Challenges in HbA1C Level as a Diagnostic Tool of Diabetes and Pre-Diabetes in Middle-Aged Population: The Bangladesh Study.

Worldwide prevalence of diabetes is found to be the human health at an alarming rate. However, large numbers of patient remain undiagnosed. Oral glucose tolerance test (OGTT) still is regarded as gold standard in diagnosis of blood glucose abnormality. Although the less number of bodies are considering measurement of HbA1C as an alternate tool to identify risk group. This study was undertaken to evaluate the role of measurement of HbA1C in the diagnosis of diabetes and pre-diabetes in middle-aged Bangladeshi subjects and carried out in the department of Biochemistry, BIRDEM from July 2013 to June 2014. A total 177 subjects of age within the range of 30-45 years were selected for the purpose and classified into healthy control (n=62) pre-diabetes (n=69) and diabetes (n=46) groups based on the values of OGTT. Middle aged Bangladeshi subjects attending Bangladesh Institute of Research & Rehabilitation in Diabetes, Endocrine and Metabolic Disorders (BIRDEM) hospital, the HbA1C values were 5.0-5.6% in control group, 5.6-6.2% in pre-diabetes and 8.1-9.7% in diabetes group (95% CI). The optimal cut-off value of HbA1C related to pre-diabetes diagnosed by OGTT was 5.6%, which showed the sensitivity 47.8%, specificity 74.2%, positive predictive value 67.3% and negative predictive value 58.5%. Variants of hemoglobin especially Hemoglobin E (HbE) is prevalent in South East Asia including Bangladesh. The presence of genetic variants of hemoglobin can profoundly affect the accuracy of HbA1C measurements. So measurement of HbA1C may not be used as an alternate tool of OGTT to identify people of diabetes and pre-diabetes in certain situation.

Challenges in HbA1c Analysis and Reporting in Patients with Variant Hemoglobins.

Hemoglobin A1c (HbA(1)c) is a well-established indicator of mean glycemia. The presence of genetic variants of hemoglobin can profoundly affect the accuracy of HbA(1)c measurements. Variants of hemoglobin especially Hemoglobin E (HbE) is prevalent in South East Asia including Bangladesh. The objective of our study is to compare the HbA(1)c values measured on high performance liquid chromatography (HPLC) and Turbidimetric Inhibition Immunoassay (TINIA) in diabetic patients with variant hemoglobins including HbE. A total of 7595 diabetic patients receiving treatment at BIRDEM General Hospital were analyzed for HbA(1)c results within a period of two months from December 2013 to January 2014. Seventy two cases out of 7595 (0.95%) had either undetectable or below normal HbA(1)c levels (males-33 and females-39; ratio = 0.82:1) by HPLC method. In 34(0.45%) cases, HbA(1)c value was undetectable by HPLC method but was in the reportable range by TINIA method. In the other 38 (0.55%) cases, HbA(1)c levels were below the reportable range (<4%) by HPLC method but were in the normal or higher range by TINIA method. TINIA method did not agree with HPLC method on Bland Altman plot in the 38 cases with below normal HbA(1)c levels, [Mean bias -5.2(-9.3 to 1.0), 95% CI] but agreed very well [mean bias -0.21 (-0.84 to 0.42), y=1.1037+0.776X; r(2)=0.30, p<0.01] in controls. In control group mean MCV was 83.80±7.48 and in study group was 73.65±10.44. Alkaline electrophoresis confirmed the variant hemoglobin to be HbE. The fasting blood sugar levels of all the 72 cases correlated strongly with TINIA method (r(2) =0.75, p<0.0001) but not with HPLC (r = 0.24, p=0.13). In our regions where populations have a high prevalence of Hb variant, proper knowledge of hemoglobin variants which affect the measurements HbA(1)c level is essential. MCV of 80fl or below may serve as a rough guide to select samples that require analysis by TINIA method. Moreover, HPLC may be a convenient and inexpensive tool for screening of hemoglobinopathies especially among diabetic population in Bangladesh. It may therefore be helpful in improving management of complications related to both anaemia and iron overload.