (Crane et al. 2013).

Our findings related to cannabis and alcohol use suggest the need for further research in this area. Among men who had used cannabis in the past year, 65.1% (n = 179) met the criteria for possible hazardous drinking, a rate slightly higher than that of 57.3% of cannabis users found in the 2017 Canadian Tobacco, Alcohol and Drugs Survey (CTADS) to be heavy drinkers (Statistics Canada 2018). In the total sample and both CLVS groups, we found a significant association between patterns of cannabis use and possible hazardous drinking (ES = medium), specifically that more sometimes users and fewer never users than expected were possible hazardous drinkers; the number of daily users was not significantly different than expected. Comparable significant associations and ES were found for binge drinking. This finding is important because it suggests that NB men who use cannabis occasionally may be most at risk for being possible hazardous drinkers and/or binge drinkers as compared to those who use daily or never. Similarly, in a large study of college students in France, the adjusted odds of being a frequent binge drinker were increased almost 5-fold for being male and 13-fold for using cannabis occasionally (Tavolacci et al. 2016). Another important implication of our findings is that lifetime cumulative violence does not appear to affect the relationship between patterns of cannabis use and either hazardous or binge drinking. A limitation of these study findings is that we do not know whether men in the sample used cannabis and alcohol simultaneously (i.e., together at the same time to complement one another) or concurrently (i.e., one at a time as a substitute for the other). Based on United States National Alcohol Survey findings, the prevalence of simultaneous alcohol and cannabis use was almost twice that of concurrent use and associated with increased frequency and quantity of alcohol consumption (Subbaraman and Kerr 2015). A future inquiry examining simultaneous versus concurrent use might yield critical information for identifying men at risk.

Of men in the total sample who used cannabis in the past year, 86 (31.3%) had used street drugs, a rate somewhat higher than the 23.8% of male cannabis users over age 15 in the 2017 CTADS who reported using illegal drugs (Statistics Canada 2018). It is unclear how illegal drugs are defined by CTADS and it is possible that prescription drugs used for non-medical purposes are included in this category. In our sample, 69 (25.1%) of cannabis users also reported using prescription drugs for non-prescription purposes. Significantly more daily cannabis users than expected were found to use street drugs (ES = large) and also to use prescription drugs for non-medical purposes (ES = medium) in the total sample and both CLVS groups. Together these findings suggest consistency of association between patterns of cannabis use and that of other substances that is not affected by lifetime cumulative violence severity in this sample of men.