Frequency of dawn phenomenon and its associations with age, HbA1c and diabetes duration in Japanese type 1 diabetes mellitus (T1DM) using the continuous glucose monitoring system (CGMS)

compared in terms of diabetes duration (13.0 ± 9.9 vs 10.0 ± 9.7: p<0.01)yr , HbA1c(8.3 ± 1.6 vs 8.5 ± 2.4: p<0.01)%, age at the time of the study (24.6 ± 18.0 vs 20.5 ± 14.9: p<0.01)yr. The subjects with dawn phenomenon had longer diabetes duration, lower HbA1c and were older. Furthermore, these subjects experienced hypoglycemia (< 70mg/dl) during the daytime. Conclusion The frequency of dawn phenomenon in the present study was lower than that in the previous studies .This might be attributable to there being many users of CSII amongour subjects and to Japanese foodscontaining a large amount of the carbohydrate as compared with protein. The associations of dawn phenomenon with longer diabetes duration and advanced age may be based on poor glycemic control. Furthermore, excess bolusescause hypoglycemia and low HbA1c.These results suggest that change in the basal insulin rate should be considered instead of an increase in the pre-meal bolus. We conclude that CGM should be used to adjust CSII.