Cultural and Sex-Related Differences in Free-Word Associations with "Sweets": A Multinational Online Study

Background: Free-word association (FWA) captures the most accessible lexical responses to a stimulus, providing a window into automatic cognitive representations of food categories that may differ across cultures and between sexes. Objectives: To examine whether the dominant cognitive associations evoked by the word "sweets" differ across three language groups (Hungarian, English, Spanish), whether they vary by sex, and whether they relate to body mass index (BMI) and self-reported eating disorder risk. Methods: A total of 1349 participants completed an online survey including a single FWA prompt. Responses were classified into 10 semantic categories and analyzed using chi-square tests. Effect sizes (Cramér's V) were reported for all tests. Height and weight were converted to uniform metric units, and BMI was calculated. Results: The association profile differed significantly across language groups (χ²[18] = 210.05, p < 0.001, V = 0.28). Chocolate dominated Hungarian responses, while Baked goods/Desserts and Sugar/Candy dominated English, and Positive emotion and Baked goods/Desserts predominated among Spanish speakers. Sex differences were significant overall (χ²[9] = 43.72, p < 0.001, V = 0.18). BMI distributions differed markedly across nations (χ²[6] = 157.17, p < 0.001, V = 0.26), and sweets categories were significantly associated with eating disorder risk (χ²[27] = 48.04, p = 0.008, V = 0.11); however, this result should be interpreted with caution given the extreme skew toward the lowest-risk category [84.2%], with Negative/Health associations overrepresented among higher-risk participants. Conclusions: Cultural context plays a significant role in shaping automatic cognitive associations with sweet foods. The exploratory association between sweets categories and self-reported eating disorder risk warrants further investigation using validated instruments before any substantive conclusions can be drawn.