LAUSR

Abstract Food preferences are influenced by a large number of physiological, nutritional, environmental, sociocultural and genetic factors. From previous studies, survey-reported food likings may reflect habitual intake, being cognitively simple to recall than intake, and have associated with risk factor for diet-related diseases. This work was secondary analysis of a large dataset of personal, clinical and lifestyle measures from 3219 Italian adults to investigate: 1) factors associated with food liking; 2) the relationship between food liking and metabolic phenotypes. Seven food liking groups (alcoholic beverages, cheeses, fish, fruit, meat, sweets foods and vegetables) were defined based on a questionnaire completed by each participant. Multivariate regression analyses showed that factors such as age, gender, education and food behaviour were related with food liking groups. For example, males reported significantly greater liking for alcoholic beverages, fish and meat and a lower liking for vegetables, fruit and sweet foods. Greater liking for all foods was associated with greater self-reported willingness to try unfamiliar foods (food adventurousness). In multivariate regression models and independently from other measures, food liking groups were associated with adiposity, serum lipids, fasting glucose and systolic blood pressure. For instance, greater adiposity was associated with higher liking for meat and cheeses, while higher HDL-cholesterol was associated with higher vegetables liking. Our results from a large adult cohort highlight that many factors may shape survey-reported food liking and suggest their link with metabolic measures.