ABSTARCT Poor sleep quality or sleep restriction is associated with sleepiness and concentration problems. Moreover, chronic sleep restriction may affect metabolism, hormone secretion patterns and inflammatory responses. Limited recent reports suggest a potential link between sleep deprivation and epigenetic effects such as changes in DNA methylation profiles. The aim of the present study was to assess the potential association between poor sleep quality or sleep duration and the levels of 5-methylcytosine in the promoter regions of PER1, PER2, PER3, BMAL1, CLOCK, CRY1 CRY2 and NPAS2 genes, taking into account rotating night work and chronotype as potential confounders or modifiers. A cross-sectional study was conducted on 710 nurses and midwives (347 working on rotating nights and 363 working only during the day) aged 40-60 years. Data from in-person interviews about sleep quality, chronotype and potential confounders were used. Sleep quality and chronotype were assessed using Pittsburgh Sleep Quality Questionnaire (PSQI) and Morningness-Eveningness Questionnaire (MEQ), respectively. Morning blood samples were collected. The methylation status of the circadian rhythm genes was determined via quantitative methylation-specific real-time PCR assays (qMSP) reactions using DNA samples derived from leucocytes. The proportional odds regression model was fitted to quantify the relationship between methylation index (MI) as the dependent variable and sleep quality or sleep duration as the explanatory variable. Analyses were carried out for the total population as well as for subgroups of women stratified by the current system of work (rotating night shift/day work) and chronotype (morning type/intermediate type/evening type). A potential modifying effect of the system of work or the chronotype was examined using the likelihood ratio test. No significant findings were observed in the total study population. Subgroup analyses revealed two statistically significant associations between a shorter sleep duration and 1) methylation level in PER2 among day workers, especially those with the morning chronotype (OR = 2.31, 95%CI:1.24-4.33), and 2) methylation level in CRY2 among subjects with the intermediate chronotype, particularly among day workers (OR = 0.52, 95%CI:0.28-0.96). The study results demonstrated a positive association between average sleep duration of less than 6 hours and the methylation level of PER2 among morning chronotype subjects, and an inverse association for CRY2 among intermediate chronotype subjects, but only among day workers. Both the system of work and the chronotype turned out to be important confounders and modifiers in a number of analyses, making it necessary to consider them as potential covariates in future research on sleep deficiency outcomes. Further studies are warranted to explore this under-investigated topic.
