That sounds very interesting, so you have a source for that? I could not find anything on Google searching for "life expectancy, gender and size".
Personally, I see this is less as a real research discovery and more like a cautionary tale on the pitfalls of statistics and the need for multivariate regressions with control variables.