Ultra-processed food intake and risk of type 2 diabetes: a pooled analysis of three prospective cohorts of Korean adults and an updated meta-analysis

Studies suggest a positive association between ultra-processed food (UPF) intake and type 2 diabetes risk. However, studies were primarily conducted in Western populations with diets and disease profiles different from populations living elsewhere. In addition, the dose–response relationship needs to be further elucidated. We conducted an individual-level pooled analysis of three Korean prospective cohorts (n = 72,776). UPF intake (in the percentage of g/d as the main UPF unit) was assessed using validated food frequency questionnaires and categorized according to Nova classification. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). We subsequently conducted a meta-analysis of prospective studies, including recent studies from both Western and non-Western populations, to assess the dose–response relationship. In a pooled analysis of Korean cohorts (up to 18 years of follow-up), the highest (vs. lowest) quartile of UPF intake was associated with an increased type 2 diabetes risk (pooled HR [95% CI] = 1.11 [1.02, 1.21] , p-trend = 0.03). The positive association persisted after additional adjustment for BMI, nutritional factors (fiber, sodium, and carbohydrate intakes), or diet quality score. Among individual UPF subgroups, processed meats, ready-to-eat/heat mixed dishes, and ice cream were positively associated with diabetes risk. In a meta-analysis of 17 prospective cohorts, every 10% (of g/d) increment in UPF intake was associated with a 10% (summary RR [95% CI] = 1.10 [1.08, 1.12] ) higher risk in a dose–response fashion. Our meta-evidence supports that higher UPF intake may monotonically increase type 2 diabetes risk.