Neurobehavioural correlates of body mass index and eating behaviours in adults: a systematic review

Abstract

The worldwide increase in obesity has spurred numerous efforts to understand the regulation of eating behaviours and underlying brain mechanisms. These mechanisms can affordably be studied via neurobehavioural measures. Here, we systematically review these efforts, evaluating neurocognitive tests and personality questionnaires based on: (a) consistent relationship with obesity and eating behaviour, and (b) reliability. We also considered the measures' potential to shed light on the brain mechanisms underlying these individual differences. Sixty-six neurocognitive tasks were examined. Less than 11%, mainly measures of executive functions and food motivation, yielded both replicated and reliable effects. Several different personality questionnaires were consistently related to BMI. However, further analysis found that many of these questionnaires relate closely to Conscientiousness, Extraversion and Neuroticism within the Five-Factor Model of personality. Both neurocognitive tests and personality questionnaires suggest that the critical neural systems related to individual differences in obesity are lateral prefrontal structures underpinning self-control and striatal regions implicated in food motivation. This review can guide selection of the highest yield neurobehavioural measures for future studies.

Figure 1. Visual overview of neurocognitive measures and their possible links with obesity and weight-related appetitive behaviours

The results of the systematic search are depicted in Figure 1. The major domains are positioned in the centre of the circle. Some of the domains are further broken into subdomains, when necessary. Each rectangle corresponds to a single neurobehavioural task. The numbers correspond to task numbers in Table 1. The length of the rectangle reflects the number of studies conducted with this task, and the colour reflects the overall outcome. Studies with replications have a separate colour scheme from studies with no replications (see legend). Asterisks indicate tasks that use food stimuli, as opposed to generic stimuli, and rectangles in bold indicate tasks that are discussed in more detail in this paper. Arrows indicate if task has been tested in a longitudional design. * = task uses food stimuli; ↑↓= Outward arrow – task performance has been tested as a predictor of BMI change. Inward arrow – BMI change has been tested as a predictor of task performance; GNG= go/no go; IAT = Implicit Association Test; IGT = Iowa Gambling Task; maze = Austin Maze; ns = not significant; RRVf = Relative reinforcing value of food; s = significant; span = Computational span; SST = stop-signal test; WCST = Wisconsin Card Sorting Test