A persistently low estimated glucose disposal rate over six years of follow-up was associated with a substantially higher risk of developing type 2 diabetes in a Chinese community cohort, according to a prospective study published in Diabetology & Metabolic Syndrome. Adults whose eGDR stayed on a low-stable trajectory had an adjusted hazard ratio of 1.72 (95% CI 1.21–2.44) for incident type 2 diabetes compared with those whose eGDR remained high and stable, while a smaller group following a moderate-curvilinear, U-shaped course showed an even higher point estimate of 2.82 (95% CI 1.76–4.53), although the researchers themselves flagged that second association as preliminary.
The estimated glucose disposal rate is a validated surrogate measure for insulin resistance, the metabolic disturbance that sits at the centre of type 2 diabetes. It is calculated from routinely available clinical variables, which makes it attractive for large community studies where direct measures of insulin sensitivity are impractical. Most previous evidence linking eGDR to new-onset diabetes has relied on a single baseline measurement, which cannot capture whether a person’s insulin sensitivity is stable, deteriorating, or fluctuating over time. The new study set out to fill that gap by tracking repeated eGDR values and modelling them jointly with the time to a diabetes diagnosis.
The analysis drew on 2,140 adults who were free of diabetes at baseline and completed four biennial examinations between 2017 and 2023. Rather than treating eGDR as a fixed exposure, the investigators used a joint latent class mixed model with a Weibull baseline hazard, a statistical framework that simultaneously describes how eGDR evolves within individuals and how those evolution patterns relate to the hazard of incident type 2 diabetes. The number of latent trajectory classes was selected using the Bayesian information criterion, entropy, posterior probabilities, and interpretability, a combination intended to balance statistical fit against clinically meaningful groupings.
Three distinct eGDR trajectories emerged. The majority of participants, 69.2%, followed a High-Stable pattern, maintaining consistently high estimated glucose disposal rates throughout follow-up. A further 20.5% stayed on a Low-Stable trajectory, with persistently low values, and 10.3% followed a Moderate-Curvilinear course, a U-shaped pattern in which eGDR started close to the reference level and then dipped before recovering. Because the joint model links the longitudinal trajectory directly to the time-to-event outcome, each participant’s class membership could be related to their risk of a new diabetes diagnosis within a single coherent analysis.
After adjustment for age, sex, education, income, body mass index, smoking, alcohol use, physical inactivity, and family histories of hypertension and diabetes, both non-reference trajectories carried elevated risk relative to the High-Stable group. The Low-Stable association proved the more robust of the two. It held after the researchers excluded participants with baseline HbA1c of 6.0% or higher, under a two-class model, and across the pre-specified subgroups defined by age, sex, BMI, baseline glycaemic status, and hypertension. It was, however, attenuated to non-significance in two sensitivity analyses: one that lagged the exposure relative to the event, and one that redefined incident diabetes using fasting glucose alone.
The Moderate-Curvilinear trajectory told a more cautionary tale. Despite carrying the highest point estimate in the main model, its members had a baseline eGDR close to that of the reference group, and the association reversed direction when early events were excluded from the analysis. The authors concluded that this association was concentrated in the early follow-up period and was not sustained once early events were removed, and they explicitly recommended treating it as preliminary rather than as an established risk signal.
The study team subjected their findings to an unusually thorough battery of sensitivity checks. Beyond the HbA1c exclusion, the lag analysis and the fasting-glucose-only outcome definition, they tested the robustness of the trajectory classification itself, varied the handling of missing data, and altered how post-diagnosis measurements were treated. None of the pre-specified subgroup analyses showed significant effect modification, with all P-interaction values above 0.05, meaning the Low-Stable association did not differ detectably across age, sex, BMI, baseline glycaemic status, or hypertension strata.
The authors summarised their conclusions carefully. In this community-based cohort, eGDR exhibited three distinguishable longitudinal trajectory patterns, and within the joint modelling framework a persistently low trajectory and a U-shaped trajectory were both independently associated with incident type 2 diabetes. Yet they stressed that the Moderate-Curvilinear association was concentrated early in follow-up and did not survive the exclusion of early events, and they called for validation in independent populations with longer follow-up, more diverse ethnic compositions, and interval-censored analytical frameworks before such trajectory-based measures are used to inform clinical risk stratification.
For now, the practical takeaway is methodological as much as clinical. The study demonstrates that repeated eGDR measurements, modelled jointly with survival time, can reveal risk patterns that a single baseline reading would miss, and it identifies a persistently low trajectory as the signal most likely to withstand further scrutiny. Whether that signal translates into a usable screening tool for community populations will depend on the independent validation the authors say is still required.
Subject of Research: Longitudinal trajectories of estimated glucose disposal rate and their association with incident type 2 diabetes in a prospective Chinese community cohort.
Article Title: Longitudinal trajectories of estimated glucose disposal rate and risk of incident type 2 diabetes in Chinese adults: a prospective community-based cohort study
Article References: Luo, X., Hu, J., Liu, K., Lan, J., Liao, W., & Wu, Q. (2026). Longitudinal trajectories of estimated glucose disposal rate and risk of incident type 2 diabetes in Chinese adults: a prospective community-based cohort study. Diabetology & Metabolic Syndrome. https://doi.org/10.1186/s13098-026-02317-7
Image Credits: AI Generated
DOI: 10.1186/s13098-026-02317-7
Keywords: estimated glucose disposal rate, insulin resistance, type 2 diabetes, longitudinal trajectories, community cohort, joint latent class mixed model, hazard ratio, Chinese adults, prospective study, risk stratification, Longitudinal, trajectories
News Source: Ophelia Keating. (October 30, 2026). Persistently low glucose disposal rate trajectory flags higher type 2 diabetes risk in Chinese adults. Scienmag.
