The results of the present prospective observational epidemiological
study do not prove that chocolate consumption causes a
lower diabetes risk. It is possible, for instance, that the inverse association
we observed between chocolate consumption and diabetes
risk is due to a genetic factor which induces both high
chocolate consumption and low diabetes risk. Rigorous randomized
placebo-controlled trials are needed to establish causality.
However there is evidence from prior studies which suggests that
the flavanols in chocolate could be responsible for the significant
inverse association between moderate levels of long-term chocolate
intake and diabetes risk observed in the present analysis.