AIMS: To investigate, in a large population in primary care, the relationship between fasting plasma glucose and HbA1c measurements, as well as the clinical implications of anaemia or chronic kidney disease for the interpretation of HbA1c values.

METHODS: From a primary care resource, we examined HbA1c and fasting plasma glucose as well as haemoglobin and estimated GFR. We stratified observations by chronic kidney disease stage and anaemia level. The estimation of the mean fasting plasma glucose level from HbA1c alone, and from HbA1c , haemoglobin and estimated GFR, respectively, was evaluated.

RESULTS: In 198 346 individuals, the fasting plasma glucose-HbA1c relationship mimicked the regression described in the A1c-Derived Average Glucose (ADAG) study, which was based on average capillary and interstitial glucose. The fasting plasma glucose-HbA1c relationship was unaffected in mild to moderate chronic kidney disease and in mild to moderate anaemia. The correlation changed only in severe hyperglycaemia and concurrent severe anaemia or when estimated GFR was <45 ml/min/1.73m², so that glucose concentration was underestimated by HbA1c in anaemia and overestimated in chronic kidney disease. The prevalence of estimated GFR <30 ml/min/1.73m² was 0.82%, while the prevalence of haemoglobin <81 g/l (5.0 mmol/l) was 0.11%.

CONCLUSIONS: The relationship between fasting plasma glucose and HbA1c mimics that of the people with diabetes included in the ADAG study. Mild to moderate anaemia and CKD do not have a significant impact on the interpretation of HbA1c as a marker of retrograde glycaemia. Hence, it seems justified to use HbA1c without adjustment in primary care.