Strategies to reduce risky alcohol use among underage girls - PDF Document

Presentation Transcript

  1. Strategies to reduce risky alcohol use among underage girls AN EVIDENCE REVIEW McCreary Centre Society

  2. Strategies to reduce risky alcohol use among underage girls: An evidence review McCreary Centre Society Prepared for British Columbia Ministry of Health

  3. Project team Youth Research Academy Annie Smith, Executive Director The Youth Research Academy (YRA) is a group of youth aged 16 to 24 with government care experience who, with the support of McCreary staff, develop, analyze, and disseminate research projects of interest to youth with government care experience and service providers. Female members of the YRA contributed to this project through their involvement on the project advisory committee, undertaking a review of media messaging, and conducting a search of McCreary literature about risk and protective factors for harmful underage alcohol use. Colleen Poon, Research Associate Maggie Woo Kinshella, Research Associate Karen Forsyth, Research Associate Stephanie Martin, Community Development Manager Katie Horton, Community Research Coordinator Cheylene Moon, Youth Research Academy Fialka Wolfblade, Youth Research Academy Jessica Hackbarth, Youth Research Academy Omyma Kafi, Youth Research Academy Sarah Kothlow, Youth Research Academy Sierra Harrison, Youth Research Academy Citation Advisory committee Smith, A., Poon, C., Woo Kinshella, M., & McCreary Centre Society (2017). Strategies to reduce risky alcohol use among underage girls: An evidence review. Vancouver, BC: McCreary Centre Society. Art Steinmann (Manager, Substance Abuse Prevention, Vancouver School Board) Betsy Mackenzie (Manager, Alcohol Harm Reduction, BC Ministry of Health) Dan Reist (Assistant Director, Knowledge Exchange, Centre for Addictions Research of BC) Gerald Thomas (Director, Alcohol, Tobacco, Cannabis & Gambling Policy & Prevention, BC Ministry of Health) Jessica Hackbarth (Youth Research Academy) Nancy Poole (Director of Research and Knowledge Translation, BC Centre for Excellence in Women’s Health) Paul Gordon (Program Director, PLEA Community Services) Sierra Harrison (Youth Research Academy) 2 Strategies to reduce risky alcohol use among underage girls: An evidence review

  4. Table of contents Executive summary ............................................................................................................................................ 4 Introduction ....................................................................................................................................................... 6 Methodology ...................................................................................................................................................... 7 Search criteria ...................................................................................................................................................... 7 Search results ....................................................................................................................................................... 8 Evaluation of key articles ...................................................................................................................................... 8 Calculation of effect sizes and Numbers Needed to Treat (NNT) ......................................................................... 9 Limitations ............................................................................................................................................................ 9 Terms used in the report .................................................................................................................................... 10 Risk and protective factors for harmful underage alcohol use among females .................................................. 11 Risk factors ........................................................................................................................................................ 11 Protective factors .............................................................................................................................................. 15 Interventions to reduce harmful underage alcohol use among females ............................................................. 18 Family-based interventions ............................................................................................................................... 18 Interventions targeting personality traits ........................................................................................................... 22 Knowledge and skills training ............................................................................................................................. 24 Culturally responsive interventions .................................................................................................................... 29 Brief Alcohol Interventions (BAIs) ...................................................................................................................... 30 Mentorship ......................................................................................................................................................... 31 Multi-component interventions ......................................................................................................................... 32 Intervention summary ........................................................................................................................................ 34 Effective policies for reducing harmful underage alcohol use among females .................................................... 36 School policies and regulations .......................................................................................................................... 37 Minimum legal drinking age laws ....................................................................................................................... 38 Zero tolerance drink driving laws and graduated licenses ................................................................................. 39 Price increases .................................................................................................................................................... 40 Restricting commercial availability ..................................................................................................................... 41 Reducing exposure to advertising/media .......................................................................................................... 42 Policy summary .................................................................................................................................................. 44 Messaging to reduce harmful alcohol use among females ................................................................................ 45 Media review methodology ............................................................................................................................... 46 Summary of findings .......................................................................................................................................... 46 Youth Research Academy’s reflections ............................................................................................................... 47 Building resiliency ............................................................................................................................................ 49 Discussion ....................................................................................................................................................... 50 Recommendations............................................................................................................................................ 51 Youth Research Academy’s recommendations .................................................................................................. 53 References ....................................................................................................................................................... 54 Appendix A: Interventions to reduce risky underage female alcohol use................................................................. 67 Appendix B: Policies to reduce risky underage female alcohol use .......................................................................... 73 Appendix C: Addressing risk factors for risky alcohol use ......................................................................................... 74 Appendix D: Media messaging sources from Youth Research Academy analysis ..................................................... 75 3 McCreary Centre Society

  5. Executive summary Risk and protective factors for alcohol use among female adolescents are often different from those for males, as are their motivations for drinking. For these reasons it is important to carefully consider which strategies are the most promising for addressing underage girls’ problematic alcohol use. An extensive search of the academic and grey literature was performed to identify policies and inter- ventions which addressed risky drinking among underage girls. Although most studies did not separate out data for underage girls, 27 rigorously evaluated interventions and nine policy studies were identified which provided gender-specific data or were targeted only at female adolescents. In addition, our search identified a number of promising programs that had adequate (non-gender specific) evidence to include them in this report. Additionally, McCreary’s Youth Research Academy reviewed published and unpublished data about risk and protective factors for harmful female alcohol use and investigated messaging aimed at reducing harmful alcohol use. Underage female youth at risk for problematic alcohol use included those who were economically advan- taged, experienced victimization and trauma, had negative body image, and were going through peri- ods of transitions. The strongest protective factors were those which built resilience and strengthened healthy relationships. Family interventions appeared to be the most consistent in reducing risky alcohol use among girls. Com- puter- and audio-based programs delivered in the home environment created opportunities for caregiv- ers and their daughters to build skills together, and may be particularly effective at reaching those who cannot or are reluctant to attend in-person sessions. Interventions that were tailored to specific cultural contexts; took into account personality traits asso- ciated with increased risk for problematic substance use; and addressed girls’ previous experience with alcohol also appeared helpful. Universal school-based interventions were generally more effective with younger girls than older ones, and with those who had not yet tried alcohol in comparison to those who were already using alcohol. Interventions with a motivational interviewing component which invited young women to self-reflect on how their alcohol use was affecting their ability to meet their goals were effective with older youth. Interventions that addressed motives for alcohol use and discussed healthier alternatives appear to be especially effective for girls identified as ‘at-risk’ for harmful alcohol use. 4 Strategies to reduce risky alcohol use among underage girls: An evidence review

  6. Policy interventions such as raising the legal drinking age and drunk-driving laws were somewhat effective for females, but appeared to be more consistently effective for males. However, other policy changes such as alcohol pricing increases and those that contribute to creating a more inclusive school environment appeared to be helpful in reducing risky underage alcohol use among girls. As studies have shown increased exposure to alcohol advertising is related to alcohol initiation and in- creased consumption and related harms, policies aimed at reducing youth’s exposure to advertising may be a promising approach. Messaging targeting risky alcohol use among underage girls was less effective if it focused on health con- sequences and harms, than if it focused on social consequences and the benefits of moderated drinking. Finally, as harmful alcohol use is often not seen in isolation from other risk behaviours and challenges (e.g., mental health problems), interventions that focus on promoting protective factors may be more effective than those which focus on eliminating or reducing a specific health risk behaviour. Additional- ly, policies and interventions targeting underage female alcohol use should take into account girls and young women’s experiences of systemic and individualized trauma and discrimination. 5 McCreary Centre Society

  7. Introduction getting into trouble at school, experiencing violence, feeling anxious or depressed, having unwanted sex, or getting drunk to the point of passing out (Evans-Whipp, Plenty, Catalano, Her- renkohl, & Toumbourou, 2013). Female-specific risks include unwanted pregnancy and inadvertent alcohol consumption during early pregnancy. In addition, alcohol is the most common substance associated with sexual violence (Atlantic Collab- orative on Injury Prevention, 2014). For example, young adult females who were heavy drinkers were found to be at greater risk of sexual intimate partner violence (Waller et al., 2012). Alcohol is the most commonly used substance in Canada (Canadian Centre on Substance Abuse [CCSA], 2014). It is estimated that alcohol-related harms cost Canadians billions of dollars a year in health care services, law enforcement, loss of productivity, and other social costs (Rehm, Patra, & Popova, 2006). Although the legal drinking age in British Colum- bia is 19, many male and female youth begin to use alcohol before this age, and 72% of BC high school students have tried alcohol by the age of 18 (Smith et al., 2015a). Despite recommenda- tions that females should drink less than males, in 2013, 39% of BC females who had tried alcohol engaged in heavy sessional drinking in the past month (a similar rate to their male peers), and younger females were more likely to drink at this level than younger males (Smith, Stewart, et al., 2014). Females may be more vulnerable to the effects of alcohol than males. For example, females’ body composition and metabolism mean they process alcohol at a slower rate than their male peers (National Research Council and Institute of Medicine, 2004; SAMHSA, 2015). Heavy alcohol use has been shown to affect bone development in young women (National Research Council and Institute of Medicine, 2004). As well, females may progress to alcohol dependence or abuse at a faster rate than males, and are more vulnerable to brain, heart, and liver problems related to alcohol (The National Centre on Addiction and Substance Abuse [CASA], 2003). A recent cohort study found that girls who start using alcohol early in high school are prone to harmful use by graduation (Flegel, 2013). In 2013, the Canadian Medical Association ex- pressed concern that girls were no longer waiting longer than boys to have their first drink and were drinking at the same rate as boys from a young age (Flegel, 2013). A narrowing of the gap be- tween males and females in alcohol consumption and heavy sessional drinking has also been seen in the United States (Substance Abuse and Mental Health Services Administration [SAMHSA], 2015). Underage alcohol use has been associated with many negative consequences during adolescence and later in life, including motor vehicle acci- dents and other unintentional injuries; suicides, homicides, and violence; risky sexual behaviour; impaired academic performance; and alcohol-re- lated problems and dependence later in life (SAM- HSA, 2015; Smith et al., 2015a). Reducing alcohol-related harms is therefore especially important for adolescent females. This evidence review identifies the most promising strategies to prevent risky drinking behaviour in underage girls. Adolescent females have a higher likelihood than males of alcohol-related harms such as getting into arguments with their family, getting injured, 6 Strategies to reduce risky alcohol use among underage girls: An evidence review

  8. Methodology Search criteria A structured search of the published literature was conducted using relevant research databases including Academic Search Complete, CINAHL, Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane Database of Systematic Reviews (CDSR), Communication & Mass Media Complete, ERIC, CABI Global Health, Medline, Psy- cINFO, PubMed Central, and Social Sciences Full Text. Research studies were also obtained by look- ing through the reference lists of relevant articles and using the “similar articles” feature available in some databases. A search of the grey literature, including govern- mental and non-governmental reports, was also conducted using search engines available through the Canadian Best Practices Portal for Health Pro- motion and Chronic Disease Prevention, Canadian Institute of Health Information, the New York Academy of Medicine Grey Literature Report, the SAMHSA registry of evidence-based programs and services, and Google. The search was limited to English language articles published between 1990 and February 2017. Eval- uation studies that assessed the impact or effec- tiveness of approaches, resources, and messages regarding the prevention of risky and harmful alcohol use among underage girls were of primary interest (e.g., binge drinking, drinking and driving, alcohol-related consequences). Search terms included combinations of terms such as: Heavy alcohol use Intervention Best practices Girls Polysubstance use Problematic alcohol use Policy Evidence Females Mixing alcohol Harmful alcohol use Resources Effective Underage Adverse childhood Risky alcohol use Messages Prevention Adolescent Trauma Binge drinking Strategies Protective factors Youth Drinking age Alcohol misuse Young people Drinking and driving Alcohol abuse Young women Zero tolerance Alcohol dependency Teen Alcohol outlet High school Alcohol availability 7 McCreary Centre Society

  9. Search results Evaluation of key articles The 27 interventions and 9 policy papers that included gender-specific data were reviewed for their methodological quality, relevance, and applicability to the BC context. A review scoring sheet with criteria for ranking the strength of the research design; sampling (e.g., including com- parison or control groups); outcome measures reported; and data analysis was used to evaluate the evidence. In the case of quantitative research, the statistical analyses used to assess effects (es- pecially effect sizes) were also evaluated. The search yielded 5,805 original articles. The titles, abstracts, and keywords of these studies were reviewed. Articles that did not focus on risky or harmful adolescent alcohol use or did not in- clude females in the sample were eliminated. This left 855 results. A brief full text review of these articles identified over 100 that discussed the effects of alcohol policies and messaging on female adolescents, and more than 200 that focused on youth alcohol misuse interventions. Among these articles, 27 interventions and 9 policy papers were rigorously evaluated and provided gender-specific informa- tion. Few of these interventions and policy papers had been implemented in Canada, so additional studies of interventions were sought which had not been as rigorously evaluated but showed promising results. Of the 27 high quality interventions (detailed in Appendix A), those with the most positive results for female participants were seen in 7 interven- tions targeted specifically at girls (Fang & Schinke, 2013; O’Donnell, Myint, Duran, & Stueve, 2010; Schinke, Cole, & Fang, 2009; Schinke, Fang, & Cole, 2009; Schinke, Fang, Cole, & Cohen-Cutler, 2011; Schwinn, Schinke, & Di Noia, 2010; Watt, Stewart, Birch, & Bernier, 2006), 6 mixed-gender interventions that reported more positive results for females than males (Longshore, Ellickson, McCaffrey, & St. Clair, 2007; Mason et al., 2009; Smith et al., 2008; Vogl et al., 2009; West et al., 2008; Wurdak, Wolstein, & Kuntsche, 2016), and 2 interventions that appeared equally effective for girls and boys (Shope, Copeland, Maharg, & Dielman, 1996; Stafström & Ostergren, 2008). 8 Strategies to reduce risky alcohol use among underage girls: An evidence review

