Knights et al. (1) reported on the sexes of the offspring of 216 women with gestational diabetes (GDM) and 216 control subjects without GDM, all of whom underwent cesarian section. The sex ratio (proportion male) for the control offspring was significantly higher than that for the offspring of the GDM women. The birth weights and gestational ages of the offspring from the two groups of women showed no appreciable differences, even when offspring sex was controlled. These authors claimed that their data do not support the suggestion that the high sex ratio generally associated with cesarian section is due to the greater weight of male fetuses. They asserted that although the male babies were heavier than the female babies in both groups of women, the male babies of the women with GDM had the same mean birth weight as the male babies of the control women.
I suggest that they misinterpreted their data; there can be no reasonable doubt that one cause of the high sex ratio associated with cesarian section is that male babies are bigger, on the average, and consequently are associated with failure to progress at delivery (2). The rationale of this practice is that cesarian section is associated with lower injury rates in large infants than are forceps, vacuum, or spontaneous vaginal delivery (3).
I suggest that the high sex ratio generally associated with cesarian section is due to strong selection for fetal weight. If I am correct, this sex ratio is a statistical artifact of the circumstance that weight is one criterion for cesarian section. In 1993, the sex ratios of U.S. live births in the weight ranges of 3,000, 3,500, 4,000, 4,500, and ≥5,000 g were 0.482, 0.554, 0.626, 0.679, and 0.703, respectively (4). So, the higher the birth weight criterion for cesarian section, the higher the sex ratio associated with cesarian section. In contrast, GDM women 1) have heavier infants on the average (5), and 2) are more weakly selected for fetal weight (in the sense that a higher proportion of these women are sectioned) (3).
These features would result in roughly the same mean birth weight in the GDM women and control subjects but a lower sex ratio in the GDM cases. In short, cesarian section subjects control women to strong selection on an average birth weight distribution, whereas cesarian section subjects GDM patients to weaker selection on a high mean birth weight distribution. If I am correct, these circumstances explain the features of their data, which mystified Knights et al. (1).
References
Address correspondence to William H. James, The Galton Laboratory, University College London, Wolfson House, 4 Stephenson Way, London NW1 2 HE, U.K.
