Association between sleep and serum hemoglobin A1c in nondiabetic population in Beijing

Objective: To understand the status quo of sleep and its associations with serum hemoglobin A1c (HbA1c) among nondiabetic people of 18-79 years old in Beijing. Methods: Data was gathered from the 2017 Beijing Non-communicable and Chronic Disease Surveillance Program. Multiple classified clusters sampling method was used while the 18-79 years old were sampled from the 16 districts of Beijing. Questionnaires would include information on demographic characteristics, chronic diseases and related risk factors, sleep duration and related problems (snore/asphyxia, difficult to get to sleep, waking often during the night, waking up early or taking sleeping pills) within the last 30 days. Complex sampling logistic regression models were established to analyze the association between sleep-related problems and serum HbA1c. Results: A total of 11 608 non-diabetic participants were involved in this study, with average age, reported sleep duration and median of serum HbA1c level as (43.36±15.27) years old, (7.49±1.29) h/d and 5.30%, respectively. 47.38% of them reported having sleep problems within the last 30 days. With the increasing time of sleep, serum HbA1c level was fluctuating significantly (F=413.06, P<0.01). Significant differences appeared in serum HbA1c levels among different age groups (t=358.3, P<0.01). Among participants with several kinds of sleep problems, the serum HbA1c levels were significantly higher than those without, through the single factor analysis (U=15.11, P<0.01). After adjusting for potential confounding factors, the combination of one sleep-related problem (OR=1.21, 95%CI: 1.03-1.41) and snore/asphyxia were associated with higher serum HbA1c levels (HbA1c≥5.7%) (OR=1.37, 95%CI: 1.16-1.61). People under 60 years of age were with higher risk of having higher serum HbA1c levels. Conclusion: Duration and sleep-related problems might affect the serum HbA1c levels, especially among those younger than 60 years of age.