Hmm, you're lucky, you see at my uni, if you don't cite any references they throw you out. Where did you go? Dreamland University?
you're one of those guys that doesn't trust his senses, I see. if you get ill, and remove factors, and find out which is doing what, then I would know more about the substance than any research, since individual results do not generalize. furthermore, corporate or government funded science is usually one of omissions of data, turning things to their own advantage and interest. even if there is no corruption going on, there is still a thing like 'publication bias', which means that only significant results usually get published.
Publication bias is a real effect, and you're right to consider it. However, the placebo effect, confirmation bias and correlations with random fluctuations in health are also real effects, and they are huge. They are the reason that people go to so much trouble and spend so much money on large randomized double-blind studies. With N=1, it's really easy to get fooled.