Laurence J. Denholm BVSc(Hons) PhD(Cornell)
NSW Department of Industry and Investment
Orange, NSW Australia
24 September, 2010

Estimated Accuracy of A Diagnostic Test for Contractural Arachnodactyly (CA) in Angus-infused Cattle


Estimates were made of the diagnostic sensitivity and specificity of a new DNA test developed at the University of Illinois by Prof. Jonathan Beever and his colleagues to identify carriers of a simple autosomal recessive mutation causing congenital contractural arachnodactyly (CA) in Angus and Angus-infused beef cattle, a heritable developmental disease formerly called “fawn calf syndrome”. Using a test panel of blood samples (n=397) assembled in Australia and New Zealand from CA affected calves (n=17), obligate CA carriers with normal phenotype (n=16) and cattle assumed to be free of the CA defect on the basis of their pedigrees or known ancestry (n=210), sensitivity of the test was estimated to be 100% using the Maximum Likelihood Estimate (MLE) method, with a 95% Adjusted Wald confidence interval of 90.96%. Specificity was also estimated to be 100% by the MLE method, with a 95% Adjusted Wald confidence interval of 98.47%. The study demonstrated that this new DNA test for CA will accurately detect carriers of the causal mutation with high precision and validity.


In July 2010, Prof. Jonathan Beever of the University of Illinois at Urbana-Champaign reported that he had developed an accurate DNA diagnostic test for detection of the causal autosomal recessive mutation in congenital contractural arachnodactyly (CA), a heritable developmental disease of Angus and Angus-infused cattle which has also been called “fawn calf syndrome” since it was first identified in the late 1990’s in Australia amongst the descendents of the registered American Angus cow Freestate Barbara 871 of Kaf (AAA9163301, USA9163301) born in 1978 in Indiana, USA.  [For a description of CA, see

Prof. Beever proposed an independent “blind” validation of this new DNA test for CA mutation carriers, using appropriate blood samples collected from cattle in Australasia, including cattle in the CA research herd established in 2004 by the NSW Department of Industry and Investment with assistance from the Angus Society of Australia.

All of the animals in this CA research herd are descendants of two CA affected cows from Western Australia, both descendents of Freestate Barbara 871 of Kaf, and two US Angus sires whose semen was imported into Australia in the 1990’s, also both descended from Freestate Barbara 871 of Kaf.  These two US sires, Bon View Bando 598 (AAA11104267, USA598) and Rambo 465T of JRS (AAA10682868, USA465T), were first identified as suspect CA carriers in the late 1990’s by the Angus Society of Australia when CA affected newborn calves were reported amongst their progeny born in the state of Victoria. All animals in this CA research herd in NSW are either heterozygous but phenotypically normal CA carriers (CAC) or homozygous and phenotypically affected CA carriers (CAA). There are no animals in the herd that do not carry the CA mutation. Samples from some animals in this CA research herd were previously sent to Prof. Beever and used in development of the DNA test for CA that is evaluated in this study.


In August 2010, a “blinded” panel of 478 blood samples in EDTA anti-coagulant was sent to the University of Illinois from NSW, each sample labelled only with a code number from 1 to 478. At the same time a double coded list of the code numbers for these 478 samples was sent to Dr. Peter Parnell, CEO of the Angus Society of Australia, together with identification details of the individual animals from which blood samples had been taken and their expected genotypes – CAF (free), CAC (carrier), or CAA (affected) – based on their pedigrees and phenotypes. Decoding of the double coded list held by Dr Parnell required an “offset number” that was only sent to Dr. Parnell after the CA test results were received from Prof. Beever. This double coding procedure with independent “umpire” was designed to ensure integrity of the test validation procedure in the event of any dispute about the data.

The test validation sample panel consisted of 397 bloods taken from 243 animals with an assumed CA genotype based on their pedigrees or ancestry and phenotypes. Randomly included amongst these 397 validation samples there were an additional 81 bloods taken from animals of unknown CA genotype but with pedigrees that included at least one descendent of Freestate Barbara 871 of Kaf known to have sired CA affected calves, making a total of 478 samples. These additional 81 bloods were “test” samples and their CA test results were not included in the assessment of CA test accuracy reported below. DNA results from 11 of the 81 animals indicated the CAC genotype; the rest were CAF.

