Collie eye anomaly (CEA) in dogs: does it blind a dog, what the NHEJ1 DNA test predicts, and what an Indian owner can actually do

Signs of Collie eye anomaly

The 8 signs to watch for are listed below. Any one of them is a reason to consult your nearby vet; none of them is a diagnosis.

• Usually nothing at all, because most affected dogs see normally for life
• A pale patch beside the optic nerve, visible only on an eye examination
• One eye clearly worse than the other, since the lesions are rarely symmetrical
• Bumping into furniture in an unfamiliar room, in a severely affected dog
• Sudden loss of sight in one eye, which suggests a detached retina
• An unusually small eye, called microphthalmia, in a severely affected puppy
• Visible blood inside the eye
• A cloudy or chalky patch on the cornea from mineralisation

Breeds on this site that document this condition

Each breed page below carries the figure for this condition with its denominator and its source. The 14 breeds listed are the ones on this site that document it, not a ranking of risk.

What collie eye anomaly is, and why one word of the name is wrong

Collie eye anomaly is a congenital malformation of the back of the eye, present from birth and not progressive, in which the choroid under the retina fails to develop properly Clark JA, Anderson H, Donner J. The malformations are there on the day the puppy is born, which is the single fact that governs everything else on this page, from the examination window to what a DNA test can and cannot promise.

Three lesions carry the name and they are not equal. Choroidal hypoplasia is the primary defect, and colobomas and retinal detachment are secondary lesions built on top of it. The choroid is the vascular layer that feeds the retina, so where it is thin the retina is short of blood supply, and the thin patch sits in a predictable place, temporal to the optic nerve Parker HG, Kukekova AV, Akey DT.

The severity is not predicted by the parents. Dogs with mild signs can produce puppies with severe collie eye anomaly, and the registry that sells the test says so on its own page. The Royal Kennel Club puts the same point in more words, that the clinical effects vary greatly among affected dogs within one breed, between parent and offspring and even within a single litter. Nobody understands why one mutation produces a dog that sees normally for fourteen years and a littermate that is blind, and the Orthopedic Foundation for Animals states plainly that the basis for this variability is not understood.

One old mutation, carried out of the herding breeds by outcrossing

The reason a spaniel from South Carolina gets a collie disease is that there is only one mutation and it is very old. Parker and colleagues mapped a 7,799 base pair deletion in intron 4 of the NHEJ1 gene on canine chromosome 37 and showed it cosegregating with the condition across the Collie, Border Collie, Australian Shepherd and Shetland Sheepdog, concluding that the primary mutation arose as a single disease allele and was transmitted to multiple herding breeds through outcrosses in early breed development. The deleted stretch is intronic but it contains a 124 base pair segment conserved across every available mammalian genome including the opossum, with binding sites for several regulatory proteins.

The Boykin Spaniel is how the researchers found that out. Carriers of the haplotype turned up by chance in a hunting breed nobody had associated with the disease, an affected Boykin Spaniel was then tracked down and genotyped, and it was homozygous. The same happened in the Nova Scotia Duck Tolling Retriever, a gun dog whose ancestry reportedly includes farm collies. As the authors put it, without deliberate selection for any herding trait, the only signature of herding ancestry left in those two sporting breeds is the eye defect.

Not every collie-looking eye lesion is this mutation. A pedigree of Soft Coated Wheaten Terriers with a broadly similar phenotype carried neither the deletion nor any segregating marker in the region, and the authors concluded a different gene in the same developmental pathway. Two Berger des Pyrenees dogs diagnosed with coloboma did not carry the deletion either Parker HG, Kukekova AV, Akey DT.

The name is wrong and has been challenged twice in print

Two separate attempts have been made to rename this condition, and both came from people who found it outside the collies. Bedford described an ocular disease ophthalmoscopically identical to collie eye anomaly in the Lancashire Heeler, a terrier, and argued that the established appearance of the disease outside the collie breeds makes the nomenclature inappropriate, proposing congenital posterior segment anomaly instead. In 2024 a consensus guideline on the nomenclature of companion animal inherited retinal disorders gave it the symbol CEA-NHEJ1, and OMIA now files the entry under choroidal hypoplasia rather than collie eye anomaly. Both changes say the same thing, that the breed in the name is a historical accident.

