Negative Affect and Drinking among Indigenous Youth: Disaggregating Within- and Between-Person Effects

Negative affect, such as anxiety and depression, often co-occur with alcohol use during the developmental period of adolescence (Marmostein, 2009). Some research indicates the prevalence of depression and anxiety among Indigenous youth may be similar, lower, or higher when compared with non-Indigenous youth (Andrade et al., 2006; Hop Wo et al., 2020; Scott et al., 2022). In a 2021 survey by Statistics Canada (2021), one in five and almost one in four Indigenous youth reported being diagnosed with a mood disorder or an anxiety disorder. However, the prevalence rates may reflect only those individuals who are help-seeking and therefore were able to get a diagnosis. In an American study of the development of depressive symptoms among Indigenous adolescents between the ages of 9 to 13 years old who were followed for 8 years, Martinez and Armenta (2020) found that 63% of Indigenous adolescents experienced depressive symptoms at some point. Moreover, at least 21% of these adolescents demonstrated consistently elevated levels of depression over time and were more likely to develop an alcohol use disorder. Depression and anxiety are well-established comorbid conditions that affect Indigenous peoples and put them at a greater risk for problems with substances, including alcohol (Kenney & Singh, 2016; Rieckmann et al., 2012; Warne et al., 2017; Warne & Lajimodiere, 2015).

Alcohol use rates and patterns vastly differ across Indigenous communities in North America, with high incidences of abstinence among some (Cunningham et al., 2016; Rieckmann et al., 2012) and earlier onset of drinking among others (Cheadle & Whitbeck, 2011). Of concern is the disparities in the consequences associated with alcohol use including high rates of substance use disorders, health complications related to alcohol use, and alcohol-related deaths (Indian Health Service, 2018; Singh et al., 2017; Substance Abuse and Mental Health Services Administration, 2019).

Two dominant models for understanding the risks that lead youth in general to alcohol use include negative reinforcement motives for drinking and the Dual Failure model. Both theories highlight the role of internalizing and externalizing symptoms. Negative reinforcement motives for drinking occur when individuals use substances as a strategy to dampen unpleasant emotions and psychological suffering (Cooper et al., 1995; Kuntsche et al., 2005), suggesting that internalizing symptoms such as negative affect may precede alcohol use. Brockie et al. (2015) found that Indigenous youth who reported high rates of historical loss were at an increased risk for depressive symptoms, PTSD symptoms, and substance use. Additionally, Stewart et al. (2011) found that depressive symptoms were directly linked to drinking to cope, which in turn was linked to heavy alcohol use among Indigenous youth in Canada. Conversely, the Dual Failure Model (Capaldi, 1991, 1992) would suggest that adolescent drinking occurs in the larger context of externalizing symptoms such as rule breaking and aggressive behaviours, which lead to negative consequences and in turn can lead to internalizing symptoms including depressive symptoms (Colder et al., 2013; Paige et al., 2021). Indeed, externalizing symptoms have been found to precede and overlap with substance use among Indigenous youth (Greenfield et al., 2017; Whitbeck et al., 2014).

In this paper, we introduce the use of analytic approaches that allow for the examination of reciprocal associations between negative affect and alcohol use over time with a group of Indigenous youth. In addition, we will disaggregate between- and within-person effects (e.g., Martinez & Armenta, 2020). Disaggregating within- and between-person effects is in line with the self-medication theory (Khantzian, 1985, 1997), as it invokes individual differences and the utility of considering both between- and within-person-level change over time. According to self-medication theory (Khantzian, 1985, 1997), individuals who have increased symptoms of negative affect (depression/anxiety) tend to drink more (e.g., between-person association). Additionally, if an individual experiences an increase in symptoms of negative affect (depression/anxiety), relative to their typical level of negative affect at one point in time, they are likely to increase their alcohol use, relative to their typical level of alcohol use, at a subsequent point in time (e.g., within-person association). Moreover, reciprocal associations between two constructs imply the disaggregation of between- and within-person effects, as earlier changes in one construct can influence later changes in the other, and vice versa (Curran et al., 2014). We used a longitudinal design and Latent Curve Model with Structured Residuals (LCM-SR; Curran et al., 2014) to examine prospective reciprocal associations and to distinguish within- and between-person associations among a group of First Nations youth from a northern community.

