This link is going to be big news over the next few days. The headline says OH MY GOSH WE’RE ALL GONNA DIE.
Here’s the abstract of the actual study, by the way.
Here’s what the real story is:
As you should know, UPenn tests hips according to how far out of the socket they will move when the dog is in different positions. This result, called the “distraction index,” is expressed as a percentage of the hip itself. If the hip moves 30% of its own diameter, the DI is 0.3. If it moves 60%, the DI is 0.6. And so on.
DI correlates to the chances of the dog developing osteoarthritis of the hip joint later in life. The lower the DI, the lower the chance of developing osteoarthritis.
UPenn has determined that a DI lower than .3 is associated with virtually zero arthritis.
OK, that’s the backstory.
What this study did was quite simple: It compared OFA ratings and DI scores on about 400 dogs.
It found that a huge number of dogs who had Good and Fair ratings had DI scores over 0.3.
– This is NOT new news. Every study that’s ever been done to compare OFA and PennHIP scores shows that they don’t correlate very well, because they’re looking at COMPLETELY DIFFERENT things.
– You are seeing the outworkings of a longstanding feud between the two camps – expect to see some kind of answering research from OFA.
– The headline of the study makes it sound like OFA is some kind of crashing disaster, and all OFA results are invalidated or something. That’s actually not what the study showed. It set an extraordinarily high cutoff – 0.3 – and considered a “failure to predict” anything that OFA passed that turned out to have a DI higher than that number.
– Numbers above 0.3, in real life, don’t automatically get arthritis. IN THE FOUR BREEDS STUDIED (and this is critical), which are Labs and Goldens and Shepherds and Rotties, the risk of arthritis is around zero for 0.3 and rise to close to 100% for a DI of 1.0 (obviously, since that indicates a hip that comes completely out of its own socket). A dog with a DI of, say, .33 is listed as a “failure to predict” in this study, but UPenn’s own numbers would give it only a very, very small chance of developing arthritis (and, critically, in all of those four breeds a .33 would be a “do breed this dog” result).
– This study looked at UPenn’s four target breeds. The extent to which this study applies to you as a breeder honestly depends on how closely your breed resembles one of these four. UPenn does not have (repeat: DOES NOT HAVE) any published data reliably linking distraction index scores and arthritis risk for any other breeds. The extent to which, say, a bulldog’s DI influences arthritis is not known.
– Once you get out of the sighthounds and the very leggy dogs, you’re not going to find very many scores smaller than 0.3 regardless of breed. This actual study showed an average DI across all breeds of .435 +/- .14. In other words, not only did the average dog fail this “predictive” score, the sizeable majority of the entire study failed.
– According to this study, every single Cardigan on earth would “fail,” because our mean DI is higher than 0.6. We’re in absolutely hideous shape, according to UPenn.
– This study wasn’t really designed for breeders, except to imply that nobody should use OFA and everybody should use PennHIP. If the 0.3 cut-off was used to actually drop dogs from breeding programs, we’d be wiping out a good hundred breeds and decimating most of the rest.
For a fantastic look at what average DIs are across breeds, look at http://www.marshfieldkennel.com/PennHIP.html.