Of the 243 animals from which the other 397 “validation” blood samples were taken, 33 were animals known to carry the CA mutation, either as homozygous affected animals (CAA) or as obligate heterozygous carriers (CAC). Seventeen (17) were affected (CAA) animals in which the CA disease was diagnosed by a veterinarian at or shortly after birth and the other 16 were animals that were not suspected at birth, or at any time after birth, to be affected by the CA disease, but which were all necessarily carriers of the causal mutation as either the parent of a CA affected calf or the progeny of a CA affected parent.

Of the 17 CA affected (CAA) animals, 11 were from the CA research herd in NSW and 6 were CA cases reported from the field by cattle breeders in NSW, Victoria and South Australia. One of these 6 CA field cases was an Angus x Murray-Grey calf.  The others were purebred Angus calves. The field CA cases were not closely related by pedigree to CA cases in the research herd. Of the 16 assumed heterozygous carrier (CAC) animals, 9 were from the CA research herd and 6 were parents of CA cases reported from the field.

Of the remaining 210 animals which comprise the negative (assumed CAF) validation sample cohort, 85 were registered Angus cows from a herd in New Zealand with no history of any infusion of American Angus bloodlines and 97 were registered Murray Grey animals in a herd with no history of any Angus infusion since the herd was founded in the 1960’s. Blood samples were also obtained from 7 animals sired by a great-grandson of Freestate Barbara 871 of Kaf that was known to not have sired any CA affected calves when joined to known CA carrier cows, with no descendents of Freestate Barbara 871 of Kaf in the maternal side of the pedigrees of these 7 animals. A further 19 bloods were taken from registered Angus cattle in Australia that were either imported as embryos from the USA or were the progeny of such imported animals but which were not descendents of Freestate Barbara 871 of Kaf.  Bloods were also taken from two Holstein crossbred cows.

A summary of animals used in the test validation is shown in Table 1.

Affected (CAA)17
Carrier (CAC)16
Free (CAF)Registered New Zealand Angus without any American Angus infusion85
Registered Murray Grey without any
Angus infusion
Registered American Angus without any
descendent of founder cow
(Freestate Barbara 871 of Kaf)
Registered American Angus with non-CA
carrier descendent of founder
(Freestate Barbara 871 of Kaf)
Holstein Xbred cows2
Sub Total (CAF)210
Table 1. Animals used in the CA test validation sample panel

Replicate samples were  taken from many of these animals (including all assumed CAA and CAC animals and about half of the assumed CAF animals) and included randomly through the validation sample panel. In the case of samples from assumed negative (CAF) animals, replicate bloods were taken on the same day. For the assumed positive (CAC and CAA) animals, many of the replicate samples were taken more than a year apart and the earlier samples were stored at -4oC for at least a year.  A summary of all blood samples used in the CA test validation analysis is shown in Table 2.

Affected (CAA)46
Carrier (CAC)37
Free (CAF)314
Table 2. Blood samples used for the CA test validation


Of the 397 “validation” blood samples, 394 (99.24%) returned the DNA test result expected from the assumed genotype determined in advance by pedigree analysis and clinical examinations of newborn calves by a veterinarian experienced with the CA syndrome. All three samples which returned an unexpected DNA test result were replicates taken on the same day from the same animal. At the time of sampling in July 2010 this animal was a newborn calf in the CA research herd. For this calf, the expected genotype was CAC but the actual test result indicated the CAA genotype.

It is important to note that, even for this particular calf, the test did not fail to successfully detect the CA causal mutation and the mutation was detected in all three samples taken from the calf.  The only anomalous results in this study therefore involved the differentiation of homozygous and heterozygous carriers of the causal mutation, not the differentiation of carriers of the causal mutation from non-carriers. Possible explanations for this anomalous result and its limited significance are discussed in more detail below.

Hence, despite the anomalous results in this single animal, of the 83 samples taken from 33 animals known to carry the CA mutation on the basis of pedigree or phenotype, all 83 samples (100%) gave a test result (CAA or CAC) which indicated that the animal carried the causal mutation – either one or two copies of the mutant gene.

In other words, the DNA test detected this CA causal mutation in 100% of 83 blood samples from 33 animals in which it was expected the CA mutation was present.

