Multidimensional elements of impulsivity as shared and unique risk factors for food addiction and alcohol misuse

Food addiction (FA) and alcohol misuse tend to co-occur, which suggests shared factors in the etiology and persistence of these health behaviors. One shared factor that has been linked to both is impulsivity, a multidimensional construct reflecting multiple facets of self-regulatory capacity. However, co-occurrence also raises issues of possible confounding if both domains are not considered concurrently, and the intersection between FA, alcohol misuse, and impulsivity has not been well characterized empirically. Therefore, the current study examined the intersection of FA, alcohol consequences, and multiple indicators of impulsivity. Participants were emerging adults reporting regular heavy episodic drinking recruited from Hamilton, Ontario (N = 730; ages 19.5–23). Participants completed measures of FA, alcohol problems, impulsive personality traits (i.e., Barratt Impulsiveness Scales, UPPS-P Impulsive Behavior Scales), impulsive choice (i.e., delay discounting), impulsive action (i.e., Go/NoGo task). Findings revealed a significant association between FA and alcohol-related consequences and both shared similar associations with specific impulsive personality traits (Positive and Negative Urgency, Lack of Premeditation, Motor and Attentional Impulsivity). However, alcohol-related consequences were uniquely associated with other impulsive personality traits (Lack of Perseverance, Sensation Seeking, Non-planning impulsivity) and impulsive choice, and FA was uniquely associated with impulsive action. Beyond alcohol-related consequences, FA was associated with additional variance in measures of impulsive personality traits (Positive and Negative Urgency, Lack of Premeditation, Motor Impulsivity, and Attentional Impulsivity) and impulsive action, but not impulsive choice. Overall, the current study reveals several common self-regulatory processes associated with both adverse drinking consequences and FA, and that the risk of inadvertent confounding appears to be limited