The aim of this project is to provide information to breeders that will aide in decisions to improve hip and elbow dysplasia while safeguarding the long term health of the breed by maintaining genetic diversity.
To learn what Estimated Breeding Values are please read on. For breeding advice, please scroll down.
- Inheritance of Complex Disorders
- How to Test for Complex Inherited Disorders
- What are Estimated Breeding Values?
- Why Use EBV's
- How Are EBV's Calculated?
- What Do EBV Values mean?
- Breeding Advice
- How To Make the Most of the EBV Database
- Why Does the EBV Value Change?
- Does This Mean an End to Hip Scoring?
The risk of a dog developing a complex inherited disorder, such as hip or elbow dysplasia, is influenced by both environmental factors (e.g. diet, exercise or factors when in the womb before birth etc.) and genetic factors (the genes a dog has). Each dog will have a mix of both “good” and “bad” versions of genes that individually increase, or decrease the risk of becoming affected by the condition. The impact one version of a gene has might only be slight, but lots of genes having a small influence, both positive and negative, will have a complicated combined effect, making it very difficult to predict whether, or to what degree, a dog will be affected. Only a dog’s genes can be passed on to its offspring.
Some complex inherited conditions have screening schemes or health programmes, which are normally a clinical assessment of a dog, assessed and scored or graded by experts, using a standardised protocol. Examples of official, standardised screening schemes are the CHEDS hip and elbow dysplasia schemes. These schemes have been developed by veterinary and canine health specialists. Your vet submits standardised X-rays of your dog’s hips and/or elbows to an expert, who then assess the X-rays to determine a grade. The grade reflects to what degree a dog is affected (or not), assigning the dog a score on a scale from least to most affected.
Estimated Breeding Values, or EBVs, are a well-established way of estimating a dog’s genetic risk of developing a specific health condition or its severity. The GRCWA is interested in developing EBVs for two conditions; hip dysplasia and elbow dysplasia. Each dog’s EBV is calculated by linking pedigree information with data from the CHEDS health schemes, allowing the genetic risk to be calculated for every individual in the pedigree. EBVs can then be compared across the breed to determine which animals have a higher or lower genetic risk than breed average.
Complex inherited disorders are influenced by genetic and environmental factors. By only looking at a dog’s genetic risk, EBVs strip away environmental influences to help estimate the type of genes that a dog may pass on to any offspring. If used effectively, EBVs can help reduce the risk of puppies inheriting hip and elbow dysplasia more effectively than by only using the sire and dams’ individual hip or elbow scores (which are partly influenced by environmental factors), leading to faster progress in reducing the prevalence of disease.
EBVs can be used to calculate genetic risk by looking at the hip or elbow scores that are available for each dog and its relatives (which share genetics). By analysing the results of related dogs, and by using statistical techniques, it removes possible environmental factors, increasing a breeder’s ability to predict the quality of the genes that a dog may pass on to any puppies. EBVs help breeders to use robust data from an established scheme, to make sensible and informed health conscious breeding choices. EBV’s will be recalculated approximately four times a year as more dogs are entered into the database.
The breed average is always set at zero. Dogs with a higher than average risk of passing on genes for hip/elbow dysplasia will have an EBV higher than zero. Dogs with a lower than average genetic risk of hip/elbow dysplasia will have an EBV lower than zero (i.e. a negative number, e.g. -10). The further a dog’s EBV is from the average, the higher or lower its genetic risk. A dog’s EBV can change during its lifetime, either upward or downward, as more information becomes available, either about the dog itself or its relatives. At birth a puppy’s EBV will be the average of its parents’ EBVs. For example a sire with and EBV of -5 and a dam with an EBV of +5 will produce a litter or puppies with an EBV of 0. Remember – these are not hip scores; this info is not telling you that a dog has better hips than another dog; it’s telling you that the a dog has better GENES for hips than another dog.
The confidence is an indication of how much scoring information has been used to calculate the EBV. The more scoring information available, from the dog itself and/or its relatives, the more confident we are that the EBV is close to the actual genetic risk. The confidence of the EBV can increase if more relatives are, or the dog itself is scored.
- A dog with just its own hip score, and no relatives scored will have a confidence of about 60%
- A dog without its own hip score, but with the score of both parents will have a confidence of around 40%
- A dog without its own hip score, but with only one parent scored will have a confidence of around 30%
Table1. Confidence of breeding values
less than 50%
Low accuracy. EBVs are preliminary and could change substantially as more performance information becomes available.
Medium accuracy, usually based on the animal’s own records and pedigree.
Medium-high accuracy. Some progeny information included. EBVs may change with addition of more progeny data.
High accuracy estimate of the animal’s true breeding value.
Ideally breeders should use dogs that that have an EBV which is lower than average (i.e. a minus number), and preferably with a confidence rating of at least 60%. EBVs with a confidence less than 60% can still be used, but the higher the confidence, the more accurate the EBV will be.
The lower the EBV the better, but breeders do not need to search out the dogs with the lowest risk EBV. Selecting animals with a lower risk EBV than average will still lower the risk of hip/ elbow dysplasia.
