Should hemoglobin A1c thresholds differ by BMI classification?

By Will Boggs MD

NEW YORK (Reuters Health) - Because of poor agreement between hemoglobin (Hb) A1c levels and oral glucose tolerance tests (OGTT) in obese patients, HbA1c thresholds for diagnosing prediabetes should differ according to body mass index (BMI) class, researchers from China say.

"Our results suggest that a fixed HbA1c cut-off value is not suitable for screening prediabetes in the population with large BMI variation," Dr. Changhao Sun from Harbin Medical University in Harbin told Reuters Health by email. "In other words, when the American Diabetes Association (ADA) HbA1c criteria are used for glycemia classification, A1c between 5.7-6.4% should be interpreted considering BMI."

Obesity and the systemic oxidative stress it brings can increase glycation of hemoglobin independently of glucose levels, so A1c concentrations may be disproportionately elevated for a given glycemia level in obese patients.

Dr. Sun's team evaluated the effects of obesity on the agreement between A1c and OGTT for diagnosing prediabetes and diabetes and sought to identify the optimal A1c cut-off values in different BMI classifications.

Among the 4,325 participants in the study, 16.8% had prediabetes and 6.3% had diabetes according to OGTT criteria. When classified by A1c criteria, though, 25.8% of the 72.7% overall who had normal glucose tolerance by OGTT criteria were misclassified with prediabetes and 1.5% were misclassified with diabetes.

The correlations between OGTT and A1c criteria decreased with increasing BMI, according to the report, online March 9 in The Journal of Clinical Endocrinology & Metabolism.

For prediabetes, the accuracy of A1c criteria was significantly lower in obese subjects than in normal BMI and overweight subjects. The optimal A1c for diagnosing prediabetes with 80% specificity would be 5.6% for normal bodyweight individuals, 5.7% for overweight individuals, and 6.0% for obese individuals, according to the researchers' calculations.

The best combination of sensitivity and specificity for diagnosing diabetes would be A1c levels of 5.7% for normal bodyweight, 6.0% for overweight, and 6.4% for obese populations, although BMI was not significantly associated with A1c levels in the diabetes population.

"Diabetes mellitus is a disorder of glucose, not HbA1c, metabolism," Dr. Sun said. "Non-glucose factors largely affect the HbA1c level in normal glucose tolerance. Such variability limits the use of A1c for screening unrecognized diabetes, and particularly for prediabetes."

"In order to avoid over-diagnosis for pre-diabetes, we should pay attention to the BMI of subjects in clinical examination; traditional OGTT may be more suitable for population with large BMI variation," Dr. Sun concluded. "That is the most important point we hope for physicians to take away from this report."

Dr. Sun added, "For a given BMI, the diabetes risks are different among different ethnic populations, and race and ethnicity affect A1c level independent of blood glucose. Therefore, the racial and ethnic disparity should be considered when extrapolating our results to other populations."

Dr. Elizabeth Selvin, an expert in the epidemiology of diabetes who was not involved in the study, said its cross-sectional nature is a major limitation.

"We know that HbA1c strongly predicts complications," Dr. Selvin, from Johns Hopkins School of Public Health in Baltimore, Maryland, told Reuters Health by email. "The lower the HbA1c value, the lower the risk of complications. We also know from clinical trials in persons with diabetes that lowering HbA1c prevents complications."

"It is the strong link of HbA1c to complications -- including incident diabetes, retinopathy (eye disease), heart attack, stroke, kidney disease, and death -- that make it an incredibly important test in clinical practice," Dr. Selvin said. "As with any test, HbA1c results need to be interpreted within the context of the individual patient. But there is no doubt that HbA1c is a major predictor of diabetes and its complications."

SOURCE: http://bit.ly/19F2LjH

J Clin Endocrinol Metab 2015.

(c) Copyright Thomson Reuters 2015. Click For Restrictions - http://about.reuters.com/fulllegal.asp