This article is a consolidation and slightly updated version of two or three that have appeared in the canine press over the past decade or so, because a couple of things have changed during that time. It concerns the more well-known hip registries operating principally in the U.S. and Canada. Revising and combining this way should give you a better “meal” at one sitting. I hope to not only bring you up to date on methods and organization, but also stress again the importance that an open registry would be to progress in reducing incidence of HD in the better strains or breeding lines.
In the course of my judging and lecturing all over the globe, I naturally see variability in how much emphasis breeders and buyers place on orthopedically sound hips. I also see some variation in attitudes and techniques in the veterinary profession in different countries. For example, when I taught one vet in Pakistan how to position dogs and read the films to look for degenerative joint disease (DJD), we had to sneak in through the side door of an M.D.’s clinic so we could use his radiography equipment. Hardly any vets there have their own. We had to do it surreptitiously because that is officially a Muslim country where dogs are religiously “unclean” and some believe they are cursed or defiled if they have any contact with dogs.
Since the 1950s, there has been a growing nearly-worldwide awareness of the genetic nature of HD (hip dysplasia, often called CHD—canine HD—by vets in order to distinguish it from HD in man and other species, although very little difference exists). In the mid-1960s, several organizations almost simultaneously arose to register and attempt to control the disorder. In North America, the OFA (Orthopedic Foundation for Animals) was established with Dr. Wayne Riser as first director; in Germany, the effort was spearheaded by the SV, their club for German Shepherd Dogs; the BVA (British Veterinary Association) took over the job in Great Britain. Shortly thereafter, Canada, Australia, and others jumped onto the bandwagon. Before the 1930s, only the ability of a dog to work all day was a key (though a small hint) to its hip joint quality.
In Canada, there is also the OVC, Ontario Veterinary College, which has maintained a database and diagnostic method similar to OFA’s. Certification is available at age 18 months (vs. OFA’s 24-month minimum and the 12-month minimum required by the SV and other countries), and the reports are slightly different. But the technique, i.e., the hip-extended view, is the same. In Germany and its sphere of influence, the Zuchtwert (breed worthiness) statistics system has been added to aid breeding choices.
Back in dogdom’s dark ages, longevity of ancestors, ability of parents and the individual to work long hours without pain or perform certain utilitarian functions were all that owners could use. In the 1950s and `60s, realization that hip dysplasia could often be detected with radiography, and that it was an identifiable genetic disorder, led to the establishment of many hip detection and registration schemes around the world. In the U.S. (and serving much of Canada’s needs), the Orthopedic Foundation for Animals was established at the U. of PA (“Penn”), with Wayne Riser at the helm. Breeders now had a standardized means of listing potentially valuable breeding animals that had no evidence of severe or moderate hip dysplasia. This was a definite advance over the performance tests, giving more information on this specific portion of a dog’s phenotype. Dogs that were considered normal for age and breed were given a certification number, and as time went on, people could fairly safely assume that a dog without this certification were dysplastic. The certificate thus became a valuable commodity. It marked dogs as more valuable than similar dogs without it.
Unfortunately, the inertia that came with great numbers of dogs in the OFA registry may have contributed toward an unwillingness to consider technological improvements. After several decades of service to breeders, this stagnation and protection of the old method had left breeders on a plateau without offering any ladder to the next level of progress. It was time to examine new paradigms.
Initially, it was easy to see that the most severe cases, which involved pain and loss of utilitarian function, were paralleled by a radiographically demonstrated deterioration of the articulating surfaces in the ball-and-socket hip joint. It was also discovered almost immediately that excess joint space was another indication of the pathology. But laxity (looseness) in the joint, especially as estimated early in the dog’s life, was not as directly proportional or parallel to the eventual worsening of the disease and symptoms. Some younger dogs look normal and tight in the “standard” AVMA (American Veterinary Medical Association) position in which they lie on their backs with legs extended horizontally, yet have a bad genotype that only comes to light phenotypically in later maturity, or else in the progeny. In that respect, the disorder is developmental and progressive.
In humans, the hip dysplasia disorder is obvious at birth, and babies are put into a type of splint that keeps the legs far apart and the femoral heads pushed deep into the acetabular sockets until the joint begins to develop more or less normally. In quadrupeds, this is not feasible, perhaps even impossible. Animals start walking and putting weight on the joints almost immediately, and are simply not carried around during such a long adjustment period. Furthermore, humans are much more developed skeletally than puppies during those first days. With dogs, our emphasis should be primarily on prevention through genetic selection, culling from the gene pool those that do not look better than average, and therefore a superior early detection method than what has been practiced in the past. Breeders must be held responsible for their “products.”