It is recommended that breeders make well balanced breeding decisions. Each puppy will have an EBV that is the average of its parents. Therefore a dog with an EBV which is higher than average can still be bred from, providing that it is mated to a dog with an EBV which is well below average (assuming that the confidence for both dogs is high).
Previously, the best advice was to use dogs with hip scores below the breed average score, or elbow scores that were ideally zero, which meant that many dogs could have been excluded from a breeding plan if their scores were a significant consideration. Excluding dogs from a breeding plan can have an impact on genetic diversity. By using EBVs, it is reasonable to use a dog with an individual BVA/KC score over the breed average, as long as the EBV indicates low genetic risk with good confidence. In such cases, the hip and elbow condition of the offspring should be carefully monitored and preferably hip/ elbow scored themselves.
As well as considering the implications of a dog’s EBV or hip score, there are other equally important factors to consider when deciding whether two dogs should be mated together, such as temperament, genetic diversity, conformation, other available health test results, the general health of the dogs etc. Your breeding decisions should always be well balanced between, and take into consideration, the qualities and compatibility of both the sire and dam that you are considering.
Can the results of the scoring scheme or EBVs be used to precisely predict if future puppies will be affected?
Hip and elbow dysplasia are conditions which are inherited in a complicated way which is not yet fully understood by scientists. Due to the complex nature of inheritance of these conditions, it is still possible that affected offspring may arise from parents which have good EBVs. It is hoped that breeding appropriately from screened dogs will reduce the risk of producing affected offspring, and using EBVs reduces this risk even further, but it must be stressed that this is not a guarantee.
EBVs link all available pedigree information with data collected through the health schemes. The more breeders that make use of this scheme, the more confident the estimation of the risk of passing on the genes for either condition. By continuing to hip and elbow score, breeders are securing the future for countless other dogs by providing the information needed to continue Estimated Breeding Values.
EBVs rely on good quality data – the best way to ensure effective EBVs is to get your dogs scored, and use EBVs to indicate genetic risk in your breeding decisions. Although EBVs are a more effective way of using the hip/elbow score information, they are not a replacement. Their calculation relies on a large quantity of good quality score data. EBVs are regularly recalculated to make use of new score data and to provide them for newly registered dogs, so it is essential that scoring continues.
Scoring individual dogs has tremendous value in indicating the actual degree of dysplasia present (or not) in an individual dog. The EBV estimates genetic risk – which is helpful for breeding – but does not take account of non-genetic factors which influence the severity of dysplasia. Hip/elbow scores remain the best diagnostic measure of hip and elbow dysplasia, and will allow/help owners to adjust known non-genetic influences (such as exercise intensity or duration) to minimise the effects of these diseases where they occur.
Although a dog’s genes do not change during its lifetime, the EBV is an estimate of genetic risk, and will change as more information becomes available. For example, at birth, a puppy’s EBV is based on the hip scores of its parents and other relatives. As the puppy grows it may be hip scored itself, as may some of its siblings. This extra information will be used in the regular calculation of EBVs. Eventually that dog may be used for breeding, and some of its progeny may be hip scored too. This information will also be used. All this extra information will increase the confidence of the estimate and may result in changes in EBV.
EBVs will also change in another way too. The breed average is always set to zero. Therefore if the breed average changes and hip/ elbow scores improve or worsen, then an individual dog’s score may become closer to, or further away from, the breed average. For example, a dog, whose hip score, and the hip score for all its family, is 10, will have a low EBV if the rest of the breed has a hip score of 20. If the breed improves and all dogs in the breed have a hip score of 5, then the dog with the score of 10 will have a high EBV.
No! EBVs are simply a more effective way of using the information we already have. It is important to remember that the estimates are only as good as the data used in their calculation. No more hip scores would mean no more EBVs. Thus, the availability of EBVs does not mean an end to participating in hip scoring schemes, but does mean that greater progress can be made in genetic selection for low hip scores through the more effective use of score information. Furthermore, stopping scoring is not a way to improve a dog’s EBV; the confidence will decrease. The best way to ensure reliable EBVs is to continue to score and use EBVs as the indicator of genetic risk in your breeding decisions. Finally but importantly, a dog’s individual hip score still has tremendous value in indicating the degree of dysplasia in that particular dog. The EBVs estimate genetic risk, which is more useful in breeding strategy, but to determine the management of non-genetic factors known to influence the severity of dysplasia the hip score remains the best diagnostic measure.
About the Database
EBV calculations are using data last updated June 2021. The current database as a breed average of 11.2 for hip scores.
EBVs using newly entered data will be calculated yearly.
Thank you to the breeders who have submitted their test results enabling the initial calculation of ebv for Hip and Elbow Dysplasia in Australian & New Zealand Golden Retrievers.
Please Click here to add you dog’s information to the database.
Estimated Breeding Values were determined using the animal model and a generalised Linear Mixed Model. The MCMCglmm R package was used for the calculations.