It isn't a good point. This was explicitly controlled for in the study. People are just making wild guesses about methodological limitations that don't exist.