Detailed FDA Vet-LIRN analysis of recovery diet patterns and dog breed distribution in DCM cases. Extracted via Tavily from the 2020 KSU Science Forum slides — provides much finer-grained detail than the consumer-facing Outbreaks page.
Group One: Fully and Partially Recovered Dogs (n = 107)
23 dogs fully recovered + 84 partially recovered = 107 total.
Definitions used by FDA:
- Grain-free: food without wheat, corn, ancient grains
(spelt, quinoa), barley, rice, rice bran, or soy (soy is a legume but grouped with grains for analysis).
- "Peas top": peas in the top 3-4 ingredients after
animal protein/meats (same for lentils).
- Pea fractions: pea protein / pea fiber / pea starch /
pea flour (same for lentils).
Pre-diagnosis diet (n = 23 fully recovered)
| Characteristic | % cases |
|---|---|
| Ate grain-free | 96% |
| Ate whole pulse ingredients | 96% |
| Ate product with by-products | 0% |
| Ate lamb meal + brown rice (1 case) | 4.5% |
Recovery phase diet
| Characteristic | % cases |
|---|---|
| Ate grain-free | 4.5% (1 of 22) |
| Ate whole pulse ingredients | 5% (1 of 21) |
| Ate product with by-products | 38% |
| Ate hydrolyzed protein diet | 3 cases (1 brewer's rice, 2 corn starch) |
Key insight: When diet changed to grain-containing + by-product-based + non-pulse, recovery occurred. 96% of the fully recovered group switched from grain-free to grain-containing in the recovery phase.
Signalment (n = 107)
| Group | Mean Age (yrs) | Mean Weight (lbs) | Sex |
|---|---|---|---|
| Fully Recovered (n=23) | 6.4 | 62 | 52.4% F / 47.6% M |
| Partially Recovered (n=84) | 6.4 | 62 | 52.4% F / 47.6% M |
Breed distribution (Group One, n = 107 fully + partial)
Most common breeds in DCM case group:
- Golden Retriever (ties most common)
- Mixed-breed dogs (ties most common)
- Labrador Retriever
- Doberman Pinscher
- Basset Hound
- French Bulldog
- Pomeranian
- Other large + small breeds (diverse)
Note: Many breeds listed are in the AKC top 10 breeds of 2019 (marked with * in original slides), suggesting popular breeds are over-represented simply due to breed popularity, not genetic predisposition alone.
LV Hypokinesis Cases (n = 30) — separate analysis
Dogs with LV hypokinesis (not full DCM): 30 total (20 fully recovered + 10 partial).
Pre-diagnosis diet (n = 30)
| Characteristic | % products |
|---|---|
| Grain-free | 73% |
| Had peas, whole peas, or peas/lentils in top | 80% |
| Had by-products | 0% |
Note: 73% is lower than the Outbreaks page 91% — this is a sub-sample (LV hypokinesis subset, not all DCM cases) and uses different window.
Recovery diet (n = 30)
| Characteristic | % products |
|---|---|
| Grain-containing | 87% |
| Peas or pea fractions | 7% (none whole) |
| Sweet potato/potato | 7% |
| By-products | 53% |
| Chicken/chicken meal | common |
LV Hypokinesis (n = 10) — case-based
| Diet type | Before diagnosis | Recovery |
|---|---|---|
| Grain-free | 8/10 (80%) | 1/9 (11%) |
| Grain-containing | 2/10 (20%) | 8/9 (89%) |
| Whole pulses | 10/10 (100%) | 0/9 (0%) |
| Meat/poultry by-products | 0/10 (0%) | 7/9 (78%) |
Key: 100% of LV Hypokinesis cases had whole pulses before diagnosis; 0% had whole pulses during recovery. This is the strongest dietary signal in the entire dataset.
Product basis vs case basis
FDA analyzed on both bases because some dogs ate > 1 diet:
- Product basis: characteristics of each unique product
(e.g., 91% grain-free across all products in Outbreaks page).
- Case basis: characteristics of the dog's reported
diet(s) (e.g., 96% grain-free across cases in recovery group).
Numbers differ slightly between bases — FDA reports both.
Sufficient data cutoff
Sufficient data was available for >92% of products. Products with missing label information were excluded proportionally.
Source
Extracted via Tavily on 2026-07-20 from ksvdl.org/resources/documents/dcm-forum/FDA_KSU-Science-Forum-slides_09-29-2020.pdf (2020-09-29 forum hosted by Kansas State University; FDA CVM presented). Data on Group One n=107 recovery cases.
Related topics:
FDA_DCM_investigation_timeline_2014_2024(overview timeline)