Sleep complaints in early pregnancy. A cross-sectional study among women attending prenatal care in general practice

Cross-sectional study comprising questionnaires and Pregnancy Health Records for pregnant women participating in the first prenatal care visit to their GPs.

The healthcare system in Denmark is tax-funded, and care is free of charge for the patient. The majority of Danes (99%) are registered with a GP who functions as gatekeeper to secondary care. A minimum of three prenatal care visits and one postpartum are offered by the GP. The first consultation is offered at gestational age 6–10 weeks to all women who wish to go through with their pregnancy. This consultation, which precedes other contacts to the healthcare system, is accepted by almost 100% of women. In this consultation a thorough and structured record is established (the Pregnancy Health Record) and sent to midwives and hospital departments. The second and third prenatal care visits usually take place at pregnancy weeks 25 and 32, and the postpartum examination 8 weeks after birth.

GPs were recruited from two of the five Danish administrative Regions (The Capital Region of Denmark and Region Zealand) with a total of 1561 practices, organised in 53 geographical units. Among these, 19 units were randomly selected. The 308 practices situated in the selected geographical area were individually contacted and asked to participate in the study. A total of 192 practices agreed to participate. The practices could be either single-handed or partnerships. Among the 190 participating practices, 117 were active in the study and recruited one or more pregnant women to the study. The 117 active practices comprised a total of 294 GPs.

All pregnant women booking an appointment for the first prenatal care visit with one of the participating GPs were eligible for inclusion in the project. The women received oral and written information on the project and were consecutively included after signing a consent form. The inclusion period was from 1 April 2015 to 15 August 2016. Women were excluded if they did not complete the electronic questionnaire (all in Danish), if they withdrew consent or if the pregnancy ended in abortion.

Data were collected from the Pregnancy Health Record and from an electronic patient questionnaire. The electronic questionnaire was sent to the women after the first prenatal care visit. The questionnaire could only be returned when it was fully completed. Nonrespondents were sent two reminders and some were contacted by phone.

Characteristics of the study population were compared between women without sleep complaints and those with sleep complaints. Multivariable logistic regression analysis was used to test the associations between “moderate or severe sleep complaints” and pregnancy-related physical symptoms. Women were labelled as having “moderate or severe sleep problems”, if they had reported a moderate or severe problem in response to one or more of the three sleep questions. The associations were assessed in five ways: unadjusted, adjusted for each of the three blocks individually, and adjusted for all three blocks simultaneously. The purpose of these adjustments was to see whether the additional information could explain the associations apparent in the unadjusted assessment. Only women with complete records and questionnaires were included in the regression analysis.

On average, more than one third of pregnant women experienced some degree of poor sleep in early pregnancy, although most symptoms were rated as minor. Waking up too early was reported by 47% of the women, but few were worried. On the other hand, 46% of the women who had been “taking a long time to fall asleep”, and 40% of the women who had been “lying awake most of the night” were worried to a moderate or severe degree. Sleep complaints showed association with other pregnancy-related symptoms, such as back pain, pelvic girdle pain and pelvic cavity pain, but only the association with pelvic cavity pain stayed significant after adjustment for depression.

No exclusion criteria were used, but with regard to representativity for the general population of women in early pregnancy a number of factors must be considered. The questionnaire was in Danish and results may be less representative for non-native women. The GPs participating in the study were sampled by a systematic procedure based on random selection from two representative regions of Denmark, including urban and rural areas, and areas with low as well as high social status. This strengthens the geographical representation of the study, but the GP practices that voluntarily chose to contribute constitute only approximately 40% of those asked. There were very few nonrespondents among women who accepted to participate, and complete data were obtained from almost all participants. However, not all eligible women may have been asked, and not all women managed to have had their first prenatal care visit between pregnancy weeks 6–10, where the inclusion was supposed to take place. This delay in receiving the signed consent form with the women’s e-mail from the GPs led to a delay in answering among the participating women. Those women were asked to recall how they experienced symptoms and feelings around week 9–10.

We assessed sleep complaints by means of questions to the pregnant women. Such subjective measures of sleep duration and sleep efficiency have previously been shown to be correlated with objective measures [10, 21], but may obviously be affected by the individual’s reporting style. The degree of trouble and the level of worries can, however, only be assessed by the women themselves. Unfortunately, we do not have pre-conception information about sleep complaints and, therefore, cannot be sure that problems are different post-conception. Significantly more mid-sleep awakenings have, however, previously been found in pregnant women compared to non-pregnant women [22].

The most commonly used postpartum instrument is the Edinburgh Postnatal Depression Scale (EPDS), which was developed to measure postpartum depression. For our study we used the MDI because it is more often used in clinical practice by GPs in Denmark. MDI has been demonstrated to have good validity and reliability, and only includes is a single question about sleep [23]. Cronbach’s Alpha value for the MDI in our study was 0.84 and for the ASS was 0.80. Factors associated with sleep problems were measured with validated scales when possible (e.g. sociodemographic status (https://​www.​dst.​dk/​da/​).

Problems with sleep in late pregnancy have been described in many studies and may affect more than 2/3 of women [24, 25]. The enlarged uterus and movements of the child may, for example, explain this. Studies from early pregnancy are fewer and the population size of our study is relatively large. We found poor sleep quality among 47% of the women. This is lower compared to 74% poor sleepers among 346 self-selected women in the first trimester in a recent web-based study [26], and other small cross-sectional studies have found frequencies between 28 and 38% [27] and 60% [28]. Sleep complaints in the non-pregnant populations seem to be less frequent than among first trimester pregnant women. A national sleep survey of Australian adults reported that 33–45% had inadequate sleep [29], and a Japanese study found the frequencies to be around 29% [30].

We found associations between sleep complaints and some pregnancy-related physical symptoms, leading us to explore possible mechanisms through which sleep in pregnancy may impact health or vice versa. A study exploring mid-sleep awakenings found no associations with nausea and indigestion [22]. Similarly, we found no association with nausea. Another study found that women experiencing distressing physical symptoms during pregnancy more often suffer from poorer sleep quality and more symptoms of depression [31], which corresponds well with our result showing association with sleep complaints and mood. Furthermore, our women had higher depression score (MDI score > 20) 8 weeks postpartum, if they experienced physical discomfort in early pregnancy, such as back pain and pelvic cavity pain [32]. Pain as well as depression affect sleep, and depression may interact with pregnant women’s subjective impression of pregnancy-related pain and their sleep experiences. This could be described as maternal stress, inasmuch as several studies have found that sleep problems relate to maternal stress. These heterogeneity studies show associations with elevated blood pressure [11], childhood abuse [33], intimate partner violence [34] and higher levels of symptoms of anxiety and depression, both during pregnancy and postpartum [13, 35]. However, it is debatable how strong this relation is, as it seems that our women were not so distressed that the sleep complaint worried them. Nevertheless, some of the pregnant women may have been worried, because those who experienced poor sleep quality had a high use of health care [36]. Overall, GPs have a unique opportunity for reducing maternal distress by introducing sleep intervention, which represents a potential low-cost, non-pharmacological prevention and treatment strategy for postpartum mental illness, by articulating that the pregnant women should be aware of the connection between sleep and symptoms of depression and anxiety [13, 37, 38]. In addition, the women should be encouraged to engage in physical activity and exercise during pregnancy, as studies have shown that this has an overall beneficial effect on sleep characteristics and sleep continuity and on the woman’s mental state [39, 40].

Sleep complaints in early pregnancy are common and should be addressed by the GP and midwife at the first prenatal care visit. Anxiety and mood may be related and relevant topics for discussion in relation to sleep problems.