Suggestions on based on PubMed studies. I try to keep to the terms used in the study as much as possible.

When you go to the suggestions page, you may see for diet something like:

First, note that there ate SIXTEEN shifts that we are trying to correct. None of the diets impact more than 5. A lot of the suggestions appear to disagree with each other. The reason is simple – one impacts 3 of the 16 bacteria, another impacts 5 of the 16 bacteria. We do not have sufficient studies published to know more.

For the above, a little common sense and reasoning helps:

• A low animal protein diet would appear to combining the first two items (plus hits the 5th item dead on) and actually agrees with the last one NOT a high animal protein diet.
• Similarly, increase dietary fiber, decrease low fiber diet and decrease high processed food diet also work together nicely.

We do not know exactly what is a “low animal protein diet” – does it include eggs and milk product only, does it include fish? pork? beef? Are the sources organic or commercial? To answer that question, you need to read the actual studies, and in some cases, contact the authors.

How are suggestions computed?

Suppose you have 4 shifts (A and B is too high, Y and Z is too low). And we have a study on apples. This study found that it increases A, decreases B, Z (nothing about Y)

We would report it as Increase Apples: 2, Decrease Apples: 1. Why, because it improves 2 items and makes 1 item worst.

We find a study for peaches, but it only mentions A decreases. We would report it as Increase Peaches: 1, Decrease Peaches:0Because B,Y,Z are not mentioned — we really have no idea if it impacts them.

We discover another study on apples, and it mentions that Y and Z decreases. This study by itself, would be Increase Apples: 2

When we go to the net effect, we end up adding each study together:

• Increase Apples: 2, Decrease Apples: 1
• Increase Apples: 2

The net result is: Increase Apples: 4, Decrease Apples:1. (I do not actually add the numbers, I use a fuzzy logic aggregation). The value for increase apples is higher — because we have greater confidence. Studies often disagree, and this is how we handle the disagreement.

We are working with studies that are often vague and may be silent about things that were either not tested for, or had no impact. We are in the world of fuzzy logic.