Mean follow-up period was 6.1±0.34 years (median 5.8 years). Subjects in the first tertile showed a significantly lower BMI and NAFLD prevalence, and a more favourable NAFLD score (Table 3). The incidence of obesity, diabetes, IFG, NAFLD, and metabolic syndrome were respectively: 101/1245 (8.1%), 30/1245 (2.4%), 286/1245 (23.0%), 229/1245 (18.4%), and 172/1245 (13.8%). The incidence of obesity significantly increased from the first to third tertile; the incidence of metabolic syndrome and estimated NAFLD were significantly higher in the upper tertile (Table 3). Individuals who consumed almost half of their daily caloric intake at dinner were 2-fold more likely to become obese in a multiple logistic regression model, after adjusting for age, sex, BMI at baseline, physical activity, living in a rural area, intake of total energy and SFA, and skipping breakfast (Table 4). In the same model, other variables were significantly associated with the incidence of obesity: the physical activity level expressed as METS h/week (OR = 0.94; 95% CI 0.91–0.97; p