  10. Calculation of effect sizes and Numbers Needed to Treat (NNT) Reflecting the experience of other researchers (e.g., Blake, Amaro, Schwartz, & Flinchbaugh, 2001) who have sought to identify interventions that are effective for girls in reducing risky alcohol use, it was challenging to isolate interventions which provided results for girls under the age of 19. There were even fewer studies which had been conducted in Canada or with specific at-risk subpopulations such as Indigenous girls, those who had experienced violence and abuse, or sexu- al minority girls. For the quantitative intervention studies that contained results for females, effect sizes were calculated if they were not already presented in the article, if adequate details were available. Such effect size measures may have included odds ratios, risk ratios, relative risk ratios, Cohen’s d, Hedges g, andη2 (eta squared). Most of these measures have general guidelines to identify what counts as small, moderate, and large effect sizes. They can also often be transformed from one to another. In addition to these more general measures of ef- fect size, there is a practical extrapolation of these measures that is often used to interpret health care interventions, called the Number Needed to Treat (NNT). The NNT can be interpreted as the number of people who need to be exposed to the intervention for one person to have the specific improvement (or reduction in negative outcome). Where applicable, for each effect size in a study that could be transformed to be included in the calculation of NNT, these calculations were performed. Unfortunately, few studies provided adequate information for female adolescents to perform the NNT calculations. Additionally, most interventions which had been evaluated focused solely on alcohol use rather than on polysubstance use, even though mixing alcohol and other drugs can be particularly risky for girls (Office of National Drug Control Policy, 2006). Most of the literature that was available came from the U.S. where the legal drinking age is older than in Canada (e.g., 21 years of age compared to 19 in British Columbia). Finally, some interventions such as the prescrip- tion of periods of abstinence from alcohol or vol- untary abstinence pledges did not appear to have any evaluation results which would be applicable to underage female youth. However, there were some evaluations completed in the 1980’s with older adolescents and young adults which would suggest these interventions may not be effec- tive. For example, one study of male and female college students found that following a 10 day period of abstinence, participants consumed more alcohol in the weeks after the intervention than they had prior to the intervention. Additionally, the heaviest alcohol consumers did not complete 10 days of abstinence (Burish, Maitso, Cooper, & Sobell, 1981). A subsequent study focused exclu- sively on female social drinkers found that a peri- od of abstinence did not negatively or positively affect future drinking behaviours (Carey, Carey, & Maitso, 1988). Limitations This study has several limitations, most notably a single intervention delivered in isolation is unlikely to reduce risky alcohol use among underage girls, however positive that intervention. Although not the primary focus of this review, it is important to note that other literature suggests that compre- hensive, multi-component interventions that fo- cus on building resilience and developing support- ive, healthy relationships for young people are more effective in reducing health risk behaviours than interventions that focus exclusively on the behaviour we wish to reduce or extinguish. 9 McCreary Centre Society

  11. Terms used in the report The following definitions were used in this report, unless otherwise noted: Alcohol abuse Alcohol abuse refers to a pattern of drinking too much alcohol or drinking alcohol too often which is often associated with risky behaviour and negative consequences. It does not necessarily imply alcohol dependence. Binge drinking Binge drinking refers to having multiple drinks on a single occasion. For females, this means having three or more alcoholic drinks, and for boys refers to having four or more alcoholic drinks. Heavy sessional drinking Heavy sessional drinking or heavy episodic drinking refers to having five or more alcoholic drinks within a couple of hours. (Note that some authors cited in this report may have referred to this level of drinking as binge drinking.) Harmful/risky alcohol use Harmful/risky alcohol use is use that can damage health. It does not include trying alcohol or drinking within low risk guidelines. Early adolescence/younger youth Early adolescence/younger youth refers to youth between the ages of 10 and 14 years old. Older youth Older youth refers to young people aged 15–18. NNT or Numbers Needed to Treat NNT or Numbers Needed to Treat is a statistical calculation which estimates how many young people need to receive the intervention for one youth to report a positive effect on their risky substance use. Cochrane Reviews Cochrane Reviews are systematic reviews of primary research studies that investigate the effects of interventions. RCT or Randomized Control Trial RCT or Randomized Control Trial is a scientifically rigorous way of evaluating an intervention, which compares participants in an intervention with a matched group who did not receive the intervention. 10 Strategies to reduce risky alcohol use among underage girls: An evidence review

  12. Risk and protective factors for harmful underage alcohol use among females This section reviews the published literature about risk factors for harmful alcohol use among underage girls and identifies protective factors that can buffer against such use. Members of the Youth Research Academy also reviewed information provided to McCreary from young people across BC through surveys, focus groups, and workshops. Many of the same risk and protective factors that were identified for risky alcohol use in the academic literature were discussed by youth in BC and their recommendations are included on page 17. Risk factors Transitions A national review of pathways to substance abuse among females in the U.S. identified transitions as key risk periods. For example, girls who frequently move from one home or neighbourhood to an- other are at increased risk of harmful alcohol use (CASA, 2003). School transitions may also be associated with increased alcohol use. For example, high school entry has been found to be a critical event linked with more rapid increases in alcohol use than before entry (Andreas & Jackson, 2015); and for girls, specifically, the transition was associated with initiation of alcohol use (Jackson & Schulen- berg, 2013). Transitions from elementary or middle school may contribute to risky alcohol use as girls try to manage their anxiety or attempt to fit in with older or different friend groups (CASA, 2003; Longshore et al., 2007; Shortt, Hutchinson, Hutchinson, Chapman, & Toumbourou, 2007), deal with increased academic pressure, and be immersed in an environment that involves less adult monitoring and more individual freedom (Andreas & Jackson, 2015; Jackson & Schulenberg, 2013). In BC, young women who had experienced harms from their alcohol use confirmed that the transition from elementary to high school was a time when their use increased because of feeling overwhelmed and stressed and wanting to fit in with older students (McCreary Centre Society, 2015). 11 McCreary Centre Society

  13. The transition to college includes many similar stressors including changes in peer and family relationships and academic pressure, as well as dealing with being an independent adult (Morton, Mergler, & Boman, 2014). This transition can also be a vulnerable time because underage drinking is typically normative for young adult women (LaBrie, Feres, Kenney, & Lac, 2009). In addition, this is a time when they associate alcohol with dealing with boredom and depression (CASA, 2003). Research has indicated an increase in young women’s experience of alcohol-related sexual consequences in the first year of college compared to the last year of high school (Or- chowski & Barnett, 2012). This study found that the onset of these types of consequences in the first year of college was associated with motives to enhance positive affect and with parents not knowing about alcohol use during high school. Young women in BC with post-secondary expe- rience identified the availability of alcohol com- bined with its social acceptability as increasing their use (McCreary Centre Society, 2015). Mental and emotional health challenges Underage females are more likely than males to use alcohol to manage their emotions, such as to improve mood or to cope with depression, anxi- ety, eating disorders, or body dissatisfaction (e.g., Kuntsche & Muller, 2012; Owens & Shippee, 2009; Schinke, Fang, & Cole, 2008). In BC, females aged 12–19 were more likely than males to use alcohol to manage challenges such as stress, anxiety, and depression (Smith et al., 2011; Smith, Stewart, et al., 2014; Smith et al., 2015a). Females as young as 11 years who reported high levels of depression were over twice as likely to use alcohol as their female peers who did not report depression (Schinke, Fang, & Cole, 2008), and those who reported drinking to manage their mental health challenges were more likely to drink at risky levels (Kuntsche & Muller, 2012; Lammers, Kuntsche, Engels, Weirs, & Kleinjan, 2013). In BC, girls aged 12–19 with mental health challenges were more likely to engage in risky drinking. For example, among those who had tried alcohol, those with a mental health condition were more likely than those without such a condition to binge drink (42% vs. 30%; Smith et al., 2015a), as well as engage in heavy sessional drinking (46% vs. 36%) and consume alcohol on three or more days (35% vs. 25%) in the past month (McCreary Centre Society, 2013). An early entry into puberty is another transition associated with increased alcohol consumption (CASA, 2003; Hummel Shelton, Heron, Moore, & van den Bree, 2012). Specifically, girls who ma- ture early are more likely than those who mature on time to report early initiation of alcohol use, increased alcohol use and abuse, and advanced drinking including multiple instances of drunken- ness, frequent drinking, and high consumption levels (Bratberg, Nilsen, Holmen, & Vatten, 2005; Hummel et al., 2012; Lee et al., 2014). There is some evidence, however, that suggests that these relationships may not persist into late adoles- cence (Kaltiala-Heino, Koivisto, Marttunen, & Fröjd, 2011). 12 Strategies to reduce risky alcohol use among underage girls: An evidence review

  14. Trauma Family challenges A history of trauma or abuse is a risk factor for harmful alcohol use among girls and young women (e.g., CASA, 2003; Poole & Dell, 2005; Weiss & Nicholson, 1998). For example, girls in BC who had been physically or sexually abused were more likely to report heavy sessional drink- ing in the past year than their peers who had not been abused (Smith et al., 2015a; Tourand, Smith, Poon, Saewyc, & McCreary Centre Society, 2016). They were also more likely to start drinking before they were 13 years old, and to report last drinking because they were stressed or feeling sad (Smith et al., 2015a). Factors such as family alcohol abuse, maternal alcoholism, maternal drug abuse, insecure at- tachment, low parental monitoring, having older siblings, unstructured home environment, and family conflict are all predictors of risky substance use among girls (e.g., Kumpfer, 2014; LaBrie et al., 2009;Paschall, Grube, Black, & Ringwalt, 2014; Schinke, Fang, & Cole, 2008; Soloski & Berry- hill, 2016; Toumbourou, Douglas Gregg, Shortt, Hutchinson, & Slaviero, 2013). Additionally, female youth whose mothers smoked or drank heavily during pregnancy are more likely to use alcohol than underage males with similar experi- ences (CASA, 2003). BC females aged 12–19 who had been bullied in person or cyberbullied were more likely to report heavy sessional drinking in the past month than their peers without these experiences (Smith et al., 2015a). These youth were also more vulnera- ble to negative peer influence. For example, those who had been bullied in the past year were more likely than their peers to report drinking because their friends were doing it (Smith et al., 2015a). In BC, female youth aged 12–19 whose parents did not know what they were doing with their free time were more likely to report heavy sessional drinking than those whose parents monitored their leisure time (e.g., Tourand, Smith, Poon, Saewyc, & McCreary, 2016). In addition, homeless and street-involved female youth in BC who had someone in their family with a substance use chal- lenge were more likely to have used substances themselves (Smith et al., 2015b). Discrimination, such as experiences of racism or because of a disability, can also be a risk factor for harmful alcohol use (Schinke et al., 2011; Smith et al, 2015a; Weiss & Nicholson, 1998) as are trau- matic experiences such as entering government care, living in extreme poverty, and being the victim of dating violence (Smith et al., 2011; Smith et al., 2015a). 13 McCreary Centre Society

  15. Peer pressure Higher family and neighbourhood income It has been suggested that female adolescents may be more susceptible to the pressures and social influences to use alcohol than their male counterparts (Amaro et al., 2001; CASA, 2003). Having a best friend or romantic partner who mis- uses alcohol can be a stronger predictor of risky alcohol use for female youth than males (Amaro et al., 2001; Schinke, Fang, & Cole, 2008; Schinke et al., 2011). For example, middle school girls in the U.S. who experienced high peer pressure to drink were twice as likely to use alcohol as those who did not experience this level of peer pres- sure. This was not the case among boys (CASA, 2003). Studies from the U.S., Europe, and China have linked heavy sessional drinking or drunkenness to adolescents growing up in families with high so- cioeconomic status (CASA, 2003; Humensky, 2010; Legleye et al., 2013; Lu et al., 2015;Lu et al., 2016; Melotti et al., 2013; Song et al., 2009; Tomcikova, Geckova, van Dijk, & Reijneveld, 2011; Zaborskis, Sumskas, Maser, & Pudule, 2006). For exam- ple, U.S. studies have shown that female youth growing up in affluent suburban neighbourhoods are at particularly high risk for heavy alcohol use (Botticello, 2009; Reboussin, Preisser, Song, & Wolfson, 2010). Other studies have indicated that economically disadvantaged or unsafe neigh- bourhoods are not associated with problematic alcohol use among underage girls (Fagan, Wright, & Pinchevsky, 2013; Svensson & Hagquist, 2010; Tucker et al., 2013). Among youth who completed the BC Adolescent Health Survey, females were more likely than males to report using substances because they wanted to have fun, their friends were doing it, and they felt pressured into doing it (e.g., McCreary Centre Society, 2015; Tourand et al., 2016). Data from the survey also showed that girls who had six or more friends were more likely to report heavy sessional drinking in the past month than their peers with fewer friends (Smith et al., 2015a), as were those who regularly participated in organized sports such as hockey and soccer teams (Smith, Stewart, Poon, Saewyc, & McCreary Centre Society, 2011b). It has been suggested that upper-middle class females constitute an “at-risk” population be- cause they face multiple pressures with educa- tional expectations, family dynamics, and peer norms while having the resources such as a car and money which allow them to obtain and use alcohol frequently and heavily (Luthar, Barkin, & Crossman, 2013). Drinking among this population is often seen as socially and culturally acceptable. For instance, parents may not impose serious consequences for risky drinking, and excessive use can be seen as normative in a culture of affluence with a ‘Work hard, play hard!’ mentality (Luthar et al., 2013). Personal income as well as family income can also be a risk factor in BC. Girls aged 12–19 who were employed were more likely to engage in risky alcohol use than their same age peers who had not worked in the past 12 months (Smith et al., 2015a). 14 Strategies to reduce risky alcohol use among underage girls: An evidence review

