NUTRIENTS DATA ANALYSIS
I analyzed the recipes in this book using a nutrient analysis program that I developed and extensively tested with commercial laboratory chemical testing. The program includes over 750 ingredients, and analyzes recipes for macronutrients, vitamins, minerals, fatty acids, and amino acids. Most of the data come from the USDA National Nutrient Database for Standard Reference, Releases 18–20. For the foods not listed in the USDA database (primarily chicken wings, necks, backs and turkey necks) I used results of the macronutrient and mineral tests I conducted for Steve’s Real Food and other pet food companies, supplier data, and published data from the BARF-world website.59
The USDA database does not provide sufficient data for chloride (Cl), iodine (I), and vitamins D and E. In all of the chemical tests on foods I formulated, lack of Cl was never a problem; nonetheless I include ample sources of Cl in some of the recipes. Likewise, I include sources of iodine and vitamins D and E. The USDA data on choline content are not complete; the values presented should be considered minimums.
The nutrient analyses are general guidelines only. The vitamin and mineral content of natural foods varies up to 30% based on variety, soil conditions, growth enhancers used, post-harvest treatments, and other factors. Nutrient bioavailability (how well the nutrient is used by the body) is not considered, just gross intake. Bioavailability varies according to many factors, including the food consumed with the nutrient. Some plant foods contain anti-nutritional factors, including oxalates, phytates, and tannates, that inhibit the absorption of some of the minerals.
How accurate is chemical testing?
I’ve had more than 100 ingredients or full recipes tested at five different testing laboratories in the USA. I have found mistakes, some significant, from every lab. I’ve concluded that laboratory chemical testing is not sufficient in analyzing diets, unless confirmed by computer analyses. Here are just two examples of why I write this.
In 2007, I sent mechanically skinned chicken necks to a nutrient-testing laboratory in California. When I got the results of the first test, I called the lab and told them that the results did not make sense, especially the 0.02% calcium (Ca) in chicken necks without skin. I asked for a retest. I sent them more skinned chicken necks, from the same batch. The second results came back meeting my calcium and fat expectations (column 2, Table B.1), but the phosphorus still did not make sense. Bone contains 2 to 3 parts Ca to 1 part phosphorus (P); the Ca:P ratio on chicken necks with meat should be less than 2:1. So I asked for a third test and, finally, this test came back with credible results.
Table B.1 Test results for mechanically skinned chicken necks
The next time I wanted to test products I went to a different laboratory, this one in the middle of the US. A client wanted me to fine-tune his formulas. His company has five varieties of food. All have the same meat to vegetable percentages, and used the same vegetables. Therefore, I expected the protein and fat content of the various diets to differ, based upon the leanness of the meats, but the carbohydrate and fiber content should be identical with all five diets. All the carbohydrate and fiber comes from the plants. Table B.2 shows the first results from the chemical testing by one of the most respected labs in the country. The laboratory results showed that food D has more than three times the carbohydrate content of food A and E. That is not possible.
Table B.2 Reported percentages of protein, fat, fiber, and carbohydrate
Needless to say, nutrient testing is an inexact science. We must make the best reasoned decisions we can with the data available, which is what I’ve tried to do in this book.