The education administration of the First Nations community approved the battery of measures that were administered to the participants of thi study. Additionally, a subcommittee in the community reviewed and approved this manuscript. The study was approved by the ethics committees at Concordia University and McGill University.

Nearly all of the students in grades 6 to 11 (i.e., the end of secondary education in Quebec) were recruited to participate. While most of the students identified as First Nations, some youth identify as Métis or Inuit. Participation was premised on parental consent and participant assent. The parents or legal guardians were provided the option to inform the school if they did not want their child to participate. Additionally, the students were told that they could withdraw from the study at any time. The data were drawn from multiple cohorts during the academic years of 2011–2018. The final dataset included 110 students (44% male; M_age=12.46–16.28; grades 6–10). Retention across grades 7 to 11 was 83.6% (n = 93), 77.3% (n = 85), 82.7% (n = 91), 72.7% (n = 80), and 56.4% (n = 62), respectively. Given attrition and the small number of participants in grade 11, only data from grades 6 through 10 (W1-W5) were used. A full-information maximum likelihood estimation was used to minimize the impact of missing data. This approach allows for inclusion of all of the participants, even those with some missing data.

We attempted to disentangle the temporal association of negative affect (depression/anxiety) and alcohol use among Indigenous youth (grades six to grade ten) from a community in northern Canada with a longitudinal design. We used an LCM-SR model to test bidirectional relationships and examine differences in hypothesized associations at between- and within-person levels. Exploring this risk pathway is essential as Indigenous youth mental health challenges are derived from intergenerational trauma and colonization.

With respect to covariates, the finding that, on average, girls reported higher initial levels of negative affect (depression/anxiety) is consistent with the literature on gender differences and adolescent depression/anxiety in both Indigenous and non-Indigenous populations (e.g., Ames et al., 2015; McLaughlin & King, 2015; Salk et al., 2017; Walls et al., 2021). Girls also reported higher initial levels of alcohol use. Epidemiological data suggest that alcohol use has been declining among adolescents and more rapidly for boys, with the exception of drinking alone, which is on the rise and increasing more rapidly for girls (White, 2020). These trends may be leading to a reversal of historical gender differences of males reporting greater alcohol use. In addition, girls may experience internalizing problems, including anxiety and depression earlier than boys (Leve et al., 2005). Future research should be focused on gender differences in negative affect (depression/anxiety) and alcohol use among Indigenous youth.

Our first hypothesis that alcohol use and negative affect (depression/anxiety) would be related at the between-person level was not supported. Initial levels of depression/anxiety and alcohol use were not significantly associated after accounting for gender. This finding is in contrast with evidence of a positive association between depression symptoms and alcohol use among Indigenous peoples, including youth (Schick et al., 2022; Walls et al., 2021). It also differs from Martinez and Armenta’s (2020) finding that when Indigenous youth experienced elevated levels of depression across their adolescence, they were more likely to meet the criteria for Alcohol Use Disorder. We expanded on past work in this area by utilizing a longitudinal LCM-SR model to distinguish between- and within-person associations. Surprisingly, adolescents’ initial levels of depression/anxiety were not significantly related to changes in their alcohol use across time. This contrasts with evidence of an association between depression, anxiety, and later alcohol use among non-Indigenous youth (see Dyer et al., 2019; McCarty et al., 2012, 2013; O’Neil et al., 2011).

