Objective To evaluate the relationship between first-trimester crown-rump length (CRL) and birthweight (BW) Z scores.Design Retrospective cohort study.Setting Two tertiary centres in France.Population Three hundred and seventeen pregnancies conceived through assisted reproductive techniques between April 2001 and December 2008.Methods We used CRL and worked forward to BW. Only pregnancies examined during the first trimester by an Fetal Medicine Foundation-certified operator were included. CRL was expressed as Z scores, and BW was transformed into Z scores by taking gestational age and gender into account. The influence of abnormal first-trimester CRL Z scores on BW was examined.Main outcome measures Weight and gestational age at birth.Results Birth weight was significantly greater in babies with larger CRL: BW Z scores (±SD) were )0.36 (±1.05), )0.27 (±0.97), )0.10 (±1.04) and 0.13 (±0.96) in the first, second, third and fourth quartiles of CRL Z scores, respectively (P = 0.01). In contrast, there was no difference in gestational length according to the quartiles of the CRL Z scores. The CRL Z score was a significant predictor of the BW Z score (b = 0.17, P = 0.001). After adjustment for maternal body mass index, a one-point increase in the first-trimester CRL Z score (i.e. 3.6 mm) was associated with a 39% decrease, 64% increase, 114% increase and 62% increase in the risk of having a BW below the 10th centile [odds ratio (OR), 0.61; 95% confidence intervals (95% CI), 0.39; 0.95; P = 0.03], above the 90th centile (OR, 1.64; 95% CI, 1.03; 2.60; P = 0.02), above the 95th centile (OR, 2.14; 95% CI, 1.25; 3.68; P = 0.006) and above 4000 g (OR, 1.62; 95% CI, 1.04; 2.51; P = 0.04), respectively.Conclusions Variations in BW may be partly explained by differences in growth trajectories that may express as early as the first trimester.

No abstract

We wish to thank Dr Tiran Dias and Professor Basky Thilaganathan 1 for their comments on our recent publication about the association between crown-rump length (CRL) and birthweight. 2 They comment on our method to date pregnancies following in vitro fertilisation (IVF). Indeed, gestational age (GA) was calculated by adding 14 days to the number of days between the date of oocyte retrieval and the date of ultrasound examination. This seems an appropriate method because fertilisation was performed on the day of oocyte retrieval in all women. Moreover, this method remains reliable whatever the timing of embryo or blastocyst transfer later on. Moreover, we must emphasise that there were no ovum donations or frozen embryos in our cohort.Second, and more importantly, they hypothesise that such misdating, rather than early variation in fetal growth, could create an artificial association between early and late growth. This has already been suggested by Smith. 3 To prevent such spurious associations, they suggest that we should look at the association with bodyweight once pregnancies have been dated by the CRL measurement. This is a good point but was already performed in our study, as stated in the second paragraph of the Results section 'This relationship persisted when GA was calculated from CRL (b = 0.13, P = 0.02)'.We understand that our work challenges the traditional view stating that physiological variations in fetal size emerge during the latter half of pregnancy. However, we believe that our data add to a growing body of evidence that complications of late pregnancy, such as abnormal growth, may be the ultimate consequence of conditions that have their origins in the very earliest weeks of gestation and precede the first ultrasound examination. j For the first time our report also included an assessment of whether or not the perinatal deaths were considered to have been potentially avoidable. We have also recently published our system for classifying contributory factors and potential avoidability for 4 years of data on maternal deaths (49 deaths in total). 4 Our classification differs from other approaches as it includes barriers to accessing and engaging in care, as well as management and organisational issues, personnel issues, geography and environmental factors. The maternal deaths were assessed by a central working party of the PMMRC, whereas the 721 perinatal deaths were assessed by local teams of clinicians, which typically included senior midwives, obstetricians and neonatologists or paediatricians. We are currently undertaking further research to validate this classification system by comparing local review with an expert review panel. Using this approach we found that onethird of maternal deaths and 14% of perinatal deaths (15% of stillbirths) were potentially avoidable. The most common categories of contributory factors found in both maternal and perinatal death reviews were barriers to accessing and engaging in maternity care, and lack of knowledge and skills among personnel. We can now direct...