Conversely, the DNA test gave a negative result for presence the CA causal mutation in all 314 (100%) of the blood samples from 210 animals for which a negative result was expected on the basis of their pedigrees or ancestry.

Put another way, no “false negative” or “false positive” test result was observed in this study in relation to the detection of any animal assumed a priori to carry or to not carry the CA causal mutation on the basis of phenotype, pedigree or ancestry.

A high level of repeatability for this DNA test was also demonstrated, with all replicate samples returning the same result in each case for all samples taken from the same animal (including three replicates taken from the animal which returned the unexpected results).

Statistical assessment of the accuracy of the U Illinois DNA test for CA

Two parameters used to describe the precision of diagnostic tests are sensitivity and specificity, defined ( as:

Sensitivity (Se): The estimated sensitivity (synonym: true positive rate) of a diagnostic test is the estimated (or assumed) proportion of animals with the disease (or infection) of interest which test positive. It is a measure of the probability that a diseased individual will be correctly identified by the test.
Specificity (Sp): The estimated specificity (synonym: true negative rate) of a diagnostic test is the estimated (or assumed) proportion of animals without the disease (or infection) of interest which test negative. It is a measure of the probability that an individual without the disease of interest will be correctly identified by the test.

Estimates of sensitivity and specificity assume random sampling of the relevant population. There are several statistical methods available for calculating “best point estimates” of diagnostic sensitivity and specificity. For diagnostic tests with binomial outcomes (that is, diagnostic tests in which only two results are possible – a positive result or a negative result), the simplest and most commonly used statistical method is called the maximum likelihood estimate (MLE). This method was used for analysis of the results obtained with the CA DNA test under evaluation in this study and the MLE analysis indicated that this particular DNA test for the causal mutation in CA is 100% accurate with respect to both sensitivity and specificity, when the test is used for the detection of animals carrying at least one copy of the CA mutation.

[The MLE is derived from p = x/n where p is an estimate of the true frequency of the parameter in the population, n is the total number of samples and x is the number of samples in which the parameter was observed. In this study, x/n is the proportion of assumed positive (or negative) samples that gave a positive (or negative) test result.]

Although widely used for this purpose, the MLE methodology does have some well-known limitations when estimating the standard error and hence the confidence intervals for sensitivity and specificity of diagnostic tests with small sample sets (where n < 150) and particularly with datasets where x = 0 or x = n, as was the case with the dataset used in this study where all samples gave the result expected with respect to presence or absence of the causal mutation. In this situation, it is appropriate to analyse the dataset by an alternative and more conservative statistical methodology, as described below.

In the first part of the assessment of the CA test accuracy in this study, the test results for assumed CAC and CAA animals were combined. Accordingly, the estimated diagnostic sensitivity and specificity reported below apply to the detection of animals carrying at least one copy of the CA mutation, regardless of whether the particular animal carries one copy (heterozygous) or two copies (homozygous) of the causal mutation.

This approach is consistent with the design of the DNA test for CA, which uses a PCR (Polymerase Chain Reaction)-based approach for differential detection of the normal and mutated chromosomes. In this type of assay, one DNA primer differentiates the mutated chromosome (present in the CAA and CAC genotypes) from the normal (unmutated) chromosome. Similarly, a second primer differentiates the normal chromosome from the mutated chromosome. A third primer common to both normal and mutant chromosomes is then used to generate specific amplicons corresponding to either the normal or mutated DNA sequence. For CAF animals, only one amplicon is produced, corresponding to the normal DNA sequence of the unmutated chromosome. Alternatively, for CAA animals only one amplicon is again produced, but in this case an amplicon corresponding to the mutated DNA sequence. For CAC individuals both amplicons are produced, indicating the presence of both the normal and mutated DNA sequences. (J Beever, pers. comm.)

Accordingly, with the dataset obtained in this study it is also possible to calculate the sensitivity and specificity of the CA test with respect to differentiation of heterozygous carriers (CAC) from homozygous affected (CAA) animals, using the subset of animals in the dataset in which the CA mutation was detected (ie. the CAC and the CAA animals). Although this differentiation of CAC from CAA aspect of the CA test accuracy has less practical significance per se than the differentiation of all carriers (homozygous or heterozygous) from non-carriers, the results are reported below.