In the canine, what has been the “standard” view for the first half-century and more is a picture taken from above, with the dog on its back and legs stretched out parallel to each other and the table, and the film cassette in a drawer under the dog’s pelvis. This is the best view for seeing most of the deterioration that occurs as a result of the loose wobbling around done in actual motion, so it is an excellent screening view in older dogs. But it is a highly inaccurate method of determining current or relative risk of later deterioration in the individual, or probable genotype. But it was all we had, radiographically speaking, for those first couple of decades. And all too often, a dog that looked OK in this position was bred and passed along its bad genes before its true hip quality became apparent. Only some wiser vets and breeders, looking for more information, used “puppy palpation” at roughly 8 weeks, and the “wedge X-ray” at some age over about 6 months. These latter two are now obsolete, since a far better pre-diagnostic technique has arisen.
Attempts at improving early detection/recognition/prediction of hip quality included those techniques of palpation (feeling the joint while manipulating the limb), the two main and similar methods being the Ortolani and the Bardens procedures. The Ortolani maneuver involves pushing the femur in such a way that the loose femoral head slides up and out of the socket, then is allowed to “snap” back into place. The “click” is audible and can be felt easily with the fingers. This Ortolani sign is still sometimes used as one of several procedures done on an animal brought in for a full evaluation. The Bardens technique differed slightly, the practitioner lifting the 8-week-old puppy’s femur laterally while it was lying on its side, and the thumb and forefinger of the other hand used to estimate laxity in millimeters and compare one pup to the others. It was very useful in the early days, but is only rarely used in some veterinary college research today.
For an uncomfortable number of years, the barely adequate hip-extended technique was used almost exclusively by vets in those countries named above, as well as others. Today, the hip-extended method (although widely used—and stubbornly, if solely) is nearly obsolete as far as a useful early predictor of HD is concerned. In the late 1980s, a great improvement over the “wedge” or “fulcrum X-ray” was developed at the same university where OFA was born, the University of Pennsylvania in Philadelphia. It became known as the PennHIP® distraction technique, the acronym’s capital letters standing for “Hip Improvement Program). Before long, scores of thousands of dogs had been analyzed by this improved technique, and a reliable early (4-6 months age) prediction of relative risk of later developing DJD was a reality. A blessing for breeders who wanted to know before they put much time and money into training and waiting for the young dog to grow old enough to breed and/or work.
That was a much-needed route to the next pinnacle. By the early 1990s, research had already established this as a considerably improved diagnostic or predictive technique that promised great benefits to those who want to produce or buy sounder dogs. As Solomon said, there is nothing truly new under the sun, but there can be advances and better combinations or arrangements of data, genes, methods, etc. than what has been seen before. The concept behind the better method for hips was not entirely new, but was based on earlier experiments in puppy palpation, wedge radiography, and other positions utilizing stress to push femurs apart and demonstrate joint laxity that was hidden in the older view. What we have now is a more highly refined approach, standardized for accuracy and repeatability, and is more quantitative than the limited and almost entirely subjective OFA or other hip-extended approaches.
While the older position with the dog’s legs stretched out as if resembling those of a bipedal human lying supine on the table is best for identifying mineralization (calcium deposits and arthritis) in the hip joint, it misses too much covert (unnoticed) laxity and thus is not a reliable indicator of things to come. Since in many or most cases a loose hip will develop degenerative joint disease (DJD) during the dog’s useful years, it is wise to detect early as much laxity as possible make comparisons based on numerical, measurable standards of excellence and risk. The looser the joint, the higher that risk that the dog eventually will show DJD, regardless of breed. Website articles that treat the differences between PennHIP and older methods in more detail include those on , , and other sites.
One of the important advantages of the PennHIP stress-radiography method is that it allows accurate and much earlier prediction of an individual dog’s later hip status; it is a vastly improved estimate of risk for developing DJD. And the scientific data is more than sufficient to be a convincing case for breeders to use PennHIP. As I have said in another place, those who quote the very old claim that says school is still out on the PennHIP method simply have been skipping classes for the past many years.
