Parental Age Effects on Man 27 



therefore to be preferred as an indication for the distribution 

 of births in this population. Since on an average only 29 years 

 separate the deaths occurring during early infancy from this 

 "standard distribution" it is not conceivable that differences 

 in external conditions could affect the result. The conclusion 

 would therefore seem to be that a maternal age of over 35 

 years increases the mortality of the progeny during the first 

 two years of life. The optimurn age with the least deaths is 

 that of 25-29 years of age. Beyond early infancy, namely 

 between the ages of 2 and 4 years, the effect of advanced 

 maternal age is already quite small or non-existent, and it 

 cannot be demonstrated in older groups, either because there 

 is no influence or because of methodological difficulties 

 arising out of incompatibility of life-expectancy data col- 

 lected from an over-long (at least 100 years) time period. 



No mention has so far been made of the possible role of 

 paternal age effects. Since maternal and paternal ages tend 

 to be correlated the question is rather intricate. If the 

 material is divided into paternal age groups and a dependence 

 of life expectancy or infant mortality on paternal age is 

 recorded, this may simply reflect the effect of maternal age. 

 If on the other hand no difference is noted, then the paternal 

 effect possibly acts in the opposite sense to the maternal age, 

 high paternal age being favourable for the progeny. Curiously 

 enough, in the total material, in spite of the bias shown to 

 arise from the different distribution of young and old mothers 

 in the material collected before and after 1830, no paternal 

 age effect upon the total expectation of life could be demon- 

 strated. The most probable explanation would be that 

 maternal and paternal ages may after all not be strictly 

 correlated. The rather large mean difference found between 

 the age of spouses (7-10 years) obviously allows fairly wide 

 discordant variations in the material considered here. The 

 same objections can, however, be made against calculation of 

 the total expectation of life in the paternal age series as in the 

 maternal series. Therefore the same procedure for elimination 