As the results in this study give an observed (estimated) sensitivity of 100% and an observed (estimated) specificity of 100% for detection of the causal mutation using the MLE method, estimation of the true sensitivity and true specificity of this test with this dataset involves the usual problems with binomial point estimators when = 0 or 1.0 in a small sample set. This is a greater a problem for the estimation of true sensitivity than for the estimation of true specificity as the sample size for sensitivity is unavoidably smaller.

Although the calculated MLEs of 100% for both sensitivity and specificity are valid estimates of the true sensitivity and true specificity of this DNA test for CA, they are not necessarily the ‘best estimates” biometrically for a small sample set where x = n. For small samples such as the sample of CA positive animals used in this study, Laplacian and other adjusted estimates are more appropriate, albeit that these methods shift the estimate away from the MLE values of 100% as a consequence of their allowance for smaller sample sizes (Lewis & Sauro, 2006). Therefore, in the analysis of the data from this study, “best point estimates” for sensitivity and specificity were calculated by the Jeffreys method (Jeffreys, 1961) and reported together with those from the MLE method, for both sensitivity and specificity. It should however be recognised that a “best point estimate” of something less than 100% for the sensitivity and specificity of this test for CA does not mean the true values of the sensitivity or specificity are necessarily less than 100% to the extent of these reported “best estimates”. The more conservative “best point estimates” obtained with the Jeffreys method simply mean that the unavoidably small size of the sample of animals carrying the causal mutation (CAA and CAC) that was available for estimation of test sensitivity is a limitation to obtaining a more accurate estimate of the true value. The true value of the sensitivity of this test could well be closer to 100% than the “best point estimate” obtained from this dataset using the Jeffreys method.

Perhaps of greater importance is the methodology used for calculation of the confidence intervals for these estimates of sensitivity and specificity. The estimates of sensitivity and specificity obtained in this study and in any other study, no matter how large or how well designed, are always just estimates of the true values of the parameters studied – estimates always subject to error. It is therefore normal practice to quote any estimated point value together with confidence intervals that indicate the boundaries within which the true value of the parameter will lie with a chosen level of probability (most commonly 95% or 99% probability). The boundaries of the confidence intervals for the estimated value of any parameter are just as important as the estimate itself.

A problem arises however when the standard method for calculating confidence intervals (which uses the normal distribution approximation of the binomial distribution) is applied in a situation where the observed frequency of the characteristic of interest in the sample is zero or 100%, (ie. when no false positives or false negatives are observed and p = 0 or p = 1.0). In this situation, the calculated standard error that is required to calculate the confidence intervals, and hence the estimated confidence intervals themselves for sensitivity and specificity, will be zero. A confidence interval of zero would suggest the true value of sensitivity and specificity is 100% exactly, with no likelihood whatsoever of any error in that estimate, a rather meaningless and unsatisfactory proposition in any normal biological context.

A nonstandard alternative to the normal method is therefore more appropriate for the calculation of confidence intervals when p = 0 or p = 1.0.  In situations where 0.9<pLaplace (Bayesian) or Jeffreys transforms provide the best point estimates of true binomial proportions from small sample sets (n<150), a non-standard adjustment such as the Clopper-Pearson Exact method or the Adjusted Wald method should be used for calculation of the confidence intervals (Lewis & Sauro, 2006). Although the Clopper-Pearson Exact method produces confidence intervals with coverage that always exceeds the specified (eg 95%) confidence level, for small samples this method is known to be overly conservative and the Adjusted Wald method is preferred (Lewis & Sauro, 2006).

Accordingly, in this study best point estimates were calculated using the Jeffreys transform method in addition to the MLE method and binomial confidence intervals were calculated using the Adjusted Wald method. It should be noted that the best point estimates obtained by such adjusted methods are not necessarily centred within the confidence interval range.

Calculation of these statistics was undertaken using the online calculator for Binomial Proportion Confidence Intervals ( published by Jeff Sauro of Oracle, Denver CO and accessed on 17 September 2010.

