Understanding biological pathways preceding type 2 diabetes may support precision prevention strategies. An analysis of four U.S. prospective cohorts evaluated whether seven pathophysiological biomarkers were associated with type 2 diabetes and its clinical subtypes before diagnosis. The analysis was recently published in the Diabetes Care.
The study included 69,725 observations from 9,661 adults without diabetes, followed for a median of 10 years. During follow-up, 1,569 individuals developed type 2 diabetes. Time-dependent Cox models assessed associations between biomarkers—including BMI, systolic blood pressure (SBP), HbA1c, LDL cholesterol, homeostatic indices of β-cell function (HOMA2-%B) and insulin resistance (HOMA2-IR), and estimated glomerular filtration rate (eGFR)—and incident diabetes and subtypes.
Higher BMI (HR 1.03 [95% CI 1.02, 1.04] per 1 kg/m²), systolic blood pressure (HR 1.09 [95% CI 1.05, 1.13] per 10 mmHg), HbA1c (HR 2.46 [95% CI 1.73, 3.52] per 1%), HOMA2-IR (HR 1.92 [95% CI 1.78, 2.07] per unit), and LDL cholesterol (HR 1.02 [95% CI 1.00, 1.04] per 10 mg/dL), along with lower HOMA2-%B (HR 0.89 [95% CI 0.87, 0.91] per 10 units), were independently associated with incident diabetes. Estimated glomerular filtration rate (HR 1.02 [95% CI 0.98, 1.07] per 10 mL/min/1.73 m²) was not significantly associated. Associations were broadly similar across diabetes subtypes.
Models using single-time-point biomarker measurements demonstrated good discrimination (C-index 0.81–0.90), modest F1 scores (0.26–0.51), and variable calibration with slope estimates ranging from 0.22 to 1.24.
As an observational analysis, these findings demonstrate association rather than causation and indicate that single-time-point prediagnostic biomarker measurements could not reliably distinguish type 2 diabetes subtypes at diagnosis.
