Identifying risk factors for stillbirth

BMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f416 (Published 24 January 2013) Cite this as: BMJ 2013;346:f416
  1. Lesley M E McCowan, professor,
  2. Katie M Groom, senior lecturer
  1. 1Department of Obstetrics and Gynaecology, Faculty of Medical and Health Science, University of Auckland, Private Bag 92019, Auckland
  1. l.mccowan{at}auckland.ac.nz

Efforts to improve detection of fetal growth restriction must be intensified

In developed countries stillbirth is one of the few potentially avoidable maternal and child health complications that have not declined in recent decades.1 Stillbirth rates vary between and within countries, suggesting that it might be possible to reduce this devastating problem.2 A linked study by Gardosi and colleagues (doi:10.1136/bmj.f108) adds important new insights into risk factors for non-anomalous singleton stillbirths after 24 weeks.3 The findings highlight the important contribution of the modifiable risk factors of fetal growth restriction (FGR) (especially when it is unrecognised before birth), smoking, and obesity.

In this English retrospective population based study, FGR is defined using a customised standard that correlates better with stillbirth than population birthweight references.4 5 FGR was recognised antenatally in 31% of women in the general population but in only 18% of women who went on to have a stillborn infant. Among pregnancies that ended in stillbirth, unrecognised FGR was associated with 32% of all deaths (relative risk for stillbirth 8.0, 95% confidence interval 6.5 to 9.9). About 90% of these stillbirths were born after 26 weeks and 50% after 33 weeks; if FGR had been recognised and the babies born alive many of these small babies would have survived.

Identification of FGR is an important goal of antenatal care, and the rates of detection in both live born and stillborn infants in this study are disappointing. Gardosi and colleagues point out that rates of detection of FGR varied between obstetric units (12.5-50%), and they attributed this variation to differences in training and adherence to protocols regarding use of customised antenatal growth charts (GROW). A recent Australian report found that antenatal detection of FGR increased from 24.8 % to 50.6% after implementation of GROW charts as unit policy.6 However, the report found that, even with the use of GROW, half of all FGR pregnancies remained undetected, highlighting the need for robust strategies for prediction and improved detection.

Pregnancies with antenatal detection of FGR in Gardosi and colleagues’ study still had a fourfold increase in the relative risk of stillbirth, and the reasons for this are not discussed. Detection of FGR must be accompanied by careful fetal surveillance and timely delivery, the combination of which has been associated with reduced perinatal mortality.7 Guidelines shortly to be released by the Royal College of Obstetricians and Gynaecologists for the management of FGR may help clinicians apply a consistent and evidence based approach to the management of this condition.

The linked study also explored the association between active and passive smoking and the risk of stillbirth, as well as their interactions with FGR. The authors emphasise that both maternal and passive smoking are important modifiable risk factors for stillbirth, with a combined population attributable risk due to smoking of 24.6%. The results of multivariable analyses have previously suggested that the effect of passive smoking is mediated only through FGR, whereas Gardosi and colleagues found that women who were active smokers but did not have FGR still had an increased risk of stillbirth.

Cessation of maternal smoking early in pregnancy is associated with reduced spontaneous preterm birth and stillbirth,8 as well as lower rates of small for gestational age babies.9 Clinicians must take a smoking history in early pregnancy, including a history of passive smoking. They should stress the importance of a completely smoke-free environment for the developing fetus and make appropriate referral or provide extra support to achieve early smoking cessation.

Consistent with other reports, the current study found that obesity was associated with an increased risk of stillbirth.10 11 The mechanisms of stillbirth in obese women are likely to be multifactorial. However, it is a particular concern that obesity has recently been reported to be associated with an increased risk of FGR by customised centiles12 and, not surprisingly, a lower rate of antenatal detection of FGR.13 Randomised trials are needed to determine whether third trimester scanning can increase detection of FGR in obese women. Unlike smoking, the adverse effects of obesity cannot easily be reversed during pregnancy. A recent systematic review of nutritional interventions demonstrated a 3.8 kg reduction in gestational weight gain; it also found reduced risk of stillbirth but this was not statistically significant.14 Large randomised trials of lifestyle interventions in overweight or obese women will provide further data regarding their potential impact on stillbirth.15 16

Contrary to the findings of a recent meta-analysis, advanced maternal age was not significantly associated with increased risk of stillbirth in the current study.17 The National Institutes of Health Stillbirth Collaborative Research Network and the Auckland Stillbirth Study also found no such association for singleton infants without congenital abnormalities.8 11 Gardosi and colleagues suggest that some of the previously described association may be explained by the relation between advanced age and congenital anomalies. Although this may be true, the debate is still open. Further studies that consider variations in antenatal management and timing of delivery for older women are needed before advanced maternal age can be excluded as a risk factor for stillbirth.


Cite this as: BMJ 2013;346:f416


  • Research, doi:10.1136/bmj.f108
  • Competing interests: We have read and understood the BMJ Group policy on declaration of interests and have no relevant interests to declare.

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