Two things to note that totally invalidate the contention of this article: (a) the sample size is meaningless. Any result whatsoever coming from a pairing of tiny sample & control groups has zero statistical validity, and (b) there’s no confirmation of the results from any more adequate study performed in the 30 years since publication of the original article. In other words, like more than 50% of studies, the results are worthless garbage. Only someone with no comprehension of experimental design and no grasp of even basic statistics would draw any conclusion whatsoever from this totally meaningless paper.