  16. Other risk factors Protective factors Results from the BC Adolescent Health Survey showed other youth at risk for problematic sub- stance use, such as binge drinking, heavy sessional drinking, and frequent or early alcohol use. These included girls who identified as lesbian or bisex- ual, lived in rural areas, or were born in Canada (Smith et al., 2015a). In contrast, immigrant girls were generally less likely to engage in problematic alcohol use in BC (Smith et al., 2015a). However, high acculturation has been found to weaken girls’ bond with their parents, which increases their risk for potentially harmful alcohol use (Schinke et al., 2011). The presence of protective factors can reduce the risk of female adolescents engaging in problem- atic alcohol use, even if those girls also experi- ence identified risk factors (Smith, Stewart, et al., 2014). Family connectedness Having positive relationships with family and spending time with them can be protective against early and problematic substance use for girls (CASA, 2003). It can also mitigate risk factors for heavy episodic drinking, such as spending time with peers who drink and having parents who pro- vide alcohol (Danielsson, Romelsjö, & Tengström, 2011). Family does not necessarily have to be bio- logical. For example, youth with government care experience in BC who felt highly connected to the people they thought of as their family were less likely to engage in regular heavy sessional drinking than their peers who felt less connected (Smith, Stewart, Poon, Saewyc, & McCreary Centre Soci- ety, 2011a). Experiencing very high social self-esteem has also been associated with problematic substance use among girls (Fisher, Miles, Austin, Camargo, & Colditz, 2007). Additionally, certain personality traits or characteristics, such as negative think- ing, impulsivity, and sensation seeking have been identified as being associated with problematic alcohol use for girls more than boys (Lammers et al., 2013). Young women who identify as being sensitive to high levels of anxiety are more likely to use alcohol to cope, to conform with peers, and because they expect that drinking helps reduce tension. Additionally, compared to low anxiety sensitivity females, they experience more nega- tive consequences related to alcohol such as going to school drunk, or not being able to complete homework or study for a test (Watt et al., 2006). Female youth in BC who felt their family paid attention to them and that they had fun together were less likely to drink heavily than those who did not feel this way (Smith et al., 2015a). Hav- ing a helpful adult in their family to turn to and feeling understood by their family was especially protective against girls’ binge drinking and heavy sessional drinking (Smith et al., 2009; Smith et al., 2015a). BC youth with both mental health and substance use challenges who were living outside the family home reported that reconnecting with fami- ly could lower the likelihood of substance use among youth, if that was a positive experience (Cox, Smith, Peled, & McCreary Centre Society, 2013). However, BC youth with FASD who ex- perienced substance use challenges found that reconnecting with their biological family was only protective if their family of origin did not engage in harmful substance use (Peled, Smith, & Mc- Creary Centre Society, 2014). 15 McCreary Centre Society

  17. The relationships between girls and their moth- ers can be particularly protective against harmful alcohol use (Schinke, Fang, & Cole, 2008), and has been associated with less frequent alcohol use among middle school girls with high-risk peer networks (Kelly et al., 2011). Having a strong bond with their mother reduced the risk that female youth would engage in substance use due to stress, low self-esteem, and body image issues (Schinke et al., 2011). Additionally, while some studies have shown that parental disapproval of alcohol is unlikely to lower alcohol use (Kelly et al., 2011), younger adolescent girls who have not yet tried alcohol may wait longer to do so if their mother urges them to abstain (Schinke et al., 2008). Positive peer relationships Positive peer influence has been linked to a lower likelihood of binge drinking among youth in BC. For example, females with friends who would be upset with them if they got drunk were less likely to engage in heavy sessional drinking in the past month and to drink at risky levels on the Saturday before completing the survey (Smith et al., 2009; Smith et al., 2015a). Community and cultural connectedness Connection to community and culture and feeling safe in their neighbourhood can all be protective against risky alcohol use for girls in BC (Smith et al., 2015a). For example, feeling connected to their community or culture was associated with a lower likelihood of binge drinking among BC females aged 12–19 (Smith et al., 2015a). In the U.S., frequent attendance at religious services in the community was associated with a lower like- lihood of alcohol use or heavy sessional drinking, especially among younger adolescent girls, and feeling connected to religion and spirituality was particularly protective among older girls (CASA, 2003). School connectedness School connectedness, including feeling like a part of school, feeling that teachers care, and having positive relationships with peers at school, was as- sociated with lower rates of heavy sessional drink- ing and binge drinking among girls in BC (Smith et al., 2015a). This was true for youth in general and for vulnerable and disadvantaged groups such as girls with government care experience (Smith, Peled, et al., 2015) and Aboriginal girls (Tourand et al., 2016). Youth with several risk factors for problematic alcohol use can benefit from connecting with their community. For example, females with government care experience who volunteered in their community were less likely to report heavy sessional drinking than those who were less en- gaged (Smith, Stewart, Poon, Saewyc, & McCreary, 2011a). Feeling connected to school has also been associ- ated with higher academic performance, which in turn has been linked to lower alcohol use (CASA, 2003). It has also been associated with a willing- ness to look after intoxicated friends, more so for girls than for boys (Chapman, Buckley, Reveruzzi, & Sheehan, 2014). The structure provided by engaging in extracur- ricular activities has been shown to be protective against alcohol use (CASA, 2003; Smith et al, 2015a). However, in BC finding those activities personally meaningful was particularly important for girls and was associated with a lower risk of binge drinking and heavy sessional drinking (Smith et al, 2015a). Having post-secondary plans can also reduce the likelihood that girls aged 12–19 use alcohol fre- quently or binge drink (Smith et al., 2009). 16 Strategies to reduce risky alcohol use among underage girls: An evidence review

  18. Other protective factors McCreary research also shows that having posi- tive body image, post-secondary aspirations, and volunteering in the community are all protective factors against risky alcohol use for girls in BC (Smith, Stewart, et al., 2014). Youth’s recommendations Several McCreary reports include recommendations from young people to reduce the risk of excessive drinking among youth: 9 Having youth-friendly spaces and a variety of extracurricular activities as alternatives to substance use (Tourand et al., 2016). 9 Supporting youth to have goals for the future (Sadler, 2006). 9 Having access to positive and affordable alternative activities, such as programs to help develop skills in sports and art (Sadler, 2006). 9 Having access to meaningful work and employment programs (Peled et al., 2014). 9 Helping youth feel connected to their community and culture (Peled et al., 2014). 9 Providing more funding for counselling so that youth do not rely on drugs and alcohol to cope with problems (Martin et al., 2007). 9 Hiring more outreach workers to specifically address youth’s drug and alcohol use (Martin et al., 2007). 9 Offering community-based substance-free girls’ groups (Smith, Warren, Cox, Peled, & McCreary Centre Society, 2014). 9 Having sober mentors (Smith, Warren, et al., 2014). 9 Having positive relationships with adult family members and other important adults (Smith, Warren, et al., 2014). 9 Having structure and routine (Smith, Warren, et al., 2014). 17 McCreary Centre Society

  19. Interventions to reduce harmful underage alcohol use among females A review of the literature shows that many older interventions focused on prevention of alcohol use, emphasizing moral and legal objections to alcohol and the detrimental effects it could have on users. By the 1980’s more emphasis was put on self-esteem building and decision-making skills, and in the 1990’s programming began to include a focus on social influences and the life skills needed to resist social pressures (Weiss & Nicholson, 1998). More recent interventions have expanded to include social media and online interventions. Family-based interventions Interventions that strengthened family connected- ness and communication appeared to be consis- tently among the most effective against risky ado- lescent alcohol use (Mason et al., 2009; O’Donnell et al., 2010; Schinke, Cole, & Fang, 2009; Schinke et al., 2011; Toumbourou et al., 2013), and could also reduce alcohol use into adulthood (Foxcroft & Tsertsvadze, 2011a). The interventions that were most successful in reducing harmful alcohol use among girls often combined supporting the development of parent- ing skills—including nurturing behaviours, estab- lishing clear boundaries or rules, and parental monitoring—while also supporting girls to devel- op positive relationships with family and peers (Foxcroft & Tsertsvadze 2011a). Interventions that placed too much emphasis on family rule setting and monitoring without also strengthening family communication and relationships were not successful (Koning, van den Eijnden, Verdurmen, Engels, & Vollebergh, 2012). Relatively few interventions that were reviewed considered the contextual risk and protective factors associated with girls engaging in health risk behaviours such as risky alcohol use. However, those that did appeared to have the most success, such as those that strengthened healthy family relationships, promoted positive peer interactions, and included community engagement (See Appendix C). One successful example was a rigorously evaluat- ed three-year Swedish multi-component interven- tion that included 13 activities which promoted family connectedness and communication, and encouraged parents to maintain restrictive atti- tudes towards alcohol use and to meet their chil- dren’s friends. Post intervention results showed lower intoxication rates among participating male and female youth, and delayed alcohol initia- tion by an average of a year in comparison with matched control group adolescents(Pettersson, Özdemir, & Eriksson, 2011). 18 Strategies to reduce risky alcohol use among underage girls: An evidence review

  20. Another successful example was the Dutch ‘Pre- vention of Alcohol use in Students’ (PAS) program which engaged parents of youth aged 13-16 to set boundaries around adolescent alcohol use and included a brief program with youth on healthy attitudes and refusal skills. A randomized control trial (RCT) reported delayed onset of weekly heavy drinking and lower rates of heavy weekend drink- ing, which persisted until young people were four years over the legal drinking age (Koning, van den Eijnden, Verdurmen, Engels, & Vollebergh, 2011; Koning, van den Eijnden, Verdurmen, Engels, & Vollebergh, 2013). Even among at-risk males and females who had begun drinking before 12 years old, there were no increases in alcohol frequency and quantity, as was seen among peers in the con- trol group (Koning, Lugtig, & Vollebergh, 2014). analysis indicated that if 48 boys and girls received this intervention, one would be prevented from engaging in heavy alcohol use. Interventions which included opportunities for girls and their female caregivers to build skills together appeared particularly effective. These interventions focused on building communication skills (particularly around the topic of alcohol) including emphasis on trust, respect, empathic listening, sharing of feelings, and conflict man- agement; increasing the daughters’ self-image and self-esteem; establishing parental monitoring practices; and spending quality time together (Fang & Schinke 2013; Mason et al. 2009; O’Don- nell et al., 2010; Schinke, Cole, & Fang, 2009; Schinke, Fang, & Cole, 2009). Some family interventions appear to work across cultures and settings. For example, a U.S. program called Strengthening Families offered separate workshops for Grade 6 students and their parents, followed by joint training sessions. Parent sessions taught participants how to clarify expectations, manage their emotions, and develop appropriate disciplinary practices; while the children’s sessions included peer resistance and relationship skills training. The combined workshops focused on conflict resolution, communication, and develop- ing a strong family bond. An RCT with over 600 The ‘Resilient Families’ intervention in Australia was delivered to 12-year-olds who received a social relationship-based curriculum in schools which covered family rules, communication, prob- lem solving, peer resistance, and refusal skills. At the same time, parents received written informa- tion about practical parenting strategies and par- ticipated in parent education events. At two-year follow-up, an RCT with 2,000 students found a small but significantly lower risk of heavy alcohol use in the previous two weeks (Toumbourou et al., 2013). The number needed to treat (NNT) 19 McCreary Centre Society

  21. participants found a 40% reduction in harmful alcohol use among male and female youth at four- year follow-up (Spoth, Redmond, & Shin, 2001). Foxcroft and colleagues (2002) completed an NNT analysis of this data which showed that for every nine young people who went through the program, there would be one who was positively impacted to not use alcohol, one who would not use alcohol without permission, and one who would not get intoxicated. Online interventions Attending in-person family programs can be a challenge for low income, single parent, or working households. There is some evidence that online interventions that can be completed by female caregivers and their daughters at home can be effective across different cultures including Black, Hispanic, and Asian American girls (Fang & Schinke, 2013; Schinke et al., 2011). These inter- ventions can reach those who might be reluctant or unable to attend in-person sessions (Schinke, Cole, & Fang, 2009; Schinke et al., 2011). Addi- tional evidence suggests that girls may prefer the more anonymous web-based programs to discuss- ing alcohol-related behaviours in a group (D’Amico et al., 2006). This same intervention was adapted for use with African American youth, and was effective in reducing past month alcohol use in comparison to a control group (Brody, Chen, Kogan, Murray, & Brown, 2010). It has also been adapted for use in BC with 6- to 14 -year-olds although no evalu- ation data was available (Sea to Sky Community Services, n.d.; Communities that Care Squamish, n.d.). Nevertheless, based on the evidence of the Strengthening Families programs in the U.S., this is a potentially promising BC intervention. One U.S. program delivered 14 modules over three weeks to girls aged 10–13 and their moth- ers. Each module included voice-over narration, skill demonstrations by animated characters of an adolescent girl and her mother, and interactive exercises to teach refusal skills, build respect, and strengthen bonds. Evaluated with over 200 mother-daughter pairs, intervention participants reported better communication with their moth- ers, more household rules against drinking, more parental monitoring, greater conflict resolution skills, more normative beliefs about adolescent al- cohol use, higher self-efficacy in avoiding alcohol, lower alcohol consumption, and lower intention to use alcohol at two months follow-up in compar- ison with a control group (Schinke, Cole, & Fang, 2009; Schinke, Fang, & Cole, 2009). A similar nine- week program evaluated with an RCT showed that 12 months after the intervention, partici- pants reported better communication with their mothers, more knowledge of family rules around alcohol use, more awareness that their parents were monitoring their free time, increased ability to refuse peer pressure, and reduced frequency of self-reported alcohol use in the past 30 days in comparison to a control group (Schinke, Fang, & Cole, 2009). Although family-based interventions are often offered to both boys and girls, they may be more effective for girls. For example, an RCT with over 400 families of youth in Grade 6 in the U.S. pro- gram ‘Preparing for the Drug Free Years’ found females who had participated in the intervention had significantly lower rates of alcohol abuse (met DSM-IV criteria) than females in the control group at 10 years follow-up, a result that was not seen among males (Mason et al., 2009). The inter- vention focused on promoting prosocial family bonding including increased family interaction, clear rules about drug and alcohol use, parental monitoring, refusal skills for youth, and reducing family conflict (Mason et al., 2009). 20 Strategies to reduce risky alcohol use among underage girls: An evidence review

