Investigating Potential Bi-directional Relationships Between Sleep and Melatonin Among Day and Rotating Day/Night Shift Workers
Objectives: The first objective of this study is to describe measures of sleep in day and rotating shift workers. The second objective is to assess the bi-directional relationships between sleep and melatonin among: 1) female hospital day workers; and, 2) female rotating day/night shift workers, in two models: Model A includes the assessment of sleep on work days on melatonin the following day and Model B includes the assessment of melatonin on work days on sleep the following day. These models were determined in a stratified analysis based on shift work status and cumulative shift work duration in years. Methods: In a cross-sectional study of female hospital employees (n=266), sleep parameters were collected using Actical® accelerometers and assessed using the Cole-Kripke algorithm. Urinary 6-sulfatoxymelatonin concentrations were collected over 24 hours and examined using cosinor analysis to obtain average concentrations (mesor) and time of peak concentration (acrophase time). All relationships were analyzed using multiple linear regression. Results: Day workers slept about seven hours per night in both temporal models, whereas rotating workers slept less than seven hours between two night shifts and more than seven hours following their second night shift. When sleep precedes melatonin, for every hour increase in main sleep time, a 13-minute decrease in melatonin acrophase time was observed among rotating shift workers (95% CI: -26.5– -0.3). Effect modification by cumulative shift work duration (< 10 years vs. ≥ 10 years) was apparent in the relationship between main sleep time and mesor for the time period of Model A (p-value for interaction = 0.02). There were weak associations with sleep latency in both models. Conclusion: In comparison to day workers, rotating shift workers were more likely to not meet the recommended seven hours of sleep per night over the two temporal periods of interest, which is a public health concern. Sleep and melatonin may be related temporally in both directions in the relationship between main sleep time and melatonin acrophase time. The effect of sleep duration on melatonin may depend on the total length of exposure to shift work and not on current day or rotating shift status.