This PennHIP method is used to forcibly distract (separate the components of) the loose hip joints while a picture is taken to compare with the deeply-seated view. Its early accuracy (around 4-6 months is basically the same as throughout life in most cases) is a great advantage, and the demonstration of passive laxity gives far more information and predictive value. This gives an objective picture of the true looseness, what I call covert laxity, measurable in millimeters. A formula is used that makes measurements on a Beagle proportional to the measurements on a Great Dane. The flexed position of the legs mimics the posture of a dog standing or walking with hip joints in a neutral-natural mode. Much more a true approximation of “functional laxity” (the actual looseness in a real-life situation). Using this distraction force and measurement is far superior to the non-stressed, subjective estimate of laxity. See the illustration below.
Thousands of vets have taken the qualifying course to become certified, and can be found by geographical listing on the PennHIP website. Breeders are increasing turning to this for the most accurate, reliable, and early determination of probable genotype. And if you can combine it with progeny testing like the SV’s “Zuchtwert” program or The Seeing Eye, Inc.’s Breed Value system (the same at heart), PennHIP can give breeders and buyers an enormous advantage in safety, savings, and improvement of their breeds.
Its early accuracy (around 4-6 months is basically the same as throughout life in most cases) is a great advantage, and the demonstration of passive laxity gives far more information and predictive value. The procedure adds another position, one that mimics the posture of a dog standing or walking with hip joints in a neutral-natural mode. Much more like that which entails a true approximation of functional laxity (the actual looseness in a real-life situation).
The use of Breed Value (progeny and sibling results being made part of a dog’s likely genetic picture) has not been practiced much by dog people in this country. It is an old, established, and critical feature of dairy cattle breeding, but in the U.S., only two groups are known to utilize its very helpful advantages for canines. One is the breeding program at The Seeing Eye, Inc.®, the famous school for guide dogs for the blind in Morristown NJ. The geneticist there, Dr. Eldin Leighton, has used PennHIP for diagnosis/prediction and BV data on relatives (mostly siblings, ancestors, and progeny) to make great strides in reducing HD incidence far below that possible in any other breeding program.
The other is the community of fanciers devoted to the imported German Shepherd Dog. In Germany, the SV (Schäferhund Verein, the parent breed club for GSDs worldwide) has adopted a somewhat similar data analysis system called Zuchtwert (breed worthiness). Through a very complicated formula, the numbers and values of relatively acceptable and unacceptable hip ratings of siblings and progeny are calculated. Stud dogs and brood bitches with low ZW numbers are more desirable as breeding partners and the parents of a pup you would want to buy. The average hip quality of GSD in Germany is fast-normal, translated as “nearly normal,” and the ZW number of 95 correlates to that status in a statistical way. That is to say, a dog with 95 might not have fast-normal hips itself, but on average, there is such a correlation across the breed population. A dog of a year or more of age whose hip radiographs are superior will get a lower ZW number than the calculated number he was born with, i.e., the average of his parents’ numbers. Furthermore, if he has previous siblings and half-sibs that have been radiographed, their results can also move his number one way or the other. His sire or dam may produce better or worse hips with a different parent, and that will also affect his official number. For the non-statistician, it is enough to look for families and individuals with low numbers, as the higher the number, the more HD that particular dog has produced or is likely to produce.
I have been working in the field of canine orthopedic disorders since 1966, and am convinced that the best course a breeder can take is multi-faceted:
- 1. Select the very best (companion, working dog, or foundation stock) you can find.
- 2. Use the most accurate and advanced diagnostic techniques available and economically feasible.
- 3. Base breeding decisions partly on progeny results; one such tool is BV (Breeding Value) or ZW (Zuchtwert).
- 4. Favor open registries; these are those in which the failures as well as the dogs with better hips are identified publicly. If there are none, promote open sharing of information among your peers.
We might as well consider these items in the order summarized above. Whether you intend to breed dogs or not, the first one is very important. You and/or your puppy customers certainly would not purposely select defective dogs, but in effect, not selecting against a problem can be thought of as selecting for it. Think of the analogy in training. If you do not train for correct behavior, you are actually training incorrect behavior. Picking a dog to share your life, home, and time could be a ten-year or longer investment. Besides the features of temperament, size, coat, general health, and so many more, you owe it to yourself and the dog to choose an animal that has good odds of having serviceable hips and other joints. Likewise, you ought to use all the tools available to you, such as those in numbers 2 and 3 in the above list.
The fourth major tool in a successful fight against proliferation or prevalence of hip dysplasia, or almost any disease for that matter, is knowledge. This time, knowledge in the form of open and honest information on the hip status of the dog you are interested in breeding to or getting a pup out of.