  22. Online interventions that build positive peer relationships using a similar format can also be ef- fective. For example, ‘RealTeen’ combined the use of an older female animated character to guide participants through knowledge and skill building sessions relating to alcohol with opportunities for peer to peer communication among participants. Participants showed small but significant increas- es in self-efficacy and lower frequency of past month alcohol use in comparison to girls in the control group at six-month follow-up (Schwinn et al., 2010). intervention with 268 families when the girls were in Grade 7, and found a lower likelihood of sexual risks, alcohol use, and alcohol intoxication among the girls in the intervention group in comparison to the control group (O’Donnell et al., 2010). Similarly, using a comic book or a cartoon-based teenage drama that parents and youth worked through together at home were both methods of teaching about alcohol harms and ways to address or avoid these harms through stories which were more effective among girls than boys (Vogl et al., 2009; West et al., 2008). The success of family-based interventions appears to be based on the opportunity they provide for female youth to build positive relationships with their caregivers (or peers), as studies of univer- sal online interventions without this relational component have not generally found reductions in substance use among adolescent girls, and in some cases, have been associated with increased use (Spijkerman et al., 2010). Using story telling An audio-based program also showed some success, including with urban Black and Lati- no girls. The ‘Specially for Daughters’ program involved Grade 6 girls and their caregivers listen- ing at home to audio CDs containing stories that increased awareness of alcohol use and its risks; taught parents supportive skills; and taught girls about peer pressure, handling emotions, and sen- sation seeking. An RCT was used to evaluate the 21 McCreary Centre Society

  23. Interventions targeting personality traits The ‘Preventure’ group intervention has been delivered in BC and Nova Scotia to high school students aged 14–17 who exhibited sensation seeking, anxiety sensitivity, and hopelessness. Most of the participants were female. The inter- vention focused on motivation for alcohol use and maladaptive coping and was evaluated with an RCT of 297 participants. At four-month follow-up, the intervention appeared effective in reducing rates of binge drinking (more than three or four drinks for girls) but not frequency of drinking (Conrod, Stewart, Comeau, & Maclean, 2006). Conrod and colleagues (2006) suggest that this is because quantity of alcohol is associated with internal motivational factors whilst frequency of drinking is more related to external factors such as opportunity and peer influences. The Vernon School District in BC is currently implementing the program (Turcato, 2017). Certain characteristics such as negative thinking, impulsivity, and sensation seeking have been more strongly associated with risky alcohol use for girls than boys (Lammers, Kuntsche, Engels, Wiers, & Kleinjan, 2013). For example, young women who identify as being sensitive to high levels of anxiety are more likely to drink to cope, to con- form with peers, and expect that drinking would help reduce tension. Additionally, compared to low anxiety-sensitivity females, girls prone to anxiety experience more negative consequences relating to alcohol, such as going to school intox- icated or being unable to complete homework or to study for a test (Watt et al., 2006). Although most interventions targeting personality characteristics did not include specific results for girls, many did include a majority female sample. The few that did report gender-specific results showed some success, particularly for girls consid- ered to have high anxiety sensitivity. Australian and UK versions of the ‘Preventure’ program were delivered to 14-year-old youth by trained therapists in two 90 minute sessions focusing on personality styles, cognitive thoughts, and personality-specific coping strategies. An RCT found the strongest effect was among youth with high anxiety-sensitivity but there was also a mild effect among high sensation seeking youth (Conrod, Castellanos-Ryan, & Mackie, 2011). Among sensation-seeking youth who used alco- hol, those who participated in the UK program were 50% less likely to binge drink (four or more drinks for girls; five or more drinks for boys) at 12 months follow-up than their peers in the control group, and at two-year follow-up they reported lower rates of problem drinking than the control group (Conrod, Castellanos, & Mackie, 2008). A three-year follow-up in Australia found a signif- icant reduction in the likelihood of consuming alcohol, binge drinking, and experiencing related harms during the past six months for youth in the intervention group compared to the control group (Newton et al., 2016). 22 Strategies to reduce risky alcohol use among underage girls: An evidence review

  24. The Preventure program is delivered by skilled therapists and this makes it difficult to implement widely (O’Leary-Barrett, Mackie, Castellanos-Ry- an, Al-Khudhairy, & Conrod, 2010). However, an adapted version (the UK based ‘Adventure pro- gram’) which was offered by trained teachers and school counsellors also showed some promising results. Two years after delivery of the interven- tion, an RCT of over 1,000 participants (mean age 13.7 years, over 50% female) found a 55% decrease in binge-drinking (four or more drinks for girls; five or more drinks for boys) among participants compared with controls for those who had already drunk alcohol before the start of the intervention (O’Leary-Barrett et al., 2010). An NNT analysis indicated that one case of binge drinking could be prevented if seven participants received the intervention. Additionally, there was a mild effect on reducing risky alcohol use among the general population of students who were in the intervention schools but did not participate themselves (Conrod et al., 2013). One of the few interventions that provided gender-specific results was a brief cognitive behavioural therapy program delivered in three 50 minute sessions to first year female undergrad- uate students in Canada who experienced high anxiety sensitivity. The program included educa- tion about anxiety; cognitive restructuring training to change coping strategies, attitudes, and beliefs; and a physical activity component (Watt, Stewart, Birch, & Bernier, 2006). An RCT with over 200 students found the intervention produced a small but significant reduction in conformity motives for drinking and expectations that drinking would help provide emotional relief. It also reduced problem drinking (drinking resulting in negative consequences) by 50% at 10 week follow-up, although this result was only marginally signifi- cant statistically. However, as with other similar interventions, it did not reduce overall drinking frequency rates (Watt et al., 2006). Evidence of the effectiveness of interventions targeting certain personalities or characteristics among Indigenous youth is inconclusive although available studies are based on small sample sizes. For example, a program developed for Mi’kmaq First Nations in Nova Scotia offered teachings from the Mi’kmaq culture to 29 youth aged 14–18 (69% females) who were categorized as sensation seeking, anxiety sensitive, or experienced hope- lessness/negative thinking. At four-month fol- low-up, the intervention group reported a lower but not statistically significant frequency of heavy sessional drinking. The authors maintain that with a larger sample size the results would have been significant (Mushquash, Comeau, & Steward, 2007). 23 McCreary Centre Society

  25. Knowledge and skills training Increasing knowledge Knowledge building appeared to be a successful component of a program when it provided girls with information about the effects of alcohol, the consequences of risky use, the benefits of non- use; dispelled myths about alcohol; discussed the potential for risk and harm in common teenage drinking scenarios; and offered specific infor- mation such as gender-specific low-risk drinking guidelines and where to access help (Longshore et al., 2007; Vogl et al., 2009; Wurdak, Wolstein & Kuntsche, 2016). Interventions that focused ex- clusively on the negative consequences of heavy alcohol use were ineffective for girls (D’Amico et al., 2006; Sloboda et al., 2009; Spijkerman et al., 2010). Interventions that focus on building knowledge and skills to address risky alcohol use are most commonly delivered in a school setting (Sussman et al., 2004). The interventions typically discuss attitudes towards alcohol, offer information about harmful alcohol use, and build skills. These skills include problem solving, decision-making, goal setting, critical thinking, communication, and as- sertiveness with the aim of reducing susceptibility to negative social influences, increasing resilience, and decreasing motivation to use alcohol (Botvin & Griffin, 2014). Vogl and colleagues (2009) suggest that school- based knowledge and skill building programs are often not effective for youth who have already started using alcohol, especially if they focus on abstinence. Interventions can also be ineffective when there are problems with implementation re- sulting from poor teacher training, poor program adaptation, teacher characteristics and program characteristics not being a good fit, and lack of integration of the program into the syllabus (Vogl et al., 2009). Offering “social norming” education to girls that addresses their misperceptions about the prev- alence and social acceptance of alcohol use can be effective for girls (Caria, Faggiano, Bellocco, & Galanti, 2011; Vigna-Taglianti et al., 2009). This may be because girls’ risky alcohol use is more strongly associated with social norms and situa- tions than boys (Longshore et al., 2007). When sustained changes occurred in girls’ normative beliefs, this led to lower past month alcohol use at 6-month follow-up (Schwinn et al., 2010). One study found that when a component that addressed misperceptions of alcohol use was removed from the intervention, there was an in- crease in drinking quantity among young women who participated (Spijkerman et al., 2010). This finding reinforces the value of addressing misper- ceptions through social norming interventions. One program that was ineffective was the police led ‘Take Charge of Your Life’ program delivered to close to 20,000 Grade 7 students in the U.S. RCT results showed that when participants reached Grade 11, female students who had participated in the intervention were 5% more likely to en- gage in heavy sessional drinking in the previous 14 days than their peers in the control group. The increased risk among intervention students compared to control group students was strongest among those who did not use alcohol at baseline (Sloboda et al., 2009). 24 Strategies to reduce risky alcohol use among underage girls: An evidence review

  26. Building social competence and resistance skills Increasing social competence and resistance skills were components present in effective interven- tions for girls. The more successful interventions included components that strengthened self-effi- cacy, such as the confidence to resist social pres- sures to use alcohol (e.g., Schwinn et al., 2010), built self-esteem and positive body image (Fang & Schinke 2013; Schinke, Fang, & Cole, 2009; Schin- ke et al., 2011), and offered training including recognizing social pressures, refusal skills, decision making, goal-setting, assertiveness and commu- nication skills (Fang & Schinke 2013; Longshore et al., 2007; Mason et al., 2009; Schinke, Cole, & Fang, 2009; Schinke, Fang, & Cole, 2009; Schinke et al., 2011; Schwinn et al., 2010; Smith et al., 2008; Vogl et al., 2009; West et al., 2008). Some of the more effective interventions included the opportunity for young people to practice de- cision-making and problem-solving skills through role playing different scenarios which allowed them to reflect and share experiences (Schwinn et al., 2010; West et al., 2008). Furthermore, inter- ventions that addressed the context and situa- tions in which girls use alcohol (e.g., at parties or to alleviate boredom) and identify safer enjoyable alternative ways to spend leisure time have shown promise when combined with life skills training (O’Donnell et al., 2010; Smith et al., 2008; West et al., 2008; Wurdak et al., 2016). However, such interventions may only be successful for baseline non-drinking girls in reducing regular use but not for those already engaging in heavy use (Smith et al., 2008). However, such interventions only appeared to be effective with younger girls, and primarily with those transitioning to high school (Jansen et al., 2016; Smit, Cuijpers, Lemmers, Jonkers, & De Weerdt, 2003; Smith et al., 2004; Vigna-Taglianti et al., 2009; Weiss & Nicholson, 1998). Interven- tions that did not address motivations for drinking and building self-esteem, or focused solely on building resistance skills or stress management skills were not effective (Graham, Johnson, Han- sen, Flay, & Gee, 1990; Vigna-Tagliant et al., 2009). 25 McCreary Centre Society

