Generalization of food devaluation following food-specific go/no-go training

The benefit of food-specific inhibition training on modulating food valuation and eating behaviors has been established, but generalization to untrained foods is seldomly examined. This study investigated whether stimulus variability and practice order, found to effect generalization in motor learning, can improve generalization following food-specific inhibition training. Ninety-three young adults practiced the Go/No-Go task online in three training conditions: 1) Constant (N = 30): inhibition practiced on one food stimulus; 2) Variable-Blocked (N = 32): inhibition practiced on 6 food stimuli, each in a separate block; and 3) Variable-Random (N = 31): inhibition practiced on 6 food stimuli in random order. Consistent with our hypothesis, the Variable-Random group showed better generalization of inhibition to untrained foods than the Constant and the Variable-Blocked groups immediately after training, demonstrating the benefit of stimulus variability and random practice order. This effect was not present 24 h after training. The Variable-Random group also showed decreased desire to eat untrained foods, exhibiting generalization of food devaluation. However, this effect was only present 24 h after training. The Constant group showed increased desire to eat untrained foods immediately and 24 h after training. The Variable-Blocked group did not differ from either group in the desire to eat to untrained foods, suggesting that random order is important for exposing the benefit of variability. The findings illustrate that presenting various training items in random order can improve generalization of food-specific inhibition training. However, inconsistencies found in the timing of generalization effects and modest effect sizes warrant additional investigation into generalization principles of food-specific inhibition training.