The scale of the drift is in the registries. OMIA documents the phene or its likely causal variant in 22 breeds, from the Australian Kelpie and the Koolie to the Jack Russell Terrier, the Siberian Husky and the Chinook. The Orthopedic Foundation for Animals lists 17, the University of California Davis laboratory lists 15 breeds appropriate for testing, and the three lists do not agree Veterinary Genetics Laboratory Online Mendelian Inheritance in Animals. Parker's team, working from the American College of Veterinary Ophthalmologists compilation, counted nine breeds and varieties known to cosegregate the phenotype and the haplotype, a further 18 in which choroidal hypoplasia and optic nerve coloboma are infrequently reported, and sporadic isolated coloboma case reports in another 77 breeds.

Does collie eye anomaly blind a dog? The proportions, with denominators

Most dogs with collie eye anomaly see normally for life, and a minority go blind. The Royal Kennel Club states that most often the disease presents in a mild form detectable only on ophthalmological examination, in a dog that retains normal vision throughout life, and the largest genotype study calls the condition non-progressive and says the majority have minimal changes to vision.

That is not what the pages ranking for this query in India tell an owner. VCA's page opens its clinical signs section with the sentence that the most common sign of collie eye anomaly is blindness. Four other sources disagree with that sentence in print. The study that mapped the locus records that many dogs exhibit no obvious clinical consequences and retain apparently normal vision throughout life. OMIA records that mildly affected animals will have no detectable vision deficit. The Orthopedic Foundation for Animals records that some affected dogs exhibit no apparent visual impairment. The University of California Davis laboratory selling the test records that some dogs may have no obvious clinical signs and present with normal vision all their lives. Wisdom Panel, which sells a panel including this variant, hedges in the middle, saying the disorder can vary in severity with some dogs showing only mild blood vessel abnormalities while others experience vision loss. This page sides with the four, because four independent bodies confirm the same thing and the fifth does not cite anything.

The proportions come from one retrospective series large enough to split the lesions apart. Over eight years from 1999 to 2007, 3,527 eye examinations in six affected breeds in Switzerland were sorted into choroidal hypoplasia alone, choroidal hypoplasia with coloboma, and coloboma alone.

Read the last column and the blindness question answers itself. In Swiss Rough Collies, about 8 affected dogs in every 100 had a coloboma at all, and in Shetland Sheepdogs about 13 in every 100. The other nine in ten had the flat, non-progressive defect that costs a dog nothing. The abstract prints a denominator only for the Smooth Collie, 101 dogs, so the percentages for the other breeds are shares of unstated group sizes inside the 3,527 examinations, and the last column is Bonbeno's arithmetic on the published percentages rather than a figure the authors printed.

Detachment is rarer still. The Copenhagen review of this disease, citing Bedford's United Kingdom survey, reports retinal detachment in 6 percent of affected dogs and calls it a relatively rare complication, while noting that it matters out of all proportion to its frequency because of what it does to vision. Detachments are usually seen in more severely affected eyes and usually alongside colobomatous defects, they are often unilateral, and when they occur they most often appear within the first year of life.

What owners of affected dogs actually report

One study asked them. In a Danish questionnaire study of Shetland Sheepdog and Rough Collie owners, vision in affected as well as unaffected dogs seemed normal on the owners' perceptions, and the only report of inferior vision in the whole study was in a dog with bilateral coloboma. That is the same split again, from a completely different instrument.

The author of that study then argued against her own result, which is why it is worth quoting. Clinical observation of vision is not a direct measurement of visual function; histopathology in some affected dogs has shown thinning of the outer nuclear layer, absence of rods and cones and a reduced number of ganglion cells, changes that must affect retinal function; and dogs rely so heavily on smell and hearing that they can have inferior vision without an owner noticing. She also flags the obvious bias, that owners who have been told the disease is usually harmless may answer accordingly. So the honest claim is that most affected dogs behave as though they see normally, not that their retinas are normal.

How common collie eye anomaly is, breed by breed and country by country

Collie eye anomaly is not rare in the breeds that carry it, and in the Rough Collie it is close to universal. Prevalence is estimated at 70 to 97 percent in rough and smooth collies in the USA and Great Britain, 68 percent in rough collies in Sweden, and 2 to 3 percent in Border collies in the USA and Great Britain. Those are the widest figures in canine inherited disease and they are for one condition in one family of breeds.

Every figure below is a measurement of a different thing, so read the population before the number. An ophthalmoscopic survey counts dogs with a visible lesion. A DNA register counts genotypes in dogs whose owners chose to test. A national club's published list counts only dogs whose results the club records. The three do not converge and should not be averaged.

