CHSD is an inherited condition
This study provides strong evidence for the inherited nature of CHSD in the Australian Cattle Dog. Clustering of deafness amongst dogs from the same litter, but minimal clustering amongst dogs from different litters in the same kennel, is consistent with an inherited aetiology and our results were consistent with an association between hearing status of the sire and CHSD. Our estimate of the heritability of having at least one ear affected was 0.21. This result is similar to that found in Jack Russell Terriers of 0.22
, but lower than estimates of 0.49 in Border Collies
 and 0.73 in Dalmations
. Based on these three sources of evidence, we conclude that, in the Australian Cattle Dog, as in other breeds, CHSD has a hereditary component to its aetiology, and that this condition may be controlled by a major common locus, as it appears to be in the Australian Stumpy-tail Cattle Dog where the trait appears to be likely to be autosomal recessive
As with other genetic diseases, the prevalence of CHSD in the Australian Cattle Dog could be reduced by only breeding from stock with normal hearing on BAER testing and that have no deaf offspring or deaf parents. However, if the prevalence of the locus for CHSD is high, the removal of a large number of animals from the breeding programme could reduce genetic variation and lead to the exposure of other hereditary problems in the breed. The development of a molecular genetic diagnostic test could allow the gradual elimination of carriers and affected dogs from the breeding program if the frequency of the disease allele was low, thus minimising the problems associated with a small gene pool
The higher prevalence of CHSD found in this study in Australian Cattle Dogs with deaf parents has also been reported in the Dalmatian
[1, 15, 36], and in the English Setter and English Cocker Spaniel
 Another study reported a significantly higher prevalence of deafness in offspring of untested parents (23%), compared with normal hearing parents (15%)
, presumably because the untested population included some deaf animals. Similarly, in a study in Border Collie puppies
, the prevalence of CHSD was significantly higher in offspring from unilaterally deaf dams (10%) compared with those from normal hearing dams (2%). In contrast, there was no significant association between parental and offspring hearing status in Australian Cattle Dogs and Bull Terriers
, although in that study, only a few subjects had parents of known hearing status
Prevalence of CHSD
The overall prevalence of deafness in complete Australian Cattle Dog litters tested in this study was 11.1%, and the overall deafness prevalence in the 899 dogs tested was 10.8% with 7.5% unilaterally deaf and 3.3% bilaterally deaf. The prevalence of CHSD in the general Australian Cattle Dog population may be higher than this, as our study did not include individual dogs presented for suspected deafness and most of the breeders of study dogs were avoiding the use of affected animals as breeding stock. This may not be the case in the general Australian Cattle Dog population. In support of this, a higher deafness prevalence was observed in a study in the USA with 296 Australian Cattle Dogs
 (14.5%, with 12.2% unilaterally deaf and 2.4% bilaterally deaf). The higher prevalence recorded in this clinical study could have been due to the inclusion of dogs presented for suspected deafness
Unilaterally deaf dogs are difficult to identify unless they are BAER tested. Thus, the proportion of deaf dogs that are unilaterally deaf provides an estimate of the percentage of affected dogs that would probably be undetected in the absence of BAER testing
. The proportion for the USA study was 84%
. For our study, the proportion of deaf dogs that were unilaterally deaf was 0.69 (7.5/10.8), suggesting as many as 69% of affected dogs would go undetected if they were not BAER tested.
In the current study, female dogs appeared to be at increased risk of deafness compared to males, including following adjustment of results for any confounding due to any associations between gender influence and each of mask type and pigmented body spots (Table
4). The literature is divided as to the effects of sex on the prevalence of CHSD. In some studies, no association was observed between sex and CHSD in a variety of breeds
[1, 8, 11]. This included a study in Australian Cattle Dogs that used both binary (ie deaf/not deaf) and ordinal (normal/unilateral/bilateral) outcome variables
. However, in other studies, female Dalmatians were at increased risk of CHSD
[14, 15, 17, 36]. A multi-breed study found no significant sex difference in the prevalence of deafness (at least one ear affected) in Dalmatians, Bull Terriers and a small sample of Australian Cattle Dogs (n = 296), although the prevalence of CHSD was a little higher in females in all of these breeds
. In the same study, in English Cocker Spaniels, there was some evidence that the prevalence of CHSD differed significantly by gender (P = 0.035) when treated as a trichotomous trait (normal/unilateral/bilateral). When treated as a dichotomous trait (deaf/normal hearing), the p-value was higher (P= 0.067)
. In the same study in English Setters, the prevalence of deafness treated as a dichotomous trait, did not differ significantly by sex (P = 0.601) and distributions of the three category trait did not differ significantly by gender after accounting for data source (two subsets of English Setters were studied). Comparisons in this multi-breed study were not adjusted for potential confounders, and clustering of deafness within litter was not accounted for. The differences in the findings of these various studies may be due to variations in statistical analysis methods used, or in the relationships specific to the study populations.
