Abstract
BACKGROUND: Ultraprocessed food (UPF) intake is associated with increased type 2 diabetes (T2D) risk. No controlled feeding trial has investigated UPF exposure and T2D risk or performed chemical analysis of UPF study diets. OBJECTIVES: To design and chemically validate nutritionally matched high- and non-UPF diets and to examine the effects of a 6-wk high-UPF diet on T2D risk in midlife adults. METHODS: High-UPF (81% UPF energy) and non-UPF (0% UPF energy) diets were designed and chemically analyzed to validate energy, macro- and micronutrients, and other dietary components. Plant-based ingredients in the diet were also assessed by FoodSeq. After a 2-wk standardized lead-in diet (59% UPF), 18 adults aged 40 to 65 y were randomly assigned to a eucaloric high-UPF or non-UPF diet for 6 wk. Insulin sensitivity and 24-h glycemic control were measured at baseline and post intervention. Serum global metabolomic profiles were evaluated. RESULTS: The high-UPF and non-UPF diets were well-matched and consistent with planned energy and nutrient targets. FoodSeq indicated that the high-UPF diet showed more frequent detection of UPF ingredients, such as guar (gum) and corn (corn starch and corn meal). There were no changes in Matsuda Index or HOMA-IR in the pilot trial, however glucose AUC (high-UPF: 13431 ± 3914 mg·min/dL to 13656 ± 4005 mg·min/dL; non-UPF: 15349 ± 4068 mg·min/dL to 14,221 ± 3722 mg·min/dL, P = 0.054; ES = 0.52) and mean amplitude of glycemic excursions (MAGE) (high-UPF: 37.6 ± 10.1 mg/dL to 40.2 ± 7.3 mg/dL; non-UPF: 44.5 ± 11.0 mg/dL to 39.3 ± 9.5mg/dL, P = 0.055; ES = 0.51) tended to worsen in the high-UPF vs non-UPF groups. In the non-UPF group, reductions in the food contact chemical 2,4-ditert-butylphenol and in the thermal food processing by-product N6-carboxymethyllysine were observed. CONCLUSIONS: Findings provide preliminary evidence that reducing UPF may reduce T2D risk. Large-scale trials are warranted to evaluate causal effects.