Parents’ sleep multi-trajectory modelling from 3 to 36 months postpartum in the SEPAGES cohort

Child birth is known to alter parental sleep. However, few studies considered both parents, mostly focusing on maternal and child sleep; and few had follow-up after 1-year post delivery. Then, our main questions were how does postpartum sleep evolve for mothers, fathers and within the couples; is there a correlation between the sleep patterns of mothers and fathers; what are the factors associated with sleep within couples?

To answer those questions, we analysed parental sleep as individual and as couple across multiple measures of sleep (nighttime, daytime, and perceived sleep deprivation) between 3 and 36 months postpartum providing groups of mothers, fathers and couples with similar sleep evolution over time (groups of sleep multi-trajectories). We then explored the interrelation between maternal and paternal groups of sleep evolution, and the factors associated with couples’ sleep multi-trajectories groups.

From the SEPAGES cohort (Grenoble, France), which included 484 pregnant women (<20 weeks, singleton pregnancy) and 410 fathers, we analysed data from 188 mothers and fathers who had sufficient sleep information available (nocturnal (NSD) and diurnal (DSD) sleep duration and subjective sleep deprivation (SSD)), collected via self-administered questionnaires at 3, 18, 24 and 36 months postpartum. Factors considered in the analyses concerned the parental socio-demographic factors (education level, working status and work qualification), the parents’ health and habits (chronotype, maternal and paternal anxiety or depressive symptoms during the pregnancy, breastfeeding duration) and the child (birth order, sex and birth seasonality).

Three maternal, paternal, and couple sleep multi-trajectory groups were identified using an unsupervised longitudinal cluster modelling method. These three groups had similar characteristics: 1) a group with long NSD, mid-short weekend DSD and low SSL prevalence (i.e., No sleep problem group); 2) a group with long NSD and weekend DSD but having high frequency of SSL (i.e., Subjective sleep problem group); and 3) a group with short NSD, mid-short weekend DSD and high frequency of SSL (i.e., Global sleep problem group). In all groups, fathers had shorter but more consistent sleep durations between 3 and 36 months postpartum with lower and more consistent subjective sleep loss than mothers.

The investigation of the interrelations between maternal and paternal sleep evolution showed that the fathers belonging to “Subjective sleep problem group” were less likely to have their partner belonging to “Global sleep problem group” compared to “No sleep problem group” (OR CI95% 0.29 [0.10-0.80]). Maternal sleep patterns were differently characterized as individual and as couple while fathers were more consistent, suggesting that women were more likely to adopt their male partners’ sleep patterns so that the couples tended to have similar sleep characteristics.

The analysis of the factors associated with couples’ multi-trajectories showed that, compared with couples within the “No sleep problem group”, couples with poorer sleep quantity/quality (“Global sleep problem group”) more often included a father with an evening (OR=3.86 [1.14-7.17]), and that their child included in the SEPAGES cohort was less often a first child (ORfirst vs siblings=0.27 [0.12-0.61]) born in autumn (ORautumn vs summer=0.49 [0.19-1.29)]). Parental age and symptoms of anxiety or depression during pregnancy, as well as duration of breastfeeding, were not associated with couples’ multi-trajectory sleep patterns.

Later chronotype has been reported to be associated with shorter sleep duration, and in our result, later paternal chronotype may have affected themselves and their partner. Also, having multiple children increases the burden of childcaring and it might be reflected in sleep outcomes. Finally, people tend to sleep longer during the winter season than other seasons, so this seasonal effect may translate into better sleep quantity/quality at age 3 months in children born in autumn.

By Mihyeon Kim, Sabine Plancoulaine



Back to top