A really good example of this was the paper that kicked off the whole "omega-3 fatty acids for heart health" thing. It ultimately got retracted.
The gist of the paper was that they observed that Inuit communities have really low rates of heart disease, and hypothesized that it could be because their traditional diet is very high in omega-3 fatty acids. The problem is, they don't actually have low rates of heart disease. They just have low rates of heart disease diagnosis, because they also have limited access to health care.
The gist of the paper was that they observed that Inuit communities have really low rates of heart disease, and hypothesized that it could be because their traditional diet is very high in omega-3 fatty acids. The problem is, they don't actually have low rates of heart disease. They just have low rates of heart disease diagnosis, because they also have limited access to health care.