However I haven't read the original report yet.
These confunding factors could be managed either in the expermental design or in data analysis.
Before this trial (which includs two experiments), the French did an independant trial, and the result was largely the same.