Further, nutrition is unique among other medical sciences in that, historically, its doctrine is not based primarily on randomized controlled trials (RCT), the gold standard of scientific research. RCTs are primarily done on people with a disease where one group is given the treatment, and the other group is not, and who got the treatment is unknown. Every treatment has a risk, but the risk is offset by the potential to improve disease consequences. But if you don’t have a disease, it’s unethical to put people at risk of treatment. Unfortunately, the early stages of obesity and metabolic disease are typically not considered diseases. Therefore, RCTs were generally not considered appropriate for nutrition research. Also, RCTs work best for diseases that present in a relatively short timeframe. Diseases that present over decades do not lend themselves well to study by RCT. Hence, RCTs are few and far between for the diseases of nutrition.

Therefore, almost all existing nutrition research has been based on population studies (what scientists call epidemiological studies), which observed the health effects, behaviors, and environments of large populations. Population studies have low reliability because it is almost impossible to control all effective variables in a large population of people.

A famous example is a series of population studies in the 1940s and 1950s that correlated higher meat consumption in countries with more heart disease. As a result, it was long considered a scientific fact that meat consumption caused heart disease. This was later disproved, although other reasons may remain to minimize or avoid meat. It turned out that the countries that ate more meat did so because they were wealthier, and they were wealthier because they were more industrialized. Being more industrialized meant these countries had more air pollution. Air pollution, not meat consumption, was later proven to be the primary cause of increased heart disease in these countries.

There are also N-of-1 studies and small-group studies that fall between population studies and RCTs in terms of reliability. These are controlled observation studies of one person (i.e., test population Number of 1 or N-of-1) or small groups of test subjects (what scientists call cohorts) who are observed and measured as variables are changed. Often, an N-of-1 study is the researcher making controlled treatments on themselves and observing and recording the results. Small group studies are typically very similar in approach, only having a few more people being tested. The main concern is statistical techniques that control for bias, anomalies, and errors, don’t work well with small numbers of test subjects.

Typically, N-of-1 and small cohort studies are not as good as RCTs but generally are better than population studies, especially if the researchers are alert to the risks of confirmation bias and the placebo effect. Where RCTs are unavailable, we use predictive modeling, N-of-1, and small group studies to connect the dots of the new science. Interestingly, obesity and metabolic disease are getting so widespread that RCTs are now becoming more common for nutrition but will likely always be limited due to the long presentation time of these diseases.

Further, we don't always know in strictly RCT-proven scientific terms why every facet of Actionable Knowledge and predictive modeling appears to have helped the people of Nutrimatters. We will remain vigilant as RCTs and other data become available to revise and update any new understandings that don't match our model. Finally, we very much embrace Dr. David Allison's mantra, "Don't mistake experience for science and data." Yes, we are sharing the experience of our journey of avoiding and reversing obesity and metabolic disease. But it's a journey that uses the latest science and data as its primary navigator. Nevertheless, we are not offering our experience as science and data, and it should not be taken as such. There is more discussion on this topic throughout the Courses and Articles.