Genetic evaluation of hip score in UK Labrador Retrievers - PubMed (original) (raw)

Genetic evaluation of hip score in UK Labrador Retrievers

Thomas W Lewis et al. PLoS One. 2010.

Abstract

Hip dysplasia is an important and complex genetic disease in dogs with both genetic and environmental influences. Since the osteoarthritis that develops is irreversible the only way to improve welfare, through reducing the prevalence, is through genetic selection. This study aimed to evaluate the progress of selection against hip dysplasia, to quantify potential improvements in the response to selection via use of genetic information and increases in selection intensity, and to prepare for public provision of estimated breeding values (EBV) for hip dysplasia in the UK. Data consisted of 25,243 single records of hip scores of Labrador Retrievers between one and four years old, from radiographs evaluated between 2000 and 2007 as part of the British Veterinary Association (BVA) hip score scheme. A natural logarithm transformation was applied to improve normality and linear mixed models were evaluated using ASREML. Genetic correlations between left and right scores, and total hip scores at one, two and three years of age were found to be close to one, endorsing analysis of total hip score in dogs aged one to three as an appropriate approach. A heritability of 0.35±0.016 and small but significant litter effect (0.07±0.009) were estimated. The observed trends in both mean hip score and mean EBV over year of birth indicate that a small genetic improvement has been taking place, approximately equivalent to avoiding those dogs with the worst 15% of scores. Deterministic analysis supported by simulations showed that a 19% greater response could be achieved using EBV compared to phenotype through increases in accuracy alone. This study establishes that consistent but slow genetic improvement in the hip score of UK Labrador Retrievers has been achieved over the previous decade, and demonstrates that progress may be easily enhanced through the use of EBVs and more intense selection.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: Hip score and pedigree data was collated and provided by the UK Kennel Club.

Figures

Figure 1

Figure 1. Distribution of hip score from all UK registered Labrador Retrievers.

Dogs radiographed from 2000–2007 inclusive (where age at radiograph was above the stipulated 1 year old, recorded sex and coat colour were from the permitted classes male and female, and black, chocolate and yellow respectively, and scores within the defined boundaries of 0 to 106).

Figure 2

Figure 2. Distribution of EBV over phenotypic scores.

Subsets defined by phenotypic scores H≤4, H = 8, H = 10, H = 13 and H≥37 respectively. These subsets represent the lowest (best) 5% (top), individuals with phenotypic values equal to the lower quartile, the median and the upper quartile, and the highest (worst) 5% (bottom). Vertical dashed lines indicate the 5th, 25th, 50th, 75th and 95th percentile for the EBV. EBV are obtained from analysis of loge(1+H).

Figure 3

Figure 3. The simulated response to selection.

Calculated as the difference in simulated generation mean EBV from parental generation mean EBV, due to selection from sampling within a cohort of dogs with the lowest p proportion phenotypic hip score (red line, closed circles) or EBV (blue line, open diamonds). Response is negative since lower EBVs indicate lower disease liability.

Figure 4

Figure 4. Predicted hip score over date of birth.

Undulations within year indicate the seasonal effect on hip score. The blue line is the predicted value and the red lines indicate±1 standard error. Values shown were obtained from the fitted values for loge(1+H) by transforming back to the observed scale. The standard error curves were obtained by first adding ± s.e. to the fitted value on the transformed scale and then transforming back to the observed scale.

Figure 5

Figure 5. Relationship of offspring to midparent.

Hip dysplasia scores (H) untransformed (left) and after transformation of individual scores to loge(1+H) (right). The points displayed are for the subset of 5115 records for which score data was available on the individual and both parents. The fitted values from regression on mid-parent are also shown using smoothing splines with 4 d.f. for smoothing the fitted curves (Genstat).

Similar articles

Cited by

References

    1. Flint AP, Woolliams JA. Precision animal breeding. Philos Trans R Soc Lond B Biol Sci. 2008;363:573–590. - PMC - PubMed
    1. Higgins A, Nicholas FW. The breeding of pedigree dogs: time for strong leadership. Vet J. 2008;178:157–158. - PubMed
    1. Bateson P. University of Cambridge; 2010. Independent inquiry into dog breeding.65
    1. Lohi H, Nicholas FW. Unlocking the genetic make-up of canine hip dysplasia: we can work it out. Vet J. 2009;181:77–78. - PubMed
    1. Lust G. An overview of the pathogenesis of canine hip dysplasia. J Am Vet Med Assoc. 1997;210:1443–1445. - PubMed

Publication types

MeSH terms

LinkOut - more resources