No reason for higher HbA1c threshold to diagnose diabetes in blacks

Retinopathy occurs at lower hemoglobin A1c (HbA1c) levels in blacks than in whites, so there is no reason to recommend a higher HbA1c threshold for diagnosing diabetes in blacks, a study concluded.

Recent studies have indicated that at the same blood glucose levels, black patients have higher HbA1c levels than white patients. This distinction has led some experts to propose setting a higher HbA1c threshold for diagnosing diabetes in black patients

To compare the relationships between HbA1c level and the prevalence of retinopathy in non-Hispanic black and white adults, researchers conducted a cross-sectional study from data in the National Health and Nutrition Examination Survey (NHANES) from 2005 through 2008. The study included 2,804 whites and 1,008 blacks age 40 years or older who were examined for the prevalence of retinopathy. Results appeared in the Aug. 7 Annals of Internal Medicine.

The weighted crude prevalence (±SE) of retinopathy was 6.3% ± 0.5% for whites and 13.1% ± 1.1% for blacks. In whites, the adjusted prevalence of retinopathy was significantly higher beginning at HbA1c levels of 6%-6.4%, compared with HbA1c levels less than 5.5% (the reference category).

Among black adults, those with HbA1c levels of 5.5% to 5.9% had significantly higher risk for prevalent retinopathy than those with HbA1c levels less than 5.5%, according to the study. The adjusted risk difference for retinopathy for blacks with HbA1c levels of 5.5% to 5.9% was similar to that of whites with HbA1c levels of 6% to 6.4%.

Researchers noted that if the diagnostic threshold of HbA1c were lowered from the current adopted level of 6.5% to 6%, an estimated additional 1.8 million black and 7.6 million white U.S. adults aged 40 years or older would be diagnosed with diabetes, of whom 0.2 million blacks and 0.8 million whites would be anticipated to have diabetic retinopathy at the time of examination. For both blacks and whites, this translates to diagnosing nine additional cases of diabetes to detect one additional case of prevalent diabetic retinopathy.

If the diagnostic threshold for diabetes were lowered from 6.5% to 5.5%, an additional 5.7 million blacks and 39.0 million whites would be diagnosed with diabetes, among which there would be 0.5 million cases of retinopathy among blacks and 2.2 million cases among whites. This new diagnostic threshold would result in diagnosing 11 additional cases of diabetes among blacks and 18 additional cases among whites to detect one additional case of diabetic retinopathy.

The authors wrote, “The results of our study suggest that the risk for diabetic retinopathy is higher for blacks at any given HbA1c level between 5.0% and 7.0%...Therefore, our findings argue against increasing the HbA1c diagnostic threshold for blacks.”