Short answer: Soy protein isolate (SPI) is the highest-purity soy protein concentrate at ≥90% protein on a dry-matter basis per AAFCO 2024 ingredient definition. Per Hill 1996 (J Nutr), canine ileal digestibility is 96–99% — among the highest of any plant protein in pet food. Per Mueller 2016 (Vet Med Int) systematic review of 297 dogs with adverse food reactions, soy ranks 5th among canine food allergens behind beef, dairy, chicken, and wheat. Per Setchell 2002 (Am J Clin Nutr) and Cerundolo 2004 (Vet Dermatol), phytoestrogen exposure at typical inclusion is below clinical-effect thresholds. The KibbleIQ rubric does not penalize SPI inclusion at population level but flags it as a watch-ingredient for individual dogs with confirmed soy allergy or for novel-protein elimination diets.

What soy protein isolate is — extraction and AAFCO definition

Soy protein isolate is produced through a multi-step refining process. Soybeans are dehulled and the oil is extracted (yielding soybean oil and defatted soy flakes). The defatted flakes are then washed in alkaline water (typically pH 8–10), which solubilizes the protein while leaving most fiber and oligosaccharides behind. The protein-rich extract is acidified to the isoelectric point (pH ~4.5), where the protein precipitates out and can be separated, neutralized, and spray-dried. The result is a near-pure protein powder — per AAFCO 2024 Official Publication ingredient definition 84.7, soy protein isolate must contain ≥90% protein on a moisture-free basis.

Three commercial soy protein products appear on pet food labels and they differ in concentration. Soybean meal at 44–48% protein dry-matter is the defatted soy residue after oil extraction (the cheapest form, the form most common in pet food). Soy protein concentrate at 65–72% protein is alcohol-extracted defatted flakes (intermediate cost, removes most oligosaccharides but retains some fiber). Soy protein isolate at ≥90% protein is the alkaline-extracted, isoelectric-precipitated form (highest cost per kg, lowest carbohydrate/fiber load). See our soy in dog food explainer for the broader context on soy as a pet food ingredient.

Digestibility — the Hill 1996 reference

Per Hill 1996 (Journal of Nutrition) canine ileal digestibility study, soy protein isolate showed 96–99% true ileal protein digestibility in dogs — among the highest of any plant protein measured. The same study reported soybean meal at 88–92% and soy protein concentrate at 92–96%. For comparison, Hill 1996 reported chicken meal at 87–91% and lamb meal at 71–85% — SPI digestibility is comparable to or higher than typical animal-protein meals.

The functional implication: a 10% inclusion of SPI delivers approximately 9.0–9.9 g of digestible protein per 100 g of food, with relatively little carbohydrate or fiber load. This is why SPI is used in formulations targeting high protein density at moderate cost — particularly veterinary therapeutic diets, performance diets, and weight-management formulations where calorie density is constrained but protein targets must be met. Per NRC 2006 canine protein requirements, adult-maintenance minimum is 18% crude protein on a dry-matter basis; performance and growth targets run 25–32%.

Allergy ranking — the Mueller 2016 systematic review

Per Mueller 2016 (Vet Med Int) systematic review of 297 dogs with confirmed adverse food reactions across published case series, the cumulative ranking of canine food allergens was: beef (34%), dairy (17%), chicken (15%), wheat (13%), and soy (6%). Soy is the 5th-most-common canine food allergen by this measure — clinically meaningful for dogs with confirmed soy allergy on elimination diet trial, but not a population-level reason to avoid soy in pet food. For context, per Olivry 2015 (BMC Vet Res) ICADA practice guidelines, an 8-week elimination diet trial with strict novel-protein selection is the diagnostic standard — not ingredient avoidance based on label scanning.

The relevant operational rule for KibbleIQ scoring: a dog with confirmed soy allergy needs a soy-free diet, but the population of dogs with soy allergy is small enough that SPI inclusion does not warrant a population-level rubric penalty. The same logic applies to other moderate-prevalence allergens: per ICADA 2015, pet-food formulators avoid stacked common-allergen exposure in elimination diets but do not exclude soy from maintenance formulations on allergy grounds alone. See our best dog food for allergies guide and hydrolyzed protein explainer for the elimination-diet pathway.

Phytoestrogens and the Brown 2008 DCM context

Soybeans contain isoflavones (genistein, daidzein, glycitein) at approximately 1–3 mg per gram of soy protein. These isoflavones have weak estrogen-receptor binding affinity (~0.001 of estradiol per Setchell 2002 review). Per Setchell 2002 (Am J Clin Nutr) pharmacokinetic review, the canine clinical-effect threshold for isoflavone exposure is approximately 4–8 mg per kg body weight per day. A 25 kg dog eating 300 g/day of a kibble containing 10% SPI ingests approximately 0.3–1.0 mg isoflavone per kg body weight per day — well below the clinical-effect threshold.

Per Cerundolo 2004 (Vet Dermatol) controlled canine study, daily soy isoflavone supplementation at typical pet-food exposure did not produce measurable estrogen-receptor or thyroid-axis effects in dogs over a multi-week feeding trial. The pet-food population-level concern is therefore low. Per Brown 2008 (J Nutr), soy and other legume-class ingredients have been examined as one component of the FDA-CVM 2018–2022 DCM Updates investigating diet-associated dilated cardiomyopathy — but per ACVIM 2022 nutritional cardiomyopathy consensus, the proposed mechanism centers on legume substitution for animal protein and taurine bioavailability, not phytoestrogen activity. See our pea protein explainer for the full DCM-investigation context.

How KibbleIQ scores soy protein isolate

The KibbleIQ Dry Kibble Rubric does not impose a population-level penalty on soy protein isolate. SPI is a high-digestibility, AAFCO-defined plant protein and is widely used in veterinary therapeutic, performance, and weight-management formulations without nutritional concerns at typical inclusion. The rubric flags SPI as a watch-ingredient in three specific scenarios. First, in elimination-diet protocols (when an owner is on an active 8-week novel-protein trial per ICADA 2015), the analyzer surfaces SPI for inclusion review against the trial protocol. Second, in low-cost formulations where SPI is the primary protein source above any animal protein, the rubric notes the imbalance toward plant protein and flags amino-acid completeness questions. Third, for confirmed soy-allergic dogs (operator-flagged), SPI inclusion is a hard exclusion.

For dogs without soy allergy on a maintenance diet, an SPI-containing formulation that meets AAFCO Dog Food Nutrient Profiles for the appropriate life stage is rubric-acceptable. The transparency-tier penalty is small: SPI is named, AAFCO-defined, and traceable on the label, unlike species-anonymous animal-by-product meal. See animal by-product meal explainer for the contrast and best dog food for sensitive stomachs for diet-pathway context. To check what your dog’s food contains, paste the ingredient list into the KibbleIQ analyzer.