It should also be noted that when the estimated binomial proportion is 100%, as is the case with the dataset in this study, the confidence interval is necessarily one-sided as the true value cannot be above 100%.  In this situation, it is appropriate to use the z-critical value for a one-sided confidence interval in transforms such as the Adjusted Wald that require the z-critical value. For the 95% confidence level, the z-value of 1.64 appropriate to a one-sided interval is used rather than the z-value of 1.96 for two-sided intervals.

Following this approach, estimates were obtained for the sensitivity and specificity of the DNA PCR test for CA developed by Prof. Beever and his colleagues, using test results from the validation sample panel described above.  These estimates are shown in Table 3.

  Point Estimates Confidence Intervals
  n x Maximum
Estimate (95%)
Adjusted Wald Estimate (95%)
  Low High
Sensitivity 33 33 100% 98.53% 90.96% 100%
Specificity 210 210 100% 99.76% 98.47% 100%
Table 3.  Sensitivity and Specificity of CA DNA test for the CA mutation (1 or 2 copies)

As no false positive or false negative results for detection of carriers of the CA causal mutation were observed, the estimated Positive Predictive Value (PPV) and the estimated Negative Predictive Value (NPV) of the DNA test for CA are equal to the sensitivity and the specificity estimates respectively when estimates of sensitivity and specificity calculated by the MLE method (p = x/n) are used, regardless of the disease prevalence.

Assuming an estimated CA carrier prevalence of 3% in the population (J. Beever, pers. comm.), the Jeffreys method estimates of sensitivity and specificity provide a PPV estimate of 92.70% and an NPV estimate of 99.95%.

Although of less practical importance, as stated above, using the same dataset it is possible to estimate the true sensitivity and specificity of the CA DNA test for differentiating heterozygous CA carriers (CAC) from homozygous affected (CAA) animals amongst those animals that are positive for the CA mutation (one or two copies).

In the original validation test results as described above, one obligate CA carrier animal with normal phenotype at birth (and therefore expected to have the CAC genotype) gave the unexpected result of CAA genotype in the DNA test. Subsequent and more detailed investigation of the DNA from this calf confirmed that the calf does not have the normal (unmutated) chromosome and therefore, by definition, has the CAA genotype as originally reported by the testing laboratory for the validation test sample.

Additional investigations of this calf were based on detection of DNA sequences that are deleted as a result of the CA mutation. Four separate amplicons distributed throughout the sequence deleted from the normal chromosome by the CA mutation were assayed; none of the four amplicons were detected in this calf. Furthermore, two additional PCR amplicons were used to detect the DNA sequences adjoining both the 5′ and 3′ deletion breakpoints. These sequences were detected, indicating there was DNA of sufficient quality in the sample to detect the four amplicons if present. (J. Beever, pers. comm.)

After the initial unexpected DNA test results for this calf were received, the calf was re-examined (at the age of 9 weeks) by a veterinarian experienced with the CA syndrome, but no abnormalities suggestive of the CA phenotype were detected in this detailed visual and physical clinical examination. The calf appears to have the normal phenotype.

The necessary conclusion from the additional investigation of DNA from this apparently normal phenotype calf with anomalous CA test results is that this calf does not carry the normal chromosome and hence does have the CAA genotype and not CAC genotype assumed from its normal phenotype at birth. In the data analysis in Table 4 below, this calf with the unexpected test result is therefore counted as CAA rather than CAC.

Estimates of the diagnostic sensitivity and specificity for detecting heterozygous (CAC) from homozygous (CAA) carriers using this particular DNA test are given in Table 4.

Table 4. Sensitivity and Specificity of CA test for distinguishing the CA mutation in the heterozygous state (CAC) from the CA mutation in the homozygous state (CAA)

  Point Estimates Confidence Intervals
  n x Maximum
Estimate (95%)
Adjusted Wald Estimate (95%)
  Low High
Sensitivity 15 15 100% 96.88% 81.98% 100%
Specificity 18 18 100% 97.37% 84.53% 100%


Despite the statistical limitations of most datasets for rare genetic diseases where there is normally only a small sample of true positive animals available, the study reported here does demonstrate that the DNA diagnostic test for CA developed by Prof. Beever and his colleagues is both highly sensitive and highly specific for the detection of carriers of one or two copies of the DNA mutation that causes the CA disease in the homozygous state, with the sensitivity and specificity of this test both well within the ranges required for practical and effective use of any diagnostic test for control of a genetic defect in cattle.

