Questionnaire and Behavioral Task Measures of Impulsivity are Differentially Associated With Body Mass Index: A Comprehensive Meta-Analysis

Although impulsivity has been implicated in the development and maintenance of obesity, evidence linking impulsivity to obesity has been mixed. These mixed findings may be related to differences in the type of impulsivity measures used and the varied domains of impulsivity assessed by each measure. The present meta-analysis aimed to examine the impact of measurement selection on the relationship between impulsivity and body mass index (BMI). A total of 142 articles met inclusion criteria and were comprised of 315,818 participants. Effect sizes consisted of Fisher's z-transformed correlation coefficients, which were weighted by the inverse variance to establish the grand mean estimate of the relationship between impulsivity and BMI. Overall weighted mean effect sizes also were computed for each type and domain of impulsivity measure. Moderator analyses were conducted using a mixed-effects approach to determine if the relationship between impulsivity and BMI varied between the types of impulsivity measures used. On average, participants were 32.25 (SD = 12.41) years of age, with a BMI of 26.63 (SD = 5.73) kg/m(2). The overall relationship between impulsivity and BMI was small but significant (r = .07). Behavioral task measures of impulsivity produced significantly larger effect sizes (r = .10) than did questionnaire measures of impulsivity (r = .05). Domains of impulsivity that assessed disinhibited behaviors (r = .10), attentional deficits (r = .11), impulsive decision-making (r = .10), and cognitive inflexibility (r = .17) produced significant effect sizes. These meta-analytic findings demonstrate that impulsivity is positively associated with BMI and further document that this association varies by the type of impulsivity measure used and the domain of impulsivity assessed.