PT - JOURNAL ARTICLE AU - John F Mulley TI - Quantification of miscarriage sex bias in England and Wales, 1993-2017 AID - 10.1101/418186 DP - 2018 Jan 01 TA - bioRxiv PG - 418186 4099 - http://biorxiv.org/content/early/2018/11/29/418186.short 4100 - http://biorxiv.org/content/early/2018/11/29/418186.full AB - Study question How extensive is miscarriage sex bias?Summary answer Girls are more susceptible to miscarriage, with 941-986 males miscarried per 1,000 females in England and Wales between 1993-2017.What is known already The human sex ratio at birth is skewed towards males, with on average 1,053 boys born for every 1,000 girls in England and Wales between 1993 and 2017. Stillbirth also shows a male bias, averaging 1,112 boys lost per 1,000 girls over the same period. If the sex ratio at conception is equal, and more boys are lost to stillbirth, then more girls must be lost during pregnancy.Study design, size, duration Here I use data on live births, stillbirths and legal therapeutic and elective abortions in England and Wales from 1993-2017 to determine relevant annual numbers of conceptions if 10%, 20%, 25% or 33% of conceptions result in miscarriage.Participants/materials, setting, methods Subtracting known numbers of boys and girls in live births and stillbirths from the predicted conceptions, and predicting sex ratios of aborted fetuses, allows calculation of the sex ratio of miscarried products of conception.Main results and the role of chance There were 23,616,601 to 31,723,793 conceptions in England and Wales between 1993-2017, resulting in 16,656,203 live births (8,114,739 female and 8,541,464 male); 86,714 stillbirths (41,059 female and 45,655 male); and 4,512,024 legal abortions. There were 2,361,660 to 10,468,852 miscarriages, averaging between 94,466 per year (941 males per 1,000 females) if 10% of all conceptions result in miscarriage to 418,754 per year (986 males per 1,000 females) for a miscarriage risk of 33%. more girls were aborted than boys, more boys were born live and stillborn, and significantly more girls were miscarried (P < 0.00001, Pearson’s χ2 test)Limitations, reasons for caution The abortus sex ratio was determined on the assumption that early (≤12 weeks) abortions are biased towards females (55:45) and later abortions (≥13 weeks) are biased towards males (45:55). If the abortus sex ratio is balanced, the miscarriage sex bias is exacerbated, with between 691-921 males lost per 1,000 females. The miscarriage rate is also currently unclear, but my predictions of the number of missing conceptions suggests that the values used here are not unreasonable.Wider implications of the findings Girls are more susceptible to miscarriage than boys, and there is a greater loss of girls throughout pregnancy. This study quantifies the extent of this miscarriage sex bias, and provides a starting point for understanding why this bias exists, how and when it varies, and when it is established during pregnancy. In the absence of evidence to support a male-biased primary sex ratio, it seems likely that the greater loss of girls is the result of parental conflict and imprinting, perhaps because of a trade-off between an elevated risk of early loss of female embryos versus a greater robustness of survivors, as evidenced by the greater number of boys that are stillborn, and the higher infant mortality rate of boys.Study funding/competing interest(s) None to declare.Trial registration number Not applicable.