Progressive retinal atrophy (PRA) in dogs: DNA tests, breeds and what blindness looks like

Signs of Progressive retinal atrophy

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.

• Hesitating or refusing to go out after dark
• Bumping into furniture in dim light but not in daylight
• Brighter than normal eye shine in a torch or phone flash
• Pupils that stay wide open even in bright light
• Staying unusually close to you on evening walks
• Missing a thrown ball it used to catch easily
• Clumsiness in a new house or after furniture is moved
• A cloudy look developing in both eyes together

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 10 breeds listed are the ones on this site that document it, not a ranking of risk.

What progressive retinal atrophy is, and how common it is

Progressive retinal atrophy in dogs is not one disease. It is an umbrella term for a group of inherited disorders in which the photoreceptor cells of the retina die early, and in almost every affected dog it ends in total blindness. The disease is painless on its own, both eyes go together, and the loss is gradual enough that most owners notice it late Cornell University College of Veterinary….

The retina works like the film in a camera, and it holds two kinds of photoreceptor. Rods handle night vision and motion. Cones handle day vision and colour. In true progressive retinal atrophy the rods degenerate before the cones do, which is why the dark goes first and daylight follows months or years later. A less common group of forms affects puppies whose rods or cones never develop properly, and those dogs go blind far faster.

Nobody has measured how common this is in the general dog population of any country, so every number below is a clinic caseload rather than a prevalence. Read each row for what it actually counted.

The Brazilian service is the one with a real denominator. Across 2,055 privately owned dogs seen between January 2014 and December 2018, 130 were diagnosed with progressive retinal atrophy, which is 6.33 percent, spread across 18 breeds plus 27 mixed-pedigree dogs. The authors say plainly that this is not an incidence figure for Brazil, because dogs presented to an ophthalmology service are not a random sample and healthy dogs are under-represented. A referral clinic in Chennai or Bengaluru would be no different.

How many breeds and how many mutations

The honest answer to 'which breeds get it' is that the list is long and still growing. More than 100 dog breeds have been described with some form of progressive retinal atrophy since the first report in Gordon Setters in 1911. Around 19 mutations have been identified that are associated with the disease in around 49 breeds, and for the majority of breeds the responsible mutation has still not been found. Those two sentences together are the shape of the whole problem: the disease is common enough to matter in dozens of breeds, and genetics has caught up with only about half of them.

One form dominates the commercial testing market. Progressive rod-cone degeneration, written prcd, is a late-onset recessive disease caused by a single change in the second codon of the PRCD gene, and the same mutation showed complete concordance with the disease in 18 different dog breeds and breed varieties when it was identified in 2006. The identical mutation was found in a human patient from Bangladesh with recessive retinitis pigmentosa, which is why this dog disease has a research industry behind it.

None of that arithmetic appears on the pages that currently rank for this query in India. Cornell's page runs 1,102 words and carries no percentage, no study and no number of dogs; VCA's runs 1,431 words and carries none either; Veterinary Partner's article body runs 1,189 words and carries none. All three are accurate and none is quantified, which is the gap this page is written into.

What it looks like at home, before anyone says the word blindness

The first sign of progressive retinal atrophy is almost always night blindness, and it shows as behaviour rather than as an eye that looks wrong. A dog that will not go down to the compound after dark, hesitates at the top of the stairs, or bumps into a chair in a room it has crossed a thousand times is describing its rods Cornell University College of Veterinary….

In the Indian series, almost every dog arrived with the same story. Of the 31 affected dogs, 27, or 87 percent, had difficulty avoiding objects in dim light, with dilated pupils, slow pupillary light responses and a characteristic shine from the eye. In Brazil, owners reported night blindness as the presenting sign in 93 of 130 dogs, 71.54 percent, and general vision impairment in only 6, 4.62 percent. Owners notice the dark before they notice the dog is going blind.

