The Foundation has collected over 10 million data points on symptom severity and influencing factors from over 10,000 people. The Foundation develops and applies predictive machine learning algorithms to the data to reveal effectiveness and side-effects of treatments and the degree to which hidden dietary and environmental improve or exacerbate chronic illnesses
These analytical results have been used to freely publish 90,000 studies on the effects of various treatments and food ingredients on symptom severity.
Although 10 million data points sound like a lot, currently, the usefulness and accuracy of these Outcome Labels are currently limited. This is due to the fact there are only a few study participants have donated data for a particular food paired with a particular symptom. In observational research such as this, a very large number of participants are required to cancel out all the errors and coincidences that can influence the data for a single individual.
For instance, someone with depression may have started taking an antidepressant at the same time they started seeing a therapist. Then, if their depression improves, it’s impossible to know if the improvement was a result of the antidepressant, the therapist, both, or something else. These random factors are known as confounding variables. However, random confounding factors can cancel each other out when looking at large data sets. This is why it’s important to collect as much data as possible.