Genetic studies in mental subnormality.

Date:
1973
Catalogue details

Licence: Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Credit: Genetic studies in mental subnormality. Source: Wellcome Collection.

    34/100 (page 20)
    20 GENETIC STUDIES IN MENTAL SUBNORMALITY Since sibships with two affected members yielded the most information were estimated for different values of The minimum value of was obtained when p = о • 66 but this was significant at the 5 per cent level = 6-08 0-05 > P > 0-02). This value was also the best estimate for 'j6' as obtained from the data. It is concluded that the significant difference between the observed and expected numbers of sibships with two affected members makes it unlikely that there was random association of the sexes in sibships even when males had a chance of о-66 and females of 0-34 of being affected. Multifactorial hypotheses involving quasi- continuity of distribution of traits, i.e. a threshold either biological or imposed, have been the subject of considerable attention over recent years (Penrose, 1953; Edwards, i960; Falconer, 1961 and 1965; and Morton, 1970). It is possible to make predictions on several models where either multifactorial liability is defined as including polygenic and environmental factors and is assumed to be normally distributed or the polygenic liability is presumed to be normally distributed and the environmental distribution is monotonie and increasing. How¬ ever, the complexities of making predictions based on the condition of a relative and allowing for sex have so far eluded the mathematicians, so that any examination of the present data using predictions based on a male or a female in the sibship being affected, is not feasible. Nevertheless, it seems legitimate in a rather cruder fashion, to test whether the cases are likely to be accounted for predominantly by multifactorial liability (in this case liability includes both environmental and genetic factors). In this situation, as females are the less frequently affected sex, it could be presumed that, on average, the parents of an affected female have a higher potential of transmitting additive deleterious genes than those of an affected male. It is not necessary to consider sibships of varying size separately as there is no fraction representing expectation with which observations can be compared. If it is supposed that the hypothesis is correct, we should expect that the presence of an affected female in a sibship should deter¬ mine a higher risk for male sibs than the presence of only a male index case. This could be tested on data where families were identified by a single propositus, but cannot be tested in this way in the present data. It is difficult to know whether evidence compatible or incompatible with such a hypo¬ thesis can be derived from the data available. The information which appears relevant is set out in Table VIII. Then, ignoring any other hypothesis, it might be expected that, irrespec¬ tive of the families from which they were derived, when any females in a sibship were affected, the overall frequency of affected of either sex would be greater than when only males were affected. Sibships which included both males and females must of necessity include at least two affected, so when setting out the sibships where only males or only females were affected, the sibships where only one was affected have been omitted. The most valid comparison appears to be between sibships with only males affected and those with only females affected plus those with sibs of both sexes affected. It will be noted that there appears to be no significant difference in the frequencies of affected sibs in these two groups of sibships. It may also be noted that in the mixed sex sibships of those at risk, a higher proportion of females than of males are affected. Without being dogmatic on the subject, the writer inclines to the opinion that there is no evidence for a multifactorial/threshold hypo¬ thesis from the data available.