Should A1c remain the gold standard for routine monitoring of glucose control in people with diabetes, or is it time to shift to “time in range” and other metrics of continuous glucose monitoring (CGM)? 

This question was debated at the 19th International Conference on Advanced Technologies and Treatments for Diabetes (ATTD) 2026, with Elizabeth Selvin, PhD, director of the Welch Center for Prevention, Epidemiology and Clinical Research and a professor of epidemiology and medicine at Johns Hopkins University, Baltimore, arguing in favor of A1c.

“Hemoglobin A1c should remain the primary metric for evaluating glucose control in diabetes,” said Sevin, who pointed out that the advantages of this method include low cost, wide availability, strict standardization, no need for fasting or timed samples, and imperviousness to acute factors.

But Richard M. Bergenstal, MD, executive director of the International Diabetes Center, a division of HealthPartners Institute in Minneapolis, countered that while A1c has served well, CGM technology — by providing metrics including “time in range” and “time below range” — offers the opportunity to improve glycemic control rather than simply monitor it. 

A1c: Fulfilling the Criteria for a ‘Gold Standard’

A1c is “the standard measure that we have used for decades to guide and adjust diabetes treatment,” with “very robust associations with long term macro- and microvascular outcomes,” Sevin told meeting attendees. It has also been a surrogate endpoint in clinical trials of glucose-lowering therapies for both type 1 and type 2 diabetes, she noted.

Importantly, she added, A1c fulfils five key criteria for a “gold standard” measure: It predicts clinical outcomes, is validated in randomized clinical trials, is standardized and reproducible, is accessible across healthcare systems, and allows for population surveillance. 

Evidence linking A1c to long-term outcomes spans decades, said Selvin, citing the following trials: the landmark Diabetes Control and Complications Trial (DCCT), in type 1 diabetes (T1D); and the United Kingdom Prospective Diabetes Study, the Action to Control Cardiovascular Risk in Diabetes study, and the ADVANCE trial in type 2 diabetes (T2D). All of these trials showed that lowering A1c reduces the risk for microvascular complications, including retinopathy, nephropathy, and neuropathy. 

In addition, there are “oodles” of observational studies in support of the clinical trial data for A1c, while “no CGM metric has this level of long-term outcome validation,” she noted.

Due to the evidence linking A1c to outcomes, its wide availability, and global standardization, it is the basis for clinical guidelines worldwide, said Sevin. It also enables population-level surveillance and facilitates national and international surveillance of diabetes control, allowing for risk factor stratification of populations and benchmarking across healthcare systems.

The same can’t be said of CGM, where there is limited accessibility in many countries, lack of standardization, and device heterogeneity, she pointed out. “A1c is a simple blood test, whereas for CGM, you need devices and sensors…It requires patient training, device management, and data monitoring.”

And, while A1c is routinely covered by health insurers, “coverage for CGM is highly variable.” 

Lack of Evidence for CGM Metrics 

Selvin acknowledges that CGM is “a revolutionary technology that has immensely improved the lives of many, many people.” However, she said, “there is still a lack of key evidence in terms of linking CGM metrics to long term outcomes. We do not have clear complication thresholds, and…results are still very device dependent.” 

Regarding the CGM’s “glycemic management indicator (GMI),” designed to estimate A1c based on continuous glucose values, Selvin noted that it “is not a great estimate of hemoglobin A1c. If you look at two different devices, you will get two different GMIs… So GMI varies substantially, even when two different CGMs are worn on the same person at the same time,” she said, citing a study from her group. 

Of course, A1c also has limitations. It doesn’t reflect hypoglycemia or glycemic variability, and it is affected by conditions that affect red blood cell turnover, hemoglobinopathies, or that involve blood loss. “However, these limitations are not sufficient to replace a validated gold standard…Hemoglobin A1c does not have any more limitations than any other laboratory test that we generally use in clinical practice,” Selvin said. 

While she believes CGM metrics are promising, “the evidence is not yet equivalent to hemoglobin A1c… A gold standard should reduce ambiguity, not introduce competing metrics and new measures.”

“CGM is a powerful complementary tool,” said Selvin. “My argument is that CGM can augment the interpretation of A1c,” but should not be used to “replace” it. 

CGM Metrics: Moving Forward to Improve Management 

The main goal of treatment should be to improve glycemic management and patient well-being, said Bergenstal. CGM does that while A1c does not. 

Relying on A1c hasn’t improved glycemic control, he pointed out. “We’re not doing well with A1c as our guiding principle…And if you’re on insulin, you’re doing even worse with A1c as your guide.”

Bergenstal noted that A1c is an average, so a person can have the same average glucose yet have vastly different A1c levels, depending on the glycemic variability. “A1c has not guided us in the right path, plus the glucose variability and hypoglycemia detection, I think, is critically important,” he said. 

Tests of A1c can also mislead in many situations or conditions in addition to hemoglobinopathies, including iron deficiency or iron replacement, chronic kidney disease, liver disease, or genetic factors affecting glycation of red blood cells, he pointed out. 

He showed examples of three patients with the same A1c value but with completely different levels of glycemic variability and hypoglycemia, requiring different treatment approaches. 

“That’s not a gold standard to me,” said Bergenstal. “It’s a marker of risk, but it’s not a marker that gets you anywhere towards management or improving the care of people with diabetes.” 

CGM Improves A1c, Reduces Hospitalizations 

A newly-updated GMI will soon be published that better aligns with A1c than the one currently used in the Dexcom and FreeStyle Libre CGMs, by incorporating the curvilinear relationship between the two measures rather than assuming a straight line, he noted. 

There’s a great deal of evidence showing that use of CGM improves A1c and also reduces hypoglycemia, hospitalizations, and hospital visits, as well as long-term diabetes complications, he told meeting attendees. In one study of 808 adults with T1D over 5 years, time in range and time below range correlated with micro- and macrovascular complications. 

In an observational database study involving nearly 42,000 patients with either T1D or T2D, CGM was associated with reduced A1c and less hypoglycemia compared to intermittent glucose testing. 

And, in a recent study of nearly 3000 adults with diabetes (65% T2D) from the Veterans Affairs Healthcare System, CGM metrics — not A1c — were correlated with 5-year mortality. 

For 20 years after the DCCT, the average A1c among participants, who are now followed long-term in the Epidemiology of Diabetes Interventions and Complications study, remained “stuck” at about 8%. “CGM is what caused the curve to bend,” Bergenstal said. 

“We have been living in the A1c era since 1993,” he said. “But it’s time to cross the bridge from A1c to CGM…There are times when paradigms do change, and I think it’s time for CGM to be this guide.” 

Indeed, he pointed out, that in 2022, the American Diabetes Association’s Standards of Care changed its recommendation from performing A1c at least twice a year to “glycemic assessment” using A1c or CGM time in rage, or GMI. And the National Committee for Quality Assurance now allows for either A1c or GMI as a Healthcare Effectiveness Data and Information Set quality measure.

With the emergence of automated insulin delivery systems, there has been further improvement. “Now we have effective ways to use these data and personalize the care in an equitable manner,” said Bergenstal. 

He closed by saying: “Let’s work together for expanded access to accurate and affordable CGMs and team care to support their use. I’m not opposed to having A1c available where CGM isn’t, until we get CGM even more widely distributed.” 

Miriam E. Tucker is a freelance journalist based in the Washington DC area. She is a regular contributor to Medscape, with other work appearing in the Washington Post, NPR’s Shots blog, and Diatribe. She is on X (formerly Twitter) @MiriamETucker and BlueSky @miriametucker.bsky.social