CGM metrics may be better for determining diabetes intervention efficacy than HbA1c
Click Here to Manage Email Alerts
Change in mean glucose may be a better metric for measuring the efficacy of diabetes interventions than HbA1c, time in range and time above range, according to a speaker.
At the International Conference on Advanced Technologies & Treatments for Diabetes, David Rodbard, MD, president and chief scientific officer for Biomedical Informatics Consultants LLC in Potomac, Maryland, analyzed correlations between HbA1c, mean glucose, time in range and time above range in seven studies of people with type 1 and type 2 diabetes. Across the presented studies, Rodbard noted that HbA1c was weakly correlated with all three continuous glucose monitoring metrics. With mean glucose a presumed mechanism for long-term diabetes-related complications, Rodbard argued, change in mean glucose within individuals may be the best primary efficacy point in diabetes clinical trials.
“Mean glucose, time above range and time in range are tightly correlated among themselves, but all three of those have only relatively weak correlations with HbA1c,” Rodbard said during the presentation. “Mean glucose has a stronger correlation with time above range than with HbA1c, and mean glucose had a better correlation with time above range than time in range. So I would like to argue that mean glucose is better than time above range, time in range or HbA1c.”
Mean glucose, time above range tightly correlated
Rodbard discussed data from seven diabetes studies published from 2019 to 2022. Across the studies, Rodbard noted a similar pattern, with mean glucose, time in range and time above range all more tightly correlated with one another than with HbA1c.
In a study published in the Journal of Diabetes Science and Technology in 2019, researchers analyzed correlations between the four metrics at baseline and the end of a 6-month intervention for people with type 1 diabetes. When the changes in metrics from baseline to 6 months were analyzed, change in mean glucose was most strongly correlated with change in time above range (r = 0.95) and change in time in range (r = –0.85). A weaker correlation was observed between change in mean glucose and change in HbA1c (r = 0.52).
Similar findings were observed in a study published in The Journal of Clinical Endocrinology & Metabolism in 2021, which looked at correlations between metrics for women with type 1 diabetes during pregnancy and postpartum. During pregnancy when a time in range of 63 mg/dL to 140 mg/dL was used, the correlation between mean glucose and time above range (r = 0.869) was much stronger than the correlation between HbA1c and time above range (r = 0.435). In the postpartum period with a time in range between 70 mg/dL and 180 mg/dL, the correlation between mean glucose and time above range remained stronger (r = 0.902) than the correlation between HbA1c and time above range (r = 0.675). Additionally, in the postpartum period, mean glucose and time in range were weakly correlated (r = –0.52).
“When we look at the correlations with HbA1c, time above range has a higher correlation than time in range in four of these [studies],” Rodbard said.
Findings similar during daytime, nighttime
In findings published in Diabetes Technology & Therapeutics in 2022, researchers analyzed differences in correlations during daytime vs. nighttime. The study found correlations were similar regardless of time of day, with mean glucose more strongly correlated with time above range than time in range as in the other studies. The findings were also unchanged when a tighter time in range of 70 mg/dL to 140 mg/dL was used.
When data from the studies were pooled together, the strongest correlation was between mean glucose and time above range (r = 0.95), followed by time in range and time above range (r = –0.91) and mean glucose and time in range (r = –0.82). Weaker correlations were observed between HbA1c and mean glucose (r = 0.73), HbA1c and time above range (r = 0.73) and HbA1c and time in range (r = –0.69).
“Time above range and time in range are better correlated with mean glucose,” Rodbard said. “Mean glucose has a consistently higher correlation with time above range than time in range. Mean glucose is the presumed mechanism of long-term complications and, therefore, mean glucose might be the best metric of efficacy, followed by time above range and then time in range.”
Rodbard concluded that using change in mean glucose, change in time above range and change in time in range within patients may be a better way to determine efficacy of diabetes interventions than change in HbA1c. However, he said more research is needed using large data sets from multiple studies and diverse populations.
References:
Beck RW, et al. J Diabetes Sci Technol. 2019;doi:10.1177/1932296818822496.
Ling P, et al. J Clin Endocrinol Metab. 2021;doi:10.1210/clinem/dgab502.
Shah VN, et al. Diabetes Technol Ther. 2023;doi:10.1089/dia.2022.0365.
Collapse
Rodbard D. IS 042. Presented at: International Conference on Advanced Technologies & Treatments for Diabetes; Feb. 22-25, 2023; Berlin (hybrid meeting).
Read more about
Click Here to Manage Email Alerts