Negative Affect Associations with Adolescent Psychopathology and Substance Use

Abstract

Adolescence and the transition to adulthood are critical periods of development during which to intervene upon problematic behavior and emerging psychopathology. Drawing heavily from developmental psychology perspectives, this dissertation’s overarching aim was to better understand the role of internalizing psychopathology in substances use during adolescence.

Utilizing a transdiagnostic, dimensional approach, negative affect was examined across three studies. In the first study, the measurement and structure of the negative affect (NA) construct was assessed using original data collected online via Qualtrics. Negative affect items from the Monitoring the Future (MTF) and the Adolescent Brain and Cognitive Development (ABCD) studies were validated against longer, well-established scales using item response theory (IRT). Next, the structure of NA was assessed using exploratory structural equation modeling (ESEM). The findings provide evidence of convergent validity for the MTF and ABCD NA items thus reducing measurement validity concerns commonly associated with survey data and allowing allow for a synthesis of findings between studies. Structurally, ESEM models fit the NA data best and subfacets of NA displayed unique associations across psychopathology related constructs.

In the second study, levels of NA and impulsivity were used to predict patterns of comorbidity in the first waves of the ABCD dataset. Findings from multinomial logistic regression revealed that, compared to those with no mental health condition (MHC), high levels of anxiety and anger predicted internalizing (IO) and externalizing (EO) diagnoses, respectively. Heterotypic comorbidity was predicted by higher levels of both anxiety and anger. Lack of perseverance was the only impulsivity construct to consistently predict membership in one of the three mental health diagnosis categories when compared to the no MHC condition. Next, resting state neuroimaging data was used to examine neural correlates of impulsivity in the pre-adolescent brain using Brain Base Set (BBS) modeling. Findings revealed no unique neural networks that were associated with impulsivity. Instead, the associations were weak, and widespread throughout the brain.

The third study used MTF data from 8th, 10th, and 12th grade students to identify unique patterns of NA. Using latent profile analysis (LPA), 5 NA profiles defined by levels of anger, anxiety, boredom, depressive affect, and anhedonia, emerged; low negative affect, moderate negative affect, high negative affect, high positive and negative affect, and emotionally numb. The model was then validated in an independent sample from Study 1. Profile membership was predicted by a number of covariates including school achievement and misbehavior, social interaction, and substance use. Additional tests revealed significant gender moderation.

Findings across studies are synthesized and implications are discussed. Together, the current studies make a number of unique contributions to the literature including 1) providing an in-depth psychometric analysis of survey items related to transdiagnostic indicators 2) establishing convergent validity across negative affect measures included in the MTF and ABCD datasets 3) exploring the combination of impulsivity and NA in heterotypic comorbidity 4) addressing the convergent and divergent hypotheses of NA and 5) examining gender moderation in psychopathology. Together, by providing some clarity regarding construct validity, heterotypic comorbidity and, gender moderation, the findings presented here may contribute to more efficient and tailored efforts to prevent and intervene upon at-risk youth during adolescence.