Competing risk analysis on visit-to-visit glucose variations and risk of depression: The Taiwan Diabetes Study

Patients with diabetes have higher rates of depression than does the general population, but diabetes management mainly aims to maintain glucose stability. For this reason, our study assessed the relationship between 1-year variations in fasting plasma glucose (FPG) and risk of depression in Chinese patients with type 2 diabetes (T2D). This retrospective cohort study was conducted on 32,829 patients aged ≥30 years who were diagnosed with T2D and who participated in the National Diabetes Case Management Program in Taiwan. Their 1-year FPG variation as a predictor was determined by coefficient of variation (CV), whereas depressive events were analyzed by Cox's proportional hazards models. During a mean 8.23 years of follow-up, 1041 new cases of depression were diagnosed. When patients were grouped based on quartiles of FPG-CV, incidence rates were 3.23, 3.49, 3.96 and 4.80 per 1000 person-years in the first, second, third and fourth quartile subgroups, respectively. After adjusting for traditional risk factors, baseline fasting glucose and HbA1c levels, and diabetes complications, FPG-CV was independently linked with incident depression. Hazard ratios of depression for FPG-CV in the fourth vs first quartile subgroups was 1.33 (95% CI: 1.11–1.59), respectively. Patients whose 1-year FPG variations were>42.6% had an increased risk of depression, thus suggesting that FPG variations may be a predictor of depression in patients with T2D. Also, glucose variation during outpatient visits may be an indicator for individualized diabetes management in clinical practice.