Dr. M.B.Willis GSD. National Magazine. July 1997.


Epilepsy means fitting and thus is a vague definition. As was listed in my book on the breed (Willis 1991) there are many reasons why a dog can fit and most of these are connected to nutrition, disease or trauma. However, it is known that primary or ideopathic epilepsy exists in many breeds usually shows no signs of abnormality in biochemical tests  of the blood, enzymes, CSF etc. In other words no obvious anomaly can be seen but never the less it fits. There is ample evidence that primary/ idiopathic epilepsy is inherited and this article will confine itself to that aspect of the disease. Interested parties should, however, read what I have to say then in 1991 (written 1990) because they will find warnings and explanations there which are repeated now.


Primary epilepsy has been reported in a variety of breeds, some of which are more seriously affected than the GSD. This is no consolation to GSD breeders but does illustrate the extent of the problem and, of course, evidence from other breeds can be of value to the GSD breeders. Breeds mentioned in the literature or known personally to the author are the Beagle, Belgian Shepherd Dog, Blood Hound, Boxer, BMD, CKCS, Cocker Spaniel, Dachshund, Golden Retriever, Horak, Irish Setter,Keeshound, Labrador Retriever, Miniature Schnauzer, Poodle, Rough Collie, St. Bernard, Shetland Sheep Dog, Siberian Husky, Welsh Springier and Wire-haired Fox Terrier.

In publicity terms the condition became widely known in the GSD in the 1950's to 1970's and I have in my possession over 700 pedigrees of fitters from this era. A scientific paper was published by Falco et al (1974) at Cambridge University and much publicity was generated in breed literature. A list was published by Hein (1987) of dogs which had sired one or more fitters and I added to it in my book (Willis 1991). Actual incidence in percentage terms is unknown and only in Welsh Springers in Holland did incidence figures get reported.

Evidence suggests that a majority of animals fit between 12 and 36 months but a large percentage may fit outside this range. Most dogs fit when at home and quite often when asleep and the duration may last a short time with the majority of dogs being unconscious at some stage. The fitting tends to by cyclic with dogs fitting at regular intervals or increasingly shorter intervals. Some dogs have been known to run into a series of almost continuous fits during which they have needed euthanised. I have had a dog which fitted at 6.5 years and again 3 months later then never again before he died at 12.5 years. As a house dog who slept in the bedroom his status would be known. He had mysolin only for a few days over the incidents and never again. The dog was never used and did not carry Quadrille lines being mainly German. Such types of fitter are the exception rather than the rule.


It is fairly well established that the idiopathic type of epilepsy is inherited and polygenic. Although some authors have noted a higher proportion of males than females affected it is not universal. The condition does not appear to be sex linked (like haemophilia) and inheritance is unclear beyond being complex but it is known to occur in families

In the GSD Hendrawens Quadrille of Eveley (1964-68), his son Hendrawens Charade of Charavigne (1966-69) and his son Eclipse of Eveley (b1967) all produced fitters and the first named  probably was a fitter while Eclipse had a sister which fitted. As a successful sire of winners Quadrille would have been widely used but he died young. His lines are seen in British pedigrees and the recent list published in the National Magazine has resurrected the problem and caused much anxiety, and, sadly, some irresponsible delight in some quarters.

While accepting the influence of these 3 dogs it must be understood that Quadrille was inbred 4.5.5/5.4.4. on Avon Prince of Alumvale (1948-47). Charade was inbred 3/2 on Ludwig of Charavigne and had 8 lines to Avon Prince while eclipse was inbred on Avon Prince. These 3 named dogs above must have derived their genes from somewhere else unless we assume, improbably, that it started with Quadrille. Falco et al (1974) suggested that  inbreeding reduced the age at onset of fitting and that line breeding to F18 increased the risk. Inbreeding to a suspect dog is going to increase the incidence in accordance with genetic theory and it is important to realize that F18 was Avon Prince of Alumvale. We therefore have to accept that Avon Prince is a risk dog whether through Quadrille or not. Avon Prince was inbred 2/3 on Ingosohn of Erol (1938-51) who was son of Ingo v Piastendamm (b1934) who is found behind many of Germanys best lines.

To single out Quadrille and ignore other lines is dangerous and self defeating since it ignores facts. Just after the recent listings in the UK GSD National MAgazine I received a pedigree of a fitter sired by an English born, German bred dog devoid of Avon Prince "blood" (not on the list) and a bitch with one line to Quadrille (generation 11) and to Eclipse (generation 11). One can not definately say that these 2 lines are the sole cause of the problem and the remainder of the pedigree is innocent - genetics does not operate like that!