The spread across those rows is not noise, and the authors of the largest study say so. Across the published literature the allele frequency of the deletion has been calculated at anywhere from 0.43 percent to 79.7 percent depending on breed and country. A figure for the Rough Collie tells you nothing useful about a Border Collie and a figure from an American register tells you little about a European one.

One of those rows carries a disagreement worth flagging, because it is the sort of thing that propagates. OMIA introduces the Italian carrier frequencies as coming from a sample of 465 Italian dogs from five breeds. The paper it is summarising says 496 dogs, and its own five breed counts, 334 Border Collies, 74 Shetland Sheepdogs, 52 Australian Shepherds, 20 Nova Scotia Duck Tolling Retrievers and 16 Rough Collies, sum to 496 rather than 465. OMIA also rounds the paper's carrier figures, printing 30 percent, 17 percent and 13 percent where the paper prints 29.6, 17.3 and 12.5 Marelli SP, Rizzi R, Paganelli. This page uses the paper's numbers. Anyone quoting the database rather than the study will publish a denominator 31 dogs short.

The Rough Collie, counted three ways, all of them high

The Royal Kennel Club publishes every dog on its register with a recorded CEA/CH DNA result, as a breed-by-breed PDF, and nobody appears to have counted them. Bonbeno counted the result column of the report generated on 13 May 2026. Of 244 Rough Collies with a published result, 142 are Affected (58.2 percent), 46 are Carrier (18.9 percent) and only 56 are Clear (23.0 percent) The Royal Kennel Club. Those three counts sum to the total, and Bonbeno's arithmetic on them gives a mutant allele frequency of 330 of 488 chromosomes, which is 67.6 percent.

Three other datasets land in the same place. In Italy 81.3 percent of Rough Collies were affected recessive homozygotes, on a sample of only 16 dogs. In Thailand 66.7 percent of Rough Collies were homozygous mutated and the remaining 33.3 percent were heterozygous, so not one Rough Collie in the Thai sample was clear. On the American register, 41.3 percent of 416 Collies are abnormal and 25.7 percent are carriers. Four countries, four instruments, and the answer is between two fifths and all of them.

The direction of travel is the surprise. Across fifteen years of global commercial testing, Border Collies showed a strong decline in affected genotypes, carrier genotypes and the disease allele, while Collies showed an increase in the frequency of homozygous-affected genotypes and of the disease-associated allele, and the heterozygous genotype rose in Shetland Sheepdogs too. The United States, one of the five countries supplying most of the samples, showed a striking increase in homozygous variant dogs over the same period. Testing is not the same thing as selecting.

The same breed reads differently in different countries

The Shetland Sheepdog is the clearest example of why a single breed figure is meaningless without its country. On the Royal Kennel Club's own published list of 429 Shetland Sheepdogs, 153 are carriers, which is 35.7 percent, and 11 are affected. On the American register, 1,582 evaluations give 4.7 percent carriers and 0.4 percent abnormal. In Italy, 74 Shelties gave 50 percent carriers and 10.8 percent affected. In Thailand, 3.3 percent heterozygous and 1.1 percent homozygous. Older clinical surveys reported 48.3 percent in the Netherlands, 72 percent in the United Kingdom and 15.1 percent in Switzerland Bedford PGC. Collie eye anomaly… Walser-Reinhardt L, Hassig M. That is a range from 3.3 percent to 72 percent for one breed, and every figure is real.

The Border Collie is the reassuring one. On the Royal Kennel Club's list, 1,481 Border Collies carry a published result, 1,288 Clear, 190 Carrier and 3 Affected, a carrier rate of 12.8 percent and an affected rate of 0.2 percent. The American register agrees closely at 11.6 percent carriers and zero abnormal on 413 evaluations, and Norway's litter study put the allele frequency at 6.3 percent. Bedford's 1982 figure for the breed was about 6 percent affected. This is what a breed looks like when the disease was never common in it.

One caution belongs on every row above. A DNA register counts dogs whose owners paid to test them, and breeders who test are not a random sample of breeders. The Rough Collie figure of 58.2 percent affected is high partly because the disease is genuinely near universal in the breed, and partly because a British Rough Collie breeder has a strong reason to test and a Rough Collie pet owner has almost none. Read it as the register's composition, not as the breed's prevalence.

