Inbreeding effects in northeastern Brazil

Abstract

The present study deals with the analysis of the inbreeding effects on a migrant Brazilian population and comparison with other studies. This population (comprising 1068 families with more than 9000 pregnancies) was found to have a good homogeneity between the inbred and control groups with respect to several socio-economic and biological variables. Inbreeding did not have a measurable effect on body weight, height, or hematocrit. These results are compatible with a small dominant component in the inheritance of the traits in question. However, diastolic blood pressure showed a significant increase with inbreeding. Diastolic blood pressure increased by 3.6 mm Hg per 10% of inbreeding. Although this association could well be spurious, the need for further investigation clearly emerges, not only in connection with the present results, but also with the results of Nance and collaborators of an apparent association of diastolic blood pressure and the ABO blood group system. Turning to mortality, there was a complete lack of inbreeding response of both stillbirths and postnatal deaths; however, abortions plus miscarriages showed a significant inbreeding effect (of both the mother and fetus). Our interpretation is that two opposing forces are being measured. On one hand recessive lethal genes tend to increase the risk of inbred fetuses and on the other hand genes involved in incompatibility mechanisms lead to a smaller death risk. This hypothesis is based on the observed interaction effects of inbreeding of the mother and inbreeding of the zygote, with parity. In an attempt to study the inbreeding effect on early and undetected abortions, we used the fraction of the cohabitation time during which the mother was not known to be pregnant or in postpartum infertility as a parameter that would measure a sterile period due to embryonic loss if all the other concomitant variables are randomly distributed with respect to F. "Waiting time" for a mother to become pregnant was distributed independently of F, thus providing no evidence for a large number of lethal equivalents acting at this period of life. The best approach available today to resolve segregation and mutation loads is the use of the double binomial model of segregation analysis. This was applied for the malformation and early fetal death data in the present material and showed that sporadic cases are not associated with inbreeding, which is inconsistent with the "phenodeviant" hypothesis. Also the present data did not show any heterosis in interracial crosses, as claimed by some proponents of phenodeviants. Comparison of estimates from different populations suggests that non-genetic mechanisms are confounded with effects of inbreeding and interracial crosses in some (perhaps many) studies. The B/A ratio of inbreeding effect to panmictic load has little power to isolate the mutational load unless supplemented by both reliability tests and segregation analysis. When the latter method reveals that inbreeding effects are limited to high-risk families, as in our material, this provides strong evidence for a mutational load.