Our second hypothesis—that an adolescent’s increased level of negative affect (depression/anxiety) would prospectively predict increases in alcohol use at the within-person level—was also not supported. The findings from the current study are consistent with evidence from non-Indigenous young adults of a unidirectional association between an alcohol use disorder and depression symptoms, but no reverse effect from depression symptoms to an alcohol use disorder (Fergusson et al., 2009). Moreover, we expand on past work by demonstrating that depression/anxiety does not prospectively predict alcohol use at the level of individual change among Indigenous youth. Indeed, when an adolescent reported a higher level of negative affect (depression/anxiety) than was usual for them at one assessment, this deviation was not associated with their change in drinking at the next assessment. This finding is consistent with evidence that negative reinforcement for drinking may not be as relevant during the developmental period of adolescence (e.g., Scalco et al., 2021; Colder et al., 2013). Rather, during the adolescent years, substance use is most likely to occur in a social context (Dishion & Medici Skaggs, 2000; Kobus, 2003; Oetting & Beauvais, 1991). Studies with Indigenous youth have shown that socializing with peers who engage in risky behaviours, including drinking alcohol, were more likely to start drinking at an earlier age (Boyd-Ball et al., 2014) engage in increased monthly alcohol use (HeavyRunner-Rioux & Hollist, 2010) and binge drinking (Chen et al., 2012). When youth experience internalizing symptoms, they may be protected from affiliating with peers who engage in substance use (Fite et al., 2006) and thereby are at a reduced risk for substance use (Colder et al., 2013, 2018; Mason et al., 2008). However, the results from our power analysis suggest that the power to detect the regression coefficients of depression/anxiety predicting alcohol use, especially at the later waves (W3 predicting W4 and W4 predicting W5), on the within-person side of the model was considerably lower than the power to detect the regression coefficients of alcohol use predicting depression/anxiety across all simulated models. Therefore, another possible explanation for our null findings is low power, and future research in this area should aim to utilize considerably larger groups to examine within-person associations between negative affect (depression/anxiety) and alcohol use among Indigenous youth.

Our third hypothesis—that high levels of alcohol use will prospectively predict high levels of negative affect (depression/anxiety) at the within-person level—was supported. When an adolescent endorsed a higher level of drinking than was usual for them at one assessment, they had higher levels of depression/anxiety than expected at the next assessment. This finding is consistent with findings among non-Indigenous youth whereby binge drinking was associated with subsequent depression symptoms one year later (McCabe et al., 2023). However, this is a novel finding among Indigenous youth, suggesting that alcohol use precedes negative affect.

The study of alcohol consumption among Indigenous communities must also involve the consideration of systemic factors (i.e., colonization) that contribute to mental health problems in this population (Gone, 2021). For example, colonization and assimilation policies have directly impacted Indigenous peoples’ cultural practices and identities (Chase, 2012; Sszlemko et al., 2006), which in turn may account for alcohol consumption and related consequences (Brave Heart, 2003; Ross et al., 2015; Whitbeck et al., 2004; Wiechelt et al., 2012). Specifically, drinking alcohol can pull Indigenous youth away from engaging in traditional activities and forming their cultural identities, and thereby lead to negative affect. This implies that the prospective links between alcohol use and negative affect may arise from consequences associated with problematic alcohol use. Conversely, building a sense of pride and belonging to one’s ancestral culture can be protective (Burack et al., 2014, 2017; Brown et al., 2021). For example, adolescents who value cultural activities may be protected from engaging in risky behaviours, including alcohol use, because they offer an alternative reinforcement to drinking (Spillane et al., 2020; Goldstein et al., 2021).

Another novel finding was that when an adolescent endorsed a higher level of drinking than was usual for them at one assessment, their alcohol use was lower than usual for them at the next assessment. As Indigenous youth are more likely to experience negative consequences associated with their drinking (Indian Health Service, 2018; Landen et al., 2014; Stanley et al., 2014), the consequences may deter heavy alcohol use at the following assessment. Additionally, these alcohol-related consequences may also lead other individuals in the community to help the adolescent reduce their alcohol consumption. However, this interpretation warrants further investigation.

