With data collected from the 21,684 couples and their 174,000 children, researchers concluded that increased family size was associated with decreased survival for both parents, although significantly more for mothers. According to Smith, "Larger family size was also associated with lower offspring survival, especially for later-born children. In neither case did other factors such as economic status play a role in the survival rates." He also notes that, "Our results are consistent with the idea that reproduction requires a trade-off between quality and quantity, and may help explain the evolution of menopause as a means of increasing mother survival."
That is based on a new PNAS article by Dustin Penn and Ken Smith, "Differential fitness costs of reproduction between the sexes." The sample was Utah families between 1860 and 1895.
The paper discusses the data in terms of life history theory. While I'm all in favor of life history theory, in this case, it's not clear to me that a true life history tradeoff is really in operation. For example, they find that women who have fewer than 3 children had more than a 98 percent chance of surviving a year after their last child, while women who had 12 or more children had only a 94 percent chance of surviving that year. That is an important difference, sure, but women with 12 or more children have a higher Darwinian fitness by a long shot, despite the increase in mortality risk. There's no tradeoff.
The most tradeoff-looking observation is that survivorship to age 18 is significantly lower for offspring late in the birth order in large families. So offspring born 12th or later in their families have only a 75 percent chance of living to age 18, while the 7th to 11th offspring have over an 80 percent chance. Family size and birth order confound each other in the data (is the lower survivorship of large families due to the higher mortality of high birth-order children, or vice versa?), but the observations are in the direction of a tradeoff, with further reproduction after 11 offspring apparently providing diminishing returns. But the returns are not negative, so there is still no real tradeoff.
Apparently, in 19th century Utah, the optimum fitness strategy was to have as many children as possible, no matter what. This strategy was not cost-free in terms of parental or offspring mortality, as the paper observes, but there is no way that these numbers would lead to an adaptive constraint on human reproduction, or the evolution of menopause.
But then, 19th century Utah is not characteristic of human evolutionary history, so the question is whether the direction of relationship observed in the paper would apply at the higher adult mortality rates characteristic of ancient people. Casting a higher mortality rate across all age classes would increase the tendency toward a tradeoff, and might result in reduced fitness for the largest families due to higher mortality cost.
And there is this interesting snippet:
Mothers stressed from rearing chronically ill children that require enormous investment have short telomeres for thier age, suggesting oxidative stress and cellular senescence (Epel et al. 2004).
Lots of good pregnancy cost references, in the discussion.
Penn DJ, Smith KR. 2006. Differential fitness costs of reproduction between the sexes. Proc Nat Acad Sci USA (online early) DOI link