Differential Impact of Sleep Duration on Fasting Plasma Glucose Level According to Work Timing

To quantify the differential contribution of sleep duration on fasting plasma glucose level by traditional regular daytime work and shift work in subjects without diabetes.

Self-reported sleep duration and work type and timing were determined in a cross-sectional sample of 9123 participants aged 20-65 years from the Korea National Health and Nutrition Examination Survey (KNHANES) 2010-2015. Those who responded that they worked between 6 am and 6 pm were classified as "traditional regular daytime workers; "those who worked in the afternoon, at night, or in several types of shift work were classified as "shift workers." FBG was compared between short (<6 h), "normal" (6-8), and long (>8 h) sleep duration groups according to work time.

In the traditional daytime workers group, mean FBG level showed a U-shaped trend according to sleep duration (p = 0.027), whereas in shift workers group, FBG level was significantly decreased across sleep duration (p = 0.001). In the regular daytime workers group, short sleep duration was associated with higher FBG (B, 95% [CI]: 1.33 [0.26-2.4]), whereas after adjustment for potential confounding variables, long sleep duration significantly increased the risk of higher FBG (2.01 [0.35-3.68]). On the other hand, the reverse was true in the shift workers. Long sleep duration was significantly associated with lower FBG by both unadjusted analysis and after multivariable adjustment (-3.79 [-5.97 to -1.62], -2.19 [-4.35 to -0.03], respectively).

Our results suggest that the impact of sleep duration on FBG level differs according to work shift.