Pitfalls in the Diagnosis of Diabetes: Are we too Lax with Laboratory Parameters?

The diagnosis of diabetes is associated with pre-analytical and analytical problems. Fasting glucose (FG), oral glucose tolerance test (oGTT) and HbA1c have advantages and shortcomings and have no equal diagnostic validity. oGTT is the most sensitive test, but its reproducibility is rather poor (CV +/- 15 %). FG detects only 70 - 80 % of overt diabetes. FG is falsified by inappropriate blood sampling, intra-individual fluctuations and mistakes with the oGTT. HbA1c despite IFC-standardization, but with a tolerable coefficient of variation of +/- 18 % in round robin tests and use of not commutable control material is not easy to interpret. HbA1c analysis shows also interferences and is therefore of limited diagnostic value. Its threshold value of >= 6.5 % (>= 48 mmol/mol Hb) is based on consensus and not on evidence. The diagnostic effort (FG and/or oGTT+HbA1c) with serious consequences is minimal invasive, reasonable and cheap. It prevents over- and underdiagnosis.