Our findings confirm that women can expect to feel increasingly stronger movements through the third trimester of pregnancy, and can be reassured by perception of regular periods of vigorous movements and foetal hiccups. Conversely, perception of decreased frequency of foetal movements in late pregnancy is associated with increased odds of late stillbirth at all gestations but more so early in the third trimester. Our findings also suggest that at term (37 weeks’ gestation or later), a single isolated occurrence of more vigorous movements is associated with late stillbirth.
A strength of this study is that it had a large sample size and extensive collection of pregnancy-related variables, containing data from across several countries. This has allowed exploration of the interactions between gestation, obesity and changes in foetal activity which individual studies were not powered to do, providing valuable additional information. Analysis was also able to be carried out to investigate the classification of cause of death according to RFM and episodes of increased foetal movements.
The study has by its nature some limitations; case-control studies are subject to the potential of recall bias. The risk of this was reduced in these studies by the use of interviewer-administered questionnaires in four of the studies included, which contained no hypothesis about the potential association of various patterns of movements. The potential for selection bias also exists; however, the reasons for this would likely vary across countries, yet the prevalence of foetal movement variables was relatively consistent between studies. The risk of bias assessment using the ROBINS-E tool has been reported previously, and showed that 4 of the 5 studies had a moderate risk of bias, i.e. does not provide the level of evidence of a randomised trial, and the internet study a serious risk of bias. The sensitivity analyses however did not show any significant changes in the results when various studies were excluded from the analyses, giving us confidence in the findings.
Historically, stillbirth research has focused on a decrease in the frequency of movements, with studies and campaigns based around kick counts [9, 24]. The data presented here show that women should expect that the strength of foetal movement should remain at least as strong throughout late pregnancy but should, more often than not, increase in strength until approximately 37 weeks’ gestation when it plateaus. Women perceive these changes in strength differently, and some (40–50% as gestation progresses) may not feel stronger movement. The perception of increasing strength of movement may simply be due to increasing foetal size, and relatively limited space making movement more perceptible. Nevertheless, this is regarded as part of normal foetal development. Given that increased strength of movements is the most commonly reported pattern, feeling decreased movement becomes an even more important consideration. The risk associated with RFM has previously been calculated in comparison with women who detect no change [5,6,7], rather than those who feel a gradual increase in foetal movements.
Notably, this IPD has also shown that gestational age at the time of assessment of foetal movements is important in assessing the significance of changes in foetal movement patterns. Increasing strength of movements is most important before 37 weeks’ gestation as the magnitude of the reduced odds of late stillbirth in the presence of increased strength of foetal movements is greater than after 37 weeks. This has clinical implications for the assessment of women attending with RFM before 37 weeks, as the association with stillbirth during this period of gestation is stronger than at term. The association between RFM and FGR suggests particular attention should be made to excluding FGR and placental dysfunction in those pregnancies before 37 weeks’ gestation who present with RFM.
Our findings suggest that maternal perception of foetal hiccups is common and associated with reduced odds of late stillbirth. The phenomenon of foetal hiccups was first reported by Ferroni in 1899  and considered to be part of normal foetal development . The prevalence of hiccups in this analysis increased through to approximately 37 weeks’ gestation, which is divergent to other reports that suggest foetal hiccups are more prevalent earlier in pregnancy and decrease as pregnancy progresses [27, 28]. As with general foetal movements, increased maternal perception of foetal hiccups near term may be due to greater foetal size, changes in foetal breathing or neurodevelopment or may reflect increased maternal recognition of the sensation. The data from this IPD indicates that foetal hiccups are a normal part of pregnancy and are not associated with increased odds of stillbirth.
The AFFIRM and Mindfetalness trials have raised debate around the clinical usefulness of foetal movement awareness to prevent stillbirth. Critically, the AFFIRM study protocol which reported no reduction in the rate of late stillbirth (aOR 0.90, 95% confidence interval (CI) 0.75–1.07)  recommended intervention for RFM after 37 weeks’ gestation, particularly when recurrent at term. Our analysis demonstrates that this recommendation coincided with the lowest, though still increased, odds of stillbirth associated with RFM. The Mindfetalness trial employed a structured approach to awareness of foetal activity  and found no difference in the primary outcome of the number of babies born with an Apgar score < 7 . Importantly, both AFFIRM and Mindfetalness identified that the investigation and subsequent management of RFM reduces the proportion of SGA infants at birth compared to the control groups. This is consistent with the association between RFM and late stillbirths due to foetal growth restriction. McCarthy et al. reported that women presenting with RFM have a higher burden of care, including increased rates of induction, admission to neonatal units and higher levels of surveillance . Our findings suggest that women with RFM should be assessed to exclude foetal compromise and FGR rather than receiving intervention for RFM alone to focus intervention on those most likely to benefit.
Importantly, our analysis found no interaction between the combined strength and frequency variable and maternal obesity. A systematic review of obesity and foetal movements identified limited data and reported that maternal body size was not associated with altered ability to perceive foetal movements (4 studies of 95 women; very low-quality evidence) . In a further study, maternal reporting of foetal movement strength and frequency was not different in relation to obesity, highlighting that maternal BMI is not a barrier to the detection of foetal movements and the clinical importance of a presentation with foetal movement concerns is not diminished by maternal body size . As such, all women regardless of BMI, attending with concerns about foetal movements, should be treated the same. The increased risk of stillbirth in relation to obesity is likely to be multifactorial; previous studies have continued to show an increased risk associated with obesity even after controlling for factors such as diabetes and pre-eclampsia [19, 30].
The most difficult finding to interpret from this IPD meta-analysis has been that of foetal movements that are more vigorous than usual. Nearly half of women report this occurring on more than one occasion which appears to be protective of late stillbirth. Only women who report an isolated instance of more vigorous than usual foetal movements have increased odds of late stillbirth. The prevalence of a single event of vigorous movements in our data (18% in cases vs. 6% in controls) corresponds to the data from other studies which suggest an incidence of 10% amongst late stillbirths . It has been suggested that such excessively vigorous movements could be related to foetal seizure activity or umbilical cord compression or entanglement . Women who have a late stillbirth and report one instance of vigorous foetal movements often describe the movements as ‘crazy’ and ‘wild’ . The distinction between ongoing vigorous movements and a single episode of exaggerated foetal activity is difficult and can only be achieved in retrospect, as exemplified by two cohort studies of women presenting with increased foetal movements which both found no association with adverse outcome [32, 33]. However, our analysis suggests that the pattern of vigorous foetal movements needs to be considered in light of the gestation, with the highest odds of stillbirth in women with a single episode of increased foetal movements, after 37 weeks’ gestation. Therefore, further research would be valuable in relation to increased movements. The ideal design would come from women who had previously experienced vigorous movements both as part of a pregnancy with a live born and as part of another pregnancy resulting in a stillbirth, such a study would be difficult to undertake in relation to sample size and recall bias. A follow-up study to STARS—the Pregnancy Research Project—is currently collecting such data (O’Brien, personal communication).