HbA1c poor marker for all-cause mortality in adults with hig
HbA1c may not be a robust marker for all-cause mortality among adults with type 2 diabetes and a high degree of glycemic variability, suggesting new metrics assessing glycemic response are needed.

In a prospective study, researchers analyzed data from 6,090 hospitalized patients with type 2 diabetes in a single center. Researchers obtained glucose coefficient of variation (CV) as the measure of glycemic variability using blinded continuous glucose monitoring (CGM) for 3 days. Participants were stratified by CV tertiles: less than 21.5%; 21.6% to 28.9%, and greater than 28.9%. Researchers used Cox proportional hazards regression models to estimate HRs for mortality. The findings were published in The Journal of Clinical Endocrinology & Metabolism.

During a median follow-up of 6.8 years, 815 patients died.

HRs for all-cause mortality associated with HbA1c levels of less than 6%, 7% to 7.9%, and at least 8% were 1.64, 1.37, and 1.8, respectively, compared with an HbA1c between 6% and 6.9%. Among participants in the lowest and middle tertiles of glucose variability with an HbA1c of at least 8%; HRs for all-cause mortality were 2.36 and 1.92, respectively, compared with an HbA1c between 6% and 6.9%.

In a fully adjusted model, the four categories of HbA1c did not differ in risk for all-cause mortality among participants with the highest glucose variability.

“HbA1c is the gold-standard for assessing glucose control, as there is ample evidence supporting the tight relationship between HbA1c and diabetes-related outcomes,” researchers told. “Patients with diabetes should adjust treatment plans to maintain their HbA1c, in order to minimize the risk for adverse outcomes such as microvascular and macrovascular complications. However, our findings suggest that, in patients with high glucose fluctuations, the use of HbA1c may not be as valuable as previously thought. For these people, new glucose markers, such as CGM metrics, may be needed for more accurate assessment of glucose control and more personalized diabetes management.”

The researchers noted that 3 days of CGM data were used in the study; however, 2 to 4 weeks of monitoring may be needed to robustly assess the actual glucose variability. Data on socioeconomic and lifestyle factors were also not available.

“However, this proof-of-concept study may provide insights into the potential caveats of HbA1c in predicting diabetes-related outcomes,” the researchers wrote.

Source: https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgab532/6324063?login=true