"Leonid Gavrilov"  wrote in message
>
> The purpose of this study is to test the prediction of the
> evolutionary theory of aging that human longevity comes with the cost
> of impaired reproductive success (higher infertility rates). Our
> validation study is based on the analysis of particularly reliable
> genealogical records for European aristocratic families using a
> logistic regression model with childlessness as a dependent (outcome)
> variable, and woman's life span, year of birth, age at marriage,
> husband's age at marriage, and husband's life span as independent
> (predictor) variables. We found that the woman's exceptional longevity
> did not increase her chances of being infertile. It appears that the
> previous reports by other authors of high infertility among long-lived
> women (up to 50% infertility) are related to incomplete data, that is,
> births of children not reported. Thus, the concept of the high cost of
> infertility for human longevity is not supported by the data when
> these data are carefully cross-checked, cleaned, and reanalyzed.
>
Here's a good opportunity for a peeve about statistics.

Now that everyone has a spreadsheet on their desktop, you can play endlessly
with statistical tests until you come up with a result that you like. The
result will of course be significant in the narrow technical sense.

However in biology it is very difficult to eliminate confounding factors.
For instance some aristocratic women are infertile because of botched
abortions. Women who go for illegal abortions are not a random cross-section
of the sample, but probably less religious, less controlled by their
familes, more promiscuous, more likely to drink heavily and use drugs, and
so on. It is quite impossible to determine whether this actually makes any
difference to the results or not, because illegal abortions are
under-reported, and the sociological data is likely to be inaccurate.
This is of course only one potential confounding factor.

The point I'm making is that biology is not like physics, where you are
looking for tiny deviations from predicable measurements. If there is an
effect worth investigating it really should be so obvious that only the
crudest statistical test is required to see that women without children live
longer. This is not to say that a tiny effect can't exist (and in fact it
would be surprising if childbearing had absolutely no effect on longevity),
but it is very hard to demonstrate, and it is not likely to lead to anything
interesting.
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 7/16/04 9:37:50 PM