Relationship Between Average Glucose Levels and HbA1c Differs by Race/Ethnicity
Key findings
- It's been suggested that race should be taken into account when translating hemoglobin A1c (HbA1c) into average glucose levels
- This substudy of the GRADE trial (n=1,454) is the first to use continuous glucose monitoring to determine whether relationships between average glucose and HbA1c differ across U.S. racial/ethnic groups
- Whether glycated hemoglobin or glycated albumin was used as the independent variable, comparisons with a variety of measures of glucose levels showed consistent interracial differences
- Failure to recognize differences across racial/ethnic groups in the relationship between HbA1c and average glucose could result in overtreatment and excess hypoglycemia or, conversely, undertreatment and higher risk of vascular complications
In 2008, the international A1c-Derived Average Glucose study published in Diabetes Care provided an equation for translating hemoglobin A1c (HbA1c) measurements into average glucose levels. Whether it can fairly be applied to all races and ethnicities has since been questioned, but remarkably few studies of that issue have used reliable measurements of average glucose levels.
Subscribe to the latest updates from Endocrinology Advances in Motion
David M. Nathan, MD, director of the Diabetes Clinical Center at Massachusetts General Hospital and a professor of medicine at Harvard Medical School, and colleagues in the Glycemia Reduction Approaches in Diabetes (GRADE) Study Research Group used continuous glucose monitoring (CGM) to address the question. They report in Diabetes Care that in patients with type 2 diabetes, average glucose levels for HbA1c levels do differ across U.S. racial/ethnic groups.
Conclusion Comes From Substudy of Large, Prospective Trial
The GRADE study, reported in NEJM, compared the glycemic effects of four medications when added to metformin: insulin glargine, glimepiride, liraglutide, or sitagliptin. The study population reflected the racial/ethnic distribution of type 2 diabetes in the U.S., and participants received all medications and diabetes care free of charge.
Between January 2018 and January 2020, the research group conducted a substudy that required one extra visit before a regularly scheduled quarterly GRADE study visit. At the extra visit, HbA1c was measured and a CGM device was applied for up to 2 weeks of self-monitoring. Participants were blinded to their CGM results.
Then, at the next regular study visit, HbA1c levels and glycated albumin were measured and a standard oral glucose tolerance test (OGTT) was performed in the fasting state.
GRADE study participants were divided into four self-reported racial/ethnic groups: non-Hispanic white (NHW), non-Hispanic Black (NHB), Hispanic white (HW), and other. The primary comparison in the substudy was differences between NHW and NHB participants in the relationship between the HbA1c obtained immediately after CGM completion and the average glucose concentration based on the first 10 days of CGM (AG10).
The Non-Hispanic Black Group Differed From All Others
Data were analyzed for the 1,454 participants who had at least 70% of expected CGM measurements for a minimum of 10 days. 534 were NHW, 389 were NHB, 327 were HW, and 204 reported other racial/ethnic backgrounds. There were differences across racial/ethnic groups in age, sex, body mass index, and peak glucose levels during the OGTT, so analyses were adjusted for those factors.
The relationship between HbA1c and AG10 was significantly different between the NHB group and all three other racial/ethnic groups. For example, HbA1c levels were 0.2 to 0.6 percentage points higher in NHB than in NHW patients for AG10 levels from 100 to 250 mg/dL. Similar results were seen for the relationship of fasting plasma glucose with HbA1c and the relationship of glucose with glycated albumin.
Applying the Results to Practice
Ignoring these findings could undermine accurate diagnosis and safe diabetes management. The lower average glucose for any given HbA1c level in NHB versus NHW patients could result in overdiagnosis of diabetes in NHB individuals. Similarly, treating all races to achieve the same HbA1c target could subject NHB patients to a higher risk of hypoglycemia.
Clinicians should consider confirming a presumptive diagnosis of diabetes based on HbA1c level with a direct measure of glycemia (fasting or 2-hour glucose level). Similarly, when establishing treatment targets, clinicians should assess whether HbA1c is an accurate reflection of average glycemia.
Medical societies should consider whether HbA1c targets, glucose targets, and the threshold HbA1c for diagnosis of diabetes should be revised to be race-specific.
view original journal article Subscription may be required