SAGE KE Bulletin Board

A Discussion of Early-Life Predictors with an emphasis on Statistical Methods

16 March 2004

Rainer Spiegel

I read the well-written contributions to early-life predictors of human longevity by Drs. Gavrilova, Gavrilov, Evdokushkina and Semyonova with great interest and I would like to reply to the authors' question for comments and suggestions: 1. I thought about including "gender" as a co-variate into the statistical analysis and testing whether the co-variate has a significant effect on the overall results. 2. The authors mention that previous research on early-life effects is often based on small data sets. Therefore I thought it might be a good idea to perform a meta-analysis over these small data sets. This way a pooling of the data for both genders in an attempt to increase statistical power could be avoided. A good overview on meta-analytic reviews can be found in Rosenthal (1995). 3. I wonder what effects diseases such as Hemophilia or Duchenne Muscular Dystrophy have on longevity of males and females in considered datasets, as these diseases are linked to the X-Chromosome. Reference: Rosenthal, R. (1995). Writing meta-analytic reviews. Psychological Bulletin, 118, 183-192.

Science of Aging Knowledge Environment. ISSN 1539-6150