PETscan: measuring incidence of disease phenotypes to prioritize genetic studies in companion animals

Quantification of the health-status of the Dutch Labrador retriever population

Health issues in purebred dogs are currently considered one of the biggest problems in companion animal health. The Labrador retriever (LR) is one of the most popular dog breeds. The aim of this study was to quantify LR breed health in comparison with mixed-breed dogs (MB), by using four different data sources: a veterinary practice management system (appr. 35,000 unique individuals LR + MB), data from two animal insurance companies (appr. 15,500 and 4500 individuals respectively), and a histopathological laboratory (appr. 4000 individuals). After extensive recoding of the data, health parameters utilised to quantify breed health were longevity, frequency of practice visits and insurance expense claims, and diagnostic codes. A Kaplan-Meier univariate and multivariable Cox proportional hazard model were used to evaluate longevity. A negative binomial model was used to analyse the frequency of visits, claims, and diagnostic codes in both sets of insurance data. Logistic regression was used to look into the categorical diagnostic codes in the laboratory data. The median lifespan of the LR was similar (12 years, practice data) or longer (10 versus 8 years, insurance data) than MB for individuals with a known birth and death date. When including censored individuals, survival time in the LR was comparable to MB individuals up to 10 years of age. Above 10 years of age, the LR lived a similar length as MB with a medium to large body size, but shorter than all MB. The LR visited the veterinary practice more often (risk ratio (RR) 1.2, 95% confidence interval 1.2-1.3), and also showed a higher frequency of insurance expense claims (RR 2.2 (2.1-2.3) and RR 1.2 (1.1-1.3) respectively for the two insurance data sets). The largest difference in organ systems between the LR and MB in insurance claims was related to ears (RR 5.3 (4.8-5.8) and RR 2.6 (2.3-3.1)), followed by airways (RR 2.6 (2.4-2.8)), tendons & muscles (RR 2.4 (2.2-2.6) and RR 1.4 (1.1-1.7)), and joints (RR 1.7 (1.3-2.1)), without a difference in median age at diagnosis. The data from the histopathological laboratory suggested a higher disease burden related to oncology for the LR compared to MB (OR 1.2, 95% CI 1.0-1.3). Oncological diagnoses were made at a younger age in the LR (8.8 versus 9.4 years). The disease burden was significantly higher for the LR than MB, but these results may suffer from substantial bias such as selection bias towards the database, and different behaviour of LR versus MB owners with regards to veterinary care. In the future, longer term population data can corroborate these results.