What are the implications of sampling to attain a given distribution in one variable? My guess is that the reasons are more psychological than statistical. The people conducting this study are MDs with very basic statistical expertise. I've observed several times that the normal distribution is the a priori "right distribution" for the statistically less adept.