The go-normal trap, and why a six-month eye certificate is worth less than an eight-week one

To catch collie eye anomaly on an eye examination, look before the puppy is about eight weeks old, because pigment fills in over the lesion as the eye matures and hides it. Ophthalmological examination within the first 10 weeks of a puppy's life is paramount, after which the retina changes colour and minor changes can be masked, an effect breeders have called go-normal since at least 1982.

The mechanism is ordinary development. The retinal pigment epithelium is immature at birth and matures over roughly the first 12 weeks of life, and the choroidal hypoplasia lesion is visible precisely because that patch lacks pigment, which lets the abnormal choroidal vessels show through. As pigment arrives, the window closes. The lesion has not healed. The examiner can simply no longer see it.

OMIA records the phenomenon and draws a genetic conclusion from it. Some affected dogs have been observed to go normal later, most of which were homozygous for the likely causal mutation but a small minority were heterozygous, which suggests other loci affect how the disease is expressed and that the disorder is multifactorial. A carrier that looks affected at seven weeks and normal at seven months is not a mistake in the record. It is a hint that the genetics are not as clean as the test implies.

The Swiss series measured the effect rather than describing it. Across 3,527 examinations, there were statistically significant differences in the occurrence of collie eye anomaly between dogs first examined younger or older than 8 weeks, and the disease was diagnosed more often in dogs younger than 8 weeks within the Rough Collie and the Shetland Sheepdog. The same authors note that masked dogs are then reported as clear at later examinations, which is how a go-normal dog gets a clean certificate. In the same study there were no significant differences between males and females and no relation to coat colour.

Different bodies name slightly different windows, so here are all of them. The Orthopedic Foundation for Animals says diagnosis can be made by an ophthalmologist as early as 6 to 8 weeks of age. VCA says breeders screen with examinations by a veterinary ophthalmologist between 6 and 8 weeks. The Copenhagen review says within the first 10 weeks. The Norwegian litter study examined puppies from 5.1 to 8.9 weeks. The British Veterinary Association litter screening service takes puppies at 5 to 12 weeks and requires any appeal to be re-examined before the puppies reach 12 weeks. The largest genotype study says the lesions can be identified in the first two to three months. Where the sources disagree, the earliest window is the safe one.

Merle coats break the examination independently of age. Lesions in a merle dog blend into a fundus that is already short of pigment, which is why the identification can be difficult in a merle puppy and why the DNA test was specifically proposed as a way to identify the condition in merle dogs where pale eye colouring can lead to a false diagnosis Parker HG, Kukekova AV, Akey DT. In the Danish study, all five dogs whose clinical and genetic diagnoses disagreed were blue merle. If you own a merle Australian Shepherd or a blue merle Rough Collie, the examination alone is the weakest of the two instruments on your dog.

How often breeders actually miss the window

One study went and counted. Reviewing every Shetland Sheepdog litter registered with the Danish Kennel Club in 2013, the author found that of 79 litters, 20, about 25 percent, were not eye-examined within the first 12 weeks of life. In a country with a national scheme, a recommendation and an engaged breed club, one litter in four missed the window. India has none of those three, so treat the Indian rate as unmeasured rather than better.

What the NHEJ1 DNA test predicts, and the two things it does not

The NHEJ1 DNA test predicts one thing well and two things badly. It reports how many copies of the 7,799 base pair deletion a dog carries, which in most breeds tracks choroidal hypoplasia closely Grosas S, Lingaas F, Prestrud KW. It does not predict whether an affected dog will have a coloboma, and in at least one national population it does not predict choroidal hypoplasia either Fredholm M, Larsen RC, Jonsson M.

Start with what is on sale. The University of California Davis Veterinary Genetics Laboratory sells the single test at USD 58 per animal, with a USD 5 discount on three or more dogs and USD 27 when it is added to another health test on the same animal, on a turnaround of at least 15 business days. The Royal Kennel Club sells it as CEA/CH at GBP 80.00, which is the dearest individual DNA test in its list, where most of the others are GBP 60.00. The Orthopedic Foundation for Animals takes orders online, mails a cheek swab, has the sample processed by the Small Animal Molecular Genetics Lab at the University of Missouri, and charges USD 15 to register a DNA-based test result in its databases Orthopedic Foundation for Animals.