The eye shine is worth learning, because it is the one thing an owner can see. As the retina thins, more light bounces back off the reflective tapetum behind it, so the normal green-gold flash in a torch beam or a phone camera becomes brighter and the pupils sit wider than they should. Vets call the underlying change tapetal hyperreflectivity, and it was present in 78 of 141 affected eyes examined in Brazil, alongside retinal vascular attenuation in 120 of those eyes, 85.10 percent.

Then the dog adapts, and the adaptation is what hides the disease. Dogs memorise the layout of a home so well that vision changes go unnoticed until the dog is in a new place or the loss is nearly complete. This is why so many dogs are presented late, and why early cases are usually caught by accident when a primary vet refers a dog for cataracts. Seven changes are worth acting on.

• Refusing a dark room, a dark stairwell or the terrace at night, the single commonest first complaint.
• Staying close to you on an evening walk in a place the dog normally leads in.
• Brighter than usual eye shine in a torch beam or a phone flash, with pupils that stay wide in bright light.
• Bumping into door frames or furniture in a new house, or after the sofa was moved.
• Missing a thrown ball it used to catch, or failing to track a small object.
• A cloudy look developing in both eyes, which in this disease is usually a secondary cataract.
• Both eyes changing together, because progressive retinal atrophy is bilateral and symmetrical.

At what age progressive retinal atrophy starts

Progressive retinal atrophy usually starts in adulthood, not puppyhood, and the age depends on which form the dog has. The Kennel Club puts first signs of the prcd form at around three years of age or later, and says every puppy is born with normal vision. The rarest early forms can leave a puppy with reduced vision by twelve weeks and blind by one to two years.

The spread across forms is wide and it is documented form by form. In the earliest-onset dysplastic forms, puppies can have decreased vision by 12 weeks of age; some degenerative forms begin in early adulthood, some in middle age, and some as late as 9 to 11 years. The Poodle Club of America puts diagnosis of prcd-PRA in Toy and Miniature Poodles at around three years of age on an ophthalmologist's examination, and notes that some affected dogs keep useful vision for life while others are blind by mid-life. The Royal Kennel Club gives the average age of diagnosis for the rcd4 form as 10 years, with considerable variation inside a breed.

Owner-reported ages run later than the biology, and both Indian and Brazilian data show why. In Brazil, the mean age at which owners first noticed anything was 8.55 years, with a mean age at diagnosis of 9.06 years and a range from 3 months to 16 years. The authors add the caveat that the true onset was probably earlier, because subtle signs are easy to miss in a dog that is adapting well. In the Indian series, ages ran from 2 to 13 years, with 8 dogs of 31 in the 1 to 5 band, 15 in the 6 to 10 band and the remaining 8 in the 11 to 15 band.

The four-year-old dog with a normal-looking eye

One finding from Anand deserves a section of its own, because it changes what a normal eye examination proves. No fundus changes at all were found in any of the three examined dogs under four years of age, even though their owners had already complained that the dogs showed fear in the dark, and four dogs with apparent fear of the dark had reduced electroretinogram amplitudes with a normal-looking fundus. Reduced rod function shows on an electroretinogram before anything shows through an ophthalmoscope. A young dog that is frightened of the dark and has a clean eye examination has not been cleared.

Age of onset also varies within a single family, not only between breeds. The Indian authors traced the pedigrees of their affected dogs and found that certain families were affected earlier and progressed faster than others, and they suggest inbreeding as a possible explanation. The Poodle Club makes the same point from the other side, that age of onset and rate of progression vary among breeds, within a breed and within a single litter. A littermate's age is a hint, never a schedule.

Why a clear DNA test result does not clear a dog of PRA

A PRA DNA test checks one variant in one gene, so a clear result clears a dog of that variant and nothing else. The Royal Kennel Club sells at least six separate progressive retinal atrophy tests, each validated for a different, short list of breeds, and its own product pages say that research has not yet validated a test for breeds outside that list The Royal Kennel Club The Royal Kennel Club.