The only potential false positive or false negative results in this study were those observed in three replicate blood samples taken from one animal in the CA research herd which was assumed to have the CAC genotype on the basis of a clinical examination of the newborn calf by a veterinarian experienced with the CA syndrome, but which had DNA test results for all three replicates indicating that the animal has the CAA genotype.

At the time this unexpected result was recorded, there were two likely explanations.

First, it was possible that this calf had the CAA genotype but only a very mild form of the CA disease that was not detected at or after birth, despite several clinical examinations.

From clinical observations of newborn calves in the CA research herd over the last few years, it was already known that there is quite a wide variation in severity of the CA phenotype of newborn affected calves, although extension of this phenotypic variability to normality or near normality had not previously been suspected. An incorrect genetic classification of this calf on the basis of clinical examination alone was however quite possible in light of this known phenotypic variation.

Phenotypic variation is observed in many heritable diseases including developmental disorders of connective tissue of which CA is one type. Environmental and background genomic influences are common explanations, although Kurnit et al (1987) showed that some non-Mendelian clustering of anomalies attributed to concepts such as reduced penetrance and multifactorial inheritance can be accounted for by simple, random chance.
Subsequent studies by Cohen (1989) lead to a conclusion that we should really consider extreme phenotypic variability such as incomplete penetrance in terms of single gene mutations that predispose the organism to develop an anomaly or disease but do not always result in an abnormal phenotype.
The possibility of CAA animals in the population that do not show any overt signs of the CA disease therefore cannot be discounted.

Alternatively, this calf with the unexpected DNA test results might have had the CAC genotype but the DNA primer used to differentiate the normal chromosome from the mutated chromosome in the DNA test for CA for some unexplained reason failed to differentiate the normal chromosome of this particular animal, or alternatively, the third primer failed to generate the amplicon indicating presence of the normal DNA sequence.

In this context it is worth noting that if this particular anomalous calf really had the CAC genotype, it must have inherited the normal chromosome from its dam because the sire is an affected bull (CAA). The DNA primers of the CA test did however successfully detect the normal chromosome in the CAC dam of the anomalous calf and also in a CAC full sibling of the anomalous calf. This detection of the normal chromosome in close relatives of the calf with the unexpected test results could be taken as good evidence that failure to detect a normal chromosome present in this calf was a less likely explanation for the observed unexpected test result than misclassification of the genotype of the calf on the basis of phenotype, as was subsequently confirmed by further studies of its DNA.

This important but unexpected outcome from the study reported here raises the possibility that other homozygous carriers of the CA mutation with normal phenotype are present in the cattle population, a possibility not previously considered.  However, no other animal with the CAA genotype but normal phenotype was detected in this study amongst the 16 individuals with normal phenotype known to be CA carriers, suggesting that animals with the CAA genotype but normal phenotype are unlikely to be common in the wider cattle population.  Whether such phenotypic variation provides any explanation for the reported variation from normal Mendelian ratios in the incidence of CA cases in the field remains to be determined.

Given that the main objective in using this new diagnostic test for CA is to remove carriers of the causal mutation from the cattle population, regardless of whether they carry one or two copies of the causal mutation, accuracy of the test in distinguishing between homozygous disease affected carriers (CAA) and heterozygous phenotypically normal carriers (CAC) is of much less practical importance than the accuracy of the test in distinguishing between all carriers of the CA mutation and non-carriers.

It is worth noting here that the need to distinguish single copy heterozygous carriers from double copy homozygous carriers did not arise with DNA tests for the neonatal lethal genetic defects of Angus calves arthrogryposis multiplex (AM) and neuropathic hydrocephalus (NH), simply because no homozygous affected carriers are present in the adult test population as all homozygous carriers of the causal mutations for these diseases die before or at the time of birth.


The DNA diagnostic test developed by Prof. Jonathan Beever and his colleagues at the University of Illinois at Urbana-Champaign for congenital contractural arachnodactyly (CA) in Angus and Angus-infused cattle descended from the U.S. Angus cow Freestate Barbara 871 of Kaf is accurate with respect to both validity and precision.

