I just noticed an interesting post from John Hawks on correlations between height and longevity:
Findings based on millions of deaths suggest that shorter, smaller bodies have lower death rates and fewer diet-related chronic diseases, especially past middle age. Shorter people also appear to have longer average lifespans. The authors suggest that the differences in longevity between the sexes is due to their height differences because men average about 8.0% taller than women and have a 7.9% lower life expectancy at birth. Animal experiments also show that smaller animals within the same species generally live longer.
The authors also suggest that the added longevity due to caloric restriction in experimental animals may actually be a reflection of small body size, rather than of caloric restriction per se. Their preferred explanation is called the "entropy theory" of aging, which essentially is the argument that the bigger you are, the more things can go wrong.
The calorie restriction connection doesn't jump out at me as plausible now that we know more about the underlying biochemistry - it seems more likely to me that it's the other way around. A lower calorie intake over a lifetime leads to a smaller body size as a side-effect, while health and longevity benefits stem from metabolic and gene expression changes due to that lower calorie intake.
Looking to the future: as scientists aquire ever better tools and knowledge of genetics and cellular biochemistry - and the cost of using the tools and acquiring the knowledge falls - these sorts of speculations on statistical correlations will seem quaint. The time between noticing a correlation and understanding the precise biological mechanisms that cause it will become shorter than the time taken to write and publish a scientific paper on the correlation itself ... and that will be that. Suddenly it will be sheer laziness to talk about a correlation of biological traits without knowing why it exists - and medical science will be the better for it.