Stillbirth and birth order

1. Some stillbirth sibships are characterized by positive birth-order effects, and among them some are probably of the sort illustrated by Fig. 1; other stillbirth sibships are characterized by negative order effects, and among them some are probably of the sort illustrated by Fig. 2. 2. The ‘strength’ of these two opposing effects was roughly equal in the sample tested, so there was no over-all trend of stillbirths with birth order within sibships. 3. There seem to be a few sibships in which a curvilinear effect is present, with higher risks of stillbirth at the beginning and at the end of the same sibship. 4. Lastly, it seems possible (though no direct evidence on this point has been presented in this paper) that many stillbirths belong to sibships which do not exemplify any appreciable maternal age or order effect. 5. It seems likely that the positive and negative effects are related to specific pathological causes of stillbirth. 6. It seems possible that the occasional sibship seeming to show higher risks at both ends is due to the co-existence in a single woman of a propensity to bear two types of stillbirth—one type characterized by a positive effect, and the other by a negative effect. 7. The stillbirth ratio declines from the 1st to the 2nd and 3rd birth ranks, and thereafter increases. 8. The initial drop is due to a genuine negative birth order (or maternal age) effect. 9. The subsequent rise seems to be not primarily due to secular changes within individual women (either as a result of maternal age or birth order) but to an association of fertility with proneness to stillbirth. The cause may thus be social rather than biological. 10. The cause of this association is not that a stillbirth is likely to occasion reproductive compensation. (Indeed in this sample, at all the birth ranks up to the ninth, a stillbirth promoted birth limitation instead.) 11. The association of fertility with a propensity to stillbirth may be mediated by social class or by duration of marriage, but it is not clear that the underlying cause of this association has been uncovered. 12. One recommendation would follow from this work if its findings should be confirmed: stillbirth-prone women on the average have more pregnancies than other women. This fact is a major cause of the variation of stillbirths rate with birth rank. This being so, workers using multivariate analysis may feel the need to scrutinize the assumptions implicit in their models.