Mingyan Li

Dr. Mingyan Li


Research Focus & Metrics

Representative Publications

View All Publications
4 November 2026 in medRxiv
Verified
Preprint
Open access
PDF file
RPRong-Shuai Pan
Rong-Shuai Pan
XLXicong Li
Xicong Li
ZWZi-Hang Wei
Zi-Hang Wei
JZJie Zou
Jie Zou
ZCZe-Ru Chen
Ze-Ru Chen
MLMingyan Li
Mingyan Li
Background: Estimated glucose disposal rate (eGDR) is an accessible surrogate of insulin sensitivity and has been associated with cardiovascular disease (CVD) risk. However, whether longitudinal eGDR trajectory patterns convey prognostic information beyond cumulative eGDR exposure remains unclear. We aimed to characterize distinct eGDR trajectories and evaluate their associations with incident CVD in two prospective cohorts. Methods: We included 4,823 CVD-free adults aged 45 years or older from the Health and Retirement Study (HRS, n=3,258) and the English Longitudinal Study of Ageing (ELSA, n=1,565). Latent class mixed models identified four eGDR trajectory classes across three waves. Associations with incident CVD were examined using multivariable Cox models. We further fitted joint models including both trajectory class and cumulative average eGDR to assess their independent contributions. Results: Compared with the persistently high trajectory, the persistently low and declining trajectories were associated with higher CVD risk in both cohorts. In joint models, whereas trajectory class showed residual prognostic signal, particularly in HRS, with directionally consistent but less precise estimates in ELSA. In HRS, adding trajectory class improved model fit beyond cumulative average eGDR alone, whereas adding cumulative average eGDR to the trajectory model did not. Direct pairwise comparisons showed that the declining-then-rising trajectory was associated with lower risk than the declining trajectory in HRS, with directionally similar but imprecise estimates in ELSA. Conclusions: Longitudinal eGDR trajectory patterns may convey prognostic information not fully captured by cumulative average eGDR alone. These findings suggest that the temporal pattern of insulin sensitivity may be clinically relevant in addition to cumulative exposure, while the potentially favorable profile of recovery trajectories warrants cautious interpretation and external replication. Graphical