Hirado: temporal trends in inbreeding and fertility.

Until recently, it has been commonplace both to deplore the lack of realism in most, if not all, genetic models used in the study of human populations and to acknowledge that few data existed from which improvements might stem. Interest and energy have, however, begun to focus upon the collection of data which hopefully will serve as bases for the formulation of more insightful and presumably more realistic models. Some investigators have, for example, turned to routinely collected vital and demographic data on the supposition that these observations, if pertinent, would be especially useful because of their numbers, breadth, and depth in time. Thus far, for a variety of reasons, the results have not lived up to expectations. Other investigators have tended to emphasize ad hoc censuses. They argue that demographic units are unfortunately more apt to be defined by geographic, political, or socioeconomic considerations than biologic ones, and though this bias can be offset to some degree in specific instances by judicious selection of the units to be studied, an ad hoc census of some variety may still be necessary if the routine data are to be fully plumbed. The Hirado Health Survey, to be described shortly, was an effort to combine an ad hoc census with routinely collected vital and demographic observations; it was directed, broadly stated, toward the ascertainment of (1) the fertility of the consanguineous marriage, (2) the reproductive performance of the inbred individual, and (3) certain observations on the children born to parents, especially mothers, who are themselves the products of consanguineous marriages. The study itself represents a conscious attempt to incorporate into traditional demographic and census procedures a few readily measurable genetic parameters. Inbreeding is, in Japan, such a parameter, and its relationship to fertility and fecundity may be one of the main avenues through which we can discern and measure the contribution of genetic factors to reproduction. Numerous considerations, administrative and scientific, entered into the choice of Hirado and the design and implementation of the survey. To mount the kind of demographic study implicit in the objectives just enumerated requires a population of a size sufficient to demonstrate differences in fertility among consanguinity classes, if differences exist, and yet not so large that the population would be inordinately difficult to study. Clearly, the population must be inbreeding at a rate such that a census of persons and households would reveal a substantial number of inbred individuals and consanguineous marriages. Ideally, the population would be sharply delimited geographically, and as static as possible. 1\oreover, it would seem advantageous in view of the effect of socioeconomic factors upon fertility to select a population as homogeneous in this regard as possible. Finally, if the census observations were to supplement routinely collected data, as proposed, the population to be censused had to correspond to some Japanese administrative unit upon which vital and demo-