Tests for a Selection-Maintained Component

Selection, of course, could be acting in a different manner and could involve almost any stage in the life cycle. With virtually complete family histories of procreation and mortality available for whole populations in machine-accessible form, such tests for selective differentials may be applied in various ways with a degree of precision far beyond what would ever be possible for studies of more conventional design. 

Similar tests for selection in relation to the multifactorial conditions are likewise possible on a large scale, although the results are more difficult to interpret. In the British Columbia study, for example, a higher than expected reproductive rate was found following deaths of infants from postnatal asphyxia and atelectasis (i.e., failure of the lungs to ex- 

To interpret such correlations, additional information may be needed. In the case of postnatal asphyxia, for example, further data are required to distinguish reproductive compensation following the deaths from a simple tendency for the disease to occur in population subgroups that are characteristically more fertile than average. 

Despite such complexities of interpretation, better insight into the role of selection in maintaining the prevalence of multifactorial diseases is sorely needed in the present context to indicate those for which mutation is unlikely to be a major cause. The ultimate in refinement will only be reached with an abundance of pedigree information, sufficient to permit comparisons within data subsets homogeneous for a multiplicity of social variables. To obtain this from the linked vital and health records is not too difficult, whereas ad hoc studies of conventional design will probably always be inadequate for the purpose. 

---