Ya Dong, Xiaohong Huang, Mingyu Hao, Hongxia Li, Dewen Yan
Background The longitudinal impact of diabetes duration on the progression of insulin resistance (IR) in type 2 diabetes (T2DM) remains to be fully elucidated. It is also unclear whether this relationship is uniform across all patients or modified by specific metabolic factors. Objective To investigate the association between diabetes duration and IR, and to identify critical effect modifiers that define high-risk trajectories. Methods This cross-sectional study enrolled 3,309 patients with T2DM, categorized into non-IR (HOMA-IR < 2.31, n = 1,530) and IR (HOMA-IR ≥ 2.31, n = 1,779) groups. Multivariable linear regression was used to assess the independent association between diabetes duration and HOMA-IR. Subgroup and interaction analyses were performed to identify effect modifiers. A two-piecewise linear regression model was employed to delineate threshold effects. Results In the overall cohort, diabetes duration was independently associated with higher HOMA-IR after full adjustment (β = 0.04, 95% CI: 0.03-0.05, P < 0.001). This association was significant only in the IR group (β = 0.04, 95% CI: 0.02-0.06, P < 0.001) and absent in the non-IR group. Subgroup analysis revealed that the TyG index was the only significant effect modifier (β interaction = 0.06, 95% CI: 0.02-0.09, P interaction = 0.001). Crucially, a threshold effect was identified in the high-TyG group: HOMA-IR increased sharply by 0.08 per additional year of diabetes only after 4.0 years ( P < 0.001). Conclusions This study defines a novel, actionable high-risk phenotype in T2DM, characterized by a combination of a TyG index ≥ 9.187 and diabetes duration exceeding 4.0 years, which is associated with a markedly accelerated increase in IR. Given that the TyG index is adjustable, this framework provides a clinically meaningful, personalized strategy for targeted early intervention, enabling precise prevention for those most susceptible to worsening insulin resistance.