I received an interesting question from my Facebook Page.
Thanks for providing an awesome guide in the Genotype Diet. Been practicing it as best I can for about 7 months and feel powerful. (I'm also doing good vitamins which is definitely part of the reason I feel really good.)
When I talk to people about the Genotype Diet and the benefits I've achieved from it, the main question I get is: Exactly what research has determined the genotypes, and the superfoods/toxic foods in the book? I can't begin to answer that question as I don't recall seeing it addressed in the book or on the website.
I'd love to learn more about the types of tests that you did to determine which people are in what genotypes as well as which foods are in which categories, per type.
This is a great question, but given that the best 'scientific answer' would be to show you the data tables and computer source code I can only try to explain a bit of the process. The problem with mass-market books is that you can only provide the upper-most level of information and a simplified version of that to boot, so I understand.
What I term the 'genotypes' (really 'epigenotypes' or 'morphotypes' but try to get a publisher to agree to use these words) are semi-synthetic constructs involving a stepwise statistical analysis of variation. They stem from the phenotypic (real world) characterizations reported for the ABO groups, Rh, secretor and additional biometric markers (D2-D4, fingerprints, etc). The idea was to look for pleiotropic (sympathetic) relationships between the multi-dimensional genotype/ phenotype data, especially if they are known to exert their effects through transgenerational actions. Using multivariate analysis we then look to see how the data separates or groups together. Since, with the exception of secretor, taster, Rh and ABO, we're looking at phenotype, I felt very comfortable including data from other, traditional typing systems (Ayurveda, TCM) which were also based of physical traits.
The base data includes virtually all published scientific tabular data on variations in physiology and pathology associated with these parameters, in addition to our own profiles of roughly 3,000+ additional people. At that point the data was filtered according to degrees of three basic metabolic 'biases': 'thriftiness' (metabolic compromise), 'receptorism' (immune tolerance) and 'reactance' (auto-immunity).
The genotypes are not 'perfect' typologies (every Explorer does not look or act exactly as every other Explorer) because we cannot possibly encapsulate all variation in everybody. Two families using the same set of blueprints will most likely build two different houses, due to differing financial constraints, choice of land plot, etc. Most of the time and given the tools we might encapsulate 30-50 percent of the data variation (principal components) in any one person and what we encapsulate in one might be slightly different than what we get for another. In statistical terms this is called 'multiple inclusion criteria' and it is a keynote of factor analysis or 'fuzzy logic.'
What results are six basic 'types' that with considerable tweaking encapsulate an acceptable amount of variation. Crunching the system into six types and cramming them into a hard-coded 'book' is much less effective than dynamically generating one-to-one diets in software, but it is still a pretty good approximation of some basic phenotypic variation and is more helpful than not.
Once we get here, the next step was to match the expected physical manifestations to a large database of foods that I've been collecting for the last two decades. For each food, this database contains about 300 individual values (gluten content, vitamin A, known allergen, etc.) At this point a second set of algorithms takes over and each food is evaluated constituent-wise based on a weighed value system much like a lawyer might argue a case in court. For example, evidence of developmental instability or constrained growth (differences between left/right sides of body, certain fingerprints, short leg length) might result in limiting foods that cause excess glycation.
If no negative attributes (for example, if the food contains a lectin or is known to encourage bacteria overgrowth, etc) is recorded, then the next step is to see if a case can be built for the food having any specialized benefit (for example, sardines might become a superfood if increasing the amount of RNA nucleotides is desirable; artichokes because they encourage probiotic growth in a strain of bacteria known to be good for a certain blood type). Lacking either of these elements, the food is simply labeled 'food' and considered more or less neutral.
In the simple case of rice versus rice milk it is most likely additional gums in the milk that are the issue. Certain gums amplify the effects of problematic proteins in other foods.
People also ask a lot about peanut oil versus peanuts or cherries versus cherry juice. Usually it is a difference between one form that contains some sort of problematic protein versus the other that doesn't. Also, occasionally in the Genotype diet (unlike the BTD) with complex foods, sometimes one nutrient influences the value of another which alters the value of the food.
Here are blogs of mine tagged as 'genotype diet.' You will see some elements of the process discussed in detail in many of these entries.