Blue Eyes and Deafness

[Journey]

Can someone point me to the study of over 4000 Border Collies in the UK that showed a statistical correlation between blue eyes and deafness?

 

I found the ACK parent club promoting this as cause to change their standard wrt acceptable eye color.

[Maxi]

The scientific paper you mention is I think this one, but the full article is behind a paywall.

 

Luisa De Risio, Tom Lewis, Julia Freeman, Alberta de Stefani, Lara Matiasek, Sarah Blott.

Prevalence, heritability and genetic correlations of congenital sensorineural deafness and pigmentation phenotypes in the Border Collie

The Veterinary Journal June 2011, Vol.188(3):286–290, doi:10.1016/j.tvjl.2010.05.012

 

Abstract

The objectives of this study were to estimate prevalence, heritability and genetic correlations of congenital sensorineural deafness (CSD) and pigmentation phenotypes in the Border Collie. Entire litters of Border Collies that presented to the Animal Health Trust (1994–2008) for assessment of hearing status by brain stem auditory evoked response (BAER) at 4–10 weeks of age were included. Heritability and genetic correlations were estimated using residual maximum likelihood (REML). Of 4143 puppies that met the inclusion criteria, 97.6% had normal hearing status, 2.0% were unilaterally deaf and 0.4% were bilaterally deaf. Heritability of deafness as a trichotomous trait (normal/unilaterally deaf/bilaterally deaf) was estimated at 0.42 using multivariate analysis. Genetic correlations of deafness with iris colour and merle coat colour were 0.58 and 0.26, respectively. These results indicate that there is a significant genetic effect on CSD in Border Collies and that some of the genes determining deafness also influence pigmentation phenotypes.

 

There is also this older paper with 2,500 BCs (again most of the article is behind a paywall).

 

Platt S, Freeman J, di Stefani A, Wieczorek L, Henley W.

Prevalence of unilateral and bilateral deafness in border collies and association with phenotype.

J Vet Intern Med. 2006 Nov-Dec;20(6):1355-62.

 

BACKGROUND: Congenital sensorineural deafness (CSD) occurs in Border Collies, but

its prevalence and inheritance are unknown. This study estimated the prevalence

of CSD in Border Collies and investigated its association with phenotypic

attributes linked to the merle gene, including coat pigmentation and iris color.

HYPOTHESIS: Deafness in Border Collies is associated with pigmentation patterns

linked to the merle gene.

ANIMALS: A total of 2597 Border Collies from the United Kingdom.

METHODS: A retrospective study of Border Collies tested, during 1994-2002, by

using brainstem auditory evoked responses. Associations between deafness and

phenotypic attributes were assessed by using generalized logistic regression.

RESULTS: The prevalence of CSD in puppies was estimated as 2.8%. The

corresponding rates of unilateral and bilateral CSD were 2.3 and 0.5%,

respectively. Adjustment for clustering of hearing status by litter reduced the

overall prevalence estimate to 1.6%. There was no association between CSD and sex

(P = .2). Deaf Border Collies had higher rates of merle coat pigmentation, blue

iris pigment, and excess white on the head than normal hearing Border Collies

(all P < .001). The odds of deafness were increased by a factor of 14 for Border

Collies with deaf dams, relative to the odds for dogs with normal dams (P =

.007), after adjustment for phenotypic attributes.

CONCLUSIONS AND CLINICAL IMPORTANCE: Associations between CSD and pigmentation

patterns linked to the merle gene were demonstrated for Border Collies. Evidence

for an inherited component to CSD in Border Collies supports selective breeding

from only tested and normal parents to reduce the prevalence of this disease.

[Sue R]

Thank you for those interesting abstracts!

[Journey]

Thank you!

[KnottyClarence]

I have a unilaterally deaf split face (black ear doesn't work per Baer test). Both parents go back to Arthur Allen's Spot (dad is now 15, mom is deceased, the split faced is 8 1/2, tested at 4 months). Dad's father threw one blue eye in several litters. Always wondered about those blue eyes.

[GentleLake]

Thanks for finding and posting these.

 

Genetic correlations of deafness with iris colour and merle coat colour were 0.58 and 0.26, respectively. These results indicate that there is a significant genetic effect on CSD in Border Collies and that some of the genes determining deafness also influence pigmentation phenotypes.

 

I find it interesting that the increased correlations for deafness in blue eyed dogs and merles, though pretty small for both, show twice as high a correlation for deafness in blue eyed dogs over merles.

 

Yet I see far more references (and warnings) made by people about the dangers of producing deaf dogs in merles than in blue eyed dogs. Fascinating.

[Maxi]

Thanks for finding and posting these.

 

 

I find it interesting that the increased correlations for deafness in blue eyed dogs and merles, though pretty small for both, show twice as high a correlation for deafness in blue eyed dogs over merles.

