A note on mongolism in twins.

It has often been thought that if one member of a pair of twins is a mongol, the other member is always affected if the pair is monozygous, and rarely affected if the pair is dizygous. This view is based on the absence of any report of monozygosity in a discordant pair. However, little evidence is available, since few of the pairs that have been recorded may be accepted unreservedly as monozygous. In the absence of direct evidence the matter can be examined by considering (a) the frequency of concordance among affected pairs, and (b) the numbers of affected pairs that are of like and of unlike sex. The present report is based on 129 pairs in which sex, number affected, and diagnosis are believed to be reliably recorded. One hundred and seven were assembled by 0ster (1953) and sources of a further 22 pairs are given in the Appendix. Nineteen of the latter were reported in the literature; the remaining three pairs are now reported for the first time: (a) Table I shows the distribution according to sex of 152 affected in 129 pairs. Acceptable evidence of dizygosity is recorded for three of 23 pairs in which both twins are affected (MacKaye, 1936; Gordon and Roberts, 1938; Jervis, 1943). Even if it is assumed that all other concordant pairs are monozygous, the proportion (20/129 = 15 5 per cent.) is lower than estimates usually given for the proportion monozygous among representative twin pairs. Since the proportion of monozygous twins varies with maternal age, reaching its lowest value at the age quinquennium 35-39 years (Waterhouse, 1950), it is conceivable that this result may be due to the late maternal age at which many mongols are born. Maternal age is known for only 35 of the 129 pairs.