Introduction
Adolescents who identify as lesbian, gay, or bisexual (LGB) are at greater risk for developing mental health and substance use problems compared to heterosexual adolescents (Goldbach et al.
2014; Plöderl and Tremblay
2015). Two major theoretical frames are often used for understanding the disproportionate rates of health issues among LGB people. First, the minority stress framework identifies several types of stigma-related stressors that LGB adolescents experience in addition to general stressors (Meyer
2003). These higher rates of (minority) stress among LGB adolescents might explain their higher rates of mental health problems and substance use, where the latter may be seen as a mechanism to cope with minority stressors (Meyer
2003). Second, Hatzenbuehler (
2009) extended the minority stress framework by proposing how stigma-related stressors might negatively affect general intra- and interpersonal psychological processes, which, in turn, are related to health and substance use disparities between LGB and heterosexual individuals. This framework has been labeled the psychological mediation framework (Hatzenbuehler
2009).
Both frameworks have been applied to explain differences in mental health and substance use between LGB and heterosexual adolescents (e.g., Baams et al.
2015; Hatzenbuehler et al.
2011; Rosario et al.
2014; Woodford et al.
2012). Unfortunately, however, integrated research is rare, though the combination of both frameworks could provide a better understanding of what drives disparities in mental health and substance use between LGB and heterosexual adolescents. Therefore, the aim of this study was to test the minority stress and the psychological mediation framework in one empirical analysis of health disparities between LGB and heterosexual adolescents.
Minority Stress Framework
Minority stressors are stigma-related stressors experienced by sexual minority people because of their marginalized sexual identity, in addition to general life stressors (Meyer
2003). Minority stressors exist on a continuum ranging from distal stressors to proximal stressors. Distal stressors comprise external, objective stressful events and conditions. Examples of distal stressors are being rejected by others or being victimized because of one’s sexual identity. Proximal stressors refer to personal perceptions and appraisals of distal stressors by LGB individuals. An example of such a proximal stressor is the application of negative attitudes that exist in society against LGB people to the self, also referred to as internalized homophobia. The experience of these minority stressors by LGB people can lead to poorer mental health compared to heterosexual people (Mongelli et al.
2018) or the use of substances as a maladaptive coping mechanism (Meyer
2003). This article focuses on a number of distal minority stressors including rejection and victimization.
Peers and parents can be sources of minority stress (Russell and Fish
2016). On average, sexual minority adolescents have less positive relationships with peers and parents than heterosexual adolescents, which is linked to differences in mental health between Dutch sexual minority and heterosexual adolescents (Bos et al.
2008). For instance, in the US, homophobic victimization by peers predicted mental health problems in LGB students, especially for girls (Poteat and Espelage
2007). Moreover, sexual minority youth are more often persistently victimized than their heterosexual peers (Robinson et al.
2013) and persistently victimized LGB adolescents reported more internalizing problems (Kaufman et al.
2019). Within the family, parental rejection explained the relation between sexual identity and depressive symptoms in a Dutch sample, especially among lesbian girls and bisexual participants (la Roi et al.
2016). Similarly, the relation between a sexual minority identity and depressive symptoms was partly explained by lower family satisfaction among US adolescents (Luk et al.
2018). Additionally, sexual minority youth reported less closeness and support from parents compared to heterosexual youth, which was linked to lower mental health, especially in US girls compared to boys (Pearson and Wilkinson
2013).
Focusing on substance use, minority stressors such as poorer quality of relationships with peers and parents can account for substance use disparities between heterosexual and LGB youth as well. Victimization by peers in schools, for instance, explained disparities in substance use between LGB and heterosexual adolescents in a representative student sample (Bontempo and D’Augelli
2002). Additionally, LGB college students’ experiences of interpersonal mistreatment explained their higher prevalence of drinking problems compared to heterosexual college students (Woodford et al.
2012). Further, poor mother-child relationship quality explained the association between sexual identity and substance use for LGB emerging adults (Rosario et al.
