Breeding Strategies – a range of options: my April “Best of Health” Our Dogs article

Best of Health

Last month I discussed part of the recent paper by Farrell et al published in the KC-sponsored Canine Genetics and Epidemiology Journal [The challenges of pedigree dog health: approaches to combating inherited disease  Lindsay L Farrell, Jeffrey J Schoenebeck, Pamela Wiener, Dylan N Clements and Kim M Summers]. It can be found online at:

I said it was one of the best papers I’ve read for a long time and, taken in conjunction with the excellent analyses being reported by Dan O’Neill and his colleagues from the VetCompass project, could be a really important contribution to the debate on pedigree dog health. Of course, what’s more  important than “debate” about the subject is what action can be taken to improve things, when an issue is identified and supported by data.

Increasingly, I expect we will see Breed Club communities taking advice from the Health Team at the KC and other specialists on what exactly their strategy should be to address their health problems.  Many would claim this is already happening, but I think the majority of examples are actually just the adoption of “formal” DNA or other Screening Tests, rather than Breeding Strategies.

Clearly, there is a range of possible Breeding Strategies and the “best” choice will be dependent on factors such as:

  • how many of the breed are registered each year (in the UK and worldwide)
  • the breed’s Effective Population Size (EPS) and degree of inbreeding/genetic diversity – the former is information that has been produced by Dr. Tom Lewis who now works in the KC’s Health Team; the latter is already in the public domain via the excellent KC Mate Select tool
  • how many “health issues” exist and their impact of a dog’s quality of life – Asher et al’s 2009 Generic Illness Severity Index for Dogs (GISID) is one useful tool to assess the relative impacts of different conditions
  • whether or not there is a known genetic cause of the condition(s) and if there is, whether a DNA or other screening test is available – this may be related to whether such diseases are “simple” (e.g. due to recessive mutations) or “complex” (e.g. due to multiple genetic and environmental factors)
  • prevalence data on each condition – ideally this will be known from Health Surveys (in 2010 Collins et al proposed calculating a “Welfare Impact” score by multiplying Prevalence with a GISID ratio)
  • whether the health issue is related to the conformation of the breed (Farrell uses the term “morphology”) – so, for example issues may be known to be related to Brachycephalic breeds (Pugs, Bulldogs etc.) or Chondrodystrophic breeds (Dachshunds, Bassets etc.)

Breeding Strategies in practice

The Farrell paper discusses several examples of what they describe as breeding strategies that have been implemented with varying degrees of success.  I’ve drawn out four examples of different strategies described in the paper, to share here:

Example 1: Dutch Kooikerhondje (FCI) Patellar Luxation

In 1994, a patellar luxation screening scheme for Kooiker dogs based on orthopaedic examination was established in the Netherlands. From 1994 to 2009, the use of the orthopaedic screening results in breeding strategies decreased the prevalence of patellar luxation in Kooiker dogs from 28% to 19%.  This is still a higher level than in other susceptible breeds.  Farrell suggests that combining current screening schemes with pedigree and genotyping information could prove helpful in selective breeding programmes to reduce further the prevalence of disorders with complex inheritance such as patellar luxation.

Example 2: Myxomatous Mitral Valve Disease in Cavalier King Charles Spaniels

In 2001 breeding guidelines aimed at reducing the prevalence of MMVD in CKCS were introduced in Sweden. The voluntary protocol advised that individual dogs should not be bred from until they had been shown to be unaffected at 4 years old, or unless both parents could be shown to be unaffected at 4 in which cases the individual could be bred at 2. In the first two years, the guidelines had no impact on reducing the prevalence or eliminating MMVD disease in CKCS. Whether longer term implementation of the guidelines will make an impact has not yet been assessed.

A similar protocol exists in the UK, but the advisory age is 5 and Farrell says it remains to be seen whether restricting breeding based on the status of the parental generation will have a positive effect.

Example 3: Primary Closed Angle Glaucoma in Border Collies

PCAG is a painful eye condition that causes blindness and appears to be linked to goniodysgenesis which is an abnormality affecting the drainage pathway of the eye. This abnormality appears to be highly heritable but can be screened for using an eye examination technique called gonioscopy.

The KC/BVA recommendation has been that dogs affected by goniodysgenesis should not be bred from if they are listed on Schedule B of the Eye Scheme. However, it is known that not all individuals with goniodysgenesis go on to develop glaucoma. Additionally, a leaflet published by the BVA in January 2015 also says “It was originally believed that the degree of goniodysgenesis did not progress after birth and so a ‘one-off’ test before breeding was advised for dogs of certified breeds. It is now known, however, that goniodysgenesis can progress with time and so testing is now advised every three years.”

As Farrell points out, breeding strategies based on the presence of goniodysgenesis, or on a DNA test for it, may not reduce or eliminate glaucoma from the Border Collie population. This same caveat applies to all single-gene DNA tests: these tests will only be effective in managing diseases if there is a proven, direct, cause and effect link between the mutation and the disease.

Example 4: The use of Estimated Breeding Values (EBVs)

EBVs are widely used in livestock breeding around the world to improve things such as egg, meat and milk yield. They are also used to quantify factors such as growth rate and calving ease and even scrotal size (!) in cattle.

For dog breeders In the UK, EBVs are relatively new, but are now available for 28 breeds via the KC’s Mate Select tool.  EBV calculations use the BVA/KC hip and elbow screening data and pedigree information from individual dogs and their surrounding family, to determine the genetic risk that each dog will pass this disease to its progeny. []

Farrell says that one of the interesting things about EBVs is that they enable breeders to make informed choices about breeding without necessarily having to know much about the phenotype of a potential mate. However, in the real world, no responsible breeder is going to choose a stud dog without detailed knowledge of that dog’s appearance, temperament, conformation and movement, among other things.

A further development is genomic EBVs (gEBVs) which are based on genome-wide markers rather than pedigree information.

Don’t forget the people dimension!

The paper concludes that a three-pronged strategy incorporating screening schemes, pedigree information and EBVs/gEBVs could reduce the number and prevalence of inherited disorders, while at the same time managing genetic diversity.

It’s clear there are many factors related to the dogs and the health issues themselves when considering a Breeding Strategy and that’s before you even begin to consider the “human factors” such as:

  • what is valued in the show ring by judges, breeders and exhibitors
  • people’s willingness to acknowledge the need to address any problems
  • people’s willingness and ability to pay to participate in testing and screening programmes

The Farrell paper doesn’t really touch on the human factors and this is where we need to  broaden the discussion out from Breeding Strategies to Breed Improvement Strategies. The latter cover the whole process from leadership by Breed Clubs/Councils, through planning and analysis of survey data, to communication and engagement of breeders, owners and others who can influence the direction a breed takes. A few breeds are already working at that level, most notably some of the Breed Watch Category 3 (formerly “High Profile”) breeds because the KC requires them to do so. The Category 3 breeds simply cannot demonstrate the evidence needed to be transferred to Category 2 without a Breed Improvement Strategy, supported by data.

For those interested in developing a Breed Improvement Strategy, the KC has published a Guide for Breed Clubs/Councils and Breed Health Coordinators.  It is available online here:

So, to sum up: DNA testing and Screening Schemes are a start; Breeding Strategies are a necessary development, but Breed Improvement Strategies explain how people can make real change happen for the benefit of dogs.


Comments are closed.

%d bloggers like this: