Hyperglycemia Linked to Mortality in Older Adults With T2D

MADRID — Time spent in hyperglycemia, not hypoglycemia, was associated with mortality in older adults with type 2 diabetes (T2D), new research found.

Data from the multicenter prospective HYPOAGE study using continuous glucose monitoring (CGM) suggested that as many as a third of adults aged 75 years or older with T2D are spending a higher than recommended amount of time in the hyperglycemic range and that glycemic variability is independently associated with increased mortality.

The findings were presented at the European Association for the Study of Diabetes (EASD) 2024 Annual Meeting by Sarra Smati-Grangeon, MD, PhD, of the Department of Endocrinology, Diabetology, and Nutrition, at the Centre Hospitalier Universitaire de Nantes, Nantes, France.

“Of course, there are a lot of criteria associated with mortality, and maybe time above range is just a [marker] of frailty. But we knew time below range was a criteria, so maybe we also have to be vigilant for the time above range and not leave these patients with a very high level of hyperglycemia because it’s probably also a [risk factor] for infection or other complications,” Smati-Grangeon said in her presentation.

In 2019, recommendations for CGM metrics from the American Diabetes Association (ADA) and endorsed by EASD and other professional societies provided adjusted targets for older or high-risk adults with type 1 diabetes and T2D that focused primarily on preventing hypoglycemia. Specifically, the document recommended < 1% of time spent with glucose levels below 70 mg/dL for older or high-risk people vs < 4% for younger and healthier people and < 10% of time spent with glucose levels above 250 mg/dL for older or high-risk people vs < 5% for younger or healthier people.

Findings from HYPOAGE and other studies suggested that in reality, older adults are spending considerably more time in the high range, with detrimental effects. “Even as it is necessary to avoid hypoglycaemia in the elderly, patients should not be left with chronic hyperglycemia, as there is a high risk of complications, [such as] hyperosmolar coma, ketoacidosis, infection, and mortality. Glycemic monitoring must be continued even if the treatment is less intensive,” Smati-Grangeon told Medscape Medical News. 

‘Hyperglycemia Is Certainly not Benign’

Asked to comment, Medha Munshi, MD, director of the Joslin Geriatric Diabetes Program and professor of medicine at Harvard Medical School, Boston, told Medscape Medical News she agrees with the concern. “We really need to make sure older people are not suffering with hyperglycemia. Even if you don’t think about long-term mortality, hyperglycemia is associated with dehydration, more infection, nonhealing wounds, and decreases in cognition. Hyperglycemia is certainly not benign.”

She also agreed with Smati-Grangeon that “glycemia might be the manifestation of how frailer and sicker and comorbid is the population that they are looking at, rather than by itself a driver of mortality.”

Earlier this year at the ADA Scientific Sessions, Munshi presented data of 65 individuals with diabetes from eight long-term care facilities in the United States, showing that more than half (54%) spent over 10% of the time with glucose levels above 250 mg/dL and that 14% were spending more than 90% of the time in that range. In contrast, just 26% spent more than 1% of the time in hypoglycemic range (< 70 mg/dL).

Hyperglycemia, not Hypoglycemia, Linked to Mortality

In HYPOAGE, a total of 141 adults aged 75 years or older with insulin-treated T2D wore a continuous glucose monitor for 28 days and were followed from September 2017 to July 31, 2023. They had a mean age of 81.5 years, were 56.7% male, and 55.0% were frail. Their average A1c was 7.9% and body mass index (BMI) was 30.3, and 20.6% were living in nursing homes.

Nearly a third (30.5%) had > 10% time above 250 mg/dL, while 72.3% had > 1% time below range (70 mg/dL). Compared with those who spent < 10% of time above 250 mg/dL, those with > 10% were older (83.5 vs 80.6 years; P = .0027), had a lower BMI (28.0 vs 31.3; P = .0037), and had higher A1c (8.6% vs 7.6%; P < .0001).

Those with > 10% time above 250 mg/dL were also significantly less often healthy, more often complex, more likely to be in a nursing home, more often frail, and more likely to have cognitive impairment.

Compared to those who spent less than 10% of time above 250 mg/dL, those with > 10% of time in that range had lower 48-month survival, with hazard ratio 0.59 (P = .05). In contrast, mortality didn’t differ between those with time below range < 1% and those who spent ≥ 1% time below 70 mg/dL (P = .41).

CGM-based metrics significantly associated with mortality were time spent above 250 mg/dL (present in 7.2% of the 83 people still alive vs 11.5% of the 58 deceased; P = .0418), time in the range of 70-180 mg/dL (68.95 of those still alive vs 62.3% of those deceased; P = .0490), and glycemic variability (33.4% vs 36.9%, respectively; P = .0024).

In multivariable analysis, three factors emerged as independent predictors of mortality: Estimated glomerular filtration rate — Chronic Kidney Disease Epidemiology Collaboration value decrease of 1 mL/min (HR, 1.026; P = .0011), the geriatric status of complex vs healthy (2.719; P = .0292), and the glycemic variability coefficient of variation increase of 1% (1.062; P = .0010). Time above the range of > 10% was no longer significant (P = .6994).

Those with coefficient of variation above 36% had significantly higher mortality at 48 months than those with 36% or less variability (HR, 0.57; P = .03).

Smati-Grangeon explained, “in the multivariate analysis, hyperglycemia is no longer associated with mortality, nor is hypoglycemia, variability is. Variability undoubtedly carries more weight because it reflects both hyperglycemia and hypoglycemia.”

A New Proposed Approach

Munshi was a co-author of a recent review paper proposing a revised approach to CGM metrics in older adults, accounting for the heterogeneity of the population and the fact that current automated insulin delivery (AID) systems generally improve time in range, while reducing the risk for hypoglycemia.

The authors suggested complete avoidance of any blood glucose level below 70 mg/dL, with “buffer zones” based on health status. For healthy older adults, they recommended < 4% of time spent with levels 70-90 mg/dL, while for those with intermediate or poor health, the buffer zone would range from 70 to 100 mg/dL.

Also, the upper limit of the defined time in range could be increased from 180 to 200 mg/dL for those with intermediate health status and up to 250 mg/dL for those with poor health status. The authors also suggested changing the target duration of time spent in those upper limits to either < 10% or < 25%, depending on the health status.

“In the near future, with further improvement of AID, it is likely that the time spent in hyperglycemia can be further reduced without increasing risk of hypoglycemia,” they concluded.

The HYPOAGE study was funded by Sanofi. Smati-Grangeon received speaker’s fees from Novo Nordisk, Sanofi, Roche, Asten, and Eli Lilly and conference invitations from Sanofi, Eli Lilly, and Novartis. Munshi consults for Sanofi and Abbott, and her institution received grant funding from Dexcom. 

Miriam E. Tucker is a freelance journalist based in the Washington, DC, area. She is a regular contributor to Medscape Medical News, with other work appearing in the Washington Post, NPR’s Shots blog, and Diatribe. She is on X: @MiriamETucker.