 

Yet I see far more references (and warnings) made my people about the dangers of producing deaf dogs in merles than in blue eyed dogs. Fascinating.

 

 

GentleLake, I completely agree with your comment... in fact when I learnt stats (a long time ago), the rule of thumb I had for correlative studies was that a correlation of less than 0.3 (where the merle results reside) meant that there was very weak/no correlation while that of 0.5-0.7 (where the blue eye results reside) means that there is a 'moderate' correlation

But it is always possible that my notion of what counts as a "significant correlation" may not be true for the specific statistical test that the authors used.

 

Interestingly The second (earlier) paper by Platt et al ..presents the data as "p values" rather than correlations, and here the results presented for deafness in merles and blues eyes both seem to be significant

 

It may also be worth bearing in mind that the study by Luisa De Risio et al. specifically states that they only considered pups under 10 weeks of age. The 2nd study (by Platt et al) discusses "Congenital sensorineural deafness" in "border collies" but does not give a distinct age range for the dogs included in the study.

 

So I wonder whether "merle deafness" (if there is such a thing) tends to manifest after 10 weeks of age (ETA ..despite the fact it is often labelled 'congenital' (meaning "present from birth")).

 

HOWEVER (and importantly) My thoughts on this work should be taken with a pinch of salt because it is impossible to comment in detail on any scientific study based on just an abstract of the work...so if there is anyone out there who has read both papers in full, perhaps they would like to give an opinion on the 2 studies..

[B Point]

This new pup we have has blue eyes. I'd like to read those studies as well.

[Mark Billadeau]

Prevalence of unilateral and bilateral deafness in border collies and association with phenotype

[GentleLake]

From the above article:

"...the generalized logistic regression results emphasize the high risk of deafness for Border Collies with a merle coat (prevalence of 7.9%, upper 95% confidence limit of 13.3%), blue eyes (prevalence of 9.3% for 1 blue eye and 23.9% for 2 blue eyes; corresponding upper 95% confidence limits of 15.7% and 40.0%, respectively) or excess white on the head (prevalence of 36.3%, upper 95% confidence limit of 55.3%)."

 

So, if I'm reading this correctly (someone please correct me if I'm not), the risk of deafness in merles is still lower than it is for dogs with even only one blue eye, and this increases for 2 blue eyes and again for dogs with excessive white on the head. (I'm not suggesting that it's not still a risk, just that the greater prejudice doesn't appear to be supported by the numbers.)

 

Does anyone know whether merle dogs with brown eyes were sorted out from those with blue eyes in this analysis? IOW, would a blue eyed merle be at higher risk of deafness than a brown eyed merle?

 

Which, again, begs the question why I hear far more negative comments about (i.e. prejudice against) merle dogs in this regard than I do about blue eyed or white headed dogs.

 

Can someone please explain to me the difference between the original percentages and the higher "upper confidence" range? Thanks.

[Mark Billadeau]

These papers measure correlation in observed traits they do not demonstrate causation (i.e. blue eyes, merle coat, excessive white on head => deafness). In this study there may have been merles with brown eyes; all the researchers were doing is making observations and assessing for correlations: x occurs a certain percentage of the time when y occurs.

 

RE: confidence interval

read this 10 Things to know about Confidence Intervals

 

The researchers sampled part of the total merle population and found that within their sample 7.9% of the merles were deaf. Due to sampling (how many of the total merle population was sampled, how many were found to be deaf, etc) statistics predict that if another sample of the total merle population was taken the percentage of deaf merles would be within 13.3%. Another way to look at this is that we can be 95% confident that the actual percentage of deaf merles in the total population falls within 13.3%.

 

 

The range of the confidence interval will be based upon the sample size; the smaller the sample the larger the confidence interval (less confident the sample is predictive of the total population).

 

Play with this Sample Size Calculator (find confidence interval). Assume the total population is 10,000. Enter in the measured percentage from the study, then change the sample size to see how the confidence interval changes. As the sample size approaches the total population the confidence interval decreases (the more accurately the measured percentage reflects what the true population percentage is).

[GentleLake]

Thanks, Mark.

[Sue R]

Interesting numbers. I had no idea that blue eyes might be associated with deafness. I knew that excess white on the head could be. And I didn't realize that merle could be associated, either, at least not to remember it! Those are some pretty high percentages.

[Maxi]

Thanks Mark for the link to full article of the earlier of the 2 studies.

 

As you say, the important thing to remember is that both papers are correlative studies IOW how often deafness is associated with dogs that either have blue eyes or are merles.

They definitely do NOT say that either blue eye colour or merle coats cause deafness.

