We assessed associations of total and group-specific ultra-processed food (UPF) intake with type 2 diabetes (T2D) incidence in the Hispanic Community Health Study/Study of Latinos, overall and with a priori stratification by Hispanic/Latino heritage, US nativity, BMI categories, and diet quality measured by tertiles of the Alternative Healthy Eating Index (AHEI). Data entailed 8010 adults, completing follow-up (2014-17), and without baseline (2008-11) diabetes, cardiovascular disease, and/or cancer (mean follow-up: 5.8 years). Baseline 24-h dietary recalls and food propensity questionnaires assessed UPF intakes (% of diet weight) defined by the Nova framework. Self-report, anti-diabetes medications, and standard biomarker thresholds defined incident T2D. Poisson models adjusted for sociodemographic, acculturation, lifestyle, energy intake, and diabetes family history variables estimated incidence rate ratios (95% CI) of T2D per 10-percentage-point increase in total UPF intake and 1-SD increase in inverse normal transformed UPF group intakes. UPF constituted 19.8% of diet weight. Each 10-percentage-point increase in UPF was associated with 1.16 (1.01-1.34) higher T2D incidence overall (1273 cases). In stratified analyses, associations remained only for Puerto Ricans (1.22; 1.01-1.47), US-born persons (1.26; 1.03-1.54), and those with overweight/obesity (1.16; 1.01-1.35). Regarding UPF groups, artificially-sweetened beverages and processed poultry/fish were associated with a higher incidence of T2D in the overall sample. Among Puerto Ricans, US-born participants, and persons in the lowest AHEI tertile, ultra-processed bread/cereals, soft drinks, poultry/fish, and hard liquors were associated with a higher incidence of T2D, whereas ultra-processed sweet snacks, desserts, and spreads and savory snacks were associated with lower incidence. Limiting UPF is associated with lower T2D incidence among US Hispanics/Latinos, particularly Puerto Rican and US-born. Attention to group-specific UPF intake is warranted.
Mendoza et al. (Tue,) studied this question.