Introduction: The aim of genomics is to decipher how genes function and provide an understanding of the link between genotype and phenotype. Therefore accurate characterization of phenotype is integral to genomic advancements. The diagnosis of canine hip dysplasia (CHD) has for decades been based on an empirically accepted phenotype derived from the ventrodorsal radiograph of the canine hips and pelvis. This radiograph is scored subjectively for the presence of subluxation of the coxofemoral joint, or secondary osteoarthritis (OA). It has been generally accepted that the subjective scoring of hip phenotype at 2 years of age accurately reflects the true hip phenotype of the dog for life. This investigation compared OFA-type scores and PennHIP scores at 2 years of age, with end-of-life histopathologic hip phenotypes.

Methods: 48, 8-week old Labrador Retrievers from 7 litters were followed for life. Hips were radiographed in the hip-extended position at 30, 42, and 54 weeks of age, then yearly until end of life. Hip radiographs were evaluated by a board certified radiologist for the presence and severity of OA. All radiographs were also scored using the criteria of the OFA. Distraction radiography was performed at 2 years of age according to the PennHIP method. Gross and microscopic histopathology of the hips was performed on 45/48 dogs after succumbing to natural causes.

Results: Radiographic prevalence of OA increased linearly from 15% at 2 yrs of age to 67% at end-of-life. Of 29 dogs receiving OFA-type 'normal' scores from the hip-extended radiograph, 16 (55%) were radiographically dysplastic by end-of-life and 92% had histopathologic evidence of OA.

At 2 years of age PennHIP scores predicted all 48 dogs to be susceptible to OA (DI range 0.36 - 0.92): 43 of 45 dogs (96%) showed Histopathologic OA by end of life. Median disease free interval was 3 years for dogs with DI>0.6 and 12 years for dogs with DI< 0.4.

Discussion: The linear increase in OA over life refutes the commonly held belief that hip OA occurs either by two years of age or much later in the geriatric years. Therefore "idiopathic" or "old-age" OA of the hip are likely misnomers. The high rate of 'false negative' diagnoses associated with subjective scoring of the hip-extended radiograph impedes genetic progress via selective breeding and also confounds advancements in genomic research. Other scoring systems based on hip-extended radiography used worldwide would likely fare similarly. The PennHIP DI predicted with 96% accuracy the dogs that would go on to develop the OA of CHD later in life. Results of this study (and those previously published) indicate that rapid genetic improvement in hip phenotype will occur using the distraction index as a selection criterion. Similarly, more rapid progress in genomic research can be expected by keying on this newer hip phenotype.

Acknowledgment: Nestle' Purina Pet Care Co, St Louis, MO