Should pregnant women sleep on their left?

BMJ 2011; 342 doi: http://dx.doi.org/10.1136/bmj.d3659 (Published 14 June 2011) Cite this as: BMJ 2011;342:d3659
  1. Lucy C Chappell, clinical senior lecturer in maternal and fetal medicine1,
  2. Gordon C S Smith, professor and head of department2
  1. 1Division of Women’s Health, King’s College London, London SE1 7EH, UK
  2. 2Department of Obstetrics and Gynaecology, University of Cambridge, Cambridge, UK
  1. lucy.chappell{at}kcl.ac.uk

The suggestion that this may help to prevent late stillbirth requires further study

The United Kingdom has one of the highest rates of stillbirth in the developed world.1 More than 4000 infants were stillborn in the UK in 2009 (out of about three million worldwide), and 1200 of these stillbirths occurred at or after 37 weeks’ gestational age.2 In high income countries, there are 10 times more stillbirths than deaths from sudden infant death syndrome—the subject of a major campaign.3 Stillbirth affects four times more babies than Down’s syndrome. A sophisticated screening programme is recommended and widely offered for Down’s syndrome, yet screening for stillbirth in the general population is confined to measurement of the external size of the uterus with a tape measure at each visit, according to the current National Institute for Health and Clinical Excellence Antenatal Care guideline.4 Any simple intervention that reduces the risk of stillbirth would be extremely welcome. Could the linked study (doi:10.1136/bmj.d3403), in which Stacey and colleagues found an association between maternal sleep position and risk of stillbirth,5 constitute the basis for a “not back to sleep” campaign for pregnant women?

Evidence based guidelines on sleep in pregnancy from professional organisations are sparse. Advice on the internet abounds, but much of it is derived by extrapolation from other contexts. The theory is that if a woman lies supine the gravid uterus may compress the inferior vena cava, resulting in reduced venous return and limited uteroplacental blood flow. A study of 22 women after 36 weeks’ gestation showed that placental intervillous blood flow was about 25% lower in the supine compared with the left position.6 But in healthy third trimester women with normal size babies, there is only limited evidence that supine positioning in antenatal women,7 or in women undergoing caesarean section,8 is associated with any significant change in fetal indices. However, in clinical situations involving pregnant women, it is standard practice to favour left lateral tilt over other positions, and it is plausible that this position may be better for the baby in other contexts.

Stacey and colleagues compared women who had had a late stillbirth with pregnant control women, who were matched for gestational age but went on to deliver a healthy baby. In addition to sleep position they analysed 15 other risk factors, and some of the findings have been reported elsewhere.9 The key questions in interpreting this study are: is the observed significant association between maternal sleep position on the night before the stillbirth a chance finding? Can it be explained by bias? Could it be reverse causation?

The current study identifies eight classifications of sleep position, each containing four groups. Randomised controlled trials in which large numbers of comparisons are reported are not as reliable as those that report on a prespecified primary outcome because statistical significance will almost inevitably be seen when large numbers of outcomes are compared; trials now have to prespecify a primary outcome. The substantial number of comparisons in the linked study means that it must be considered as a hypothesis generating study rather than a hypothesis testing one.

As with any retrospective study, results may be explained by bias. Women who had experienced a stillbirth completed questionnaires 25 days after the event. A report of left sided sleeping position may be a surrogate measure of increased access to and uptake of sources of educational information, and it may act as a confounder for which adjustment for social deprivation level is insufficient. The authors did not present detailed information on cause of death in the stillbirth cases in this report, and the lack of any analysis of cause of death as an influential variable makes it difficult to assess the biological plausibility of the study’s findings. A greater association between non-left sided sleep position and stillbirth in fetuses vulnerable to impaired uteroplacental blood flow, such as those with growth restriction, would add weight to the finding. Stillbirth is a descriptor of heterogeneous events that lead to death; future research should elucidate which pregnancies are most likely to benefit.

There is a strong possibility that part of the association can be explained by reverse causation. Reduced fetal movement is one of the most common symptoms seen before stillbirth. Moreover, in many cases delay occurs between intrauterine fetal death and its confirmation by a health professional. In a proportion of cases of stillbirth in this study, the baby may have died before the last sleep night reported by the mother. Compromised babies may have reduced movements in the days leading up to the death. Hence, rather than being a cause of stillbirth, the associations between longer sleep and not rising during the night in the week before stillbirth may reflect absent or reduced fetal movements, as a consequence of the baby’s death.

A forceful campaign urging pregnant women to sleep on their left side is not yet warranted. Further research is needed before the link between maternal sleep position and risk of stillbirth can be regarded as strongly supported. If these findings are validated in a future study, advice on sleep position is an intervention that would be relatively easy to implement. The message is appealing, perhaps partly because of resonance with the campaign on changing sleep position for infants, which led to a marked reduction in rates of sudden infant death.10 However, the impact of a similar intervention in pregnancy is uncertain. A previous study in which pregnant women over 30 weeks’ gestation admitted to the antenatal ward were directly observed reported that most (77%) women slept with a left tilt,11 which is higher than the proportion of controls who reported left sided sleep in the current study. Although the message for mothers to sleep on their left is probably harmless and may be helpful, this study should be seen as one that only generates a hypothesis that needs validation.


Cite this as: BMJ 2011;342:d3659


  • Research, doi:10.1136/bmj.d3403
  • Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.