HD in the German Shepherd Dog — a Statistical Study

Very honored Professors, Ladies and Gentlemen,

I would like to submit to you today what is in all probability the most detailed study you have ever obtained in reference to German Shepherd Dogs. This study had begun with the summary of the HD-diagnoses of all dogs of one large breeder in particular, and now finds its continuation, as I have combined all Internet- and Multimedia-provided data concerning German Shepherd Dogs into only one set of statistics.

Please take into consideration that for the statistics in the following table, the diagnoses of no less than 554,596 German Shepherd Dogs were consulted.

Since “Horand von Grafrath” (born as Hektor Linksrhein) was registered in 1899 with “SZ number 1” in the “Zuchtbuch” (breeding register) as the “Stammvater” (progenitor) of all registered German Shepherd Dogs, up until July 7th, 2005, there have been 2,162,084 dogs registered in the breed book of the Shepherd Dog Association (SV). It makes me very happy to be able to offer you, with the statistics in question, the officially registered results of 25.65% of all dogs ever listed in the breed book.

I very much hope that these numbers can help you in your scientific assessment (or judgement) and in your personal work and study. These statistics should not only encourage you to be particularly critical during the evaluation of radiographs, but also to contribute in making the breeding regulations of the breed club much more stringent, so that dysplastic dogs no longer would be used for breeding, no matter how minor their ailment may be. I fervently hope that you will also personally contribute to it, and I would be very pleased to hear any opinion you may have on this subject.

Yours sincerely, Jan Demeyere, 8570 Vichte, Belgium Sept., 2005

(Chart modified by Fred Lanting)

Number evaluated & recorded:


# X-rayed








not X-rayed, (status unknown)

(see next 6 columns)



noch zugelassen

moderate dysplasia

(severe) dysplasia

foreign, equiv. to HD-1, 2


(50.38% of total)

















Percentages of radiographed males: =>

















(49.62% of total)









Percentages of radiographed females: =>







Total, both sexes


















Percentages of all radiographed dogs: =>








Legend: “HD-?” = is not available or is unknown; HD-1 = normal; HD-2 = nearly normal; HD-3 = still permitted; HD-4 = medium HD; HD-5 = severe HD; HD-6 = “a”-foreign country certification of “normal & nearly normal” (details of findings are unknown).

Summary of the statistics: In these statistics, 554,596 descendants are included.

In the documents that have contributed to the study, the following dogs were publicly listed: 279,381 males and 275,215 females. A total of 554,596 dogs were officially registered in the included period in the registry books of the SV. This quantity corresponds to 25.65% of all purebred dogs ever born and acknowledged by the German Shepherd Dog Association.

How “mathematically correct” the reproduction of this domesticated dog breed functions, is impressively documented by the fact that just as many males as females were born and registered: the distribution of 50.38% males and 49.62% females shows a marvellous natural equality. It is clear that the German Shepherd Dog no longer has natural enemies, and so nature does not have to intervene and take correcting counter-measures for one or the other sex.

HD diagnoses: evaluated for the two sexes.

Only 184,315 of those registered 554,596 dogs were X-rayed = 33.23%.

Some 66.77% have no known status! Note: please do consider that, with few exceptions, only the “promising” dogs are presented for radiograph evaluation; that is, if they have a chance at all to participate in show or sport, or if they are intended to be used for breeding. We are dealing with the elite dogs here.

The HD evaluation “HD-1” (“normal”) is given to 57.24% of the X-rayed dogs.

The HD evaluation “HD-6” (certified as approved in a “foreign country”) is given to 3.49% of the X-rayed dogs. [not many apply for an SV rating, as long as they have a hip clearance in their own country. – FL]

That means that 60.73% of the “X-rayed” dogs have made the cut as having a very good rating, but it also means that only 19.02% (“HD-1”) + 1.16% (“HD-6”) = 20.18% of the entire population attained these good ratings. In plain language, only one-fifth of all examined GSDs have “perfect” hips!

(If the rest of the dogs have not been X-rayed and/or reported, one cannot rely on them to supply excuses for this devastating outcome. Many results also do not show up in the database, because the “bad” results are simply not sent in to the SV.)

