Here our nutritionist Ed van Tegelen explains some of the trials and tribulations of accurate nutritional research and what to bear in mind when see articles in the media!
There is uncertainty about the effects of essentially all foods and/or dietary patterns on health and weight. Much of what we know about diet derives from lesser forms of evidence rather than absolute proof, lots of thing we know about diet derive from observation. If we would only agree on using absolute proof we still would not know that for example smoking causes cancer.
The gold standard for research is the randomized controlled clinical trial (RCCT) in which 2 groups of people, one intervention and one control group, are compared at the end of the study. These results have to be adjusted for confounding factors like for example smoking, body weight, activity level.
Studies on the effect of a single food on health outcomes are very expensive and hard to do. As are studies that show diet A is better than diet B or C, D, E, F. Studies in which different diets are compared are routinely designed by those who favour one kind of diet over another, and the losing diet becomes the straw man. In these studies you really have to take a closer look what these people have been eating.
To design a trial with absolute proof it has to be a long term RCCT, because the health differences between the two groups are likely to be small, the sample size has to be in the thousands. We also know that nutritional effects, and taste preferences start in utero and that diet continues to influence health across the total lifespan. Thus a study should start in utero with thousands of pregnant women adhering to their assigned diet. Breastfeeding should be standardized for all participants, because nutrients are transmitted through breast milk. The true study participants, the babies, would have to adhere to their diet for their whole life and optimally life to over 100 years old if we are looking for optimal diets. This would need a new generation of scientists to continue the trial till the health outcomes of the participants can be evaluated. Thus, you can understand that such studies are pretty much impossible to perform.
That’s why we only have RCCT studies that run for shorter terms, and can only look at certain markers of health. The big issue with this is that lots of the chronic diseases people suffer from nowadays often take decades to develop into symptoms.
Another problem nowadays is the media. They love sensational headlines and like to show you a new diet every week. There are lots of quick fix celebrity diets, but being a celebrity doesn’t equal expertise. It’s true that food is a universal experience but genuine nutrition expertise is not. This leads to false equivalence in which opinion equals expertise.
If diet books say for example “forget everything you’ve ever heard about nutrition” that’s a red flag because that’s not the way nutrition science works. Most of the fad diet books rely on success stories (anecdotes) and not on science. When they have references in their books it’s worth fact-checking the references as these all to often don’t support the things they claim.