The number behind that sentence is the one no ranking page prints. When DNA from 231 dogs already diagnosed with progressive retinal atrophy, across 36 breeds, was screened against all 17 mutations then known, 129 dogs were homozygous for one of them, 29 were carriers, and 73 were homozygous for the normal allele at every single locus tested. Worse for a buyer reading a certificate, cases with no known mutation turned up in 15 of the 21 breeds where a PRA mutation is known to segregate. Cornell says the same thing in one sentence, that a dog could still develop the disease even if it tests negative for the known variants. The denominator is what makes it actionable.

Two dogs in that screen carried a mutation from the wrong breed entirely. One Chinese Crested was homozygous for the RCD3 mutation usually found in the Cardigan Welsh Corgi, and one Standard Poodle was homozygous for the RCD4 mutation reported in Gordon and Irish Setters. Variants cross breed lines, which is why OMIA attaches a standing warning to its own variant tables, that inclusion of a variant does not mean it should be used for DNA testing and that anyone using one should examine the evidence critically, especially in breeds other than the one the variant was first described in.

The registries themselves do not agree on the labels

Two disagreements sit in plain sight, and a buyer should know about both. First, OMIA records the Golden Retriever among the 33 breeds documented for the PRCD variant, while the Royal Kennel Club does not list the Golden Retriever among the 18 breeds its prcd-PRA test is validated for The Royal Kennel Club. Second, the Royal Kennel Club sells a test headed PRA3 whose own body text calls it the PRA cord1 test, while the Orthopedic Foundation for Animals attributes cord1 to the RPGRIP1 gene across more than 25 breeds, and the PRA3 locus in Tibetan Spaniels and Tibetan Terriers was mapped to a different gene, FAM161A Downs LM, Mellersh CS. An intronic SINE…. Ask the laboratory which gene and which variant it is actually running, and ask for the report rather than a summary.

A third disagreement sits inside the result itself, and the laboratory selling the test is the one that publishes it. UC Davis grades PRCD/PRCD as expected to develop the disease, then records that a few dogs between 10 and 13 years of age have been identified that were clinically normal while carrying two copies of the mutation, in the American Eskimo, American Cocker Spaniel, English Cocker Spaniel and Toy Poodle, and postulates unidentified genetic modifiers. That cuts both ways for a buyer. An affected result is a strong prediction rather than a certainty, and a clear result is narrower than it looks, because the same page states that a normal PRCD result, N/N or N/PRCD, does not exclude the possibility that a dog may carry or be affected by another PRA mutation. The breed list is that caution in another form. Where the Royal Kennel Club validates its prcd test for 18 breeds, UC Davis offers its own for many breeds including but not limited to 46 named entries, a list that does include the Golden Retriever the Kennel Club leaves off, and that names the Poodle twice and the Labradoodle three times The Royal Kennel Club. An open list is a wider offer, not a wider validation.

Testing does work, and the size of the effect is the reason to bother. The PRCD variant frequency in Labrador Retrievers fell from 0.078 before the variant was published to 0.003 eight to ten years afterwards, a 96.5 percent decline, and in English Cocker Spaniels from 0.126 to 0.006, a 95.1 percent decline, both attributed to testing. It is not complete: in a dataset of 86,667 dogs genotyped for the prcd variant, dogs homozygous for it were still being found fourteen years after the test became available. Where nobody tests, the frequency stays where it was, and 220 English Cocker Spaniels sampled in Brazil carried the variant at 25.5 percent overall, 14.9 percent among registered dogs and 41 percent among dogs with no registration history.

Two more named tests show how narrow a validated breed list can be. The Kennel Club's early onset PRA test covers one mutation in the Portuguese Water Dog, where the disease usually appears at about two to three years of age rather than in middle life. The Orthopedic Foundation for Animals sells a NECAP1 test for a late-onset form first recognised in Giant Schnauzers at about four years old, and the breeds it lists for that variant are the Giant Schnauzer, Giant Spitz, Medium Spitz, Miniature Spitz and Dachshund. A Spitz owner reading that list should note that it is a different variant again from the one found in the Indian series.

