Maternal age and birth rank in the aetiology of mongolism.

The incidence of mongolism increases so remarkably with advancing maternal age that it hardly seemed necessary to earlier workers to know the maternal age structure of the population from which their cases were drawn (Shuttleworth, 1909; Brousseau, 1928; Engler, 1949), or to ensure that the control group was representative (Jenkins, 1933; Rosanoff and Handy, 1934; Lahdensuu, 1937; Bleyer, 1938; Benda and others, 1943; Beall and Stanton, 1945). These precautions were not neglected by Carter and MacCarthy (1951), who used a series of hospital births to establish firmly the relationship between incidence and maternal age. Association with birth rank has been studied with less satisfactory results. Conflicting opinions have been expressed by investigators who did not have access to a satisfactory control group (Shuttleworth, 1909; Brousseau, 1928; Murphy, 1936; Bleyer, 1938; Beidleman, 1945), and the need first to eliminate the effect of maternal age has usually been overlooked. Examination of the separate effect of birth rank has rarely been attempted. Malzberg (1950) used the GreenwoodYule method to investigate the influence of birth rank at each maternal age group, but his results are misleading since many of the unaffected sibs were born outside the maternal age range under consideration. A more acceptable adaptation of the Greenwood-Yule method designed to remove the influence of maternal age was used by Penrose (1934), who found that the risk of mongolism in first-born was slightly raised although the result was not statistically significant. He recognized that the method may give false results if applied to incomplete sibships, and avoided this error by selecting cases some years after birth when most of the fraternities were probably complete. This manoeuvre may be equally misleading as a result of the high early mortality of mongols. In a previous communication (Record and Smith, 1955) it was noted that the usual cause of death is infectious disease. In view of the marked increase in infant mortality from infection with birth order (Gibson and McKeown, 1952), it would be expected that a series collected some years after birth would show a deficiency in the later birth ranks. Mortality is greater for female than for male mongols, and the increase in the proportion of males with increasing birth rank exhibited by Penrose's series is consistent with this interpretation. The demonstration by 0ster (1953) of an excess of mongols at the lower birth ranks, compared with normal infants born to mothers of the same age, may have a similar explanation since his observations were confined to mongols alive at the time of the inquiry. It appears therefore that the question whether birth order is of aetiological importance in mongolism has not been satisfactorily answered. In an attempt to explore this problem we have collected data on a substantial series of mongols most of which were ascertained soon after birth, and on a control group representative of the population of births from which the affected subjects were drawn.