The prevalence of CHSD and coat characteristics
In our study, the presence of pigmented body spots was associated with a reduced risk of CHSD independent of gender and facial mask status. Similarly, it has been shown previously that Dalmatians with large pigmented patches instead of overall pigmented spotting are less likely to be deaf
[1, 8, 13, 14, 16]. Coloured patches in Dalmatians are thought to be due to weaker expression of the S gene now known to be MITF, as all Dalmatians are homozygous for the extreme white piebald allele sw. In Dalmatians, pigmented patches are present in the white coat at birth whereas spots are not. Similarly, Australian Cattle Dog puppies are born with white coats but also show all dark body and facial masks or markings at birth; red in the case of red dogs and blue/black in the case of blue dogs. No new markings appear as the puppy grows, and the size of the markings merely grows at the same rate as the dog. This might indicate a similar genetic basis for these markings in Dalmatians and Australian Cattle Dogs. However, other mechanisms for white markings may occur in the Australian Cattle Dog. Anecdotally Dalmatians may have been used to develop the Australian Cattle Dog as a breed but this is not well documented. Interestingly, the prevalence of CHSD in the Australian Cattle Dog could possibly be reduced if pigmented body spots were no longer classified as a show fault in this breed
, given the association of pigmented patches with reduced deafness prevalence.
From the univariable multilevel logistic models and logistic animal models, in the current study, dogs with bilateral facial masks had a reduced risk of deafness in at least one ear. While the reduced risk of deafness in Australian Cattle Dogs with bilateral masks may be due to a weak expression of the sw allele of MITF, it is also possible that a gene other than S, possibly a locus linked to a CHSD locus, is also involved. One candidate is the EM allelle of the MC1R gene, which is associated with the presence of a facial mask in some breeds, and only a single allele is required to produce this effect
. It is unknown whether this gene is involved in the production of a black facial mask in the Australian Cattle Dog. However, this explanation does not account for the fact that while masks in the Australian Cattle Dog are black in blue coated dogs, masks are dark red rather than black in red coated dogs. It is also interesting that it is the bilateral rather than the unilateral mask that is associated with a reduced prevalence of deafness, possibly due to increased gene dose effects.
From results of the current study, there appeared to be no strong association between the base coat colours of red or blue, and CHSD in the Australian Cattle Dog. In a previous study on 293 Australian Cattle Dogs, prevalence of CHSD also did not differ substantially among dogs with different base coat colours. In that study, coat colours were categorised as blue, blue and black and tan, blue and tan, and red
. In the current study, we analysed the base coat colours in two ways, comparing deafness prevalence between (1) colour groups which had sufficiently large numbers of dogs for use in analysis; these groups were blue, blue and black and tan, blue and tan, blue and black, and red, and (2) by base colour (blue or red). Both analyses produced similar results suggesting base coat colour is not strongly associated with CHSD in the Australian Cattle Dog. Results of this and the current study are interesting as in a recent study in the related Australian Stumpy-tail Cattle Dog, a significant association was observed between coat colour and CHSD, with dogs with red coats at increased risk of CHSD compared with those having blue coats
. This difference in such related breeds is difficult to reconcile. While this may be due to a genuine difference between breeds, it is also possible that the study population of Australian Stumpy-tail Cattle Dogs was affected by a founder effect, resulting in a higher prevalence of CHSD in red dogs.
The number of mottled dogs (n = 5) was too small to draw any meaningful conclusions about associations between mottling and CHSD. The genetic basis for speckling and mottling is as yet unclear. The speckling effect in the Australian Cattle Dog may be associated with a variant of the MITF gene
[22, 23], or a dominant ticking gene T,
. There is also a recent description
[40, 41] of a flecking gene giving a roaning effect where a trait defined as roan in English Cocker Spaniels was mapped to a specific region on chromosome 38. In subsequent studies, the trait that this group defined as ‘ticking’ in English Springer Spaniels and Dalmatians also mapped to nearby regions. The authors suggested that ticking was inherited as a co-dominant trait.
This present study has identified a possible role for pigmentation genes in CHSD in the Australian Cattle Dog, due to the negative association between CHSD and masks and dark body patches. While no relationship between CHSD and white head/body patches was found, this may have been due to imprecise effect estimates due to low numbers of dogs with white body patches.
In the current study, we found no association between the side of the mask and the side of deafness. If there is truly no association, there are interesting implications for the molecular pathogenesis of CHSD in the Australian Cattle Dog. The MITF gene on CFA20 regulates the differentiation of neural crest derived melanoblasts to melanocytes
[24–26] and MITF-M and SOX10 have been shown to be involved in melanoblast migration from the neural ectoderm to the otic vesicles and epidermis in the mouse
[27, 28], and are separately expressed in different cell types in the newborn cochlea
. Mutations of MITF can affect melanoblast survival and affect skin and hair pigmentation
, and these mutations may also affect otic melanocytes and hearing status
. However, as it is unlikely that pigment cell migration into hair and keratinised skin are entirely controlled by the same genes
, it may also be unlikely that melanocyte migration to the stria vascularis and to the skin and hair are totally controlled by the same genes. This explanation could account for the lack of association between the side of unilateral deafness and the side of a pigmented facial mask.