Abstract
Fasting blood glucose (FBG) and serum fructosamine are simple and commonly used tests for monitoring diabetes mellitus. Unfortunately, both these parameters are associated with high error rates and therefore used with caution in high-risk populations. Setting high cut-off values for these parameters increases the sensitivity but at the cost of poor specificity (more false positives). Continued efforts have been made to evaluate the efficacy of FBG and fructosamine, singly or in combination, in avoiding a large number of unnecessary oral glucose tolerance tests (OGTT). Therefore, to better understand their time-course trends, we analysed FBG and c-fructosamine in 211 blood samples from 51 Saudi pregnant women during their multiple (> or =3) antenatal visits. The mean+/-standard deviation of FBG and c-fructosamine were 5.22+/-1.07 and 2.22+/-0.25 mmol/l respectively with a significant correlation between their individual values. Using the FBG cut-off >5.3 mmol/l, 19 subjects were classified as hyperglycaemic; this frequency was reduced to 1 when a FBG cut-off of >7.0 mmol/l was used. Combined values of FBG (>5.3 mmol/l) and c-fructosamine (>2.5 mmol/l) filtered 6 high-risk subjects with a prediction of gestational diabetes mellitus (GDM). Analysis of variance revealed high within-group variance for FBG. These fluctuations were also confirmed by higher coefficient of variations (CVs) for FBG (13.27%) as compared to c-fructosamine (5.49%). The CVs of FBG were not correlated with those of corresponding CVs of c-fructosamine (R = 0.007, P = 0.962), indicating that the fluctuations in FBG were independent of fluctuations in c-fructosamine. These findings clearly suggest that the paired values of FBG and c-fructosamine would be more advantageous than their individual values in filtering high-risk patients on whom OGTT should be performed.