Conventional measures underestimate glycaemia in cystic fibrosis patients

Aims Diabetes mellitus is an increasingly important complication of cystic fibrosis (CF). The association with increased morbidity of cystic fibrosis-related diabetes (CFRD) has emphasized the need for accurate monitoring of glycaemia in all CF patients. The diagnosis has relied on conventional thresholds in an oral glucose tolerance test (OGTT) derived from epidemiological studies in non-CF subjects. However, it has not been established if these values are equivalent in CF and non-CF populations. Methods We compared glycaemia in 21 non-diabetic CF subjects with 21-age and BMI-matched non-CF controls using HbA(1c), OGTT and a subcutaneous continuous glucose monitoring system (CGMS) that measures interstitial glucose levels. Results All conventional measures of glycaemia were similar in the two groups: HbA(1c) mean CF vs. controls (5.5 vs. 5.3%P = 0.4), fasting glucose (4.8 vs. 4.7 mmol/l P = 0.7) and 2-h glucose (5.8 vs. 5.7 mmol/l P = 0.8). However, these conventional measures did not accurately reflect glycaemia 30-, 60- and 90-min. Glucose values and area under the curve in OGTT were all higher in CF subjects than controls (P = 0.01-0.0001). Mean CGMS value [5.9 (0.8) vs. 5.1 (0.5) mmol/l, P = 0.004], and the proportion of subjects with peak CGMS values > 11.1 mmol/l (33 vs. 5%P = 0.00001) were also higher in CF subjects than controls. These results remained significantly different when only subjects with normal glucose tolerance in the two groups were studied. Conclusions We have shown that overall glycaemia, as shown by both the response during an OGTT and CGMS, is higher in CF subjects who have similar HbA(1c), fasting and 2-h glucose values. These results question whether it is appropriate to use the diagnostic thresholds and OGTT time points derived from the non-CF population for a diagnosis of diabetes in patients with cystic fibrosis.