Hemoglobin A1c is not a reliable marker of dysglycemia in overweight or obese children and adolescents.…
Lead author Dr. Joyce M. Lee from University of Michigan in Ann Arbor states that, “Despite the new guidelines recommending the use of hemoglobin A1c for diagnosis of diabetes, it is not as reliable a test for identifying children with diabetes or at high risk for diabetes.”
She suggests that doctors “consider ordering alternative tests, such as a random glucose or a 1 hour nonfasting glucose tolerance test.”
Dr. Lee and colleagues compared five nonfasting screening tests in 254 overweight or obese children and adolescents aged 10 to 17 years: HbA1c, urinalysis, fructosamine, a one-hour glucose challenge, and a random blood test.
A formal two-hour oral glucose tolerance test showed that 39% of the youngsters had prediabetes and 1.2% had diabetes.
Urinalysis had a very low sensitivity (but high specificity) for detecting dysglycemia. On the other tests, higher thresholds provided lower sensitivity and higher specificity, whereas lower thresholds had higher sensitivity but lower specificity.
Discrimination was poor for HbA1c and fructosamine levels, as evidenced by relatively low likelihood ratios across test thresholds, as well as by low values for area under the curve (AUC).
With random glucose and one-hour glucose challenge tests, however, discrimination was “closer to an acceptable range,” the authors said. Both provided substantially higher AUC compared to HbA1c or fructosamine.
The researchers say their findings are consistent with other recent studies of HbA1c tests in children.
From the result it was concluded that, HbA1c had poor discrimination, which could lead to missed cases of dysglycemia in children. Random glucose or 1-h GCT may potentially be incorporated into clinical practice as initial screening tests for prediabetes or diabetes and for determining which children should undergo further definitive testing.
Dr. Lee said she and her colleagues are now trying to learn “whether a clinical risk score based solely on clinical characteristics” would help screen children for diabetes. “This would be a convenient and cost-effective way to identify high-risk children,” she said.