2014) and parental rejection explained the association between a sexual minority identity and marijuana and hard drug use for women, but not men (Needham and Austin
2010). Similarly, poor parent-child relationship quality explained higher levels of alcohol use of sexual minority youth compared to heterosexual youth, especially for girls (Pearson and Wilkinson
2013). Of note, all studies on substance use have been conducted on US samples.
Minority stressors such as LGB adolescents’ compromised relationships with peers and parents might explain their elevated risk for mental health problems and substance use. However, this research is limited in that it neglects the role of general intra- and interpersonal psychological processes as intermediate links between minority stressors and mental health and substance use. The psychological mediation framework has been proposed as a refinement of the minority stress framework (Hatzenbuehler
2009). Where the minority stress framework hypothesizes that minority stressors explain links between sexual identity and psychopathology or substance use (Meyer
2003), the psychological mediation framework examines general intra- and interpersonal psychological processes through which minority stressors might affect psychopathology or substance use (Hatzenbuehler
2009). In accordance with the minority stress framework (Meyer
2003) it posits that LGB people are exposed to increased stress resulting from stigma. This stigma-related minority stress is thought to elevate emotion dysregulation, social/interpersonal problems, and cognitive processes that ultimately result in higher risks for psychopathology (Hatzenbuehler
2009). These processes are thought to account for the relation between stigma-related minority stress and psychological problems or substance use. This article focuses on a number of these proposed processes including fear of negative social evaluation, an intrapersonal process, but also social/interpersonal processes such as social support and substance use norms.
Minority stressors such as low relationship quality with peers and parents due to sexual minority status can to a large extent explain associations between sexual identity and health outcomes. Following the psychological mediation framework, it is expected that negative evaluations by others, in turn, mediate the relation between these minority stressors and health outcomes. Although the expected negative evaluations by others are described as a possible consequence of minority stress (Meyer
2003), studies on LGB people hardly examined their role. A recent longitudinal study conducted among emerging adults in the US with concealable stigmatized identities (ranging from being a sexual minority to using drugs) revealed that the expectation to be stigmatized rather than enacted stigma predicted depressive symptoms (Chaudoir and Quinn
2016). More specifically, among US adults, sexual orientation-related rejection sensitivity explained the relation between discrimination and internalizing behaviors (Feinstein et al.
2012). Similarly, chronic expectations of rejection were related to smoking among young sexual minority men in the US (Pachankis et al.
2014) expectations of rejection were related to several internalizing problems among gay and bisexual US university students (Cohen et al.
2016). Thus, for the present study, it was expected that minority stressors such as victimization by peers and cold and unsupportive relationship with parents because of one’s sexual identity would affect one’s mental health and substance use through expected negative evaluations by others.
Social support might also mediate the association between poorer relationship quality with peers and parents, and health outcomes (Hatzenbuehler
2009). It is reasoned that sexual minority people isolate themselves to avoid minority stress experiences such as being rejected (Link et al.
1997). Self-isolation, however, further diminishes their social support, which can affect mental health negatively (Umberson and Karas
2010). In contrast, greater sexual identity-related support by peers and parents has been associated with less emotion-related distress (Doty et al.
2010). Though less studied, support might also work in the opposite direction as relationships with peers tend to be fairly socially and hedonically oriented, sometimes resulting in positive associations between peer support and substance use (Wills et al.
2004).
Permissive substance use norms of peers are a factor that could specifically mediate the association between poorer quality of relationships with peers and parents and substance use. Although empirical support is mixed, it has been argued that minority stressors might contribute to more permissive substance use norms (Hatzenbuehler
2009). Experiences of minority stress might push LGB adolescents into social circles which are characterized by more permissive substance use norms, for example LGB communities with a strong ‘bar culture’. In general, peer’s substance use increases one’s own substance use (Soloski et al.
2016). In fact, sexual minority adolescents’ social networks tend to include more individuals that use substances than those of heterosexual adolescents (Hatzenbuehler et al.
2015). Permissive social norms regarding substance use in one’s network account for the link between sexual identity and alcohol use (Hatzenbuehler et al.
2008). Thus, negative experiences with peers and parents might make a person more vulnerable to permissive substance use norms, resulting in more substance use.
