Abstract
BACKGROUND: Evidence regarding health effects of ultra-processed food (UPF) in children is scarce and contradictory. Therefore, the aim was to assess the associations between UPF intake and anthropometric measurements and biomarkers of cardiometabolic health in 8-year-old children in Sweden. METHODS: This cross-sectional analysis was conducted at the follow up of N = 167 8-year old children born within the Swedish pregnancy cohort study GraviD. The children’s diet was assessed through a questionnaire and nutritional data derived using the Swedish Food Agency Food composition database. UPF intake was defined according to the NOVA framework. Participants were categorized into sex-specific tertiles of UPF intake, based on their energy-adjusted total UPF intake expressed as servings per day. Intake of UPF from the following subgroups was also categorized: beverages, sweets/snacks, bread/cereal, and protein/meals. Anthropometric measurements included body mass index (BMI), waist circumference and fat mass measured by bioimpedance. Biomarkers of cardiometabolic health were measured in non-fasting venous blood samples and measurements of blood pressure. Total UPF intake and subgroups of UPF were assessed in relation to the outcomes using adjusted linear regression analysis. RESULTS: Mean ± SD intake of UPF was 6.1 ± 2.1 servings/day and contributed to 56 ± 11% of the children’s total energy intake. Neither intake of total UPF nor subgroups of UPF was associated with anthropometric measurements. Participants in the highest tertile of total UPF intake had lower concentrations of low-density lipoprotein (LDL) cholesterol compared to those in the lowest tertile (-0.28 [-0.51– -0.04]) and lower ApoB/ApoA1 ratio (-0.07 [-0.12– -0.03]). Higher intake of UPF beverages was associated with higher C-reactive protein, while higher intake of UPF sweets/snacks was associated with higher High-Density Lipoprotein, higher ApoA1 and lower ApoB/ApoA1 ratio. Higher intake of UPF protein/meals was associated with lower total cholesterol, LDL and ApoB/ApoA1 ratio. Higher intake of UPF bread/cereal did not show any significant associations. CONCLUSION: In this cross-sectional study of Swedish 8-year-old children, UPF intake was not associated with anthropometric measures. Higher intake of total UPF and certain UPF subgroups, particularly protein/meals, was associated with a more favorable blood lipid profile. These findings underscore the heterogeneity of foods classified as UPF and suggest that their cardiometabolic associations in children may differ substantially by subgroup. TRIAL REGISTRATION: The GraviD-Child study was registered at www.clinicaltrials.gov (NCT05228925, 2022–01-27).