 

The tables in the linked article to Platt et al (the 2006 paper) also divides the study into pups versus adults.. and my earlier musing about why the statistical significance of the results in the second larger 2011 study seems lower are obviously wrong...oh well.. I did say, they should be taken with a pinch of salt.

[Mark Billadeau]

I did look for free open access to the newer study and could not find one.

 

Association could occur if key merle(s) (i.e. popular sires) also happen to carry the gene(s) for congenital deafness; due to the small total popluation of merles this could influence the prevalence of deaf merles without out merle causing deafness.

 

[Sue R]

Thank you, Mark, for helping to explain things. There is a reason I did not do well in genetics class...

[Maxi]

Which, again, begs the question why I hear far more negative comments about (i.e. prejudice against) merle dogs in this regard than I do about blue eyed or white headed dogs.

 

Perhaps this paper has the answer to your query. It looking at the prevalance of deafness in dogs heterozygous or homozygous for the merle allele.

In the introduction of this paper, the authors state

"...Relatively few studies of the merle phenotype have been published. Among these is a study from a research colony of Dachshunds (Tekels in German) maintained in Hanover, Germany involving auditory function,with measurements taken from 38 dogs. This study reported a deafness prevalence (unilateral or bilateral) of 54.6% in double merles and 36.8% in single merles. The findings in this study were limited to a small established population of 1 breed and have, unfortunately, been extrapolated to all breeds having the merle allele….

The research described in the present report and records of many breeders suggest that the deleterious effects of the merle allele may have been overstated in some publications. In the present study, we used a population of both related and unrelated merle-genotyped dogs to assess the prevalence of deafness in heterozygous and homozygous merles, and describe differences among breeds."

(my emphasis)

 

 

[Journey]

Thank you Mark. Lots to understand!

[B Point]

Silly and perhaps obvious question but,, Solid blue and white is a dilution of black, but with blue eyes does it make the dog a merle? I would think, yes...right?

[Eileen Stein]

Which, again, begs the question why I hear far more negative comments about (i.e. prejudice against) merle dogs in this regard than I do about blue eyed or white headed dogs.

I don't think the risk of deafness in merles (Mm) has been overstated or that there is more prejudice against merle dogs in this regard than blue-eyed or white-headed dogs. I think the difference is that with merles, you know (with rare exceptions) whether the breeding pair you've chosen can produce merles or not. You generally don't know whether the breeding pair you've chosen will or will not produce blue eyes or white heads, because reproduction of those traits is not very predictable. So there is more scope for disapproval of the deliberate choice to produce merles.

[Eileen Stein]

Silly and perhaps obvious question but,, Solid blue and white is a dilution of black, but with blue eyes does it make the dog a merle? I would think, yes...right?

 

No, blue eyes are often found in dogs who are not merles, whether they have dilute coat color or not.

[B Point]

Cool, thanks Eileen.

[GentleLake]

Silly and perhaps obvious question but,, Solid blue and white is a dilution of black, but with blue eyes does it make the dog a merle? I would think, yes...right?

 

As Eileen has already pointed out, blue eye does not make a "blue" dog a merle.

 

It's very confusing that some form of the word "dilute" (e.g. "dilute," "dilution") is used to describe two entirely different genetic variations. (It doesn't help that slate dogs are also called blue and black merles are called blue. )

 

Dilute black is a blue (or slate) dog and dilute red is lilac. In this case it's like diluting black or red paint with white to create grey or a lighter shade of "red."

 

The dilution that occurs in merles is more like mixing salt and pepper. Not really a good analogy, but if you look at a blue merle dog you'll see that there are black hairs and white hairs (as in a roan horse) that make the dog appear grey. Really what it is is random individual hair follicles (would that be correct? or should it be shafts?) that would have been black have been blocked by the merle allele from producing the black pigment, which then gives it that salt and pepper appearance. (The reason that MM, or double, merles have so much white is that large areas are blocked from producing melanin, the dark pigment.)

 

I hope I'm explaining this adequately, and correctly. If not, as always, I welcome corrections.

[B Point]

Thanks GentleLake.... Genetics wouldn't be genetics without the ambiguity.

[juliepoudrier]

Actually though, merle isn't just salt and pepper; the merle gene affects coat color in patches. The patches that are affected have the "roaning" effect, but the unaffected patches will be the dog's normal base color. To confuse things even more, you can have a white factored dog who is heavily ticked in the white areas (looks roan) and the normal base color where the coat is unaffected by the white factor gene, and many people will confuse that color combination with merle, which it is not.

 

White factored, heavily ticked (Phoebe):

 

Blue merle (this is my old Indy, my first dog, who was a border collie x aussie). I wish I had a better picture that shows his whole body, but don't have one readily available.

 

Indy had one blue eye and one brown eye. Each eye had a small patch of the opposite color on it. His hearing was normal.

J.