Glycosylated hemoglobin assay among Japanese rural inhabitants and as an indicator of screening test for diabetes mellitus / 日本農村医学会雑誌

ABSTRACT

The purpose of this study is to establish the normal physiological values of glycosylated hemoglobin (HbA<SUB>1</SUB> and HbA<SUB>1C</SUB>) among Japanese rural inhabitants and to evaluate its potential as a screening tool for diabetes mellitus. Six hundred and seventy-five people aged 35-64 years (210 males and 465 females) were examined. Oral glucose tolerance tests were perfomed on 135 inhabitants of them by giving a 50-gm.<BR>Frequency distributions of HbA<SUB>1</SUB> and HbA<SUB>1C</SUB> fitted approximately to the log-normal distribution.<BR>Geometric means and standard deviations in bracket were 7.95 (1.11)% for HbA<SUB>1</SUB> in men, 8.13 (1.12)% for HbA<SUB>1</SUB> in women, 5.17 (1.10)% for HbA<SUB>1C</SUB> in men and 5.16 (1.11)% for HbA<SUB>1C</SUB> in women<BR>Group average values of HbA<SUB>1</SUB> and HbA<SUB>1C</SUB> tended to increase with age in both men and women except a male group aged 60 years and over. There were no significant differences between mean values of glycosylated hemoglobin in men and those in women. HbA<SUB>1</SUB> was positively correlated with age in women and negatively correlated with hemoglobin in both men and women. HbA<SUB>1C</SUB> was positively correlated with age in men and women, body mass index, systolic blood pressure and serum cholesterol in women, and negatively correlated with hemoglobin in men.<BR>In order to evaluate glycosylated hemoglobin as a screening tool for the identification of unknown persons with diabetes, we calculated the sensitivity (the extents to which the participants who were found to have a diabetic OGTT also had an abnormal glycosylated hemoglobin) and specificity (the extents to which the participants who were not diabetic had normal glycosylated hemoglobin). The sensitivities for the person aged 59 years or under using HbA<SUB>1</SUB> were 72.7-100% and the specificities 67.3-75.0 percent. Futhermore the sensitivities for the same aged person using HbA<SUB>1C</SUB> were 81.8-100% and the specificities 78.8-81.3 percent. The validities for both HbA<SUB>1</SUB> and HbA<SUB>1C</SUB> were higher than those for glucosuria or blood glucose. But glycosylated hemoglobin was not useful as a screening tool for the person aged 60 years and over because of the low specificity.