The mediating effect of DNA methylation in the association between maternal sleep during pregnancy and offspring adiposity status: a prospective cohort study

Abstract

Background: Childhood overweight/obesity is a global public health concern. It is important to identify its early-life risk factors. Maternal poor sleep is common in late pregnancy, and previous studies indicated that poor sleep may influence the offspring's adiposity status. However, very few studies in humans investigated the effect of the different sleep parameters (sleep quantity, quality, and timing) on the offspring's adiposity indicators, and long-term studies are even more scarce. In addition, the underlying mechanism remains unclear. The present study therefore aimed to examine the association between the three maternal sleep dimensions in the late pregnancy and the offspring adiposity indicators and to explore the potential mediating effect of the cord blood DNA methylation in the above association.

Methods: Included participants in the current study were 2211 healthy pregnant women with singleton gestation from the Shanghai Birth Cohort (SBC) and Shanghai Sleep Birth Cohort (SSBC). Maternal nighttime sleep duration, quality, and midpoint (an indicator of circadian rhythm) were assessed by the same instrument in both cohorts during late pregnancy, and the offspring's body mass index (BMI) and subcutaneous fat (SF) were measured at 2 years old. Additionally, in 231 SSBC samples, the genome-wide DNA methylation levels were measured using the Illumina Infinium Methylation EPIC BeadChip. The multivariate linear regression was used to determine the associations between the maternal sleep parameters and the offspring adiposity indicators. The epigenome-wide association study was conducted to identify the maternal sleep-related CpG sites. The mediation analysis was performed to evaluate the potential intermediate role of DNA methylation in the association between maternal sleep and offspring adiposity indicators.

Results: The mean maternal nighttime sleep duration and the sleep midpoint for combined cohorts were 9.24 ± 1.13 h and 3.02 ± 0.82, respectively, and 24.5% of pregnant women experienced poor sleep quality in late pregnancy. After adjusting for the covariates, the maternal later sleep midpoint was associated with the increased SF in offspring (Coef. = 0.62, 95% CI 0.37-0.87, p < 0.001) at 2 years old. However, no significant associations of the nighttime sleep duration or sleep quality with the offspring adiposity indicators were found. In the SSBC sample, 45 differential methylated probes (DMPs) were associated with the maternal sleep midpoint, and then, we observed 10 and 3 DMPs that were also associated with the offspring's SF and BMI at 2 years, of which cg04351668 (MARCH9) and cg12232388 significantly mediated the relationship of sleep midpoint and SF and cg12232388 and cg12225226 mediated the sleep midpoint-BMI association, respectively.

Conclusions: Maternal later sleep timing in late pregnancy was associated with higher childhood adiposity in the offspring. Cord blood DNA methylation may play a mediation role in that relationship.

Keywords: Childhood adiposity; Cord blood; DNA methylation; Late pregnancy; Sleep midpoint.

Fig. 1

Statistical analysis workflow. DMPs, differential methylated probes; EWAS, epigenome-wide association study; FDR, false discovery rate; SBC, Shanghai birth cohort; SSBC, Shanghai sleep birth cohort. ^aAdjusted for maternal age, education level, pre-pregnancy BMI, gestational weight gain, and offspring’s sex. ^bAdjusted for maternal age, education level, pre-pregnancy BMI, gestational weight gain, offspring’s sex, batch effect, and cell type

Fig. 2

Cord blood DMPs associated with maternal sleep midpoint during late pregnancy. A Manhattan plot showing the distribution of DMPs associated with maternal sleep midpoint in chromosome. The blue line represents the threshold for multiple testing (FDR-p < 0.05). DMPs, differential methylated probes. B Pie chart showing the distribution of maternal sleep midpoint-related DMPs across CpG island region. C Venn diagram showing the overlap of maternal sleep midpoint-related DMPs associated with offspring body mass index (BMI) and subcutaneous fat (SF)