One obstacle to general and rapid improvement is the complacence or non-participation of the American dog fancier. The typical owner took a long time to accept OFA or an equivalent registry of normal or acceptable for breeding dogs. To change horses in mid-stream (however advantageous) is not something most would want to attempt, or have the gumption to do, even though it is easy. The other obstacle is the OFA’s intransigence to such matters of progress. They had the opportunity to persuade their almost-captive clientele to switch to a better predictive-diagnostic method and an open registry, one in which inquirers could get the readings on potential studs, dams, or near-relatives of their own potential breeding stock. But the managers gave the inadequate excuse that it might discourage owners from having dogs radiographed and entered into the database. They refused to adopt the progressive ideas of the now-defunct GDC, as they have also refused to advance to the superior stress-radiographic technique developed at Penn. It is largely a matter of politics (power) and personalities. If cooperation were the word of the day instead, there would not have been the need for competing organizations.
I contend that the wisest breeders and shoppers will use all the tools they can to improve their breeds and lines. The ultimate would be to use PennHIP for early risk prediction and a numerical comparison value, and to select stock with lower ZW or BV numbers if such data is available. Breed clubs can do much to improve genetic health by peer pressure through awards and codes of ethics. Unfortunately, if they are affiliated with AKC, they are not allowed to put restrictions on which dogs are allowed to breed, as that monolith is in the business of appropriating not only breeds and breed clubs, but controlling all registration decisions as well. And their practice has always been to register anything as long as the money and the form are sent in, and if there is no complaint about it. In other countries, it is usually the breed club that is the sole and responsible guardian of the breed. Here in North America, we must rely on friendly persuasion and club-suggested standards of conduct to mold attitudes and encourage compliance with such common-sense practices as would improve hip quality.
What about the update on progress? The GDC floundered and was lost. The OFA shows no signs of real progress. It does little good to list only dogs that have passed their low or imprecise standards, when the solely-used hip-extended method is ambivalent in value or accuracy. What about those dogs’ relatives that do not pass, or have not been evaluated?
What of progress at Penn? This is the only one that is moving at fast-forward, and much has been published in the professional (veterinary) literature that supports the value of the PennHIP technique, including many articles in JAVMA, the largest and possibly most prestigious and demanding of the vet journals. But many other journals also carry studies that support PennHIP methods. In 2002 there was an excellent one on risk factors for DJD and its confirmed relationship to Distraction Index (DI). This is Penn’s numerical (quantitative) description of how much laxity can be revealed when a small but significant force is applied to separate the ball from the socket so that a picture can be taken while the hip is in this loosest state. Remember, OFA’s premise is also that “tighter is better” but Penn has the only well-known method of determining true tightness.
Not really new, but an update of sorts for newcomers to the discussions on hip quality analysis, is the fact that PennHIP radiographs can only be submitted by trained and certified vets and their certified/trained technicians. Trained by Penn, that is, in symposia-seminars held in connection with veterinary conferences. Vets who might withhold radiographs because there is obvious HD or DJD, would be dropped from participation. This protects the integrity of the database and prevents the type of bias found in OFA, where mostly the good hip pictures are sent in, and the bad-hip dogs hide in the shadows of the local clinic or the owners’ closets. Vets who would allow non-certified assistants to do the radiography without their personal action in holding the distractor or manipulating the dogs legs would also face expulsion, as has happened twice that I know of.
Semi-annual breed prevalence and statistical profile updates are made available to participating veterinarians and on the Penn website. Penn has made a small move in the direction of an open registry, leaving the decision up to the breeder/owner to decide whether to check off the box indicating willingness to share information. That is a small box, and since an office worker often fills out the form except for the owner’s signature, it is not checked as often as it should be. Make sure you see and mark it when you have your dog evaluated by a PennHIP practitioner, so others interested in breeding to your dog or its parents can contact you. If you want to find out about the DI of a dog or close relatives, you can network with the owners who allow their names and addresses to be made available. PennHIP is still a closed (confidential) registry, so they will not reveal to just anyone the DI of specific dogs, but they advise that when you contact the owner-participants, you ask for a photocopy of the official PennHIP report. If not, you could be fooled by disreputable owners. For example, a dog with considerable DJD may have its joints so frozen by calcium deposits that much laxity cannot be demonstrated. You also must see the part of the report that mentions the absence or presence of remodeling or DJD or evidence of HD. Just to say that a dog is in the 80th percentile or whatever, in regard to comparative laxity is not enough (a dog could have so much “calcium” build-up that you can’t distract the ball from the socket). You need all the information on the report. An owner can get his breed’s statistics on the website.