Cats get their own version of all this, and it is worth a paragraph because Indian breeders of Persians and Siamese meet it. The feline equivalent is rdAc, a splice variant in intron 50 of the CEP290 gene, inherited recessively and first reported in 2007, with a DNA test built on it. A survey of 41 cat breeds and 846 individuals found the rdAc allele in 16 of 43 breeds, 37 percent, with an allele frequency of about 33 percent in North American and European Siamese populations, and it was first described in the Abyssinian and Somali. The same rule holds as in dogs: rdAc is one variant, and separate feline retinal atrophies linked to the KIF3B and CRX genes are recorded alongside it.

How progressive retinal atrophy is diagnosed

To diagnose progressive retinal atrophy, a veterinarian examines the back of both eyes with a lens after dilating the pupils, and refers to a veterinary ophthalmologist to confirm it. Electroretinography, which measures the electrical response of the retina to flashes of light, is the test that confirms the diagnosis when the fundus is inconclusive or when cataracts block the view.

The fundus changes follow a recognisable order and a vet reads them as a stage. Early signs are a granular tapetum and a slight taper of the peripheral vessels; later come tapetal hyperreflectivity, attenuation of the retinal vessels and atrophy of the optic disc. In the Brazilian series, of 141 affected eyes with fundus changes, vascular attenuation was found in 120, 85.10 percent, and hyperreflectivity in 78, 55.31 percent, with optic disc atrophy and tapetal granulation recorded in 2 eyes each. The Indian series describes the same picture in its own words, with an atrophied optic disc showing scalloped or indistinct borders and, at the advanced stage, only ghost vessels visible.

Electroretinography is where a suspicion becomes a diagnosis, and it is also where a suspicion gets withdrawn. Of 30 dogs taken to electroretinography in Brazil on clinical suspicion, 23, or 76.66 percent, were confirmed by a subnormal or absent response, and 7, or 23.33 percent, had a normal response, which excluded the diagnosis in those dogs. Roughly one dog in four that looked like a case was not one. In the Indian series, all 13 dogs taken to electroretinography returned non-recordable flatline responses in both the dark-adapted and light-adapted states. Five things make up a full work-up.

1. History first, including whether the dog avoids the dark and when the owner first noticed it.
2. Vision testing, which in the Brazilian protocol meant an obstacle course run in both bright and dim light, plus menace response and tracking a falling cotton ball.
3. Pupil and reflex checks, since pupillary light reflexes stay partly preserved while the menace response and tracking fail first.
4. Dilated examination of the fundus of both eyes with an indirect ophthalmoscope and a lens.
5. Electroretinography by an ophthalmologist, which in most practices needs sedation and is the confirmatory test.

Why the diagnosis has to be separated from other causes of blindness

Separating this from other blinding disease is the reason to go to a vet rather than to a search engine. Other causes of blindness can be painful or carry systemic effects, which progressive retinal atrophy does not Veterinary Partner. Sudden acquired retinal degeneration syndrome, retinal detachment, optic neuritis, posterior uveitis and glaucoma all sit in the differential, and the Brazilian team excluded every one of them before counting a case. Electroretinography is exactly the tool that tells those apart.

The cataract question runs in both directions and it changes what surgery can do. Secondary cataracts were diagnosed in 72 of the 130 affected dogs in Brazil, 55.38 percent, and 61 of those 72 were bilateral, a figure the authors note is higher than an earlier study's 44 percent. A cataract can also hide the retina from view, which is when electroretinography becomes essential rather than optional. It also decides the surgical answer: because of the retinal disease underneath, no dog with progressive retinal atrophy is a good candidate for cataract removal surgery.