This study has some limitations. One, the project involves Indigenous youth from a single First Nation in northern Quebec. We consider this study to be a first step to gathering information from various communities independently, which can help identify the ways in which each respective community show similarities and differences in the developmental patterns of alcohol use and negative affect (depression/anxiety). Although the current findings are not immediately generalizable to other Indigenous youth in Canada or the United States, this approach is consistent with the methodological perspective that the inclusion only of youth from a single community is preferable in that findings from homogenous grouping are more precise and reliable than those from heterogenous groups. Two, only youth who attended school were recruited. While nearly all of the youth in this community attend the school, the relevance to those who do not regularly attend or are not enrolled in school should be examined. Three, the current study was limited by how negative affect was measured. Although depression and anxiety are highly correlated, both theory and research would suggest that they also have independent features. For example, while the items are generally used to assess negative affect and psychological distress, the physiological hyperarousal may be unique to anxiety and the low positive affect may be specific to depression (Anderson & Hope, 2008; Chorpita, 2002; Clark & Watson, 1991), which in turn may have independent pathways to predicting drinking behaviour (e.g., Schleider et al., 2019). Further, the YSR has not been adapted to local Indigenous worldviews and expressions of depression/anxiety. Defining and measuring depression and anxiety in ways that are culturally meaningful to Indigenous youth is key to ensuring construct validity (Beals et al., 2005; Whitbeck et al., 2014). Yet, this work remains limited among Indigenous populations, as the DSM-5 criteria for Major Depressive Disorder and Generalized Anxiety Disorder are based on Western conceptualizations and do not consider Indigenous knowledge and concepts of health and wellness related to depression.

Four, the small number of participants only provided power to detect moderate sized effects; and thus, we may have missed conceptually interesting small effects. Relatedly, we were unable to test moderation or include certain covariates in our model due to limited power. Namely, past findings suggest the importance of considering externalizing symptoms and how it influences risk for substance use among Indigenous youth. For example, in a longitudinal study of non-Indigenous youth, Scalco et al. (2021) found support for the role of externalizing symptoms and co-occurring externalizing and internalizing symptoms in risk for alcohol use. Conversely, internalizing symptoms alone did not increase risk for alcohol use. We tried to address this limitation; however, adding externalizing symptoms to our model as a covariate resulted in non-convergence, likely due to the small number of participants and model complexity. However, this limitation, in some ways, represents the inherent trade-off between internal and external validity. Indeed, Indigenous youth represent an understudied and marginalized population. We believe that findings from the current study are critically important due to a dearth of research on this population, and we call for researchers to better include Indigenous youth going forward.

We found that within individuals, higher-than-usual reported levels of drinking prospectively predicted higher levels of self-reported negative affect (depression/anxiety) than expected (accounting for their typical level of depression/anxiety) among Indigenous youth from a remote First Nations community in northern Quebec. In turn, alcohol use may precede negative affect. This temporal association may be explained by the Dual Failure Model, which suggests that adolescents who drink alcohol are more likely to experience negative consequences and in turn are at greater risk for negative affect. This risky pathway must be situated in the context of intergenerational trauma. For example, heavy consumption of alcohol was found to be more likely among Indigenous youth who had a parent or grandparent who attended an Indian Residential School (First Nations Information Governance Centre, 2018), demonstrating the devastating impacts of post-colonialism. Our findings should inform key audiences, such as community leaders and health-care providers about this issue, as well as the developers of prevention and interventions strategies. For example, similar to the work in the general health care setting, screening for negative affect (depression/anxiety) when Indigenous youth present with an alcohol use problem and vice versa should be considered best practice. However, cultural frameworks are important and majority-culture practices cannot necessarily be applied to Indigenous peoples. Beyond screening for alcohol use and negative affect, interventions and treatment programs for Indigenous youth must call on “culture as medicine” based on the history of colonization (Basset et al., 2012; Walters et al., 2020). In support of the resilience that is seen across many Indigenous communities, future research needs to continue to highlight the success and well-being of Indigenous youth and how these stories of success can help to support Indigenous youth navigate away from alcohol use, and in turn reduce depressive symptomatology.