Read that middle row twice, because it is a registry error rather than a subtlety. The Orthopedic Foundation for Animals page describes collie eye anomaly as autosomal recessive and then tells the owner of a carrier that the dog is at risk for developing the disease. On the same page, the affected paragraph tells the owner that the dog will pass down the risk for developing IVDD to each of its offspring, naming intervertebral disc disease on a page about an eye. Both look like boilerplate copied from another disease page and left uncorrected, and both were still there on 27 September 2026. Read the laboratory report, not the registry's summary paragraph.

Limit one, the test does not predict coloboma

This is the limit that matters, because coloboma is the lesion that leads to detachment and detachment is what blinds. A University of California Davis team collected Nova Scotia Duck Tolling Retrievers with optic nerve head coloboma and found the deletion discordant with the phenotype: only 3 of 7 coloboma-affected dogs were homozygous for the deletion. A genome-wide scan on those 7 cases against 47 unaffected controls found no region significantly associated with the coloboma once population structure was controlled for.

The authors' own recommendation is the one no ranking page repeats. Because of the complex genetic origin of optic nerve head coloboma, they write, the deletion test results should be carefully considered when making breeding decisions, and if the goal is to select for visually competent dogs, eye examinations of puppies would be more effective as a guide in selecting breeding pairs than relying solely on currently available genetic tests. The laboratory that sells the test publishes that finding on the test's own product page, which is to its credit.

Two older observations point the same way. Parker's team tested two Berger des Pyrenees dogs clinically diagnosed with coloboma and neither carried the deletion Veterinary Genetics Laboratory. And the largest genotype study states that the deletion may be linked to the true causal mutation rather than being it, because the deletion does not segregate with coloboma or choroidal hypoplasia in some cases, adding that a recent abstract suggests colobomas are inherited in a complex manner involving multiple loci.

Limit two, in one national population it does not predict the primary lesion either

A Danish group tested the assumption directly in the two breeds where it matters most. Their conclusion is blunt. The deletion in NHEJ1 is not predictive for choroidal hypoplasia in the Danish Rough Collie population, whereas the clinical and genetic diagnosis do agree in the Danish Shetland Sheepdog population, and on that basis the intronic deletion is not the causative mutation but a marker linked to the locus underlying the trait in some populations, and linked to both the wild-type and the disease locus in most dogs in the Danish Rough Collie population. OMIA carries that finding as a standing warning on the entry.

The counterweight is real and should be published beside it. In Norway, 103 Border Collie puppies from 16 litters were examined and genotyped, and the compliance was good: 85 clear, 17 carriers, one genetically affected, and one puppy with clinical findings consistent with choroidal hypoplasia and optic nerve coloboma. OMIA adds a sharp reader's note on that paper, that readers were not told whether the clinically affected puppy was the same puppy as the genetically affected one, nor whether both parents of the affected puppy were heterozygotes. In Slovakia, 75 samples across four breeds found 31 dogs carrying the mutation, 15 homozygous affected and 16 carriers, and the authors judged the PCR test effective for identifying carriers. In Poland, a Kelpie with choroidal hypoplasia proved homozygous, and genotyping 21 related dogs found the dam affected and eight more carriers.

So the test works, breed by breed, until it does not. OMIA attaches a general warning to every variant it lists, that inclusion of a variant in the table does not mean it should be used for DNA testing and that anyone contemplating its use should examine the evidence critically, especially in breeds other than the one it was first described in. For this variant it adds a specific one, pointing at the Danish and the Nova Scotia Duck Tolling Retriever results by name.

One more thing to check on the page you are buying from

The University of California Davis test page states that the 2007 study identified the 7.8 kilobase deletion in exon 4 of the NHEJ1 gene. The paper it cites, and OMIA, both place the deletion in intron 4 Online Mendelian Inheritance in Animals. It is an error in one word on the page of the laboratory selling the test, and it was there on 27 September 2026. It changes nothing about what the test measures, but it is a fair reminder that a laboratory page is marketing copy with a citation attached, and that the citation is the thing to read.

Litter screening versus the DNA test, and why you want both

Litter screening and the DNA test answer different questions, so a breeder who has done one has not done the other. The DNA test reports a genotype before any lesion could be seen, and an eye examination reports what that dog's retina looks like today, including the coloboma the DNA test does not predict Brown EA, Thomasy SM, Murphy CJ.

