Validation of a laboratory craving assessment and evaluation of 4 different interventions on cravings among adults with overweight or obesity.

ABSTRACT

Food cue reactivity (FCR) is an appetitive trait associated with overeating and weight gain. We developed a laboratory craving assessment to objectively evaluate cognitive aspects of FCR. This study examined the preliminary construct and criterion validity of this craving assessment and evaluated 4 different interventions, 2 of which incorporated cue-exposure treatment for food, on craving over treatment and follow-up. 271 treatment-seeking adults with overweight/obesity (body mass index = 34.6[5.2]; age = 46.5[11.8]; 81.2% female; 61.6% non-Latinx White) completed the Food Cue Responsivity Scale and the laboratory craving assessment, during which they alternated holding and smelling a highly craved food and provided craving ratings over 5 min. Participants were subsequently randomized to 26 treatment sessions over 12-months of ROC, Behavioral Weight Loss (BWL), a combined arm (ROC+) and an active comparator (AC), and repeated the craving assessment at post-treatment and 12-month follow-up. Linear mixed-effects models assessed associations between trial type (holding vs. smelling), trial number, pre-treatment FCR, treatment arm, assessment time point, and craving. Cravings were greater when smelling vs. holding food (b = 0.31, p < 0.001), and cravings decreased over time (b = -0.02, p < 0.001). Participants with higher pre-treatment FCR reported elevated cravings (b = 0.29, p < 0.001). Longitudinally, we observed a significant 3-way interaction in which treatment arm modified the relationship between pre-treatment FCR and craving over time (F(17,5122) = 6.88, p < 0.001). An attenuated FCR-craving relationship was observed in ROC+ and BWL from baseline to post-treatment but was only sustained in BWL at follow-up. This attenuation was also observed in ROC and AC from post-treatment to follow-up. The preliminary validity of this laboratory craving assessment was supported; however, greater craving reductions over time in ROC/ROC+ compared to BWL and AC were not consistently observed, and thus do not appear to fully account for the moderating effect of FCR on weight losses observed in the trial.