Even in pregnant women without gestational complications, sleep/wake problems were frequently observed and were more prevalent during the third trimester than during the second trimester, except for EDS. EDS was common throughout the course of pregnancy and was associated with short sleep durations and DMS, rather than SDB. Additionally, cesarean sections were slightly more prevalent in the insomnia group and one case of stillbirth with maternal RLS/WED occurred in this group. However, other sleep/wake problems did not affect delivery outcomes.
Relationships among different sleep/wake problems in pregnant women
In this study, SDB was not a significant factor for EDS among pregnant women. It has been reported that ESS global scores at pre-pregnancy and in the third trimester are higher among pregnant women with obesity than among those without obesity . Moreover, snoring during the first trimester is associated with not only continuous EDS throughout pregnancy, but also with EDS onset during pregnancy . In the present study, only 6 participants were obese before pregnancy, and almost all participants suspected with SDB were considered mild cases. Such factors might however influence the relationship between SDB and EDS.
Although, DIS and DMS were frequently observed during the third trimester in the present study, moderate to severe DMS, rather than moderate to severe DIS, was associated with EDS. Physiological changes from the second to third trimester are thought to affect both DMS and EDS. During the third trimester, frequent urination due to a rapidly growing uterus, backpain, leg cramps, and increased fetal movement can cause both DIS and DMS. However, in a study reporting subjective findings using the Basic Nordic Sleep Questionnaire during the third trimester, while the percentage of participants with DIS was only 14%, the percentage of participants with DMS was as high as 70% . As one of the causes of EDS among pregnant women without gestational complications, DMS due to physiological changes during late pregnancy may be important.
In the present study, short sleep durations of ≤ 6 hours were associated with EDS. Total sleep durations were shorter among participants with insomnia than among those without insomnia. Although there was no collinearity between total sleep durations and DMS in a logistic regression analysis, shorter sleep durations induced by DMS, but might affect EDS.
Although an earlier meta-analysis associated shorter sleep durations with preterm birth , this was not confirmed in the present study. Shorter and longer sleep durations are known risk factors for obesity, diabetes, hypertension, and cardiovascular disease, which might be risk factors for gestational diabetes and hypertensive disorders during pregnancy [12, 27]. However, in pregnant women with normal gestational course until the second trimester, the effect of these factors on later gestational complication and delivery outcomes might be small. Along with sleep/wake problems and general sleep-related health problems, sleep durations have been speculated to affect gestational complications and delivery outcomes. We discuss potential problems in the one case of stillbirth, where the maternal sleep duration was 6 hours in the third trimester, below (see section “Restless legs syndrome/Willis-Ekbom disease among pregnant women”).
Sleep disordered breathing among pregnant women
In the present study, BMI values were higher in pregnant women with a 3% ODI ≥ 5/hour than in those with a 3% ODI < 5/hour from pre-pregnancy to the third trimester; the tendency was clearly noticeable in the later period of pregnancy. These findings are consistent with those of previous studies[10, 16, 28-30]. However, there were only few participants with obesity in this study. Additionally, a lower BMI is thought to be associated with a lower desaturation index obtained from the pulse oximeter.
SDB is associated with gestational complications, delivery outcomes, and neonatal conditions [6, 8-10, 13]. However, there was no significant effect on delivery outcomes and infant conditions in the 3% ODI ≥ 5/hour group. Overall, SDB severity was relatively mild in the study participants; this is thought to have contribute to the good outcomes along with the fact that the participants had no gestational complications.
Restless legs syndrome/Willis-Ekbom disease among pregnant women
RLS/WED is prevalent among pregnant women and affects DIS, DMS, and EDS [7, 11, 31]. In the present study, the prevalence of suspected RLS/WED was increased during the third trimester. However, the presence of RLS/WED did not affect EDS, insomnia symptoms, or subjective sleep quality. In pregnant Japanese women without gestational complications, few severe cases of RLS/WED have been reported, and almost all RLS/WED patients reported only mild to moderate symptoms . Among 8 patients suspected with RLS/WED during the third trimester of pregnancy in this study, 7 reported moderate to severe symptoms. However, only 3 patients experienced these symptoms twice a week or more. The lower symptom occurrence frequency and the small sample size might have influenced the relationship between RLS/WED and EDS observed in this study.
It is worth mentioning that a stillbirth was observed in one pregnant woman with RLS/WED. This participant had complained of insomnia and EDS, which seemed to be related to her RLS/WED symptoms, and her subjective sleep quality had decreased. However, no other pregnancy-related complication was observed. While there are some reports of an association between RLS/WED and gestational complications [11, 32] as well as delivery outcomes [11, 14], other studies did not observe an association between RLS/WED and delivery outcomes [15, 31, 32]. Although we cannot assume an association between maternal sleep/wake problems and the stillbirth in this case observed here, further studies are needed to clarify the influence of RLS/WED on delivery outcomes.
The pathophysiological mechanism underlying the higher percentage of cesarean section in the insomnia group in this study is not clear. However, compared to pregnant women without RLS/WED, an odds ratio (95% CI) of 2.40 (1.03–4.42) for cesarean sections has been reported in pregnant Iranian women with RLS/WED, and insomnia has also been reported as being more prevalent among those with RLS/WED . Although such a higher prevalence of cesarean sections in pregnant women with RLS/WED was not found in the present study, insomnia was related to cesarean section in previous reports as well as in the present study. To determine the relationship between cesarean sections and insomnia, a more comprehensive study on sleep/wake problems among pregnant women is needed.
Some limitations of the present study should be noted. First, we could not assess the specific sleep/wake problems of high-risk pregnant women who are also obese. Second, SDB was evaluated using only a pulse oximeter; therefore, no information on sleep stages, respiration itself, and position can be provided. Third, symptoms of RLS/WED were examined using a self-reported questionnaire, and the diagnosis of RLS/WED was not confirmed by sleep medicine specialists.