Skip to main content

21-03-2012 | Internal medicine | Article

Prediabetes criteria lack validity in Asian Indians


Free abstract

MedWire News: Two recommended methods for detecting prediabetes fail to diagnose at least a third of Asian Indian patients with the condition, research suggests.

Use of glycated hemoglobin (HbA1c) cutoff values recommended by the American Diabetes Association (ADA) and the International Expert Committee (IEC) miss prediabetes in 38% and 64% of patients, respectively, according to a study published in Diabetic Medicine.

It has been recognized that there is a significant need for earlier detection of diabetes. But, "for diagnosis of prediabetes, subjects require fasting and 2-hour post-glucose load plasma glucose estimation, which is cumbersome and has poor reproducibility," say Anil Bhansali (Postgraduate Institute of Medical Education and Research, Chandigarh, India) and team.

In search of a more convenient testing method, Bhansali and colleagues assessed the utility of HbA1c cutoff values recommended by the ADA and IEC and oral glucose tolerance tests to diagnose prediabetes (defined by the World Health Organization 1999 capillary plasma glucose criteria) in a randomly selected sample of 1972 adults of Asian-Indian ethnicity, aged 20 years and above.

Overall, 329 (17%) of the adults screened tested positive for prediabetes. The diagnosis was based on impaired fasting glucose in 125 (6.3%) participants, impaired glucose tolerance in 141 (7.1%), and the presence of both glucose abnormalities in 63 (3.2%) of those tested.

The ADA and IEC use different HbA1c cutoffs used to diagnose prediabetes: the former recommends an HbA1c value of 5.7%; whereas the IEC uses a higher threshold, of 6.0%.

The ADA threshold gave a sensitivity of 62% and specificity of 77%, compared with a sensitivity of 36% and a specificity of 90% with the IEC threshold.

The positive predictive value, negative predictive value, and accuracy of the ADA cutoff were 34.7%, 89.5%, and 67.8%, respectively. The respective values calculated when using the IEC cutoff were 42.7%, 85.4%, and 77.0%, respectively.

Use of the ADA cutoff would have resulted in 38% of patients not being diagnosed, compared with 64% of patients if using the IEC cutoffs.

Bhansali et al conclude: "Long-term prospective studies are required to define the usefulness of one over the other."

By Cher Thornhill

Related topics