The effect of sleep duration on the risk of diabetes mellitus in an open population of men aged 45-64 years (international epidemiological studies)

Abstract

Objective: to study the effect of sleep duration on the 16-year risk of developing type 2 diabetes mellitus (DM) in an open population of men aged 45-64 years living in Novosibirsk. Patients and methods. In 2003-2005 during the IV screening (HAPIEE project), a representative sample of men aged 45-69 years (n=781; mean age - 56.48±0.2 years, response rate - 61%) was examined. A standard clinical and epidemiological examination was carried out: sleep duration was assessed using the Jenkins scale, the level of state-trait anxiety (STA) - using the Spielberger self-assessment scale, depression, life exhaustion, family and workplace stress - with MONICA-MOPSY scales. Social support was assessed with Berkman-Syme Social Network Index. The cohort was followed up for 16 years. Results and discussion. In the studied population of men aged 45-64 years, the most common sleep duration was 7 hours (44.7%), in second place - 8 hours of sleep (27.6%), in third place - 6 hours of sleep (16.4%). Among people with newly diagnosed DM, 7-hour sleep prevailed - 39.2%, 6- and 8-hour sleep - 25.3% each (χ2=7.774; df=5; p>0.05). In men with 5-6 hours of sleep, compared with men sleeping for 7-8 hours, we found a 1.72-fold increased 16-year risk of developing diabetes (95% CI 1.066-2.776; p<0.05), and in men aged 45-54 years - a 1.868-fold increase (95% CI 1.089-3.927; p<0.05). In the Cox-proportional multivariate model, an independent effect on the diabetes risk was observed for: 5-6 hours of sleep at night: hazard ratio (HR) 1.561 (95% CI 1.063-2.83; p<0.001), depression (HR 1.767; 95% CI 1.058-2.952; p<0.05), life exhaustion (HR 1.511; 95% CI 1.266-2.984; p<0.05), and low and medium-1 SNI (HR 1.956; 95% CI 1.074-3.560; p<0.05). Conclusion. Short and very short sleep duration could be defined as a major risk factor of DM.