If we want to be generous and tolerant now, and include HD-“2” (“near- normal”), we reach: 27.96 %. What a state of affairs when only 28% of the more than half-million German Shepherd Dogs have “good” hips!

Differences of the HD diagnoses between the sexes:

It is certainly remarkable that the bitches fare substantially better with the diagnoses of HD-1 and HD-6: 59.19 + 4.17 = 63.36%, compared to the males: 55.26 + 2.79 = 58.05%. That represents a difference of no less than 5.31%. Females, on average, have better hips.

Equally remarkably it is then, in addition, that with the diagnoses HD-2 and HD-3, this substantial difference is again seen: the males have: 25.07 + 11.98 = 37.05%; the bitches however only have: 21.81 + 9.92 = 31.73%. The difference of 5.32% re-surfaces here accurately. On the whole, males have worse hips.

Note: in my opinion, the 10.94% of dogs having HD-3 (“still permissible”) and higher (4 and 5), shares: 4.15 + 0.76 = 4.91%, totalling 15.85%, in no case whatsoever should be used for breeding. They nevertheless are being used and continue to keep this disorder in the breed.

There were no other peculiarities in the statistics that would have drawn my attention. Approximately just as many males (32.59% of the total population) as bitches (33.87%) are submitted for hip radiography.


This is a small supplemental study. I think that in your profession you have a substantial responsibility and ability for changing things, notably in regard to the German Shepherd Dog. I very much hope that with these statistics you can make and rally additional influence, wherever possible, so that fewer animals must suffer or be euthanized.

Yours sincerely, Jan Demeyere, 8570 Vichte, Belgium

The next section is from the 2005 Orthopedics book by Fred Lanting (www.MrGSD.com):

Some statistics compiled in the 1960s indicated the incidence in German Shepherd Dogs to be about 80%, though this figure was later believed to be unrealistic and much too high; The Eye Dog Foundation and International Guiding Eyes organizations found about 55% hip dysplasia in the breed before selection. The US Air Force, screening dogs for possible Armed Services employment, found 56.9% with HD in one series of evaluations and 63.4% in another set, most of these being GSDs. Although some investigators reported an apparent fast rise in incidence paralleling the spurt in popularity in some breeds such as Afghan Hounds and Doberman Pinschers, such drastic changes have not been well documented elsewhere. The greatest improvement, however, can be expected in breeds with low population, high heritability, and strong breed clubs concerned about genetic defects; likewise the greatest decline in joint quality can be expected in breeds experiencing the fastest growth in popularity and indiscriminate breeding.

In a one-year study of “a”-stamp radiographs made and evaluated in Germany of 5,475 German Shepherd Dogs from November 1974 through October 1975, about 13% of males and a perhaps significantly different figure of 21.6% for bitches were considered free of HD. Possibly because of their age (as young as 12 months old) and males growing faster, some of the females classified as free of HD may have shown dysplasia if radiographed at a later age. With the standard view, one year of age just doesn’t give enough accuracy. Over 33% of the males and about 36% of the females in that retrospective study of radiographs were considered suspicious. Those “free” of HD (17.6% of both sexes combined) represent a population percentage far different from the 66% “normal” statistic derived from reading the radiographs of nearly 4,000 German Shepherd Dogs of approximately one year of age. When one weighs the probable lenience built into a radiographic evaluation of a year-old dog against the stricter standard in that German classification at the time, the earlier OFA figure of 66% normals in GSDs is maybe a little too high for the hip-extended view, and the more recent German 17.6% figure too low especially when you consider the average age of those dogs. Some of those rated “suspicious” might possibly fall into the normal category at 2 years or older, but most would likely not.

© copyrighted.Permission to reprint is available from either j.demeyere@skynet.be or mrgsd@hiwaay.net

In either case, a form of the available biographical data for each is to accompany any reprint:

Biographical notes on the authors.

