Have I The Right To Sing The Blues?
Originally Presented in the Fall 1999 issue of The Dalmatian Quarterly
Thomas R. Famula, Ph.D.; Department Of Animal Science; University of California, Davis 95616
Congenital deafness, though it may be a serious problem for Dalmatian breeders, is not limited to dogs. Many species, including our own, suffer from hearing loss in part due to the genes inherited from unsuspecting parents. Several years ago, when we began our search for the genetic basis of hearing loss in Dalmatians, our hope was to identify those animals most likely to transmit these potential "bad" genes to their offspring. That, after all, was how I came to be involved. As a dairy cattle breeder, I must confess to having no knowledge, or experience, in working with dogs until that time (beyond the pair of pound puppies that my daughters fuss over). Instead, my job had been to identify those bulls that would father daughters with the highest milk production. Because you can't milk a bull, the search for superior bulls is one of data collection, statistics and computers - just the ingredients needed to mount a search for disease causing genes in Dalmatians.
So, back in 1995, when we began our genetic hunt, I knew little more about Dalmatians than what I could pick up in a cartoon. But a review of the scientific literature revealed several important characteristics about hearing loss in Dalmatians. The first, and most important, was the association of blue eyes with hearing problems (discussed and reviewed by Holliday, Nelson, Williams and Willits in the Journal of Veterinary internal Medicine from 1992, Volume 6, page 166). Other data sets, collected throughout Europe and North America, reaffirmed this association, as well as establishing others. For example, the presence of a color patch at birth is associated with a lower likelihood of suffering from deafness (presented by Greibrokk in the Journal of the American Animal Hospital Association from 1994, Volume 30, page 170). The point is, even before we began, several researchers had uncovered some clues as to which dogs might be the best for breeding, if reducing deafness was our goal.
Our initial data set had BAER (brainstem auditory-evoked response) tests on 825 dogs (457 males and 368 females). Clearly most dogs have brown eyes, and the frequency of normal hearing in these dogs is 0.68 (249/368) for females and 0.74 (336/457) for males. Now, recognizing that doubly blue-eyed dogs occur rarely, the frequency of normal hearing in dogs with two blue eyes is 0.43 (3/7) for females and 0.50 (4/8) for males. This significant reduction in the incidence of normal hearing is the source for the concern over blue-eyed dogs. Certainly blue-eyed dogs can be free of hearing loss. Nevertheless this should serve as a warning - dogs with blue eyes pose a greater risk of carrying and passing on deleterious hearing genes. It isn't foolproof, but it can, and should, be viewed as a genetic "red flag". And don't forget to notice that the frequency of normal hearing for dogs with one blue eye and one brown eye has also dropped significantly from the frequency of those with two brown eyes (9/22 = 0.41 in females and 12/32 = 0.38 in males).
Now, I italicised the word significant to drive home an important point. For the term has statistical as well as everyday meaning. Statisticians (and I must confess to being one) use this term to describe differences between groups that are larger than what might be expected by chance. For example, if we flip a coin 100 times, we would expect 50 heads and 50 tails. But what if we get 48 heads and 52 tails? Is the coin "fair" or "biased"? The role of statistics is to help us decide how far you can get from 50:50 and still consider the coin “fair”. When the difference is large enough, we use the term significant to suggest that the coin no longer conforms to our expected 50:50 result. The trick is to know just how "large" is "large". Now I don't expect anyone wants me to have him or her break out their calculus books. But, determining what is "large" requires a bit of mathematics.
Here is where you must take what I say on faith (although this work has been published in a scientific journal which has been reviewed by other researchers; Mammalian Genome, 1996, Volume 7, page 650). The difference in normal hearing between brown and blue-eyed dogs is significant - a difference so large as to suggest that it didn't happen by chance. Something about the biology of eye color and hearing development must be creating this observed difference. So, although we may not know the exact mechanism that creates this effect, there is a sizeable collection of data (that above combined with the work of others) that strongly establishes this association.
What remains is the question: "What are we to do with this information?" It is my contention that blue eye color is a flag, a warning that such dogs present an increased risk of passing on deleterious hearing genes. But, as the data above demonstrates, it is not a perfect predictor. Nevertheless, the fact that the association between blue eyes and deafness is not perfect should not permit ignoring the signal. The signal of eye color is real and significant; it can not be ignored.
If blue-eyed dogs were so common that eliminating these animals as potential breeders would cause hardship when selecting for other important trait, perhaps we could overlook the tie to deafness. But the relative scarcity of blue-eyed dogs requires we give the association of deafness and blue eye color our serious attention.