The name is a trap. “Collie Eye Anomaly” makes it sound like a problem confined to Rough Collies, Border Collies and their close relatives — and that assumption causes real harm, because owners of other affected breeds often have no idea their dog should be tested at all. The exact same genetic mutation that causes CEA in Collies appears in a surprising list of unrelated-looking breeds. If you own one of them, this is the article that tells you why, and what to do about it.

Why one mutation shows up in unrelated breeds

CEA is caused by a specific deletion in the NHEJ1 gene on chromosome 37 — the same single mutation across every affected breed. The reason it crops up in dogs as different as a herding Collie and a duck-hunting retriever is shared ancestry and founder effect: the mutation is old, predating the modern split into distinct breeds, and it travelled into multiple lineages before anyone could test for it. Once a popular sire carries it, it spreads silently through a closed gene pool, because carriers are completely healthy and most affected dogs show no obvious signs. You can read the full biology in our explainer on the NHEJ1 gene mutation.

The clinical condition itself is identical regardless of breed: choroidal hypoplasia as the mild, usually harmless core finding, with a minority of dogs developing colobomas and, more rarely, retinal detachment. What differs between breeds is the frequency of the mutation and how much attention the breed community pays to testing.

The lesser-known at-risk breeds

Beyond the well-publicised Collies, Shelties and Australian Shepherds, these breeds carry the NHEJ1 mutation at frequencies worth taking seriously:

Lancashire Heeler

This small British herding breed has one of the highest reported CEA carrier frequencies of any breed — historically a large share of the population. Because the Lancashire Heeler went through a severe genetic bottleneck (it was nearly extinct mid-20th century), inherited conditions concentrated in the surviving gene pool. Responsible breeders now DNA-test routinely, and given the high carrier rate, testing both parents before any mating is essential rather than optional.

Nova Scotia Duck Tolling Retriever

The “Toller” surprises people because it is a retriever, not a herding dog — yet CEA is firmly established in the breed. Carrier frequencies have been reported in the meaningful single-to-double-digit-percent range depending on the population. Toller breed clubs in several countries recommend CEA DNA testing, and prospective owners should ask for it specifically, since a buyer expecting a “retriever, not a Collie” may not think to.

Boykin Spaniel

This American flushing and retrieving spaniel also carries the NHEJ1 mutation, and CEA appears on the recommended health-testing panels published by the breed’s parent club alongside its other known conditions. As with the Toller, the spaniel appearance means owners rarely associate the dog with “Collie” eye disease — which is exactly why it gets missed.

Sighthound and other carriers

The mutation has also been documented in the longhaired whippet and the silken windhound (a relatively new breed developed partly from longhaired whippets, which is how the gene travelled in), and in the Japanese Hokkaido. These are small populations, but in small gene pools a single carrier sire has an outsized effect, so testing matters proportionally more, not less.

Why “rarer” breeds need testing more, not less

There is a dangerous logic that says, “It’s rare in my breed, so I don’t need to test.” The opposite is closer to the truth. The well-known affected breeds have large, well-organised testing cultures that have driven carrier rates down over decades. The lesser-publicised breeds often have:

• Smaller gene pools, where each carrier represents a larger fraction of available breeding stock.
• Less awareness, so fewer dogs are tested and carriers stay hidden.
• Owners who don’t expect it, meaning puppies are bought without anyone asking for clearances.

That combination lets the mutation persist quietly. The same carrier-to-carrier breeding math that applies to Collies applies identically here: pair two untested carriers and on average one in four puppies is affected.

What owners of these breeds should actually do

The good news is that managing CEA in any of these breeds is straightforward, because the tools are the same as for Collies:

1. DNA test before breeding. A single, inexpensive, highly reliable test reports each dog as clear, carrier, or affected for the NHEJ1 deletion. Make sure at least one parent of any litter is clear — that alone guarantees no affected puppies.
2. Have litters eye-examined at 6–8 weeks by a veterinary ophthalmologist, the same critical window used in Collies, to catch structural findings the DNA test doesn’t grade.
3. Buyers: ask for both. If you are getting a Toller, Boykin, Lancashire Heeler, or any of the above, request the parents’ DNA results and the litter eye certificate exactly as you would for a Collie. Don’t let the breed’s appearance lull you.
4. Don’t remove carriers wholesale. In small gene pools, culling every carrier would devastate genetic diversity. The correct strategy is to breed carriers to clears and select away over generations — preserving the dog’s other qualities while never producing an affected pup.

The bottom line

Collie Eye Anomaly is a misnomer that has let the mutation hide in plain sight across breeds that don’t look anything like a Collie. If you own a Lancashire Heeler, a Toller, a Boykin Spaniel, a longhaired whippet, a silken windhound or a Hokkaido, the same NHEJ1 mutation is in your breed’s gene pool — and the same simple DNA test and 6–8 week eye exam protect against it. For the wider context on how the condition behaves across populations, the ophthalmology section brings together the clinical detail behind every breed mentioned here.