Treatment for progressive retinal atrophy, in general terms

There is no cure for progressive retinal atrophy and no commercially available treatment that stops it. Management is about safety and adaptation rather than medicine, and an affected dog usually needs no medication at all, because the disease itself is painless.

Three claims circulate about treatment and each one needs its own answer. Gene therapy is real but not available. The first cures of dogs with this disease have happened with gene therapy in experimental research settings, where an abnormal gene is surgically replaced with a normal one, and that work became a stepping stone to human treatment; it is not commercially available for pets and is unlikely to be soon. Antioxidants are modest and honest about it. Some photoreceptor death is caused by oxidative stress, so oral antioxidant therapy may slow the rate of vision loss, but it will not stop a dog going blind, and Cornell puts it as specific antioxidant supplements possibly improving retinal function and delaying cataract formation. Nothing else has evidence behind it. The registries are blunt: there is no treatment for progressive retinal atrophy, of which several genetically distinct forms are recognised, each caused by a different mutation in a specific gene.

This page names no drug and prints no dose, and that is deliberate. Supplements, eye drops for a secondary cataract and anything else your dog may need are decided by the veterinarian who has looked into that dog's eyes. What an owner can usefully know is that a dog with this disease and no cataract generally needs nothing, and a dog that develops a secondary cataract may need long-term anti-inflammatory eye medication and regular examinations to watch for lens-induced uveitis and glaucoma, both of which can be painful.

What happens to the dog over time

The outcome is honest and it is not as bleak as it sounds. Nearly all dogs with this disease eventually go completely blind, and all genetic variations of it progress to complete blindness. The speed varies by form and by individual: typically an owner notices night blindness first, and then over a few months to one or two years the ability to see in daylight is lost as well. OMIA records that for the prcd form, most dogs become completely blind in one to two years, and that there is no known cure or treatment for slowing its progression.

Because the loss is slow, the dog gets time that a suddenly blind dog does not. Dogs acclimatise well precisely because most cases progress slowly, and affected dogs go on to live otherwise healthy lives. That sentence is the reason the last section of this page exists.

Progressive retinal atrophy in India, and what a dog DNA test costs here

Indian law already names this disease. The Prevention of Cruelty to Animals (Dog Breeding and Marketing) Rules, 2017 require that breeding dogs shall have been tested for genetic defects, including hip dysplasia, progressive retinal atrophy, night blindness, hypothyroidism, entropion, ectropion and a list of dental and reproductive faults. The rule is Second Schedule, Part III, rule 8, sub-clause (10), notified as G.S.R. 496(E) on 23 May 2017.

What the rule does not do is say which test, to which standard, read by whom. No Indian laboratory that this page could find sells a breed-specific progressive retinal atrophy variant test the way the Royal Kennel Club or the Orthopedic Foundation for Animals does. What is sold in India is a broad panel, and the three Indian-facing laboratories opened on 23 September 2026 price it between Rs 4,200 and Rs 16,000.

Every rupee figure above was divided by 84.38 and by 74 before publication. Only one lands suspiciously: Rs 12,600, printed on the page as ₹12,600, is USD 149.3 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 EasyDNA India, Animal DNA Testing Services…. The rest do not fall on round dollar amounts DNA Labs India, Dog DNA Test.

One Indian listing deserves a closer read before anyone spends Rs 16,000 on it. The DNA Labs India page names progressive retinal atrophy and hip dysplasia as example conditions its panel detects, prints CPT code 81479 and ICD code Z13.89, which are human medical billing codes, and records its medical reviewer as a consultant physician with an MBBS and an MD in General Medicine rather than a veterinarian. Hip dysplasia in dogs is not screened by a DNA test anywhere in the world; it is graded from radiographs under a named scheme, as our page on hip dysplasia in dogs sets out. Ask any Indian laboratory which variant, which gene and which breeds before you pay.

