Genetics and Heritability of Elbow Dysplasia: What Breeders Need to Know

Elbow dysplasia is heritable. That much has been established beyond debate for decades. But saying a condition is "genetic" tells breeders almost nothing actionable. The critical questions are: how heritable, through what mechanism, which genes are involved, and how can breeders use genetic information to make better decisions? The answers to these questions have evolved substantially over the past decade as genomic technologies have moved from research curiosity to practical breeding tools. Yet much of the breeding community still relies on a simplified understanding of ED genetics that was current in the 1990s. This article synthesizes the current evidence on elbow dysplasia heritability, the polygenic threshold model, genomic discoveries, and the practical application of estimated breeding values in selection programs.

Heritability: What the Numbers Mean

Heritability, expressed as h-squared, measures the proportion of phenotypic variation in a population attributable to genetic differences between individuals. For elbow dysplasia, published heritability estimates range from 0.10 to 0.45 depending on breed, population, and study methodology. This range has important implications for selection programs.

A heritability of 0.30, roughly the average across breeds and studies, means that 30% of the variation in ED status between individuals within a population is attributable to additive genetic effects. The remaining 70% reflects environmental factors, gene-environment interactions, non-additive genetic effects, and random developmental variation. This moderate heritability explains a frustrating reality: breeding two clear-screened parents still produces affected offspring at rates of 5-15%, depending on the genetic background of the pedigree.

The Polygenic Threshold Model

Elbow dysplasia does not follow simple Mendelian inheritance. There is no single "ED gene" that can be tested and eliminated. Instead, ED follows a polygenic threshold model: many genes, each with small individual effects, contribute to an underlying continuous liability. When the cumulative liability exceeds a threshold, the disease is expressed.

This model explains several otherwise puzzling observations about ED inheritance:

• Clear parents producing affected offspring: Both parents may carry subthreshold liability that combines in some offspring to exceed the disease threshold.
• Affected parents producing clear offspring: Some offspring inherit a favorable combination of alleles that keeps their liability below threshold, despite parental disease.
• Variable severity among affected littermates: Dogs above threshold can be mildly or severely affected depending on how far their liability exceeds it.
• Environmental modulation: Factors such as nutrition during growth, body weight, and exercise patterns shift the threshold rather than the genetic liability, explaining why environmental management reduces disease expression without changing genotype.

Component-Specific Genetics

Growing evidence suggests that the three classical elbow dysplasia components, fragmented coronoid process (FCP), osteochondritis dissecans (OCD), and ununited anconeal process (UAP), may have partially distinct genetic architectures. This has important implications for breeding programs that treat ED as a single condition.

The clinical relevance is this: selecting against overall ED grade may reduce all three components simultaneously, but the rate of progress will differ. UAP, with potentially larger-effect genes and higher heritability, may respond faster to selection pressure than FCP, where many small-effect genes make genetic progress slower. This partly explains why breed-specific prevalence data shows different component distributions across breeds, as selection pressures interact differently with each component's genetic architecture.

Genomic Tools: Where We Are in 2026

The promise of genomic testing for elbow dysplasia has been discussed for over a decade. Progress has been real but slower than initial optimism suggested. The current state of genomic tools for ED can be summarized honestly.

Genome-Wide Association Studies (GWAS)

Multiple GWAS have identified chromosomal regions associated with ED in various breeds. However, no study has identified a single locus explaining more than 15% of genetic variance. The typical finding is multiple loci, each explaining 2-5% of variance, consistent with the polygenic model. Importantly, many identified loci are breed-specific, meaning a QTL significant in Labrador Retrievers may not be relevant in Rottweilers.

SNP Panels and Genomic Prediction

Commercial SNP arrays (typically 100,000-200,000 markers) enable genomic estimated breeding values (gEBVs) that incorporate DNA information alongside pedigree and phenotype data. The Scandinavian kennel clubs and several breed-specific programs are piloting gEBV-based selection for ED.

Estimated Breeding Values: The Practical Tool

For most breeders, estimated breeding values (EBVs) represent the most immediately useful genetic tool available. EBVs estimate an individual's genetic merit for ED by incorporating information from the individual, its parents, siblings, offspring, and more distant relatives, weighted by their degree of relatedness.

How EBVs Work

A simplified example: Dog A is ED Grade 0 with three siblings, two Grade 0 and one Grade 2. Dog B is ED Grade 0 with three siblings, all Grade 0. Both dogs have identical phenotypes, but Dog B has a better EBV because its relatives provide additional evidence of favorable genetics. If Dog A's offspring data shows 18% affected versus Dog B's 6% affected, the EBV calculation adjusts further to reflect this progeny information.

EBV Availability by Country

Breeders in countries without formal EBV systems can approximate the approach by systematically collecting screening data from relatives and calculating affected rates across generations, as outlined in our guide to breeding decisions based on elbow scores. While less statistically rigorous than formal EBV calculation, this manual pedigree analysis still outperforms individual phenotype selection alone.

Gene-Environment Interaction

One of the most clinically significant aspects of ED genetics is the substantial gene-environment interaction. Dogs with moderate genetic liability may or may not develop clinical disease depending on environmental exposures during critical developmental windows. This interaction explains why identical twins (rare but documented in dogs) can have different ED outcomes when raised in different environments.

The major environmental modifiers with documented genetic interaction include:

• Growth rate: Genetically predisposed puppies raised on controlled-calorie diets show lower ED expression than fast-growing littermates. The genetic liability is identical; the environmental trigger differs.
• Body weight: Heavier dogs within a breed show higher ED prevalence, but this relationship is stronger in genetically predisposed individuals than in those with low genetic liability.
• Calcium intake: Excess calcium during growth exacerbates endochondral ossification defects in genetically susceptible puppies with minimal effect on resistant genotypes.
• Exercise during growth: High-impact exercise before growth plate closure may trigger ED expression in predisposed individuals, though the evidence is less definitive than for nutritional factors.

The practical implication for breeders is clear: genetic selection and environmental management are complementary strategies, not alternatives. A breeding program that selects rigorously for low ED genetics but provides no nutritional guidance to puppy buyers undermines its own genetic progress. Conversely, optimal environmental management without genetic selection merely masks underlying liability that will express in future generations raised under less controlled conditions. For breeds where both elbow and hip dysplasia are prevalent, coordinating genetic selection across both conditions prevents improving one trait at the expense of the other, a strategy well-documented in shepherd breed health programs.

The Future of ED Genetics

Several developments will likely transform ED genetic management within the next decade:

• Whole-genome sequencing: As costs continue falling, whole-genome data will replace SNP panels, capturing rare variants that current arrays miss.
• Multi-breed reference populations: Combining genomic data across breeds will identify shared ED risk loci and improve prediction accuracy for breeds with small screening populations.
• Epigenetic markers: Research into DNA methylation patterns associated with ED may reveal how environmental exposures during gestation and early life "program" gene expression, potentially enabling risk stratification at birth.
• Integrated selection indices: Future breeding programs will likely use composite indices that balance ED genetic merit against hip dysplasia risk, temperament, longevity, and breed-specific health priorities, preventing overemphasis on any single trait.

Conclusion

Elbow dysplasia genetics is complex, polygenic, and modulated by environmental factors, but it is not intractable. The tools available today, estimated breeding values, pedigree analysis, and emerging genomic panels, enable meaningfully faster genetic progress than individual phenotype selection alone. Breeders who invest in collecting comprehensive screening data from relatives, who access EBVs where available, and who combine genetic selection with evidence-based environmental management are making the most effective use of current knowledge. The perfect genomic test remains elusive, but waiting for perfection while ignoring available tools wastes generations of potential progress.