Abstract: Objective To analyze the characteristics and related factors of the dawn phenomenon in elderly patients with type 2 diabetes using the continuous glucose monitoring system and explore the effective therapies. Methods A total of 94 patients with type 2 diabetes aged 65 or older were selected from the Department of Endocrinology, Chuiyangliu Hospital Affiliated to Tsinghua University from July 2019 to March 2020. The general clinical data were collected, clinical biochemical level, fasting C-peptide, and fasting insulin were tested．Fundus photography, arterial ultrasonic doppler and quantitative detection of sensory nerves were performed on the patients.  The continuous glucose monitoring system was carried out for three days. According to the results of continuous glucose monitoring system, 46 cases were divided into patients without dawn phenomenon group(non-dawn phenomenon group, NDP group) and 37 cases with dawn phenomenon group(dawn phenomenon group, DP group). The patients with dawn phenomenon undertook continuous glucose monitoring system again 3 months after adjusting treatment. Results The incidence of DP in elderly patients with type 2 diabetes was 39.36% (37/94). Patients in the DP group had lower body mass index and fasting insulin levels（P=0.029 and P<0.001）, higher fasting blood glucose and hemoglobin A1c （P<0.001 and P=0.032）compared with those in the NDP group. The glycemic variation (GV) metrics in DP group were statistically higher than those in the NDP group（P value all<0.05）．The incidence of diabetic peripheral neuropathy in the DP group (86.49%) was higher than that in the NDP group (67.39%)（P=0.043）. Logistic regression analysis with the dawn phenomenon as the dependent variable showed that low fasting insulin and low body mass index were independent risk factors for dawn phenomenon. Of the 37 patients with DP, 18 were given basal insulin or titrated insulin dose, 11 were given metformin or added dose,  7 were given dipeptidyl peptidase-4 inhibitor (DPP-4i)， and one was given glucagon-like peptide-1 receptor agonist. Five of the patients given insulin and DPP-4i combined with metformin, and another five patients only adjusted their lifestyle. After 3 months of treatment, the fasting blood glucose and the glycemic variation metrics were significantly reduced (P all<0.05), and the fasting blood glucose compliance rate (≤ 7.0 mmol/L) was 72.97% (27/37). Conclusion Dawn phenomenon is common in elderly patients with type 2 diabetes and is the main cause of fasting hyperglycemia. Low fasting insulin and low body mass index were independent risk factors for dawn phenomenon. Dawn phenomenon is associated with large glycemic variation and may lead to an increased risk of diabetic peripheral neuropathy. Basal insulin, metformin, and DPP-4i are clinically effective methods to improve dawn phenomenon.

Key words: type 2 diabetes mellitus, dawn phenomenon, continuous glucose monitoring system