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

Abstract

Negative affect (depression/anxiety) and alcohol use among Indigenous youth in Canada remain a concern for many communities. Disparate rates of these struggles are understood to be a potential outcome of colonization and subsequent intergenerational trauma experienced by individuals, families, and communities. Using a longitudinal design, we examined change in alcohol use and negative affect, and reciprocal associations, among a group of Indigenous adolescents. Indigenous youth (N = 117; 50% male; M_age=12.46-16.28; grades 6-10) from a remote First Nation in northern Quebec completed annual self-reported assessments on negative affect (depression/anxiety) and alcohol use. A Latent Curve Model with Structured Residuals (LCM-SR) was used to distinguish between- and within-person associations of negative affect and alcohol use. Growth models did not support change in depression/anxiety, but reports of drinking increased linearly. At the between-person level, girls reported higher initial levels of depression/anxiety and drinking; depression/anxiety were not associated with drinking. At the within-person level, drinking prospectively predicted increases in depression/anxiety but depression/anxiety did not prospectively predict drinking. When Indigenous adolescents reported drinking more alcohol than usual at one wave of assessment, they reported higher levels of negative affect than expected (given their average levels of depression/anxiety) at the following assessment. Our findings suggest that when Indigenous youth present for treatment reporting alcohol use, they should also be screened for negative affect (depression/anxiety). Conversely, if an Indigenous adolescent presents for treatment reporting negative affect, they should also be screened for alcohol use.

Keywords: Adolescence; Alcohol; Depression/Anxiety; Drinking; Indigenous Youth.

Fig. 1

Latent Curve Model with Structured Residuals for Alcohol Use and Depression/Anxiety. Note. Solid black lines are significant and dotted grey lines are non-significant pathways. Betas are reported next to hypothesized significant associations and standard errors are reported in parentheses. Levels of significance were based on unstandardized regression estimates. For simplicity, parameter estimates for latent factor loadings are not depicted. Gender was coded such that 0 = female and 1 = male. RI = Random Intercept. W = Wave. ** = p < 0.01. * = p < 0.05