The clinical half is a real scheme with a real price. Under the British Veterinary Association Canine Health Schemes Eye Scheme, a panel of 42 expert veterinary ophthalmologists examines dogs across the United Kingdom, a routine eye examination for one dog starts from GBP 72.00 including VAT, and litter screening for puppies aged 5 to 12 weeks starts from the same figure. Every puppy must be microchipped before it is examined, results are recorded as Clinically Unaffected or Clinically Affected on a certificate issued at the appointment, and over 14,000 dogs were examined under the scheme in 2025. An appeal against a litter screening result must be notified quickly enough for the whole litter to be re-examined before the puppies reach 12 weeks.

In North America the equivalent is a register rather than a scheme, and it is cheap. The Orthopedic Foundation for Animals charges USD 15 for an Eye Certification Registry entry with no minimum age, USD 30 at the litter rate and USD 10 each at the kennel rate, on a fee schedule whose last increase took effect on 1 January 2022. A DNA-based test result is registered for USD 15.

One working registry goes further and makes the eye examination a condition of registering a litter at all. The International Sheep Dog Society requires that the sire and dam must have passed defined eye scheme requirements when the Registration Folder is submitted, alongside its other rules that both parents be registered and that puppies be registered before they are a year old. That is stronger than anything a pedigree registry asks for, and it applies to working sheepdogs rather than show dogs.

Six questions are worth asking a breeder of any of these breeds, in this order.

1. What is each parent's NHEJ1 genotype, in writing, on the testing laboratory's own report rather than in an advertisement.
2. Which laboratory ran it, because a kennel club records a result only from a laboratory on its own approved list.
3. Were the puppies examined by an ophthalmologist, and at what age, since an examination after about 8 weeks can miss an affected puppy.
4. Has any dog in the line had a coloboma, which is the finding the DNA test does not predict.
5. Is either parent merle, because a merle fundus makes the examination harder and the DNA result more important.
6. What else has been tested, since these breeds share MDR1 drug sensitivity and progressive retinal atrophy as well, and our guide to checking a dog breeder in India covers the paperwork.

Breeding: why removing every affected Rough Collie is not the advice

Breeding advice for collie eye anomaly depends entirely on how common the mutation already is in the breed, and in the Rough Collie it is so common that culling every affected dog would do more harm than the disease. It would not be possible to cull all affected dogs in a breed with a very high prevalence, since it would reduce the breeding population considerably and cause problems from inbreeding, as the Copenhagen review of this disease puts it.

That review sets out the whole logic in one paragraph, breed by breed. In a breed of very low prevalence, like the Border Collie at about 6 percent, it would be unproblematic to remove affected dogs from breeding. The Shetland Sheepdog sits at a roughly medium prevalence, so it is possible to require at least one parent to be unaffected without contributing to inbreeding. And in a breed where nearly every dog carries the mutation, the sensible target is to thin it out slowly while keeping the gene pool wide. The same review makes the uncomfortable companion point, that in Shetland Sheepdogs other conditions such as hip and elbow dysplasia and progressive retinal atrophy deserve a higher priority in breeding programmes than this one.

The Royal Kennel Club's own published list shows what that looks like in practice. Among the 244 Rough Collies with a recorded CEA/CH result, only 56 are Clear. A rule that every mating must involve a clear dog would restrict British Rough Collie breeding to fewer than a quarter of the tested population, in a breed the Royal Kennel Club already places in a conservation band on 345 registrations in 2025. That is the arithmetic behind the advice, and it is why this page does not tell a Rough Collie breeder to do what it would tell a Border Collie breeder.

The global data suggest the slow-thinning approach is not yet happening in the Collie. Across 33,834 dogs tested between 2005 and 2019, the effect of each passing year on the log odds of an affected or carrier genotype was five times smaller for collie eye anomaly than for progressive rod-cone degeneration, and the Collie itself showed rising frequencies of the affected genotype and the disease allele over the same period. The authors are careful about why, noting that the trends may reflect genetic drift rather than deliberate selection, and that the severity of a disease drives how hard breeders act, with retinal atrophy motivating more direct action than this condition.

There is a genuine argument on the other side and it deserves stating. The condition is non-progressive, the majority of affected dogs have minimal changes to vision, and choroidal hypoplasia by itself costs a dog nothing The Royal Kennel Club. Against that, dogs with mild signs can produce puppies with severe disease, so breeding two mildly affected dogs is not a safe bet on the next generation The Royal Kennel Club. The practical middle is the one the breed clubs have converged on, which is to know every dog's status, keep at least one parent clear where the breed's numbers allow it, and examine every litter.

