Category: Bipolar Disorders
Bipolar disorders (BD) are associated with high rates of comorbid substance use disorders (SUD; Merikangas et al., 2011). Among adults with BD, reward system, or the behavioral approach system (BAS), sensitivity and disinhibition traits are shown to mediate the relationship between BD and SUD (Alloy et al., 2009). Normative adolescence is characterized by developmental changes in BAS sensitivity (Urosevic et al., 2012) and disinhibition (Schissel et al., 2014). Despite these normative developmental changes during adolescence, no studies have investigated BAS sensitivity and disinhibition links to SUD comorbidity in adolescents with BD. The present study examined BAS sensitivity, using the BIS/BAS scales (Carver & White; 1994), and constraint (CON; inverse of disinhibition), as well as positive (PEM) and negative emotionality (NEM) personality traits, using the Multidimensional Personality Questionnaire (Patrick et al., 2002), in 13 to 19-year-old adolescents with BD but without SUD (BD; n = 37), adolescents with BD and comorbid SUD (BD+SUD; n = 12), and healthy controls (HC; n = 58). Age, IQ, and current bipolar symptoms were entered as covariates. In three-way ANCOVAs, there were significant main effects of group on BAS Total, F(2, 100) = 4.15, p = .019, partial h2 = .08, BAS Fun Seeking, F(2, 100) = 4.99, p = .009, partial h2 = .09, CON, F(2, 93) = 3.26, p = .043, partial h2 = .07, and NEM, F(2, 93) = 23.30, p < .001, partial h2 = .33. Follow-up analyses revealed that the BD+SUD group exhibited significantly higher BAS Total and Fun Seeking scores than the HC group, whereas there were no significant differences between the BD and HC groups. The BD+SUD group also exhibited significantly higher NEM scores than the HC and BD groups. The BP+SUD group exhibited lower CON scores at a trend level compared with the HC group, whereas the BD group exhibited significantly lower CON scores than the HC group. In a binary logistic regression predicting presence of SUD comorbidity among adolescents with BD (n = 49), a model including age, BAS Total, NEM and CON was significant, χ2= 17.58, p = .001, and accurately predicted SUD comorbidity in 90% of cases. Moreover, BAS Total had incremental predictive validity for BD+SUD comorbidity, Wald’s χ2 (1) = 4.72, p = .030, OR = 1.28, after accounting for age alone, and at a trend level, Wald’s χ 2 (1) = 2.97, p = .085, OR = 1.24, after taking into account age, CON, and NEM. The present results suggest that BAS sensitivity, disinhibition, and negative emotionality traits, combined, may be useful in identifying adolescents with BD who are at risk for SUD. Longitudinal studies of adolescent BD are needed to examine prospective associations.