Female fertility and longevity

Abstract

Does bearing children shorten a woman's life expectancy? Pleiotropic theories of aging predict that it should, and in particular, the Disposable Soma theory predicts unequivocally that this effect should be inescapable. But many demographic studies, historic and current, have found no such effect. In this context,the Caerphilly cohort study stands apart as the sole test that corroborates the theory. Why has this study found an effect that others fail to see? Their analysis is based on Poisson regression, a statistical technique that is accurate only if the underlying data are Poisson distributed.But the distribution of the number of children born to women in the Caerphilly data based departs strongly from Poisson at the high end. This makes the result overly sensitive to a handful of women with 15 children or more who lived before 1700. When these five women are removed from a database of more than 2,900, the Poisson regression no longer shows a significant result. Bilinear regression relating life span to fertility and date of birth results in a small positive coefficient for fertility, in agreement with the main trend of reported results.

Fig. 1

Age at death for women, averaged by the number of their children. There is no clear downward trend in the chart, even without correction for year of birth. The line indicates sample size, referred to the scale at the right

Fig. 2

Same as Fig. 1, limited to women born before 1700. It is in these women that Westendorp and Kirkwood claim to discern a downward trend at the right side of the chart. Note the small sample size, indicated in the scale at the right

Fig. 3

For Poisson regression to be a valid trend indicator, the underlying data must be Poisson distributed. The figure compares the distribution of women in the full sample by number of children with a Poisson distribution of the same mean