Collie eye anomaly in India: the test nobody here sells, and the five clinics that can look

In India the DNA test is not sold by any laboratory this page could find, and the eye examination is reachable in five metros. Every affected breed is kept here, from the Rough Collie and the Border Collie to the Shetland Sheepdog and the Australian Shepherd, so the gap is real rather than academic.

Start with what Indian law says, because it says less than people assume. The Prevention of Cruelty to Animals (Dog Breeding and Marketing) Rules, 2017, notified as G.S.R. 496(E) on 23 May 2017, require at Second Schedule Part III rule 8 sub-clause (10) that dogs shall have been tested for genetic defects, including hip dysplasia, progressive retinal atrophy, night blindness, hypothyroidism, entropion, ectropion, overshot and undershot jaw, wry mouth, missing teeth and cryptorchidism. Collie eye anomaly is not named. The word including makes that list illustrative rather than closed, so an unnamed eye defect is not excluded by the rule, but a breeder has nothing specific to point to and an inspector has nothing specific to ask for.

No Indian laboratory found selling the test, and the check that would have caught one

Three Indian-facing laboratories were opened on 27 September 2026 and none sells a breed-specific collie eye anomaly or NHEJ1 test. The check is worth describing because a byte search cannot prove an absence on its own. DNA Labs India's product page names MDR1 three times, progressive retinal atrophy twice and hip dysplasia twice in its visible text, so the search does find disease names on that page; it names collie, choroidal, NHEJ1, CEA and coloboma zero times. Urban Animal's page and EasyDNA India's index name none of the five either EasyDNA India, Animal DNA Testing Services…. What is sold here is a broad panel with an unnamed variant list.

Every rupee figure above was divided by 84.38 and by 74 before it was published. One lands suspiciously. Rs 12,600 is USD 149.32 at 84.38, close enough to a round USD 149 that it may be a converted dollar price rather than an independently set Indian one, and it is reported here as one. The others do not fall on round dollar amounts: Rs 16,000 is USD 189.62 and Rs 4,500 is USD 53.33 at the same rate. Rs 4,200 comes out at USD 49.77, which is near USD 50 without being it, since USD 50 at 84.38 would be Rs 4,219 Urban Animal, DNA Test Kit for dogs. The same two Urban Animal prices were recorded unchanged on 23 September 2026 by our page on progressive retinal atrophy in dogs.

The arithmetic an Indian buyer should notice is that the named test is cheaper than most of the unnamed panels. USD 58 is about Rs 4,894 at 84.38, which sits below the Rs 16,000 panel and below both EasyDNA prices, and it buys one named variant in one named gene reportable to a registry rather than a list nobody will show you. International shipping in both directions is not in that figure and no laboratory on this page prices it. The sample is a cheek brush you take yourself, the laboratory mails the brushes, the dog must have eaten and drunk nothing for at least an hour, three brushes per dog are rubbed against the cheek for 15 seconds each, saliva and tongue swabs are refused, and puppies should be at least three weeks old. Our guide to importing a dog to India covers the paperwork if the dog itself is coming from abroad.

Where a veterinary ophthalmologist can actually examine a litter

The eye examination half is reachable, in five cities. The Eye Vet, founded by Dr Kasturi Bhadsavle MVSc, lists Indian clinics with addresses and telephone numbers in Chembur in Mumbai, Pimpri-Chinchwad in Pune, Sector 29 in Gurugram, Jubilee Hills in Hyderabad and Rajarajeshwari Nagar in Bengaluru, plus one in Dubai, and its named team runs to a dozen veterinarians. The Gurugram clinic answers to a delhi address, so Delhi NCR is covered. Chennai and Kolkata are not on the list, and neither is any city outside those five metropolitan areas. Worth noting as a caution about single-source lists, the practice's own home page names only four of its five Indian cities, omitting Bengaluru, while the clinics page lists all five with addresses.

What does not exist here is the scheme around the examination. India has no equivalent of the British Veterinary Association Eye Scheme, so there is no panel of appointed ophthalmologists, no Clinically Unaffected certificate, no litter screening service at a published fee and no national database a buyer can look a dog up in. There is no Indian equivalent of the Orthopedic Foundation for Animals register either. An Indian buyer is therefore asking a breeder for private records, and our guide to KCI registration explained sets out what a pedigree certificate does and does not say about health testing.

