Ultra-processed food consumption and body mass index development from childhood to adolescence in Dutch cohorts

Ultra-processed food (UPF) consumption has been related to an increased risk of overweight/obesity in adults, but evidence in children and adolescents remains limited. This study examined associations between UPF intake and subsequent body mass index (BMI) development across childhood and adolescence in two Dutch cohorts. Additionally, the potential temporal direction of relationship between UPF intake and BMI z-scores was explored. Of 1091 children (50.2% male) from the Dutch GECKO Drenthe cohort at age 3, and of 2,970 children (50.0% male) from the Lifelines cohort at age 8–12, dietary intake was assessed using a validated Food Frequency Questionnaire. Food items were categorized according to the NOVA4 classification. Height and weight, measured by trained nurses at 10/11 years in GECKO and 9–17 years in Lifelines, were used to compute age- and sex-standardized BMI z-scores. Overweight/obesity status was categorized by IOTF criteria. Regression models examined the association with BMI development in these two cohorts. Cross-lagged panel models assessed the temporal association in the GECKO cohort. The median[IQR] UPF intake at age 3 in GECKO was 744[567–961] g/d, accounting for 52% of total food intake by weight, while in Lifelines (ages 8–12), it was 908[719–1166] g/d, representing 50% of total intake. In GECKO, children in the highest quartile of UPF consumption showed a greater increase in BMI z-scores between ages 3 and 10/11 compared with those in the lowest quartile (b = 0.22; 95%CI: 0.07 to 0.37). In contrast, in Lifelines, higher UPF intake at ages 8–12 was associated with lower BMI z-score at ages 9–17 (Q1 as reference, b[Q2] = −0.12, 95%CI: [−0.22 to −0.02]; b[Q3] = −0.20, 95%CI: [−0.30 to −0.09] and b[Q4] = −0.13, 95%CI: [−0.24 to −0.03]). Cross-lagged analyses indicated a directional association, with early UPF intake predicting later BMI z-score. Higher UPF intake was prospectively associated with greater increase in BMI z-score in young children. Although an inverse association with BMI z-score was observed in adolescents, the evidence for this association was less consistent. These findings suggest that early childhood may be a potentially sensitive period for the association between UPF consumption and weight gain and underscore the need for further longitudinal research across developmental stages.