  27. Combining knowledge and skill building One example of a successful intervention was the ‘Alcohol Misuse Prevention Study’ (Shope et al., 1996) which taught Grade 10 students in five 45 minute sessions about short-term effects of alcohol use and the risks of excessive alcohol use, in combination with the skills to deal with social pressures to drink. The intervention was evaluat- ed with an RCT of over 2,000 Michigan students. At two-year follow-up, girls had higher alcohol refusal skill rates than boys, and both boys and girls in the intervention group had lower rates of negative consequences of alcohol use, such as drinking and driving, in comparison with youth in the control group (Shope et al., 1996). A similar rigorously evaluated 12-session African inter- vention offered to Grade 8 and 9 students from low income families in South Africa (Healthwise) combined life skills training involving emotion management, decision making, and self-aware- ness with alcohol and sexual health education. The evaluation found a significant decrease in the likelihood of heavy alcohol use (defined as four or more drinks on a single occasion) among both boys and girls in the intervention group compared to control (Smith et al., 2008). A combination of increasing knowledge of alcohol harms and developing a range of skills (e.g., resis- tance skills, coping skills) appears more effective for girls than focusing solely on knowledge or skill development (e.g., Graham et al., 1990). Some such interventions have been associated with low- er negative alcohol-related consequences, includ- ing problems with parents, peers and school staff, getting injured, feeling anxious or depressed, and drinking and driving (Evans-Whipp et al., 2013), and with small reductions in weekly alcohol use for male and female adolescents (Strøm, Adolfsen, Fossum, Kaiser, & Martinussen 2014). However, while most knowledge and skills pro- grams have been effective in increasing knowl- edge of alcohol’s effects and harms, this has not always led to reduced risky alcohol use (Mor- genstern, Wiborg, Isensee, & Hanewinkel, 2009; Wynn, Schulenberg, Maggs, & Zucker, 2000), or a sustained impact (Smith et al., 2004; Weiss & Nicholson, 1998). More effective interventions have included dis- cussion about the reasons girls might use alcohol, such as peer pressure, handling emotions, and sensation seeking (O’Donnell et al., 2010); rea- sons for getting drunk (Vogl et al., 2009); coping with puberty and bodily changes (Schwinn et al., 2010); and also taught skills to manage stress and anxiety, and how to relax using alternatives to al- cohol (Fang & Schinke, 2013; Schwinn et al., 2010; Schinke, Fang, & Cole, 2009; Schinke et al., 2011). Knowledge and skill development initiatives may be more effective for younger girls than older ones, possibly because the skill enhancement component helps girls to cope better with puber- ty-related social and emotional changes (Vigna-Ta- gliant et al., 2009). For example, an RCT with over 7,000 European students who participated in ‘Un- plugged’, a substance use and life skills program found it to be generally more successful for boys than girls (Caria et al., 2011; Vigna-Tagliant et al., 2009). However, at 18-month follow-up, girls aged 12 or younger at baseline had a reduced risk for alcohol-related problems such as arguing, fighting, relationship problems, poor school performance, and hospitalization (Caria et al., 2011). Our calcu- lations indicate that for every 38 young girls ex- posed to the intervention, one would be prevent- ed from experiencing alcohol-related harms. 26 Strategies to reduce risky alcohol use among underage girls: An evidence review

  28. While universal school-based knowledge and skill building interventions may have some positive effects for younger girls, a different approach may be needed for older ones. For example, a Dutch study found that the same knowledge and skills program (Healthy School and Drugs Program) de- livered to younger youth was more effective with younger girls than younger boys but was more effective for older boys than older girls (Smit et al. 2003). additional sessions were added for students in Grade 9 which covered the challenges of transi- tioning from middle school to high school (Ellick- son, McCaffrey, Ghosh-Dastidar, & Longshore, 2003; Longshore, Ellickson, McCaffrey, St. Clair, 2007). There were no significant effects on alcohol use for boys in Grade 11. Programs that took a harm minimization approach showed some success in significantly reducing weekly alcohol use (Longshore et al., 2007; Vogl et al., 2009), heavy sessional drinking (Vogl et al., 2009), and alcohol-related harms for girls (Longshore et al., 2007; Vogl et al., 2009). Useful elements appeared to include building knowledge and skills to recognize and help people who are intoxicated or are in an unsafe situation because of their alcohol use, and reviewing how to prevent alcohol-related harms in common teenage drink- ing scenarios (Vogl et al., 2009). These programs may be more effective with girls than boys be- cause adolescent females experience alcohol-re- lated harms more than their male counterparts (Evans-Whipp et al., 2013). These findings suggest it may be beneficial to ensure any intervention is not delivered in a single instance but is age-adapted and delivered again in later grades. For example, ‘Project ALERT Plus’ is a U.S. program focusing on risky alcohol use. It initially consisted of 11 knowledge and skill build- ing sessions for students in Grade 7 and three booster sessions in Grade 8 that discussed conse- quences and benefits of non-use, identified social pressures to use alcohol, developed resistance skills, and enhanced self-efficacy. However female intervention participants showed no difference over control group participants in Grade 11 until 27 McCreary Centre Society

  29. The Australian ‘CLIMATE program’ discussed how to keep people safe who are drinking too much, and what to do in a medical emergency, including learning first aid responses such as the recovery position. At 12-month post-intervention, an RCT with approximately 1,500 13-year-olds found boys and girls showed increased knowledge. However, only girls reported any impact on their drinking behaviour, with those in the intervention group reporting lower heavy sessional drinking rates than girls in the control group (Vogl et al., 2009). Members of McCreary’s Youth Research Academy reviewed the different interventions highlighted in this review, and felt that the Fourth R, a Canadi- an developed school-based curriculum delivered over 28 hours to build healthy relationships and decrease unhealthy and abusive behaviours could be particularly effective at reducing harms for adolescent girls, as it includes the opportunity to practice negotiating skills in a series of realistic and escalating scenarios. Evaluation results of the intervention with 196 Grade 9 students who participated in scenarios where they experienced increasing peer pressure from older student actors around alcohol use found improved nego- tiation skills, increased refusal skills and delayed usage among female participants in comparison to a control group (Wolfe, Crooks, Chiodo, Huges & Ellis, 2012). This review found very few interventions that targeted or were effective for older youth. ‘Project Towards No Drug Abuse’ was an example of an intervention in mainstream schools which could be effective among youth in older grades who had used alcohol. It discussed negative stereotypes of young people who drank heavily in addition to the common components of building communication and refusal skills. When delivered to Grade 9–11 mainstream high school students, participating heavy alcohol users at baseline had a significantly lower rate of alcohol use at one-year follow-up in comparison to a control group. However, there were no significant effects for participants in alternative schools or for those who did not use alcohol or engaged in moderate use at baseline (Dent, Sussman, & Stacy, 2001). The mixed results seen for knowledge and social skills interventions delivered in isolation suggest they may be more successful when combined with other approaches such as restricting avail- ability and engaging communities (World Health Organization, 2015), and include different modes of delivery such as online learning and classroom discussion (Vogl et al., 2009). 28 Strategies to reduce risky alcohol use among underage girls: An evidence review

  30. Culturally responsive interventions intervention was seen as successfully incorpo- rating “Western and Indigenous pillars of knowl- edge” (Baydala et al., 2016, p. 74) that allowed for “two-eyed seeing”, acknowledging diverse perspectives on substance abuse prevention. Find- ings pointed to the importance of building part- nerships among the community, schools, Elders, and the university and how this benefits students (Tremblay, Baydala, Rabbit, Louis, & Ksay-yin, 2016). Interventions that were developed or modified to reflect the cultural background of the girls who participated were not common but showed some promise (Schinke, Tepavac, & Cole, 2000). Several interventions have been identified by the Canadian Best Practices Portal as successful prac- tices in Aboriginal contexts (referred to as ‘Aborig- inal Ways Tried and True interventions’). These interventions “have undergone a rigorous, cul- turally-relevant assessment process based on the following six criteria: basis in the community, ho- listic approach, integration of Indigenous cultural knowledge, building on community strengths and needs, partnership/collaboration and demonstrat- ed effectiveness. Accepted standards of evidence include both Aboriginal and academic research approaches”(Government of Canada, n.d.). Adapted from the ‘Strengthening Families’ pro- gram, ‘Listening to One Another’ is a family-cen- tred substance abuse program for youth aged 10– 14, delivered in First Nations communities in four provinces including BC (Culture and Mental Health Research Unit, 2017a). The program uses Indige- nous values to enhance communication and social skills and tools for healthy lifestyles. Early find- ings of an adaptation of the intervention in the U.S. indicated that the program delayed onset of drinking for 10- and 11-year-old girls, but had no effect on older youth or for those who had tried alcohol before the intervention (Whitbeck, Hoyt, & Stubben, n.d.). Evaluation results of the current Canadian programs are not yet available (Culture and Mental Health Research Unit, 2017b). One example is the ‘Nimi Icinohabi Substance Abuse Prevention program’ in Alberta for young people in Grades 3 to 9 in Alexis Nakota Sioux Na- tion School. The program was adapted from the Botvin Life Skills Training (LST) to include cultural elements related to beliefs, values, language, and visual imagery. The program was found to change youth’s alcohol refusal skills, self-beliefs, and knowledge about consequences of alcohol use (Government of Canada, 2016). However, it is unclear if it affected consumption. A school-based program that used a culturally adapted life skills approach was evaluated with an RCT that included 1,396 Native American youth in Grades 3 to 5. The program sessions were led by community leaders and included older peers demonstrating skills to resist alcohol. Participants were also assigned homework to gather infor- mation and testimonials on drug and alcohol use from their family and community. At 32-month follow-up, youth who had participated in the program were less likely to have four or more drinks a week than participants in the control group (Schinke et al., 2000). The intervention had a stronger effect on boys’ alcohol use, but overall alcohol use rates also decreased for girls (Schinke et al., 2000). Although not specifically identified as an ‘Aborig- inal Ways Tried and True intervention,’ an adap- tation of the Nimi Icinohabi Program was carried out with the Maskwacis community in Alberta. A three year pre- and post evaluation of the program with elementary and junior high school students indicated that students’ knowledge and attitudes about alcohol improved as did cultural knowledge and life skills including those around substance use refusal (Baydala et al., 2016). The 29 McCreary Centre Society

  31. Brief Alcohol Interventions Motivational interviewing Motivational interviewing is a non-judgemental, client-centred counselling style that helps clients explore discrepancies between goals and be- haviours and supports efficacy to change. Brief Alcohol Interventions (BAIs) aim to motivate and provide resources to moderate participants’ alcohol use. The interventions are short in length (one to five sessions) and typically include a discussion of alcohol consumption, feedback on level of use, comparison to norms, information on potential harms, and goal setting (Tanner-Smith & Lipsey, 2015). While the effectiveness of BAI’s has not yet been determined, there is some evidence that the motivational interviewing component of those interventions can be effective for some older girls across different cultures (Mason et al., 2011). For example, an RCT of 287 young women (average age of 17–18) who participated in a 90-minute motivational interview to address female specific reasons for drinking found that five weeks after the session, there were significant reductions in binge drinking (four or more drinks) among participants who had a family history of alcohol abuse in comparison to the control group, but no significant effect on young women with no family history of alcohol abuse (LaBrie et al., 2009). Three large scale systematic reviews and me- ta-analyses concluded they could be effective (Hennessy & Tanner-Smith, 2015; O’Donnell et al., 2014; Tanner-Smith & Lipsey 2015), as there was a small but significant effect on alcohol-re- lated problems such a going to school drunk and experiencing negative consequences of alcohol use (Tanner-Smith & Lipsey 2015). There was a reduction in days on which alcohol was consumed when BAIs were delivered to individuals, but there were no such effects for BAIs delivered in a group setting (Hennessy & Tanner-Smith, 2015). Motivational interviewing might be particularly effective when young women have personally ex- perienced alcohol-related harms. The ‘HaLT’ pro- gram is widely-used in German emergency rooms with youth admitted for alcohol intoxication. It uses motivational interviewing to address drink- ing motives through short exercises which teach alternative ways to fulfill the role alcohol plays in a young person’s life. At four-week follow-up, large effects were seen for female patients aged 14–16 who had received the motive-tailored interven- tion as they had decreased drinking frequency, heavy sessional drinking, and drunkenness to a greater degree than males (Wurdak et al., 2016). A similar smaller scale motivational interviewing intervention in the U.S. was offered to male and female youth aged 18 and 19 who were treated in the ER for an alcohol-related event. The program was effective in reducing rates of drinking and driving, alcohol-related problems, and injuries at six-month follow-up compared to youth in a stan- dard care control group (Monti et al., 1999). Brief interventions can be effective for those who are experiencing harmful alcohol use (Babor, 2010; World Health Organization [WHO], 2009). However, there is insufficient evidence to draw conclusions for subpopulations, including females and adolescents (O’Donnell et al., 2014) and a recent Cochrane review concluded that it was not yet possible to determine the effectiveness of BAIs among high school students (Carney, Myers, Louw, & Okwundu, 2016). 30 Strategies to reduce risky alcohol use among underage girls: An evidence review

  32. Mentorship Mentorship did appear as a component in a successful intervention that also included school- based life-skills curriculum and volunteering in the community. Evaluation results from 400 U.S. stu- dents in Grade 6 reported significant reductions in alcohol use; lower levels of impulsive behaviour; and higher levels of self-confidence, self-control, family and school attachment, and cooperation among youth who participated in the mentor- ship program, which were not seen if youth only participated in the other components (Aseltine, Dupre, & Lamlein, 2000). Mentorship interventions entail a caring individ- ual providing support, guidance, and consistency to help develop youth’s strengths and set them on a healthy trajectory (Thomas, Lorenzetti, & Spragin, 2013). A systematic review found a mod- erate effect of mentorship on adolescent alcohol use but concluded there have been insufficient well-designed studies to make a firm conclusion (Thomas et al., 2013). However, other studies have found no evidence of mentorship programs reducing risky alcohol use in underage girls (Bodin & Leifman, 2011; Herrera, Grossman, Kauh, & McMaken, 2011). Although structured mentorship programs were not generally effective, interventions which included role models for non-use of alcohol (Longshore et al., 2007), enhanced parents as role models (Schinke, Fang, & Cole, 2009; Weiss & Nicholson, 1998; West et al., 2008), and included female health educators (Longshore et al., 2007) proved to be more successful with girls than boys. 31 McCreary Centre Society

