Fight, Flight, - Or Grab a Bite! Trait Emotional and Restrained Eating Style Predicts Food Cue Responding Under Negative Emotions

Abstract

In today's society, obesity rates are rising as food intake is no longer only a response to physiological hunger signals that ensure survival. Eating can represent a reward, a response to boredom, or stress reduction and emotion regulation. While most people decrease food intake in response to stress or negative emotions, some do the opposite. Yet, it is unclear who shows emotional overeating under which circumstances. Emotion regulation theories describe emotional overeating as a learned strategy to down-regulate negative emotions. Cognitive theories, by contrast, attribute emotional overeating to perceived diet breaches in individuals who chronically attempt to diet. After consuming "forbidden foods", they eat more than individuals who do not restrict their food intake. This laboratory study investigated emotional overeating by exposing individuals to a personalized emotion induction while showing images of palatable foods. Outcome variables indexed cue reactivity to food images through picture ratings (valence, desire to eat), facial expressions (electromyography of the corrugator supercilii muscle), and brain reactivity by detecting event-related potentials (ERPs) by means of electroencephalography (EEG). The influence of emotion condition (negative, neutral) and individual differences (self-reported trait emotional and restrained eating) on outcome variables was assessed. Valence ratings and appetitive reactions of the corrugator muscle to food pictures showed a relative increase in the negative condition for individuals with higher emotional eating scores, with the opposite pattern in lower scores. Desire to eat ratings showed a similar pattern in individuals who showed a strong response to the emotion induction manipulation, indicative of a dose-response relationship. Although no differences between conditions were found for ratings or corrugator activity with restrained eating as a predictor, an ERP at P300 showed increased activation when viewing food compared to objects in the negative condition. Findings support emotion regulation theories: Emotional eaters showed an appetitive reaction in rating patterns and corrugator activity. EEG findings (increased P300) suggest a motivated attention toward food in restrained eaters, which supports cognitive theories. However, this did not translate to other variables, which might demonstrate successful restraint. Future studies may follow up on these findings by investigating eating disorders with emotion regulation difficulties.

Keywords: P300; corrugator supercilii; emotional eating; food cue reactivity; mood induction; multilevel modeling; restrained eating.

FIGURE 1

Moderators, task conditions, and outcome measures for appetitive responding. NER, negative emotional reactivity; NA, negative affect; DEBQ, dutch eating behavior questionnaire; ERP, event-related potential; EMG, electromyography; ECG, electrocardiogram.

FIGURE 2

Trial sequence exemplifying two idiosyncratic scripts of the negative and neutral condition with interleaved food/object image presentations alongside respective ratings in the negative condition (red box) and the neutral condition (blue box). The 52 different pictures were divided up evenly among the eight different sentences, resulting in six to seven picture presentations after a sentence presentation. DTE, desire to eat.

FIGURE 3

Individual negative affect scores at different time points.

FIGURE 4

Picture responses with simple slopes illustrating different patterns in the two conditions between high and low emotional eaters for (A), pleasantness ratings (difference score Food – Objects), (B), desire to eat ratings (food only), and (C), activity of corrugator supercilii (difference score Food – Objects). NER, negative emotional reactivity.

FIGURE 5

Two-way interaction of (A), restrained eating (±1SD) and condition, and (B), NER (±1SD) and condition at P300. The circled area in the head plot shows the extracted cluster, the blue bar in the waveform graph illustrates the P300 peak observed in the 200–400 ms time window.

FIGURE 6

Time course of the corrugator supercilii muscle responses to images as a function of conditions and picture type (mean amplitude of 500 ms segments). BL, Baseline.