The only Indian clinical record, and why it matters more than the price

One published Indian series exists and this page publishes its denominator. Over two years at Anand Agricultural University in Gujarat, 340 dogs were given a detailed ophthalmic examination and 31 were diagnosed with progressive retinal atrophy, 18 male and 13 female, aged 2 to 13 years. A PubMed search for progressive retinal atrophy in dogs scoped to India returns exactly one record, and this is it; the same search without the India term returns 167. That is the whole Indian clinical literature on this disease.

Its breed table is the finding that should change what an Indian buyer asks for. The largest group by far was the Spitz, 18 of 31 dogs, 58 percent, followed by the Labrador Retriever at 8, 26 percent, the Cocker Spaniel at 2, and one each of Canadian Pointer, Lhasa Apso and German Shepherd. When the team sequenced its cases, the PRCD mutation was linked to the affected Labrador Retrievers and Cocker Spaniels, but mutations in PRCD, XLPRA1 and 2, CNGB and GALK1 were not linked to the affected Spitz dogs at all; five mutations in the Phosducin gene were. India's commonest affected breed in the only Indian record has a form that the commercial prcd test does not detect. If you own an Indian Spitz or a Pomeranian, a prcd-clear certificate says nothing about this.

Geography matters to how much weight that series carries. The work was done at Anand in Gujarat, with a consulting vitreoretinal surgeon in Vadodara. A PubMed search run on 23 September 2026 for progressive retinal atrophy in dogs scoped to India returns that one record and nothing else, so there is no published series from Maharashtra, from Karnataka, from Tamil Nadu or from Delhi. An owner in Mumbai, Pune, Hyderabad, Chennai, Bengaluru, Kolkata or Delhi is reading evidence gathered in one state, and an owner who wants a named-variant test is buying it from a laboratory outside India, in pounds or dollars, and posting a cheek swab abroad.

The dollar half of that sentence has a number, and this page now prints it. UC Davis charges USD 58 for the single prcd test on one animal, USD 27 if it is added to another health test on the same animal, with 5 dollars off for three or more dogs, and a turnaround of at least 15 business days. At the 84.38 rate used elsewhere on this page, USD 58 is about Rs 4,900, which lands just above the two cheapest Indian panels in the table and buys something different from them: one named variant in one named gene, validated and reportable to the Orthopedic Foundation for Animals, rather than an unnamed panel. International shipping in both directions is not in that figure and no laboratory on this page prices it. The practical detail matters too, because the sample is posted by you: the laboratory mails its own cytology brushes, asks that the dog has eaten and drunk nothing for at least an hour, wants three brushes per dog rubbed against the cheek for 15 seconds each, refuses saliva and tongue swabs, and will not test a puppy under three weeks old.

Screening, breeding and what to ask before you buy a puppy

Screening for progressive retinal atrophy has two halves that do different jobs. A DNA test tells you about one variant before a dog shows anything, and a clinical eye examination by an ophthalmologist tells you what that dog's retina looks like today, including forms no test exists for Veterinary Partner.

The clinical half is a real scheme with a real price and a real certificate. Under the British Veterinary Association Canine Health Schemes Eye Scheme, a routine eye examination for one dog starts from GBP 72.00 including VAT, litter screening for puppies aged 5 to 12 weeks starts from the same figure, and results are recorded as Clinically Unaffected or Clinically Affected on a certificate issued at the appointment. The scheme is open to pedigree and crossbreed dogs alike and does not require kennel club registration, and over 14,000 dogs were examined under it in 2025. India has no equivalent national eye scheme, so an Indian buyer is asking a breeder for private records rather than for a scheme certificate.

In North America the equivalent is the eye registry, and progressive retinal atrophy is one of the findings that makes a dog ineligible under it: the Orthopedic Foundation for Animals lists it as ineligible code IE10 in the Companion Animal Eye Registry. Breed clubs then stack the two halves. The Poodle Club of America requires a prcd DNA result from an approved laboratory plus a yearly eye examination by a board-certified ophthalmologist for Toy and Miniature Poodles, while Standard Poodles get the yearly eye examination without the prcd requirement. Six questions are worth asking a breeder, whatever the breed.