  33. Multi-component interventions There does not appear to be a single intervention type that is universally successful at addressing underage girls’ risky alcohol use. Multi-compo- nent interventions are ones that involve multiple settings such as family and school (Foxcroft & Tsertsvadze, 2011b). The evidence for multi-com- ponent interventions was also mixed. The Na- tional Research Council and Institute of Medicine (2004) found a combination of school curriculum, parental involvement, and community mobiliza- tion may be effective. A Cochrane Review noted that although multi-component interventions were effective, there was insufficient evidence to indicate they were any more effective for prevent- ing risky alcohol use among school-aged youth than single component interventions (Foxcroft & Tsertsvadze, 2011b). The authors of this latter review noted that reporting methods of studies generally do not allow for an analysis to deter- mine what intervention content or ingredients differentiate effective from ineffective programs. There is some evidence from Europe that in- terventions focusing on community education, regulation, and enforcement have promising results for underage girls. In eight participating Dutch communities, interventions were aimed at encouraging moderation among 10- to 19-year- old youth, and activities were carried out in multiple settings including school, home, recre- ational centres, and bars. Intervention activities varied across communities and may have included parent focused activities such as parenting skills and education about the effects of alcohol on the adolescent brain; health education to youth by school nurses; and instructions to community es- tablishments that served alcohol. Younger females and males (aged 13–14) were less likely to engage in heavy sessional drinking compared to their peers in non-participating regions, although there was no effect on youth aged 15–16 (Jansen et al., 2016). An NNT analysis could not be performed for girls specifically, however, for the group of 13- and 14-year-old males and females, one case of heavy sessional drinking could be prevented if 21 individuals received the intervention. Interventions which include a community compo- nent often involve education of community mem- bers (e.g. retailers, bar staff, club members, and neighbourhood groups) combined with efforts to reduce underage alcohol availability through po- lice enforcement, educating adults not to provide alcohol to underage youth, and enforcing the legal age for alcohol sales. 32 Strategies to reduce risky alcohol use among underage girls: An evidence review

  34. A similar Swedish intervention targeted 15- to 16-year-olds in six communities that carried out community-based information and media advoca- cy programs, school-based social and emotional learning interventions, integrated motivational in- terviewing into school health services, and trained police around availability and enforcement. Most communities also offered programs to educate and support parents to maintain a restrictive atti- tude towards adolescent alcohol use and trained bar owners and servers in responsible beverage service programs (Hallgren & Andreasson, 2013). Repeated cross-sectional surveys were conducted with over 8,000 youth until they were 19 years old. Results showed the intervention was most effective among females aged 15–16 years old as binge drinking (number of drinks not specified) was reduced by 20% in the trial communities com- pared to 10% in the control communities. How- ever, there were no significant differences among female youth aged 17 and older and no differenc- es overall between trial and control communities at five-year follow-up (Hallgren & Andreasson, 2013). Langford et al., 2014). However, when delivered in Croatia, an RCT of almost 2,000 Grade 6 students found that at three-year follow-up, the interven- tion was effective in reducing intention to use alcohol and frequency of use among girls but not boys (West et al., 2008). The Croatian version of the project was adapted to address the traumatic effect the recent war may have had on young peo- ple and included a youth leadership component where youth participants conducted surveys with parents and students; interviewed representatives from the police, medical community, and local shopkeepers; and wrote letters to the President, which they presented to the community (West et al., 2008). Communities That Care (CTC) is a communi- ty-based multi-component delivery model that may help to build resilience, as its aim is to foster protective factors and reduce problem behaviours (Communities That Care, 2017). The program involves five phases centered around communi- ty consultation and partnership development, assessment of youth risks and strengths, defining outcomes and selecting evidence-based policies and programs to address community needs, and implementation and evaluation. Another Swedish multi-component intervention included student, parent, and community edu- cation as well as city and school policies about alcohol availability. The intervention was associ- ated with a significant reduction in heavy drinking and spirit consumption among 15–to 16-year-old girls and boys in the participating community at four year follow-up, which was associated with a decrease in alcohol-related accidents and violence (Stafstrom & Ostergren, 2008). In this case, one case of alcohol-related violence could be prevent- ed if 15 individuals received the intervention. The model has shown some promising results with respect to alcohol use. For example, a ran- domized trial involving 24 communities in seven states in the U.S. began with a cohort of fifth graders. By Grade 8, students in CTC communities were less likely to have initiated alcohol use than those in control communities and rates of past month alcohol use and heavy sessional drinking were lower (Hawkins et al., 2009). The results for abstinence from alcohol were sustained to 12th grade although past month substance use did not differ (Hawkins, Oesterle, Brown, Abbott, & Catalano, 2014). Furthermore, results from the trial indicated that at the time the students were in Grade 8, higher levels of protective factors were found in CTC communities compared to controls including higher levels of prosocial involvement, prosocial peer interaction, and social skills (Kim, Gloppen, Rhew, Oesterle, & Hawkins, 2015). A North American example of a multi-component intervention was ‘Project Northlands.’ It was conducted over three years in two large U.S. cities and combined school curricula, parent engage- ment, youth-led community service projects, and a community task force to reduce underage alcohol availability. It was not effective in reducing risky alcohol use among a general population of youth in the U.S. (Foxcroft & Tsertsvadze, 2011b; 33 McCreary Centre Society

  35. Interventions in British Columbia PLEA Community Services provides residential substance use treatment services for BC youth. Females entering treatment reported very high rates of negative consequences of their alcohol use. During their stay in PLEA they received indi- vidualized supports, attended school and lived in a family care home. Analyses of over 60 female youth six months after discharge indicated that girls who felt they had gained important skills, had been supported to develop healthier relationships with family, had positive relationships with PLEA staff, and had positive peer relationships during their time in treatment reported lower rates of alcohol use (McCreary Centre Society, 2012). The Communities That Care model is being used in Williams Lake, BC where they completed their risk and strengths assessment and found early onset of alcohol and drug use to be a key priori- ty (Social Planning Council of Williams Lake and Area, 2010). It is also the model being used in Squamish where they are aiming to prevent a handful of health risk behaviours including sub- stance abuse (Communities that Care Squamish, 2017b). They are using the Strengthening Families intervention as well as a multicomponent commu- nity-based intervention to reduce high-risk drink- ing (Communities that Care Squamish, 2017a). The School Age Children and Youth (SACY) Sub- stance Use Prevention Initiative is a multi-com- ponent intervention for Vancouver students in Grades 8–9, which aims to strengthen relation- ships to family, school, and community (Boute, 2013). Two key programs of SACY are S.T.E.P., a three-day alternative to suspension program for regular substance users; and capacity cafes, where youth and parents are brought together to talk about their lives. The SACY programs incorporate multi-lingual meetings for Asian parents. There are also Aboriginal Capacity Cafes which empha- size intergenerational kinship and bring together family members, Elders, and other adults to hear about lived experiences of Aboriginal youth in their community, learn about what supports are being provided, and about traditional ways which can further support youth (Arbor Educational & Clinical Consulting Inc., 2012). A four-year eval- uation of SACY completed in 2013 found 57% of participating youth set personal limits for their alcohol use and 20% reported using less alcohol (Boute, 2013). Intervention summary The most common interventions examined in this review were interactive, provided information about alcohol use, taught skills to resist social pressures to drink, and taught stress management skills. However, limiting a program to the common components of knowledge and skill building may not be the most effective for girls, and a broader more comprehensive approach may be necessary. Overall, family-based interventions and inter- ventions targeted at certain personality charac- teristics appeared to be more effective for girls than some of the other interventions. However, most programs that were effective for girls were focused on younger girls or only produced sig- nificant results for younger girls. Targeted inter- ventions which examined older girls’ motivations for drinking and exposed them to motivational interviewing techniques appeared to be more effective for them and for those who had experi- enced alcohol-related harms. 34 Strategies to reduce risky alcohol use among underage girls: An evidence review

  36. Interventions which appeared to be more effective in preventing or reducing risky alcohol use for girls included components which: 9 Engaged caregivers and strengthened relationships with families, especially youth’s mothers. 9 Increased knowledge about alcohol effects, consequences of use, misperceptions, low risk drinking guidelines, benefits of non-use, and sources of help for girls. 9 Addressed the reasons young people use alcohol. 9 Took a harm minimization approach that included education on accessing help for someone who is intoxicated. 9 Built skills for resisting social influences to use alcohol. 9 Developed self-efficacy and self-esteem. 9 Provided gender-specific information about alcohol and its effects. 9 Were tailored to address youth’s specific needs, culture, age, and experience with alcohol, and the needs of their families. 9 Identified typical situations when alcohol use occurred and offered alternatives to alcohol use. Less successful interventions: 9 Had a single component. 9 Focused on abstinence. 9 Focused exclusively on the risks and consequences of alcohol use. 9 Had facilitator or program characteristics that were not a good fit for the youth. 9 For school-based programs, were not integrated well into class syllabus. 35 McCreary Centre Society

  37. Effective policies for reducing harmful underage alcohol use among females We used a few key publications to review effective policies to reduce harmful underage alcohol use. For example, evidence-based alcohol policy has been a focus of the World Health Organization (WHO) in Europe for over 40 years. They have sponsored several reviews, including Evidence for the effectiveness and cost-effectiveness of inter- ventions to reduce alcohol-related harm (WHO, 2009), which was based on systematic reviews, reviews of systematic reviews, and meta-analyses of evidence primarily from North America and Northern Europe. However, there is some evidence from tobacco research that when health warning labels include graphics or pictures that are updated regularly, these may be effective in preventing non-users from starting to use during adolescence (Bach, 2016). On the other hand, policies regarding pricing, availability of alcohol, advertising, and drinking and driving have been considered best practices in reducing consumption and alcohol-related harms (Babor, 2010; WHO, 2009). For example, alcohol control policies may reduce sexual violence which is of particular concern for young women (Atlantic Collaborative on Injury Prevention, 2014). Some of the policy research had specific evidence for underage youth, and these are discussed in the sections that follow. (Nine papers that contained gender-specific data for policies to reduce risky alcohol use are detailed in Appendix B.) Another key publication was a summary of the book Alcohol:No ordinary commodity—Research and Public Policy (Second edition) by Babor (2010). An international group of scientists conducted a critical review and rated various alcohol policies based on the effectiveness evidence, the amount of scientific evidence, and availability of evidence across countries, and subsequently identified best practices for policies to reduce alcohol-related harms. A review of provincial programs and policies in Canada aimed at reducing alcohol-related harms and costs ranked the provinces on 10 policy dimensions based on scope (population reach) and effectiveness. These included alcohol pricing, availability, drinking and driving, marketing and advertising, legal drinking age, and warning labels and signs (Giesbrecht et al., 2016). BC ranked among the top provinces overall, as well as on the majority of the individual dimensions, although no province ranked higher than 50% when compared to ideal policies. BC did receive a bottom ranking for their alcohol control system (e.g., type of retail system, availability of alcohol through alternative means such as online, whether a government ministry has responsibility over alcohol control, etc.) and a below average ranking for the pricing policy dimension and warning labels (Giesbrecht et al., 2016). These reviews covered population level poli- cies including education, drinking and driving, alcohol availability, alcohol advertising, alcohol pricing, and drinking context or environment. Knowledge-based interventions have already been addressed elsewhere in this report. More generally, education and information strategies including school-based education, public informa- tion campaigns, and health warnings have been found to be ineffective in affecting behaviour or alcohol-related harm, although they may increase knowledge, change attitudes, and raise awareness (Babor, 2010; WHO, 2009). For example, according to a Centre for Addictions Research of BC (CAR- BC) review of international alcohol policy strate- gies, alcohol warning labels on bottles and cans are noticed and understood by the majority of adolescents but have little effect on their drinking behaviour and related harms (Stockwell, 2006). 36 Strategies to reduce risky alcohol use among underage girls: An evidence review

  38. School policies and regulations Survey, found that the presence of anti-homopho- bia policies in BC schools was associated with re- ductions in regular heavy sessional drinking in the past month among both heterosexual girls and boys (NNT=23-26) and multiple harmful conse- quences in the past year among heterosexual girls (NNT=21). The review indicated that in a typical BC school of 1,000 students, such supportive poli- cies would result in 37 fewer students engaging in heavy sessional drinking on six or more days in the past month, and 21 fewer students with problem substance use (Saewyc et al., 2016). School policies and regulations about alcohol use at school and issuing fines for alcohol-related rule breaking did not appear to be effective in reduc- ing risky alcohol use for girls (Smit et al., 2003; Jansen et al., 2016). However, policies which help to create a more inclusive and welcoming school environment could be effective. A recent evidence review provided evidence that LGBTQ-inclusive policies in schools contribute to better health among both sexual minority and heterosexual students (Saewyc, Poon, Kovaleva, Tourand, & Smith, 2016). For example, Konishi and colleagues (2013), using data from the BC Adolescent Health 37 McCreary Centre Society

