Table of Contents
Scientists make mistakes, too.
While nutrition science in general can’t be blamed for its piecemeal progress, individual nutrition scientists frequently commit avoidable errors that only increase our confusion once they are exposed. More often than you might think, poorly designed nutrition studies and poorly interpreted data yield false conclusions that must be corrected later. Common problems include small sample sizes, faulty data collection methods, lack of adequate placebo controls, and dismissal of unexpected results.
In some cases, studies are designed or interpreted badly with full awareness of the researchers, because they want to please the party (often a food industry corporation) funding the study. In other cases, researchers are so keen on seeing their pet hypothesis validated that, well, they make it right. An example of this latter scenario comes from a large, international study that sought a correlation between cholesterol levels and heart disease in 27 countries. According to the raw data there was only a weak correlation, but inexplicably, in their analysis of this data, the researchers leading the study simply threw out data from countries that defied their expectations and found a much stronger correlation in the remaining data. Years later the correlation between total blood cholesterol levels and heart disease was proven to be much weaker than we were once led to believe.