Discussion
Research has repeatedly found that LGB adolescents report more internalizing problems and substance use (smoking, marijuana use, and alcohol use) than their heterosexual peers (Goldbach et al.
2014; Plöderl and Tremblay
2015). The minority stress framework (Meyer
2003) and the psychological mediation framework (Hatzenbuehler
2009) have both been used to explain these disparities between LGB and heterosexual adolescents. However, little research has integrated both frameworks, although doing so provides a more fine-grained understanding of the drivers of disparities in mental health and substance use between LGB and heterosexual adolescents. Therefore, the aim of this study was to examine how indicators of minority stressors (Meyer
2003) and psychological mediators (Hatzenbuehler
2009) together explain disparities between LGB and heterosexual adolescents in internalizing problems and substance use. By integrating both frameworks, health disparities were not only examined by focusing on minority stress processes, but also by taking intra- and interpersonal psychological processes into account that have been proposed as intermediate links between minority stress and health outcomes.
It was hypothesized that sexual identity and internalizing problems would be related through a serial mediation process with peer victimization and negative relationships with parents as the first set of mediators (following the minority stress framework), and fear of negative social evaluation and lower social support as the second set of mediators (following the psychological mediation framework). Similarly, sexual identity and substance use were expected to be linked through peer victimization and negative relationships with parents (first set of mediators, following the minority stress framework) and fear of negative social evaluation, lack of social support, and substance use norms of peers (second set of mediators, following the psychological mediation framework). LGB adolescents reported more internalizing problems, smoked more cigarettes, and consumed more marijuana compared to their heterosexual peers. Mechanisms indicative of minority stressors partially explained these differences. As hypothesized, the association between sexual identity and internalizing problems was mediated by peer victimization and parental rejection. No substantial evidence was found for psychological mediation processes further explaining these health differences between LGB and heterosexual adolescents. Thus, only partial support was found for minority stress processes and no support for psychological mediation processes acting as intermediate links in the sexual identity – minority stress – health outcomes process.
Contrary to expectations, LGB adolescents did not report more alcohol use than their heterosexual peers. This is noteworthy given that a previous meta-analysis found higher alcohol use of LGB adolescents compared to their heterosexual peers (Marshal et al.
2008). Most studies reviewed in this meta-analysis were conducted in the US, which might suggest that sexual identity-based alcohol use disparities are more prevalent there than elsewhere. Alternatively, it might be that only some subgroups of LGB adolescents have a higher risk of alcohol use compared to heterosexual adolescents. Previous research among Dutch adults showed that disparities in substance use between LGB and heterosexual people were driven by the bisexual group (van Beusekom and Kuyper
2018). However, separating the bisexual and lesbian/gay group did not change results. That is, no differences in alcohol use were found when comparing bisexual and heterosexually identified participants (results available upon request). Further, neither peer victimization nor negative parent-child relationship accounted for sexual identity disparities differences in smoking and marijuana use. Although research has established that minority stressors predict substance use of LGB adolescents, not all research has consistently found this pattern for all types of substances. For example, incivility and hostility explained higher rates of LGB students’ drinking problems compared to heterosexual students, but not other drug use (Woodford et al.
2012). Similarly, maternal discomfort with homosexuality did not explain higher rates of smoking among lesbian and gay adolescents (Rosario et al.
2014). Together with this study’s results, these findings imply that using a concept or umbrella term as ‘substance use’ might miss nuances.
This study is not the first to not find strong support for psychological mediation processes (Austin et al.
2004; Wichstrøm and Hegna
2003; Ziyadeh et al.
2007), although prior studies used sexual identity as a proxy for minority stressors. Despite efforts to measure indicators of minority stressors, no support was found for psychological mediation processes. This might reflect that the psychological mediation factors under study are not as important in explaining differences in internalizing problems and substance use among adolescents in the current sample. With regard to more permissive substance use norms of LGB adolescents, it might be that LGB adolescents in the current sample were not able to engage in an ‘LGB bar culture’ because they were either too young to be admitted into these bars, or because no such bars existed in their surroundings (especially in rural areas). With respect to lack of social support, it is feasible that LGB adolescents substitute support. For instance, qualitative research showed that LGB youth more often seek friends online with whom they can talk about their experiences and from whom they receive support (Hillier et al.
