The FDA never settles a product marketing claim based on only one study.
Here are some possible problems with this one that jump out at us:
1). It's a pity that this has to be considered, but it's true: When a scientific study is sponsored (paid for) by anybody who has a money interest in the result of the study, there's a strong possibility that the researchers could be influenced to tilt the study in the direction that their sponsor will like. This kind of study always gets less respect than a study with an independent source of funding.
2). "50 people" is actually a very small sample for a medical investigation. Five thousand would be more like it if you want results that are statistically significant.
3). "99%" is very, very suspicious. In a properly designed and controlled experiment, you probably could not get 99% agreement that the sky is blue. This kind of experimental result is very fishy, and should be enough to toss out the whole study.