Estimating Insulin Resistance May Help Predict Stroke, Death in T2D

Sara Freeman

October 21, 2021

Calculating the estimated glucose disposal rate (eGDR) as a proxy for the level of insulin resistance may be useful way to determine if someone with type 2 diabetes (T2D) is at risk for having a first stroke, Swedish researchers have found.

In a large population-based study, the lower the eGDR score went, the higher the risk for having a first stroke became.

The eGDR score was also predictive of the chance of dying from any or a cardiovascular cause, Alexander Zabala, MD, reported at the annual meeting of the European Association for the Study of Diabetes (Abstract OP 01-4).

The link between insulin resistance and an increased risk for stroke has been known for some time, and not just in people with T2D. However, the current way of determining insulin resistance is not suitable for widespread practice.

"The goal standard technique for measuring insulin resistance is the euglycemic clamp method," said Zabala, an internal medical resident at Södersjukhuset hospital and researcher at the Karolinska Institutet in Stockholm.

"For that reason, [the eGDR], a method based on readily available clinical factors – waist circumference, hypertension, and glycosylated hemoglobin was developed," he explained. Body mass index can also be used in place of waist circumference, he qualified.

The eGDR has already been proven to be very precise in people with type 1 diabetes, said Zabala, and could be an "excellent tool to measure insulin resistance in a large patient population."

Investigating the Link Between eGDR and First Stroke Risk

The aim of the study he presented was to see if changes in the eGDR were associated with changes in the risk of someone with T2D experiencing a first stroke, or dying from a cardiovascular or other cause.

An observational cohort was formed by first considering data on all adult patients with T2D who were logged in the Swedish National Diabetes Registry (NDR) during 2004-2016. Then anyone with a history of stroke, or with any missing data on the clinical variables needed to calculate the eGDR, were excluded.

This resulted in an overall population of 104,697 individuals, aged a mean of 63 years, who had developed T2D at around the age of 59 years. About 44% of the study population were women. The mean eGDR for the whole population was 5.6 mg/kg per min.

The study subjects were grouped according to four eGDR levels: 24,706 were in the lowest quartile of eGDR (less than 4 mg/kg per min), signifying the highest level of insulin resistance, and 18,762 were in the upper quartile of eGDR (greater than 8 mg/kg per min), signifying the lowest level of insulin resistance. The middle two groups had an eGDR between 4 and 6 mg/kg per min (40,187), and 6 and 8 mg/kg/min (21,042).

Data from the NDR were then combined with the Swedish Cause of Death register, the Swedish In-patient Care Diagnoses registry, and the Longitudinal Database for Health Insurance and Labour Market Studies (LISA) to determine the rates of stroke, ischemic stroke, hemorrhagic stroke, all-cause mortality, and cardiovascular mortality.

Increasing Insulin Resistance Ups Risk for Stroke, Death

After a median follow-up of 5.6 years, 4% (4,201) of the study population had had a stroke.

"We clearly see an increased occurrence of first-time stroke in the group with the lowest eGDR, indicating worst insulin resistance, in comparison with the group with the highest eGDR, indicating less insulin resistance," Zabala reported.

After adjustment for potential confounding factors, including age at baseline, gender, diabetes duration, among other variables, the risk for stroke was lowest in those with a high eGDR value and highest for those with a low eGDR value.

The corresponding values for risk of ischemic stroke were 0.55, 0.68, and 0.75. Regarding hemorrhagic stroke, there was no statistically significant correlation between eGDR levels and stroke occurrence. This was due to the small number of cases recorded.

As for all-cause and cardiovascular mortality, a similar pattern was seen, with higher rates of death linked to increasing insulin resistance. Adjusted hazard ratios according to increasing insulin resistance (decreasing eGDR scores) for all-cause death were 0.68, 0.75, and 0.82 and for cardiovascular mortality were 0.65, 0.75, and 0.82.

A sensitivity analysis, using BMI instead of waist circumference to calculate the eGDR, showed a similar pattern, and "interestingly, a correlation between eGDR levels and risk of hemorrhagic stroke." Zabala said.

Limitations and Take-Homes

Of course, this is an observational cohort study, so no conclusions on causality can be made and there are no data on the use of anti-diabetic treatments specifically. But there are strengths such as covering almost all adults with T2D in Sweden and a relatively long-follow-up time.

The findings suggest that "eGDR, which may reflect insulin resistance may be a useful risk marker for stroke and death in people with type 2 diabetes," said Zabala.

"You had a very large cohort, and that certainly makes your results very valid," observed Peter Novodvorsky, MU(Hons), PhD, MRCP, a consultant diabetologist in Trenčín, Slovakia.

Novodvorsky, who chaired the session, picked up on the lack of information about how many people were taking newer diabetes drugs, such as the glucagon-like peptide 1 receptor antagonists and sodium glucose-lowering transport 2 inhibitors.

"As we all know, these might have protective effects which are not necessarily related to the glucose lowering or insulin resistance-lowering" effects, so could have influenced the results. In terms of how practical the eGDR is for clinical practice, Zabala observed in a press release: "eGDR could be used to help T2D patients better understand and manage their risk of stroke and death.

"It could also be of importance in research. In this era of personalized medicine, better stratification of type 2 diabetes patients will help optimize clinical trials and further vital research into treatment, diagnosis, care and prevention."

The research was a collaboration between the Karolinska Institutet, Gothenburg University and the Swedish National Diabetes Registry. Zabala and coauthors reported having no conflicts of interest.

This article originally appeared on, part of the Medscape Professional Network.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.