Jan Demeyere has been a GSD enthusiast since he was 10 years old, and is a serious student of the bloodlines used in both show and working lines. Fred Lanting got his first GSD (an import) early in 1947, has had both American- and German-line GSDs (mostly the latter), has bred other breeds, handled professionally, became an SV judge, and has judged in about 30 countries. His curriculum vitae and articles are on www.SiriusDog.com, www.MrGSD.com, and elsewhere. His translation of the SV Standard, schutzhund rules, Körung, and other documents are probably the best available in English.

Addendum: The following statistics may be added if the editor/webmaster chooses, though they are not necessary to an understanding of the problem.

In a review of 200 recent GSD bitches that had a hip rating of HD-1 (Normal, “a”-stamp first level, also called “a”-1), there were 5,272 pups (a little over 26 pups per dam) in 1,103 litters (average 5.5 litters per dam), sired by a variety of males. Some 1,738 of the offspring (33%) were evaluated (entered into the database) at age 12 months or older. Of those, there were 1,109 that got a-1, 390 with HD-2 (a2, fast-[near]-normal), 156 with a3 (noch zugelassen), 56 with mittlere HD, and 13 with schwere (severe) HD — I left out the 14 “a6” progeny because that represents foreign (non-Germany) hip registry data.

Percentages in the evaluated progeny were 64% a-1, 22% a-2, 9% a-3, and fewer with the worst hips. Of course, lower numbers are seen if one considers percentages of all progeny: 21% a-1, 7.4% a-2, etc. Incomplete participation is something we have to put up with when formulating statistics, but if the sampling is large enough and across the spectrum, the values will be valid.

Of the same number (200) fast-normal (a2) bitches used as dams, there were 5,343 pups in 1,144 litters, and 1,778 progeny (again, about a third) were evaluated — almost the same number as in the previous set. But of those, only 953 (54%) were a-1. 464 (26%) a2, 231 (13%) a3, 86 (5%) a4, and 16 (1%) severe. That means that almost the same percentage came up with Fast-normal (a2) hips as in the progeny of the a-1 dams, but fewer got the Normal rating.

Now we look at the bitches that had a3 (noch zugelassen) hips and still were allowed to breed. In their 5,054 progeny, very slightly more offspring were radiographed and entered into the database: 1,742 or 34.5%. Almost the same number and percentage got a-1: 931 dogs, 53.4%. Again, 447 or 26% got a2, 240 (14%) got a3, 5.4% were a4, and 0.8% a5.

It doesn’t seem to offer much incentive to keep the a3 (moderately but noticeably dysplastic) bitches out of the gene pool if there is no advantage to favoring the a2 bitches over them. And the Normal (a-1) bitches had only a small (though statistically significant) advantage over the a2 dams insofar as producing similar hips to their own. If this 10% spread between using a-1 bitches versus using a2 bitches for breeding doesn’t bother you, you have plenty of company. But the age of radiographic evaluation (average is 13 months) and the relative inaccuracy of the standard hip-extended position for X-raying, should drive breeders to look for better ways to make progress. Indeed, the breeders with most influence in the SV caused the Zuchtwert system (breed value calculated from radiographic data of relatives and progeny) to be adopted. It provides much more and safer information than simply the Normal or Fast-normal “a”-stamp.

Even more progress can be made when breeders use the PennHIP distraction index method to determine true laxity, because many lax hips “hide” when a dog is radiographed in the old-fashioned leg-extended ventro-dorsal view. Laxity cannot hide from the distraction view. Details on this procedure are found in the book “Canine Orthopedic Disorders” (www.MrGSD.com) or in articles on www.SiriusDog.com and elsewhere. Of course, the very best approach would be to use both these wonderful tools together: ZW and PennHIP.

Although a3 (and worse) bitches are still being used, a slightly encouraging preference for the bitches with better hips is seen in the statistics: by the time you get to the 200th dam in each set, you are further into the alphabetized list of kennel-names. Demeyere says, “I got further into the alphabet with HD-2 than with HD-1 mothers, as there are fewer in the list, and even further with HD-3 mothers, than with HD-2 mothers. Should I have polled 1,000 HD-3 bitches, it would have taken me maybe to the letter D or F (remember there were 9,251 bitches with HD-3). But with 55,187 bitches with HD-1, I probably would still not get further than B or C.”