  39. Minimum legal drinking age laws Despite the positive impacts of raising the MLDA, the results for girls are not as consistent as they are for boys. For example, heavy sessional drink- ing rates for youth generally decreased in the U.S. following the establishment of an MLDA of age 21 (Grucza, Norberg, & Bierut, 2009). However, according to data from the National Survey on Drug Use and Health, the relative risk for heavy sessional drinking did not change among females aged 12 to 20 and increased for underage minori- ty females as well as for college females aged 21 to 23 (Grucza et al., 2009). One policy aimed at reducing alcohol-related harms among youth is the establishment of a minimum legal drinking age (MLDA) which deter- mines the age at which youth are legally able to purchase, possess, and consume alcohol. Most of the evidence regarding MLDA policy comes from the U.S. For example, a review of over 130 studies conducted between 1960 and 2000 (a period when the MLDA was lowered and raised in the U.S.), concluded there was strong evidence that the MLDA was inversely related to alcohol consumption and motor vehicle crashes (Wage- naar & Toomey, 2002). Specifically, among studies identified as being of higher quality (i.e., used either longitudinal designs, comparison groups, probability sampling, or a census), a third showed evidence of reduced alcohol consumption and 58% showed a reduction of motor vehicle crashes. Lowering the drinking age from 20 to 18 years in New Zealand was associated with an increase in alcohol-related motor vehicle accidents for both males and females, but increases in hospitaliza- tions for traffic crash injuries were found only for males (Kypri et al., 2006). In Australia, having legal access to alcohol at age 18 was associated with increased hospitalizations for males and females (Lindo, Siminski, & Yerokhin, 2016). More recent U.S. reviews of MLDA laws confirm that increasing the minimum age to 21 was asso- ciated with decreases in alcohol consumption, in- cluding heavy use, underage consumption, and al- cohol-related motor vehicle accidents. In addition, they suggest long-term impacts such as heavy drinking and alcohol dependence in adulthood as well as increased risk of adverse birth outcomes, suicide, and violence for women (McCartt, Hell- inga, & Kirley, 2010; Dejong & Blanchette, 2014). Several recent Canadian studies have looked at gender differences in relation to MLDA. For ex- ample, Callaghan and colleagues (2016) found an overall increase in alcohol-impaired driving inci- dents when young people were slightly below the legal drinking age compared to when they were slightly above it. For females, alcohol-impaired driving incidents rose by 40% in provinces with an MLDA of 18. However, there was no such increase for females in provinces with a legal drinking age of 19 years, such as BC (Callaghan, Gatley, Sanch- es, Ashbridge, et al., 2016). Nationally, there were increases in severe and fatal alcohol-related collisions for males older than the MLDA compared to younger, but not for females. The stronger effects for males may be at- tributable to males being more likely than females to engage in hazardous drinking and drinking and driving (Callaghan et al., 2016; Callaghan, Sanch- es, Gatley, & Stockwell, 2014). 38 Strategies to reduce risky alcohol use among underage girls: An evidence review

  40. In Canada, inpatient hospitalization rates for alco- hol use disorders or alcohol poisoning were higher for both males (17%) and females (21%) who were slightly older than the MLDA compared to those who were slightly younger (Callaghan, Sanches, & Gatley, 2013). However, at the population-level males slightly older than the MLDA had increases in mortality compared to those slightly younger than the MLDA for all causes, which was not the case for females (Callaghan et al., 2014). A U.S. study found zero tolerance laws decreased heavy episodic drinking among young men aged 18–20 by 13% and there were corresponding increases in lighter alcohol use (Carpenter, 2004). However, these laws were not associated with reductions in number of drinks or heavy episodic drinking among young women. For young women, there was a marginally significant trend (p < 0.1) towards increased likelihood of self-reported drunk driving, though this may have been because of changing perceptions on drunk driving after en- actment of zero tolerance laws (Carpenter, 2004). It has been suggested that the legal drinking age in BC should be raised to 21 years of age (Thompson, Stockwell, Vallance, Giesbrecht, & Wettlaufer, 2013), and there is evidence for higher minimum legal drinking ages for youth in general (Giesbrecht et al., 2016). However, it may be that MLDA policies may be less effective for underage girls than for boys in reducing risky drinking and related harms. A Canadian study suggested that decreases in youth drunk driving may be attributable to de- clining trends in alcohol use and changing cultural norms rather than policy changes (Carpenter, 2006). For example, there was a 35% decrease in drunk driving among 16 and 17 year olds in Ontar- io after a graduated driver licensing program was instituted in 1994. However, these decreases were also found among older and younger groups who were not impacted by the policy changes (Carpen- ter, 2006). Zero tolerance drink driving laws and graduated licenses An evaluation of the graduated licensing program in BC found a significantly higher proportion of al- cohol-involved crashes among graduated licencing program participants as compared to before the change was enacted, but this may have been due to increased reporting by police officers (Wiggins, 2004). Zero tolerance laws make it illegal for drivers to have measurable amounts of alcohol in their blood. In places where this is an element of gradu- ated driver licensing programs for young or inex- perienced drivers, it can also mean returning to the beginning of their graduated licensing stage. Some U.S. states have zero tolerance policies for all drivers under the age of 21 (Carpenter, 2004). Although this may also be connected to MLDA laws, there is some evidence that zero tolerance laws can reduce the frequency of alcohol-related crashes. For example, enactment of zero tolerance laws among young and inexperienced drivers in the U.S. was associated with a 9–24% decline in fatal crashes and an 11% decline in crashes where it was believed the driver had been drinking (Shults et al., 2001). Additionally, enactment of a lower blood alcohol limit for driving in Japan was associated with a 64% reduction in alcohol-related crashes among youth between the ages of 16 and 19 (Desapriya, Pike, Subzwari, Scime, & Shimizu, 2007). 39 McCreary Centre Society

  41. Price increases A study of minimum pricing in BC over a 20-year period estimated that 10% increases in minimum prices reduced consumption of spirits by 6.8%, alcoholic sodas and packaged cider by 13.9% (although marginally significant), beer by 1.5%, wine by 8.9%, and all alcoholic drinks by 3.4% (Stockwell, 2011). Although these estimates were not specific to underage youth, the estimated reductions could be pertinent to girls as the 2013 BC Adolescent Health Survey indicated that a third of female drinkers binge drank on the Saturday before taking the survey (Smith et al., 2015a); and 32% of female youth who had tried alcohol con- sumed spirits on the Saturday before the survey, 27% drank coolers, 18% consumed beer, and 7% drank wine (McCreary Centre Society, 2013; Smith et al., 2015a). Evidence suggests that alcohol consumption in adults is associated with youth alcohol consump- tion and therefore policies aimed at adults (e.g., pricing, hours of sale, etc.) also affect underage alcohol consumption (SAMHSA, 2015). Price increases and taxation can be one of the most effective ways to reduce alcohol-related harms (Babor, 2010). Meier and colleagues’ (2008) systematic review concluded that price increases do reduce con- sumption and related harms, including for those under age 18; and young drinkers, binge drinkers (drinking twice the recommended daily allow- ance), and harmful drinkers are the most likely to consume cheaper drinks (Meier et al., 2008). One study with gender-disaggregated data was based on a nationally representative sample of over 17,000 college students in the U.S. (Chaloupka & Wechsler, 1996). The study found higher beer prices would slightly reduce drinking, drinking fre- quency, and binge drinking in underage females (four or more drinks) but not males (five or more drinks) (Chaloupka & Wechsler, 1996). In addition, Kenkel (1993, cited in National Research Council and Institute of Medicine, 2004) concluded that a10% increase in pricewould reduce the risk of drinking and driving by 21% among young females and 13% among young males. In 2010, Saskatchewan introduced pricing policies based on type of beverage and alcohol content leading to an increase in prices on 9% of alcoholic drinks (Stockwell, Auld, Zhao, & Martin, 2012). For the population aged 15 and older, a 10% increase in minimum price was associated with an 8.4% decrease in total alcohol consumption. Greatest reductions in consumption were seen in higher al- cohol content beer, wine, and cocktails which also had the greatest price increases. No effect was seen for higher strength coolers and spirits, per- haps due to the possibility that consumers were switching from higher strength beers. The higher effect sizes in Saskatchewan overall compared to BC were attributed to the minimum pricing policy applying to most beverage types in Saskatchewan as opposed to BC’s focus on spirits. Minimum pricing for alcohol has been adopted by several Canadian provinces. Some of these pricing policies are designed to encourage the consumption of lower strength alcohol products (e.g., higher minimum prices for stronger alcohol content). However, those which have imposed flat rates per litre of alcohol may have the opposite effect (National Alcohol Strategy Advisory Com- mittee, 2015). Minimum pricing policies may have an impact on higher risk drinkers who experience harms as well as young drinkers (Centre for Addic- tions Research of BC, n.d.). 40 Strategies to reduce risky alcohol use among underage girls: An evidence review

  42. Minimum unit pricing has also been implement- ed in parts of the UK (Stockwell, 2014). As in Saskatchewan, minimum unit pricing is estab- lished based on alcohol content and this creates a uniform price “floor” when measured as price per standard unit of alcohol. Evaluation data has found consistent evidence for reducing alcohol-re- lated harm, and that these policies would affect harmful drinkers more than moderate drinkers (University of Sheffield, 2017). For example, a 10% increase in minimum price was associated with a 9% decrease in acute hospital admissions attribut- ed to alcohol and an estimated 9% decrease in chronic admissions two years later (Stockwell et al., 2013). Additionally, most female adolescents who use alcohol get it from informal or social connections such as family members, rather than through commercial outlets (Harrison, Fulkerson & Park, 2000; Hearst, Fulkerson, Maldonado-Molina, Perry, & Komro, 2007; Treno, Ponicki, Remer, & Gruenewald, 2008). A longitudinal study in California noted that restricting alcohol outlet density in certain zip codes was ineffective among adolescents if those living in low alcohol density neighbourhoods were in medium income house- holds and had a friend with a car (Chen, Grube, & Gruenewald, 2010). However, it has been suggested that having a high number of commercial alcohol outlets available in a neighbourhood may increase the perception of normative alcohol use, especially for girls (Kunt- sche, Kuendig, & Gmel 2008; Milam, Johnson, Furr-Holden, & Bradshaw, 2016), and regulating commercial access to alcohol may still therefore have some benefits for underage girls. For ex- ample, a study with over 10,000 Australian high school students found that a higher density of off-premises alcohol sales outlets was associat- ed with parents supplying alcohol to their male and female children (Rowland, Toumbourou, & Livingston, 2015). In addition, a high density of bars was associated with lower levels of parental monitoring, which was subsequently associat- ed with higher levels of risky behaviours among youth such as substance use (Freisthler, Brynes, & Gruenewald, 2009). Restricting commercial availability Stockwell and colleagues (2011) noted a decline in adult consumption of alcohol in BC until 2002 when partial privatization of alcohol sales coin- cided with increasing rates of per capita alcohol consumption. High private liquor store density was associated with an increased number of alcohol-related deaths in BC (Stockwell et al., 2011). A study from the U.S. found fewer instanc- es of heavy sessional drinking among high school students from states with a retail monopoly over spirit and wine in comparison to students from non-monopoly states (Miller, Snowden, Birckmay- er, & Hendrie, 2006). Additionally, re-monopoliza- tion of beer in Sweden decreased rates of alco- holism, alcohol intoxication, and alcohol psychosis among youth aged 10 to 19 years old by 20% (Campbell et al., 2009). As policies related to altering the drinking con- text or environment (e.g., training bar staff and alcohol outlets to ensure responsible beverage service practices) are largely dependent on strong enforcement, they may not be a cost-effective ap- proach to reducing alcohol-related harms (Babor, 2010; WHO, 2009). 41 McCreary Centre Society

  43. Two European interventions that reduced alcohol access for youth through responsible beverage services programs in combination with a mass me- dia campaign on radio, posters, and TV commer- cials were effective in reducing alcohol use among girls aged 13 to 15, but not among older girls aged 16 to 19 (Hallgren & Andreasson, 2013; Jansen et al., 2016). One of the programs worked closely with journalists to increase the number of media reports describing alcohol-related problems and harms occurring at the local level in the interven- tion towns and was effective in decreasing the proportion of parents who offered alcohol to their children (Hallgren & Andreasson, 2013). In the U.S., the alcohol industry has voluntary exposure guidelines restricting alcohol advertising to youth whereby advertisements are not to occur when more than 30% of the audience is expected to be underage (Center on Alcohol Marketing and Youth, 2010). Even stricter guidelines have been recommended to drop the cut-off to 15%, which is reflective of the proportion of underage youth aged 12 to 20 in the population (Center on Alco- hol Marketing and Youth, 2010). According to the latter guideline, non-compliance would indicate youth “overexposure” to alcohol advertising. Noel, Babor, and Robaina (2016) conducted a review of studies evaluating content and exposure of alcohol advertising. Nineteen content studies in 19 countries found violations of content codes, most commonly including associations of alcohol with social or sexual success and those intended to protect youth. Additional content analyses revealed themes potentially appealing to youth including humour, relaxation, cartoons, animals, sports, and social and financial success. With re- spect to exposure research, 79% of the 57 studies reviewed from 18 countries reported youth expo- sure to advertising or marketing. Finally, although there were no results specific to girls, a systematic review found that alcohol bans were only effective in isolated communities. Peo- ple in less isolated communities travelled to areas of higher alcohol availability, which increased the likelihood of alcohol-related injuries (Campbell et al., 2009). Reducing exposure to advertising/media For example, much research from the Johns Hopkins Center for Alcohol Marketing for Youth in- dicates that the U.S. alcohol industry has failed to follow these guidelines in radio (Center on Alcohol Marketing and Youth, 2011) and television (Center on Alcohol Marketing and Youth, 2010b) and less so in magazines (Center on Alcohol Marketing and Youth, 2010a). For example, a study of television programs popular with youth (aged 12–21) in 25 major metropolitan markets in the U.S. found 24% of the advertisements in 2010 violated the 30% cut-off guideline and 35% violated the 15% cut- off; and if these were to be eliminated, alcohol ad- vertising on these programs would drop by a third and 54%, respectively (Jernigan, Ross, Ostroff, McKnight-Eily, & Brewer, 2013). Cable television accounts for most exposure violations (Ross, Brewer, & Jernigan, 2016), and youth exposure to alcohol advertising on cable television increased by 3% from 2014 to 2015 with 1 out of 8 adver- tisements not complying with regulations (Ross, Henehan, Sims, & Jernigan, 2016). In 1996, the Canadian Radio-television and Telecommunications Commission introduced regulations for the content of radio and television advertisements. Among these are restrictions that messages not be directed at non-drinkers or to underage youth, involve any association of alcohol with youth or youth activities, include any en- dorsement by youth role models, portray alcohol as being necessary for the enjoyment of life or an activity or event, or imply that alcohol is associat- ed with social, personal, professional, or athletic success. These regulations apply to all types of advertising in BC including print, web, and promo- tional materials (Advertising Standards of Canada, 2016). 42 Strategies to reduce risky alcohol use among underage girls: An evidence review

