As a trained epidemiologist, 47 subjects (or n=47 in epidemiology-speak) is far too low to draw any conclusions or even to speculate on any purported pathophysiology. Once again, you have morons in the media drawing conclusions based on a small research study. These types of studies are notoriously bad for drawing inane conclusions.
Dr. Dan's Laws of Biomedical Research:
1) The plural of "anecdote" is NOT "evidence"
2) Correlation or association is NOT causality
3) Small "n's" mean big disasters if taken seriously
4) Study populations are not generalizable to the entire population
5) Reporters don't understand studies--they are scientific morons
6) Every hysterical junk science study gets more publicity
7) Only believe a study that can be replicated