Abstract
Previous epidemiological studies have demonstrated associations between sleep duration and diabetes, vascular complications and cardiovascular disease in diabetic patients. However, few studies have so far explored the effect of sleep time on estimated glomerular filtration rate (eGFR) in diabetic patients. This study was performed for the purpose of exploring the relationship between sleep duration and eGFR in diabetic patients. This study analyzed 1389 patients with diabetes. Sleep duration at night was categorized into four groups: very short (< 6h), short (6-6.9h), optimal (7-8h), and long (> 8h). The association of sleep duration with eGFR was analyzed using univariate linear regression, and generalized additive models were applied to assess the nonlinear relationship between sleep duration and eGFR. Compared to optimal sleep duration (7-8h), both long sleep duration (> 8h) and very short sleep duration (< 6h) were associated with level of eGFR (β = -3.63, 95%CI: -5.54 to -1.71, P = 0.0002 and β = -4.79, 95%CI: -8.55 to -1.03, P = 0.0126, respectively). The smooth curve showed there is a U-shaped relationship between sleep duration and eGFR levels. The results of this study show that both very short and long sleep durations were associated with low eGFR in diabetics.