Strongest correlation of HbA 1c with 1-month-earlier glucose levels in type 2 diabetes

Abstract

We examined the correlations of hemoglobin A 1c (HbA 1c) with each plasma glucose (PG) level obtained at 0 (the same day), 1 and 2 month(s) prior to HbA 1c determination. Data were from glycemic profiles of four patients of type 2 diabetes mellitus treated with tablets whose HbA 1c and pre- and post-breakfast PG levels were monitored each month. There was no significant difference in the correlation coefficients in cases 1 and 2, who presented with linear glycemic time courses. In contrast, HbA 1c correlated with 1-month-earlier pre-breakfast PG level more strongly than 2-month-earlier post-breakfast PG level in cases 3 and 4, and than same-day post-breakfast PG level in case 3 ( P<0.05, ANOVA). The cases 3 and 4 presented with fluctuating glycemic time courses. Samples were separated into upslope’s and downslope’s sections according to HbA 1c fluctuation in the latter two cases. Reflecting around the 1-month lag between HbA 1c and PG, the two sections’ regression lines for PG versus HbA 1c corresponded in the only samples related to 1-month-earlier pre- and post-breakfast PG ( t-test). In conclusion, it appears that pre- and post-breakfast PG levels are the most reliable predictors of 1-month-later HbA 1c in type 2 diabetic outpatients who undergo medical examinations every month.