  44. Despite recommendations to reduce or ban alco- hol advertising directed at youth (e.g., Giesbrecht et al., 2013), young women are increasingly targeted by alcohol companies. This includes the marketing of specific products such as sweeter beverages (CASA, 2003; Ross et al., 2014) and distilled spirits and wine (Centre on Alcohol Mar- keting and Youth, 2011). Girls and young women are also susceptible to the sexualized content of marketing: There is a lack of consensus about the degree to which reducing exposure to alcohol advertising is directly linked with reduced consumption among youth (Giesbrecht et al., 2013). Nelson’s (2010) review of 20 longitudinal studies of youth alcohol use and advertising noted several methodological shortcomings, and concluded that causal interpre- tations are not warranted and that results of such studies should not be used as support for bans on alcohol advertising. Likewise, Grube (2004) conducted a review and concluded that there was little evidence that advertising influences under- age alcohol use and related harms. Alcohol marketing frequently depicts sexual content, sexism, and messages intended to promote gender norms of masculinity and femininity. The increased likelihood of social and sexual success in relation to alcohol use is a common underlying theme. Although hypersexualized and gendered marketing do not cause sexual violence, they intersect with broader social and cultural norms that sup- port sexualized violence. However, a more recent review of 13 longitudinal studies from the U.S., Europe, and New Zealand looked at the effect of alcohol advertising and me- dia exposure on underage alcohol use. All but one of the studies found that exposure was linked with initiation of alcohol use among male and female non-drinkers and increased consumption among drinkers. Further, over half the studies indicated a dose-response relation, with increased exposure associated with increased consumption (Ander- son, de Bruijn, Angus, Gordon, & Hastings, 2009). Atlantic Collaborative on Injury Prevention, 2014, p. 11 43 McCreary Centre Society

  45. A more recent study in Los Angeles followed 3,890 Grade 7 students over four years. Exposure to TV advertising was measured in the first year followed by subsequent measurement of alco- hol outcomes. Results indicated that advertising exposure and liking of ads in Grade 7 was predic- tive of increased alcohol use and alcohol-related harms in Grade 10 for both boys and girls, even after controlling for factors such as problems at baseline, age, exposure to peers or close adult drinking, and parents’ jobs and education level (Grenard, Clyde, Dent, & Stacy, 2013). Nevertheless, the current review of the published literature suggests that policies aimed at changing the environment have the potential to affect alco- hol use and related harms among underage girls. 9 Making school environments more inclusive may reduce heavy sessional drinking and problematic drinking (i.e., experiencing nega- tive consequences as a result of alcohol use). 9 Increasing the price of alcohol may reduce consumption and alcohol-related harms such as drinking and driving. 9 Minimum pricing policies could reduce consumption if applied to types of alcohol beverages favoured by girls. Policy summary Generally speaking, pricing, availability, advertis- ing, and drinking and driving regulations are best practices for reducing risky alcohol use. Howev- er, little research was found that addressed the effectiveness of policies specifically for underage girls and the available evidence was somewhat inconsistent. Furthermore, it was difficult to tease out which laws and policies may be effective in isolation, and appeared more likely that a combi- nation of approaches would be necessary to make a difference. It was also difficult to ascertain why different laws with similar objectives would have differential effects on girls. For example, although both policies aim to restrict access to alcohol, increases in minimum legal drinking age were not consistently associated with reductions in drinking and driving, but the evidence indicated that price increases could reduce this risky behaviour. 9 Restricting the commercial availability of alco- hol through de-privatization or re-monopoli- zation may help to reduce risky drinking and alcohol-related harms. 9 Reducing exposure to alcohol advertising may reduce alcohol initiation and decrease con- sumption and related harms including sexual violence. 44 Strategies to reduce risky alcohol use among underage girls: An evidence review

  46. Messaging to reduce harmful alcohol use among females Messages that have been shown to be effective at reducing risky underage alcohol use include emphasizing alternative safer ways to achieve the benefits of alcohol (Goldberg, Halpern-Felsher, & Millstein, 2002); emphasizing the negative social consequences of heavy drinking, combined with messaging about the positive health consequenc- es of not drinking (Kingsbury et al., 2015); pre- senting both the positives and negatives of drink- ing (Cornelis, Cauberghe, & De Pelsmacker, 2014); and ensuring messaging is realistic, relatable, and gendered including female-specific statistics and consideration of reasons why young women drink (Canadian Women’s Health Network, 2014). Offering targeted messaging for either binge drinkers (in excess of four drinks for women or five drinks for men) or more moderate drinkers (fewer than four drinks for women or five drinks for men) may be more effective than offering one message for both (Lee & Chen, 2013). For example, negative slogans such as “Don’t drink” increased perceptions of risks associated with heavy drinking and increased underage moderate drinkers’ intentions to reduce their drinking, but had the opposite effect on underage male and female binge drinkers who demonstrated a lower intention to change their drinking behaviour (Lee & Chen, 2013). Presenting a balanced picture of the effects of drinking alcohol was associated with a lower intention to binge drink among male and female adolescents who had already binge drank than just presenting information on the harms of drinking (Cornelis et al., 2014). A focus on shaming or blaming young women for alcohol-related sexual violence, although com- mon, was reported to be ineffective at reducing risky alcohol use by young women (age 20–25) in Ontario. These young women suggested that women respond negatively to moralized media that engages in shock-tactics and provides unhelp- ful ‘facts’. More effective messaging portrayed examples of people they could relate to, offered realistic scenarios, shared gendered information, and provided helpful advice to mitigate risks (Ca- nadian Women’s Health Network, 2014). In summary, the evidence suggests that alcohol messaging targeting underage girls is less effective if it focuses on health consequences and risks, whereas focusing on social consequences and the benefits of moderated drinking are more effective. 45 McCreary Centre Society

  47. Media review methodology Summary of findings Having reviewed the literature around messaging, six female members of McCreary’s Youth Research Academy looked at media messaging around risky alcohol use among girls and young women. They reviewed 29 different media messages (see Appendix D) and provided their assessment of what aspects of the messages were effective or ineffective. While the literature shows that focusing solely on the risks associated with alcohol use is an ineffec- tive way of framing preventative messages about harmful alcohol use for girls and young women, 15 out of 29 media messages were found to use this strategy. Over a quarter of the messages addressed the benefits of not drinking or of less harmful alcohol use, which has been found to be a more effective strategy (Kingsbury, Gibbons, Gerrard, 2015). For example, a poster campaign released by the Wesleyan University in Illinois presented harm-re- duction approaches to alcohol use and promoted awareness that not all young people drink alcohol (Illinois Wesleyan University, n.d.). The search was conducted using search engines such as Google and Ask.com, as well as the video search engine YouTube. When searching for vid- eos on YouTube, additional videos were obtained through the ‘Up Next’ feature. Combinations of the following search terms were used: girls, females, youth, underage, teens, ado- lescent, young women, young people, underage drinking, binge drinking, risky alcohol use, under- age alcohol abuse, youth heavy drinking, messag- es, campaigns, strategies, media, underage binge drinking Canada, and BC. One media message created counter-advertis- ing, which may be effective in reducing harmful alcohol use among underage girls (Banerjee, 2013). This poster campaign and website por- trayed young people having fun with their friends without drinking. Through this search the Academy members collected and analyzed a wide range of media messages including websites, poster campaigns, public service announcement videos, and videos created by independent YouTube users. Inclusion criteria included messages that targeted both underage alcohol use in general and specifically underage girls. Messages that were directed only to adults or only to males, as well as ones that had a specific focus on substances other than alcohol were excluded from the analysis. An educational website created by an organization in the UK offers strategies for moderating drink- ing and addressing risky alcohol use for people who have concerns about their use (https://www. drinkaware.co.uk/). This site was the only one found that offered different targeted messaging for binge drinkers and moderate drinkers, which Lee and Chen (2013) suggest may be more effec- tive than having one message for everyone. While the majority of these messages did not use the positive strategies discussed above, and instead focused on the negative consequences of heavy alcohol use, a number of them did address risk and protective factors and youth’s reasons for drinking. The most commonly addressed risk factors included concern for peer approval and gender differences in metabolizing alcohol. Many of the other risk factors that have been shown to contribute to harmful alcohol use, such as history of victimization, low self-esteem, weight concerns, and challenging life stages such as school transi- tions were either absent or present in only 1 or 2 of the 29 media messages analyzed. 46 Strategies to reduce risky alcohol use among underage girls: An evidence review

  48. Some of the media messages promoted protective factors such as encouraging dialogue between friends about risky alcohol use, having clear rules around alcohol consumption, and promoting positive norms with friends including an aware- ness that not all youth drink. One media message contained a portrayal of positive communication between a young person and his family, which Mason and colleagues (2009) found to be protec- tive against risky alcohol use (discussed below, titled “The Dangers of Underage Drinking: My Instead Movie”). As shown in the literature, the focus on negative consequences and health risks are not very help- ful because female youth aren’t thinking about these things when they’re drinking. For example, the drinking and driving ads often show events leading up to an accident but youth often don’t think about the consequences when they decide to get in a car after drinking or with someone who has been drinking. A lot of the messages members of the Youth Research Academy reviewed talked about peer pressure, with comments from youth like “one more drink won’t kill you,” or “don’t be lame just do it, we all have done it before.” It can be hard to resist peer pressure but youth-led messaging can counteract these and be effective because it is coming from a peer and not from a guardian/ parent. Youth Research Academy’s reflections Women have reported feeling angry about messages that place blame and shame on females for being the victim of sexual violence because of their alcohol use (Canadian Women’s Health Network, 2014). Simi- larly, members of the Youth Academy felt many of the messages they reviewed placed unfair re- sponsibility on girls and young women to protect themselves from sexual assault and other harms. For example, the “Don’t Be That Girl” poster campaign released by students at the University of Windsor showed a girl dressed provocatively asking the question “Let’s see how many guys will buy me a drink tonight. I didn’t wear this outfit for nothing” (Reid, 2013). In the experience of the Youth Research Academy, having a harm reduction approach also seems like a better approach than messages to abstain from drinking altogether, such as “just say no” or “don’t do it.” For example, a better way to say it might be “if you have a drink, wait and see how you feel in a bit before you have another.” Based on our research and our personal experi- ence, we believe that concise video ads based on real-life experience or scenarios give a more real- istic approach to the effects and risks of drinking. Interactive online games depicting what a night out drinking with your friends might be like and how your choices will lead you in many different directions in a split second are also good. “Love and accept yourself” messages help female youth remember that they don’t need to get drunk to feel good about themselves or to fit in, whereas “shame based” messages just make youth feel like they have done something wrong and won’t make the situation any better. These messages stigmatized the sexuality and agency of females and can make girls feel that it’s their fault if something happens to them if they’ve been drinking, which can lead them to not report assault or seek help. The members of the Youth Research Academy felt there should be more messages that teach young women how to support their friends if they’re drunk or show them how to get home or to another safe place instead of focusing on scare tactics. Many youth don’t watch “regular television” or read newspapers, so advertising in these places is not effective at reaching young people. Instead, video ads through YouTube, for example, would be more effective. 47 McCreary Centre Society

  49. Example of effective messaging: The dangers of underage drinking: My instead movie (https://www.youtube.com/watch?v=aY4B1wSEqlY) In the video, Andy’s male friends have encouraged him to steal a bottle of alcohol from his parents and are now wanting him to drink it with them and a female friend. Next the video gives you three options for how this story ends. 1: Andy dumps out the bottle instead. The boys get upset with him and leave. However the girl is proud of him for not caving into peer pressure. He offers to walk the girl home. 2: An adult shows up and asks if they have seen the missing bottle. They all shake their heads. The adult continues to talk about the effects of drinking and says if they happen to see it to put it back or leave it on the porch. The adult leaves and Andy and his friends run to the porch and leave the bottle. 3: Andy takes a drink and passes it around. The girl gets the bottle but before she takes a sip, Andy starts to feel sick and throws up. His friends leave. Reflection This video shows a realistic scenario of peer pressure and the different choices you can make. The video is also good because the adult knew the boy had taken the bottle, but didn’t get mad. In- stead, he told the group about the effects and gave them information to help them make the right decision. Example of ineffective messaging: Teen anti-drinking campaign (https://www.youtube.com/watch?v=Bhk4fh6Sld0) Music is playing and teens are standing around hanging out at a house party. In one version, a girl declines a drink from a boy, meets another boy who takes her number, and leaves the party with her friends. In the second version, she accepts the drink from the first boy, drinks too much, ends up having sex at the party, feels ashamed, and is judged by her friends. Words slide on to the screen “DRINKING. Where are your CHOICES taking you?” Although this video had some good points about how drinking too much can affect your choices, it was very blaming and shaming and for that reason we didn’t find it effective. 48 Strategies to reduce risky alcohol use among underage girls: An evidence review