Aims:

1. To establish whether sexual orientation (at 15) is associated with problem drinking in late adolescence;

2. To assess evidence that any link between sexual orientation and problem drinking in late adolescence may be mediated by high levels of depressed mood and/or anxiety;

3. To test whether there are gender-specific differences (females vs. males);

4. To establish the prevalence of substance use (i.e., tobacco and illicit substances) in non-heterosexual vs. heterosexual adolescents.

Adolescent alcohol use represents a main public health concern as it causes short and long-term health problems and antisocial behaviour, which can lead to self-harm or harm to others (Coleman & Carter, 2005). Alcohol use is widespread in the UK where 88% of 15 year-olds report having drunk alcohol in the past 12 months and 1% of 14-16 year olds drink nearly every day.

Adolescents who identify themselves as gay, lesbian or bisexual (i.e. sexual minority adolescents) are at increased risk of involvement in problem drinking compared to their heterosexual counterparts, with studies reporting earlier age of onset of alcohol use, more frequent and heavier (i.e., binge) drinking, and more alcohol-related problems than heterosexuals (Bergmark, 1999; Corliss et al., 2008). Although there is evidence of differences in risk of problem drinking between males and females from a sexual minority background, the pattern remains unclear. Some studies have observed more alcohol problems among females than males while others have observed the opposite pattern (Ziyadeh et al., 2007).

Evidence also shows that sexual minority adolescents experience more mental health issues (i.e., depression and anxiety) than their heterosexual peers (Fergusson, Horwood & Beautrais, 1999; Hatzenbueller et al., 2008). Greater levels of anxiety and depressive symptoms may be elicited by gay-related stressors, such as the stress of coming out (Elze, 2002; Hatzenbueller et al., 2008). Symptoms of depression and anxiety have been found to predict elevated drinking and alcohol-related problems among non-heterosexual youth. However, to our knowledge, no study has attempted to establish whether these risk factors may explain the developmental relationship between sexual orientation and drinking. Our study aims to fill this gap by assessing whether the relationship between sexual orientation and problem drinking is explained by internalising disorders. We will also include risk factors of problem drinking, identified in the general population, in order to establish if they can further help us to understand this relationship. The risk factors we intend to use are school satisfaction, relationships with friends and parents, parental monitoring and alcohol attitudes (Sareceno et al., 2010).

Few studies also identified greater substance use and misuse among sexual minorities although less research has been dedicated to tobacco and illicit drug use in this sub-group (Hefferman, 1998). Thus, our final aim is to assess substance use prevalence in the non-heterosexual group compared to their heterosexual counterparts.

Hypotheses:

1. Adolescents who identify as not heterosexual report an earlier onset age of alcohol use and greater problem drinking in later adolescences;

2. We expect that the relationship between sexual orientation and problem drinking may be stronger for those experiencing internalising disorders;

3. There are gender-specific differences in these patterns. Although there is evidence in the general population that rates of depression are higher in female than male adolescents (Thapar et al., 2012), the findings for non-heterosexual adolescents are unclear. Thus, these analyses will be exploratory in nature

4. We expect higher engagement with other substances among non-heterosexual individuals.

Exposure variable(s): Sexual orientation assessed as sexual identity and also as sexual behaviour (i.e., gender of sexual partner); depressive and anxiety symptoms experienced by the teenager; relationship within the family members; parental monitoring; relationships with peers; school satisfaction; attitudes to alcohols

Outcome variable(s): Alcohol use: age of onset; drinking patterns and alcohol-related problems. Tobacco and illicit drug use.

Covariates: Family socio-economic status (SES); parental alcohol and substance use; parental depression; sensitivity to alcohol; conduct problems (i.e., Strengths and Difficulties Questionnaire)

Analyses:

We will conduct preliminary correlations and univariate regression models between sexual orientation and problem drinking as well as substance misuse to examine the association at baseline (i.e., age 15; cross-sectional data). Moreover, we will use structural equation modelling (SEM) to assess these relationships over time while controlling for relevant covariates (e.g., family SES, parental depression and alcohol misuse). Finally, we will use SEM to test the effects of theorised moderators and mediators (i.e., depressed mood and anxiety) on this association and assess whether they explain the direction and/or strength of the relationship. The model will also be adjusted for variables which have been identified as risk factors for both alcohol and substance misuse to assess whether they further explain this mechanism (e.g., relationship with parents and peers; attitudes to alcohol). We will conduct Multiple Imputation to account for missing data.