What is the situation in certain breeds? I am listing here only those breeds for whose magazines or websites I have previously written articles. The others will be found on the website. I’ll start with my own favorite breed, the GSD, although the Golden has one-third more dogs in the database and the Lab has nearly 80% more. The GSD has a mean DI of around 0.43 which means that you should prefer breeding partners that do not have higher numbers, and especially look for ones toward the 0.3 threshold (below which there is no DJD development). For those not familiar with DI, let me suggest that you could think of a number such as 0.50 as representing a joint in which distraction allows the ball to be moved about 50% out of the socket. For a Corgi, Neapolitan Mastiff, or Newfie, this amount may not be as serious (there are breed differences as to relationship between DI laxity and later-onset DJD) or as avoidable, but for a GSD it is definitely a situation to steer clear of.
- Airedale Terriers: 0.58, but this is one of those breeds in which the DJD is not quite as bad at that level as it would be in most GSDs.
- American Bulldog: 0.57
- American Pit Bull Terrier and American Staffordshire Terrier (basically the same breed with different registries): 0.59 to 0.61
- Australian Shepherd: 0.49
- Australian Cattle Dog: 0.56
- Canaan Dog: 0.48, which may surprise some who believe that the rigors of feral life eliminate much HD; in fact, more depends on a combination of breed/group/family phenotype in those wild breeds that actually do not compete that much for food and breeding. But the truly wild dogs that rely on very demanding environments that weed out the weak, and allow only the strongest to do the breeding in the pack, are famously low in HD incidence. The wolf in its many varieties, the Dingo, and even the first generations of wolf-domestic dog blends are examples.
- Dutch Shepherd: 0.53 It is interesting to compare this with its close cousins the Belgian Shepherd varieties: the black Groenendal has 0.3, the mahogany Tervueren has 0.35, and the short-coated Malinois has 0.39. The Dutch Shepherd is phenotypically and probably historically an intermediate form but closer in ancestry to the GSD than to the Belgian Shepherds.
- Mastiff: 0.51 More commonly referred to as English Mastiff or Old English Mastiff
- English Shepherd: 0.62 Also known as Farm Shepherd, found mostly in the USA.
- Fila Brasileiro: 0.56
- Greater Swiss Mountain Dog: 0.56
- Neapolitan Mastiff: 0.65
- Nova Scotia Duck Tolling Retriever: 0.53
- Rottweiler: 0.55 It is of importance and interest that the typical Rottie with a given DI will usually have less DJD than a typical GSD with the same DI. There are even greater differences between several other breeds.
- Shiba: 0.52 Compare this with its descendant, the Akita which has 0.6, but also with breeds more frequently identified with HD symptoms. Even light-weight, small dogs’ hips should be monitored and improved.
- The Greyhound is listed as having a mean DI of 0.23, and probably included both the track and the show Greyhounds. Breeds like this and the Whippet have nearly a zero incidence of HD, although occasional examples of dysplastic individuals are seen.
Non-PennHIP Rating Schemes
In the following table, I have compiled and revised data from various hip rating schemes that do not use the superior PennHIP evaluation method. All of these below, then, represent “best guess” scenarios based on limited data input. But it is helpful as an approximate comparison of the diagnoses used by various clubs and countries.
In conclusion, let us agree that while such big advances in the field of hip dysplasia do not come along frequently, there is no excuse for not availing ourselves of them when they do. The latest great new thing (some decades past, now) was the introduction of the measurable stress radiography technique at Penn. Since then, voluminous studies there and elsewhere have confirmed the superiority of this PennHIP radiographic approach over all others in the world. I am not including such overpriced research methods as MRI, nor do I foresee any genome maps in the future that will help, partly because of the polygenic nature of HD. Those who can use the most tools will have the best chance of success. If you have friends in the few countries where there are no PennHIP certified vets, urge them to use the ZW system where they can, and radiograph dogs later in life to see if DJD has developed after the usual age of certification, which frequently happens. If you don’t have BV-ZW available in your breed or country, use PennHIP even if you have to travel. Share all your results with the public, whether flattering or dismal or anywhere in between, for the sake of the breed and your fellow dog fanciers.
HD and other orthopedic disorders are genetic (you don’t “get” it unless you have “bad genes”), and therefore the two best ways of making a close guess as to the dog’s genotype are through this vastly improved diagnostic technique (PennHIP), and (where available) the Zuchtwert calculations or similar data on near-ancestors and other relatives. Unfortunately, where one method is widely used, the other is not, and vice-versa. It is my practice to use both, and my dream that the two will in my lifetime be coupled to the advantage of the fancy, the breeds, individual dogs, and the sport.
