Metabolic heterogeneity at the onset of type 2 diabetes: phenotypes of insulin resistance and β-cell function in the Kazakh population.

BACKGROUND: Type 2 diabetes mellitus (T2DM) is increasingly recognized as a heterogeneous metabolic disorder characterized by varying contributions of insulin resistance and pancreatic β-cell dysfunction. However, data on early metabolic phenotypes at diagnosis remain limited in Central Asian populations. METHODS: In this cross-sectional study, 240 adults with newly diagnosed T2DM were recruited from primary health care facilities in the Aktobe region of Kazakhstan between May 2024 and January 2025. Islet autoantibodies (GAD, IA-2, ZnT8, and ICA) were measured to exclude autoimmune diabetes. After exclusion of seropositive individuals, 178 autoantibody-negative participants were analyzed. Insulin resistance and β-cell function were assessed using the HOMA2 model, and metabolic phenotypes were explored using unsupervised clustering based on BMI, HbA1c, HOMA2-IR, HOMA2-%B, and basal C-peptide. RESULTS: Among the 178 autoantibody-negative individuals, insulin resistance defined as HOMA2-IR ≥1.6 was observed in 59% of participants, whereas β-cell dysfunction defined as basal C-peptide <1.70 ng/mL was present in 30.9%. Unsupervised clustering identified four distinct metabolic phenotypes. Cluster 1 represented the largest subgroup (42.1%) and demonstrated a relatively balanced metabolic profile. Cluster 2 (12.9%) was characterized by obesity-associated hyperinsulinemia with elevated BMI and C-peptide levels. Cluster 3 (25.3%) showed marked hyperglycemia accompanied by reduced β-cell function, whereas Cluster 4 (19.7%) combined increased BMI with relatively preserved β-cell function and lower glycemic burden. CONCLUSIONS: Newly diagnosed T2DM in the Kazakh population demonstrates substantial metabolic heterogeneity already at the time of diagnosis. Identification of distinct metabolic phenotypes using routinely available clinical variables may provide a framework for future studies evaluating precision-based approaches to diabetes management.