Yeah, but isn't the word "may" always used when the results of ONE study are presented? Because it's bad form to suggest that one study proves...anything. The only time they drop the word "may" (or should drop it) is when multiple studies are presented at once, pointing to the same conclusion. A "study of studies"... what's that word, I'm drawing a blank.
So when I read a title like the one in question, I take it to mean,
"High Soy Intake Decreased Breast Tumor Risk by 30 Percent In This Study"
Notice they said "decreased," as in 'it happened once,' and not "decreases" as in 'it happens all the time.'