1. Which named PRA test did each parent have, by its name, prcd or rcd4 or cord1 or another.
2. Which laboratory ran it, since a kennel club records a result only from a laboratory on its own approved list.
3. Is that test validated for this breed, because a result in an unlisted breed is of uncertain value.
4. Has either parent had a clinical eye examination, which catches forms with no DNA test.
5. Show me the laboratory report, not a summary line in an advertisement.
6. Has any relative gone blind, since the Indian series traced early, fast-progressing disease through whole families.

What breeding advice actually is, once a result comes back

Most forms of this disease are recessive, which makes the breeding arithmetic simple and the wording on certificates important. A dog needs two copies of the mutation to be affected, one from each parent; a dog with one copy is a carrier that will not develop the disease but can pass it on. Carriers and affected dogs can be bred safely only to DNA-tested clear dogs, and about half the puppies from a litter with one carrier parent will themselves be carriers, so any of those puppies used for breeding should be tested in turn. A few forms are dominant or X-linked, which changes the arithmetic.

The part breeders skip is the sentence the test's own publisher writes. Because at least one further form of PRA exists in the Gordon Setter and its mutation is unidentified, the Orthopedic Foundation for Animals states that it cannot guarantee any dog will not develop PRA even if that dog is clear of the rcd4 mutation, and it advises that all breeding dogs have their eyes clinically examined by an ophthalmologist before breeding and throughout life. Affected dogs should not be bred, and for a disease with no treatment that is the only prevention there is. Our guide to checking a dog breeder in India covers the paperwork side, and KCI registration explained covers what a Kennel Club of India certificate does and does not mean.

Living with a dog that is going blind, and the question owners ask first

A dog going blind from progressive retinal atrophy does not need to be put to sleep for it. The disease is painless, it does not affect the rest of the body the way other causes of blindness can, and most affected dogs remain happy and otherwise healthy as long as they are kept an active part of the household. Cornell puts it the same way, that most dogs adapt well and continue to have a good quality of life.

The reason a blind dog copes so well is that sight was never its first sense. Vision ranks third for a dog, behind smell and then hearing, and because this disease progresses slowly the dog has time to memorise its environment and lean harder on the other two. Dogs get so good at it that it can be impossible to tell when they finally lose the last of their sight. The blunt version from the same source is worth quoting: living with a dog who has this disease is not difficult, and your reaction to the loss of vision is likely to be much stronger than the dog's.

What the dog needs is a stable map and a bit of planning. Eight changes cover most of it in an Indian flat or house.

• Leave the furniture where it is, because the dog is navigating from memory.
• Keep food and water bowls in the same place, every day.
• Teach the children not to leave toys and clothes in the dog's usual path.
• Gate the stairs, using the same safety gates sold for babies.
• Fence off any water, since swimming pools and open tanks are a hazard.
• Add lighting indoors and outdoors early, while only night vision is affected.
• Switch from hand signals to voice commands now, and teach wait, curb, down and stop while the dog can still see.
• Use a leash and harness in unfamiliar places, including on travel.

Two habits make a bigger difference than any equipment. Talk to the dog, and avoid startling it awake by calling it or tapping a foot nearby before you touch it. If there is a second dog in the house, expect it to take over as a guide, because the blind dog will follow the sighted one and read its cues. A ball with a bell inside, introduced while the dog can still see it, keeps a game alive that would otherwise end.

Age changes the frame rather than the answer. Most dogs in the Brazilian series were middle-aged or older when they were diagnosed, with a mean of 9.06 years, which means many owners are managing blindness alongside the ordinary business of an older dog, and our guide to senior dog care in India covers that side. The practical difference is simply that a dog losing its sight in a stable home, on known routes, with the same people, does well, and a dog moved between houses does less well. Keep the map still.