The published Indian evidence, and the search that found none

There is no published Indian clinical or genetic record of collie eye anomaly, and here is the search that establishes it. On 27 September 2026, with NCBI E-utilities returning HTTP 500 on everything, two India instruments were run on the Europe PMC REST service and both returned zero.

Those three controls are what make the zero mean something. The Labrador control returned 110 on-topic canine papers, so the field syntax was correct; the same query in PubMed bracket syntax returns 3 European haematology abstract books on the identical service, which is the trap this check exists to avoid. The affiliation control returned 1,523 canine papers with Indian affiliations, so Indian veterinary science is well represented in the index and simply has not published on this disease. And NHEJ1 with an Indian affiliation returns three papers, all of them human, which proves the gene name itself is live in the index.

The commercial data tell the same story from the other end. Of the 33,834 dogs in the largest CEA genotype dataset, 96.14 percent of samples came from North American and European countries, and the top five contributing countries, France, Germany, Switzerland, the United Kingdom and the United States, supplied 64.88 percent between them. India is not among them. The authors add a caveat that matters here, that another company began offering these tests in Asian and Pacific countries from 2006, so their 2006 to 2019 figures are unlikely to be fully representative for that part of the world. Read the Indian absence as unmeasured rather than as proof of a low frequency.

The nearest usable number to India is Thai. Two hundred and twenty-four dogs across five purebred breeds in Thailand gave a mutated NHEJ1 allele frequency of 83.3 percent in Rough Collies, 7.8 percent in Border Collies, 5.1 percent in Australian Shepherds and 2.8 percent in Shetland Sheepdogs, with the homozygous mutated genotype found only in Rough Collies at 66.7 percent and Shetland Sheepdogs at 1.1 percent. The study also carries a clean negative control in the same population: Thai Ridgeback Dogs were not affected by this mutation at all. A South Asian breeding population of imported collies is much more likely to resemble Bangkok than Zurich.

Living with a dog that has collie eye anomaly

A dog diagnosed with collie eye anomaly usually needs nothing done to it, because the defect is present at birth, does not progress and in most affected dogs costs no vision The Royal Kennel Club. There is no cure and no treatment for the condition itself, which sounds worse than it is, because in most dogs there is nothing to treat.

What changes is what you watch for. The primary lesion is static, so a dog that sees well at a year old is very unlikely to lose sight to choroidal hypoplasia later Online Mendelian Inheritance in Animals. The risk sits with the minority that have a coloboma, and the thing a coloboma can cause is a retinal detachment, which most often appears within the first year of life and is usually unilateral. Ask the ophthalmologist who examined your puppy whether a coloboma was seen and how big it was, and write the answer down, because it decides whether this is a finding or a follow-up.

Surgery exists for one narrow situation and only when it is caught early. Where a coloboma has caused a partial retinal detachment and the condition is detected early, laser surgery may be attempted to reattach the retina, and a coloboma may not lead to detachment at all, particularly if it remains stable during the dog's first year. This page names no drug and prints no dose, and that is deliberate. Anything your dog needs is decided by the veterinarian who has looked into that dog's eyes.

Four things are worth doing for an affected dog.

• Get the grade in writing at the first examination, including whether a coloboma is present, since the DNA test will never tell you this.
• Treat any sudden change in vision as same-day, because detachment and intraocular haemorrhage are the routes to blindness in this condition.
• Do not breed the dog on a guess, since mildly affected dogs can produce severely affected puppies.
• Check the rest of the breed's list, because these breeds also carry MDR1 drug sensitivity, progressive retinal atrophy and, in some lines, primary lens luxation.

If a dog does lose vision, the practical work is the same as for any blind dog and it is mostly about the house rather than the pharmacy. Dogs rely far more on smell and hearing than on sight and adapt to vision loss well enough that owners often do not notice it at all. Our page on progressive retinal atrophy in dogs carries the full checklist for living with a blind dog in an Indian flat, from leaving the furniture where it is to teaching voice commands early.

The wider point for an Indian owner of a herding breed is that this is one of the better items on the list. Compared with heat, which is the real constraint on a double-coated collie in an Indian summer, and compared with the MDR1 deletion, which can kill a dog on a routine visit, an eye defect that leaves most affected dogs seeing normally for life is a manageable problem Clark JA, Anderson H, Donner J. Test for it, examine the litter early, record what you find, and then get on with the dog.