If one considers a pedigree it has 2 parents, 4 grand parents, 8 great grand parents etc. with the number of ancestors doubling each generation. In generation 7 there are 128 ancestors but a dog only has 78 chromosomes. Although one can inherit part of a chromosome in real terms it means that some of the ancestors in a pedigree have given nothing at all  to their descendant. This is certainly true if one considers generation 110 (1024 ancestors) or 11 (2048 ancestors) which is where many Quadrille lines are being found.

If epilepsy was a simple autosomal gene and one had an ancestor in generation 10 that was a known carrier then there is a 1/2 chance it would be transmitted to his son or daughter and a 1/2 chance it would not be transmitted. The descendant 11 generations down the line would have one chance in every 2048 of inheriting the gene which is an 0.0488% chance. Clearly no one should be getting worked up over such a low chance and if an ancestor 10 generations back in a potential stud dog carried dwarfism it would hardly deter me from using him.

Epilepsy cannot be calculated like this because the risk is unknown and therefore both parents may not contribute equally but the same principal applies. However, if we had Quadrille in generation 10 but through Eclipse, his grand son, in generation 8 then the risk lies in generation 8 not 10. Similarly if we had an increasing number of lines to a suspect dog then we increase the risk. In simple terms 2 appearances in generation 10 would be similar to one appearance in generation 9 and 12 appearances in generation 10 similar to 3 appearances in generation 8. However even then the risk is not high though it will be for some animals

Counting appearances is misleading if one does not know where they appear and behind which ancestor. One dog in the list has, for, example all his lines to Quadrille behind a bitch in his pedigree. There is no evidence of epilepsy in that bitch or from her and thus the risk from this dog must be minimal. To my knowledge only one animal on the GSD national magazine list is known to have sired fitters (Donkin 1997) This does not mean that others may not do so in the future and obviously if epilepsy can occur up to 36 months of age or older, sires that have young offspring have clearly had little time to demonstrate epilepsy.

Epilepsy is a serious problem but it is not the only problem in the breed and it must be kept in perspective. Many people are so afraid that they might be tempted to cover up cases and we might well have breeders who are quite open about hip or other failings but cagey about revealing fitting data. If you offered me a fitter of a pancreatic insufficiency dog I would prefer neither but in reality a fitter may have problems at intervals but a PI dog is going to have problems every day of his life. Nothing will be gained by avoiding dogs that might give rise to epilepsy if in lieu of these we use dogs that produce other problems that can be just as debilitating or serious to the dog and the breed.


What I find astonishing is that some breeders use dogs without even knowing what is behind them. Any would be breeder should do his/her homework on the dogs they wish to use. Before anyone brings this up let me tell you that my wife used Marcoys Eiko at Jemness who is on the list. She used him after due discussion with me and we were aware of the Quadrille lines but I was prepared to take the risk. What I have written is not influenced by that decision because the basis of what I am saying I have said before Eiko was even born.

If one is as concerned about epilepsy as one should be then one should:
1) Report all fitting cases to the Council data base with veterinary reports where possible.

2) Not assume that all fitters are idiopathic epilepsy but try to get veterinary confirmation.

3) Remove from the breeding programme fitters and those proven producers of fitters.

4) Check the pedigree of breeding stock as far as suspect lines are concerned but do not imagine that Quadrille is the only line, others like Avon Prince are just as serious.

5) Remember that there is a difference between a dog that might produce fitters and a dog that is producing fitters. Any dog might produce something serious and we have to take some gambles when breeding but we should minimize those risks. For example a widely used descendant of a suspect sire that has never produced any epileptic problems is safer than a sire little used of the same descendant who has also never  produced the problem.

6) Remember that a dog will have more descendants as generations pass. Quadrille had only 215 direct progeny but would have had many more grand children and will appear in many thousands of pedigrees in generations 10/11 etc. Most of these long distance descendants will be totally safe as regards epilepsy because, if not, epileptics would be occurring in their thousands. The problem is identifying the safe from the unsafe and that means reporting known cases to some central body.

7) Most long distance descendants of Quadrille will not carry a single gene from him at all.

To those gloating over the fact that some stud dogs have had use curtailed remember that these may be innocent dogs with much to offer. None of us should derive any pleasure from failings within the breeed since we are all involved in one way or another. Diminishing your dog does not enhance mine!