2012). Last, focusing on the fear of negative evaluation, the Netherlands is a relatively tolerant country regarding attitudes towards LGB people (van Beusekom and Kuyper
2018), which could result in LGB people expecting or fearing negative evaluation less.
Ultimately, this study aimed to explain internalizing problems and substance use disparities between LGB and heterosexual adolescents by focusing on minority stress processes as potential mediators of these disparities, but also include mediators of the link between minority stress and health outcomes as proposed in the psychological mediation framework. Owing to limitations imposed by secondary data and sample size, only a subset of (distal) minority stress and factors as proposed by the psychological mediation framework were included. Conclusions about the empirical validity of the minority stress and psychological mediation frameworks as a whole are thus beyond the scope of this study. For instance, it could be that that proximal minority stressors or emotion dysregulation and cognitive processes as described in the psychological mediation framework would have been better able to explain health differences between LGB and heterosexual adolescents than the currently employed measures.
Using the TRAILS data enabled us to study internalizing problems and substance use in a general sample of adolescents, but it also came with a drawback. TRAILS was designed as a cohort study on the development of mental health from childhood to adulthood and therefore only general stressors which were used as proxy measures of minority stress were available. For that reason, we cannot be certain to what extent sexual identity disparities in peer victimization and negative parent-child relationships actually reflect minority stress processes. Past studies have made similar assumptions when using general measures for minority stress processes (Katz-Wise and Hyde
2012); these studies are thus comparable to the present work. What is more, LGB participants reported more peer victimization, parental angry outbursts, and parental rejection than heterosexual participants even after controlling for a long list of potentially confounding factors, which affirms that LGB adolescents experience additional (minority) stress in peer and parent-child relationships compared to their heterosexual counterparts.
Related to this, TRAILS included proxy measures of distal minority stressors such as rejection and victimization, but no measures of proximal stressors such as internalized homophobia. Therefore, the present study was unable to study all components proposed in the minority stress framework. Future studies that focus specifically on LGB adolescents should aim to measure minority stress more precise and complete.
In addition, the role of negative relationships with parents as minority stressors was assessed using three measures, guilt inducing behavior, parental angry outbursts, and parental rejection. To the best of our knowledge, no earlier studies on this topic looked at the former two of these or closely related constructs, as most existing work focused on either parental rejection or support (Bouris et al.
2010; Russell and Fish
2016). Although all three indicators of negative parent-child relationship were associated with sexual identity and health outcomes in expected directions, associations were stronger for parental rejection and angry outbursts than guilt-inducing behaviors. A tentative conclusion would thus be that experiencing parental rejection or anger are more important emotional dimensions of the family context than guilt-inducing parental behavior for explaining sexual identity disparities in health outcomes.
Further, although using data from a longitudinal study, we were not able to disentangle how minority stress mediators influenced changes in psychological mediation processes and how both influenced changes in the health outcomes over time. In order to estimate whether a causal relation exists between two variables, a study design is needed that is able to prevent bias in the estimates due to reverse causation, incorrect specification of the lag of the effect, and confounding. Methods that can take all these issues into account require at least three measurements of both dependent and independent variables (Allison et al.
2017; Hamaker et al.
2015; Leszczensky and Wolbring
2019), which were not available in TRAILS. Future research that has at least three measures of all variables at study would be able to overcome these issues with estimating how characteristics induce change in one another.
Lastly, although no significant evidence for indirect effects in line with the psychological mediation framework was found, several of the path estimates in the integrated mediation model were in line with expectations (Fig.
3/Table A2 in online supplementary A). For instance, an LGB identity was related to more peer victimization and parental angry outbursts, which were associated with fear of negative social evaluation and lack of social support, which, in turn, were associated with higher levels of internalizing problem behaviors. The indirect effects running through these paths did not reach statistical significance, however, which might be a consequence of insufficient statistical power to detect small effects.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.