Association between CumAvgTyG and clinically recognized diabetes among middle-aged and older adults with prediabetes: a prospective cohort study from CHARLS

The triglyceride-glucose (TyG) index is a practical surrogate of insulin resistance derived from fasting glucose and triglycerides. A single TyG value cannot characterize metabolic burden across repeated assessments. We examined whether the cumulative average triglyceride-glucose index (CumAvgTyG) was associated with clinically recognized diabetes in 2018 among adults with prediabetes. We analyzed prospective data from the China Health and Retirement Longitudinal Study (CHARLS). Prediabetes was identified at the 2011–2012 baseline. The exposure was CumAvgTyG, defined as the arithmetic mean of fasting TyG measured in 2011–2012 and 2015. Participants with diabetes identified by 2015 or with unascertainable diabetes-free eligibility in 2015 were excluded. The outcome was clinically recognized diabetes in 2018, defined by a participant-reported physician diagnosis of diabetes and/or use of Western/modern glucose-lowering medication or insulin. Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs), with adjustment for age, sex, body mass index (BMI), smoking status, drinking status, and baseline HbA1c. Restricted cubic splines with three knots were used to examine the dose-response association. The final analytic cohort included 1,958 participants (mean age, 58.5 years; 53.8% women), of whom 88 (4.5%) had clinically recognized diabetes in 2018. In the fully adjusted model, each 1-SD increase in CumAvgTyG was associated with higher odds of clinically recognized diabetes (OR, 1.50; 95% CI, 1.21–1.86; P < 0.001). Participants in the highest CumAvgTyG tertile had higher odds than those in the lowest tertile (OR, 2.89; 95% CI, 1.56–5.35; P < 0.001), with a significant trend across tertiles (P for trend < 0.001). Restricted cubic splines showed an overall association (P for overall = 0.001) without evidence of nonlinearity (P for nonlinearity = 0.850). Higher CumAvgTyG was associated with greater odds of clinically recognized diabetes in 2018 among middle-aged and older adults with baseline prediabetes. Serial TyG measurements may help characterize longitudinal metabolic risk in prediabetes.