Development of the Leeds Food Preference Questionnaire in Japanese: Sensitivity and reproducibility of liking and wanting for food in fasted and fed states

The Leeds Food Preference Questionnaire (LFPQ), a computer-based task for measuring reward responses (liking/wanting) and preferences for images of food, is a widely used tool. However, no cultural adaptation studies to date have addressed its validity and repeatability in a test–retest design. The present study aimed to develop a Japanese version of the LFPQ (LFPQ-J); examine its outcomes under fasted and fed states; and test its reproducability after one week. An online survey containing foods that were either low-fat sweet, high-fat sweet, low-fat savoury or high-fat savoury was first conducted among a sample of 200 Japanese adults (100 men and 100 women) to develop and validate a culturally appropriate food image database. Sixty participants (30 men and 30 women) then participated in two identical trials where they completed the LFPQ-J under fasted and fed states (immediately after a standardised meal), at least one week apart. The absolute difference within the participants in scores for explicit liking, explicit wanting, implicit wanting and relative preference between the trials was analysed using Bland-Altman plots and Pearson’s or Spearman’s correlation coefficients. In the fasted state for each food category, 91.1 to 96.4 % of the data were plotted within the 95% limits of agreement and intra-personal correlation were 0.58–0.81. In the fed state for each food category, 91.1 to 98.2 % of the data were plotted within the 95% limits of agreement and intra-personal correlation were 0.40–0.83. The present study demonstrates that the LFPQ-J is a sensitive and reproducible instrument for the evaluation of liking and wanting for food varying in fat content and sweet taste in Japanese adults.