In the study reported here, the CA test did not fail to detect the CA causal mutation in any animal (n=33) known to carry the CA mutation from pedigree or phenotypic evidence, nor did the test give a positive result for the CA mutation in any animal (n=210) that was unlikely to be a carrier of the CA mutation based on absence of the founder cow in the pedigree or known ancestry.

The estimates from this study of the sensitivity (100% or 98.5% depending on the statistical method) and specificity (100% or 99.8% depending on the statistical method) of this DNA PCR test for CA with respect to detection of cattle carrying at least one copy of the causal mutation are both sufficiently high for the test to be suitable for use in any control program to limit the spread of CA in the cattle population by detection of the phenotypically normal carriers of the causal mutation and elimination of the future genetic influence of their CA mutation carrying descendents from the population.


The Angus Society of Australia, the New Zealand Angus Association, several LHPA District Veterinarians in NSW, several Australian and New Zealand veterinarians in private practice and a number of Angus cattle breeders in Australia and New Zealand provided significant assistance in this study, for which I am sincerely grateful. I also acknowledge the assistance and advice of Prof. Jonathan Beever of the University of Illinois at Urbana-Champaign in testing the samples and resolving the unexpected test results from one animal in the study.  Support for the study was provided by the Department of Industry and Investment of the NSW Government.


(1) Cohen MM (1989) Int J Oral Maxillofac Surg 18:339-46
(2) Jeffreys, H (1961) Theory of Probability (3rd Ed), Clarendon Press pp 179-192
(3) Kurnit DM et al (1987) Am J Hum Genet 41:979-995
(4) Lewis JR & Sauro J (2006) J Usability Studies 1(3) 136-150

Previous Updates

August 10th 2010:

Congenital Contractural Arachnodactyly (CA) — formally known as: Fawn Calf Syndrome.

Dr Jonathan Beever of Illinois State University in the USA has been working for sometime to identify the mutation that causes the genetic defect Congenital Contractural Arachnodactyly (CA) in some Angus cattle.

Recently he was successful and has now developed a DNA test to identify the carriers of Congenital Contractural Arachnodactyly (CA) in the breed. At present Australia and New Zealand Associations are assisting him with blood samples from their countries to validate this test so it can be released for commercial use to give breeders the means to eliminate carriers.

This test for genetic defects along with AM and NH will be available through our DNA service provider PBBnz and the Massey Equine Laboratory at Massey University. Further information will be posted to our website as it becomes available.

Update on Congenital Contractural Arachnodactyly (CA), July 26, 2010, submitted by Jonathan Beever, PhD University of Illinois.

Previous Notices

Important notice regarding ‘Genetic Defects’

Please read

The Angus New Zealand (AngusNZ) Board took an initial stance of restricting future registration of AM (Arthrogryposis Multiplex) and NH (Neuropathic Hydrocephalus) positive carrier animals assuming that it was desirable to eliminate the conditions as soon as possible.

Subsequently, and complicated by the considerably more prevalent NH defect, it has been questioned whether this was an over-reaction in light of the low incidence and significance of the defect/s compared to existing physical and environmental challenges, plus the fiscal and emotional consequence – both individually and collectively.

After consideration, the Board now believes that facilitating the management of recessive genetic conditions by providing current and transparent individual status information on animals (rather than mandating the culling of them) is the more responsible way forward.  With this information it is possible for members to effectively manage the defects as their conditions and circumstances dictate, and at a pace that is individually achievable.

Therefore (and in knowledge of the fact that additional recessive defect tests will shortly become available) we have rescinded prior regulations regarding genetic defects, and replaced with the following;

Regulation Update:

  1. Only those progeny derived from sires carrying an AMF/AMFU and NHF/NHFU status will be eligible for registration for;
    (i) AM progeny born on or after 1st January 201
    (ii) For NH progeny born on or after 1st January 2011
  2. Only animals that are AMF, AMFU, NHF and NHFU can be sold with transfer.
  3. No clones of animals identified as carriers of AM/NH shall be eligible for registration.
  4. Only the results from AngusNZ approved labs will be recognized.  
  5. AngusNZ will review these regulations from time to time as new information becomes available.

 Please ring the office if you have any concerns over AM & NH.

AND remember: 


Will MacFarlane.
Angus New Zealand.