Development of the Highly Processed Food Withdrawal Scale.

Food addiction posits that highly processed foods may be capable of triggering addictive-like symptoms in some individuals, including withdrawal. The current study developed and assessed the psychometric properties of the first self-report measure of highly processed food withdrawal. Individuals (n = 231) aged 19-68 (51.9% female) were recruited online through Amazon Mechanical Turk and reported cutting down on highly processed foods in the past year. The Highly Processed Food Withdrawal Scale (ProWS) was adapted from self-report measures of drug withdrawal and internal consistency and validity were evaluated. Paralleling the course of drug withdrawal, symptoms assessed by the ProWS were reported as most intense between days 2-5 during an attempt to cut down. The ProWS demonstrated convergent validity with addictive-like eating (r = 0.48, p < .001), body mass index (BMI) (r = 0.16, p = .02), and weight cycling (r = 0.29, p < .001) and discriminant validity with dietary restraint: (r = -0.13, p = .04). The ProWS explained 11.2% of variance in self-reported success in last diet attempt beyond addictive-like eating and BMI. The ProWS seems to be a psychometrically sound tool for future research investigating highly processed food withdrawal in humans, and the present data may provide preliminary insight into the plausibility of withdrawal symptoms occurring in response to cutting down on highly processed foods.

Foods are differentially associated with subjective effect report questions of abuse liability.

OBJECTIVES: The current study investigates which foods may be most implicated in addictive-like eating by examining how nutritionally diverse foods relate to loss of control consumption and various subjective effect reports. Subjective effect reports assess the abuse liabilities of substances and may similarly provide insight into which foods may be reinforcing in a manner that triggers an addictive-like response for some individuals. DESIGN: Cross-sectional. SETTING: Online community. PARTICIPANTS: 507 participants (n = 501 used in analyses) recruited through Amazon MTurk. MEASUREMENTS: Participants (n = 501) self-reported how likely they were to experience a loss of control over their consumption of 30 nutritionally diverse foods and rated each food on five subjective effect report questions that assess the abuse liability of substances (liking, pleasure, craving, averseness, intensity). Hierarchical cluster analytic techniques were used to examine how foods grouped together based on each question. RESULTS: Highly processed foods, with added fats and/or refined carbohydrates, clustered together and were associated with greater loss of control, liking, pleasure, and craving. The clusters yielded from the subjective effect reports assessing liking, pleasure, and craving were most similar to clusters formed based on loss of control over consumption, whereas the clusters yielded from averseness and intensity did not meaningfully differentiate food items. CONCLUSION: The present work applies methodology used to assess the abuse liability of substances to understand whether foods may vary in their potential to be associated with addictive-like consumption. Highly processed foods (e.g., pizza, chocolate) appear to be most related to an indicator of addictive-like eating (loss of control) and several subjective effect reports (liking, pleasure, craving). Thus, these foods may be particularly reinforcing and capable of triggering an addictive-like response in some individuals. Future research is warranted to understand whether highly processed foods are related to these indicators of abuse liability at a similar magnitude as addictive substances.