DNA Testing may debunk all dog bite studies that cover breed

There is a new article that just came out today via the AVMA that discusses DNA testing and what it may mean to any type of previous study that is done based on dog breeds.

It's well worth the read, and Winograd has kindly posted a copy on his site.

Essentially the article looks at how shelters look at dogs that come into the shelter and how the dogs are classified. 

"Those of us who walk through shelters and animal control facilities compare the posted breed descriptions of the dogs to what they look like to us – with frequent differences of opinions.  Those who have worked at shelters and similar facilities are aware that as dogs move through the steps in admission or during their stay that their breed descriptions may change.  It is my impression, when visiting    animal control or adoption agencies, that most medium to large size dogs with straight, short/ medium length brown hair coats are cast as German shepherds or shepherd mixes, dogs with a black spot on their tongues are designated   Chow mixes , and most medium sized, stocky, broad headed,  small eared dogs with a short hair coats are pit-bulls or pit-bull mixes."

 

That certainly mirrors how I've seen most shelters work. In fact, in Overland Park, Kansas, when the public safety committee was hearing testimonial about their potential breed ban (which eventually passed), when questioned about the city's ability to accurately identify dogs, City Attorny Mike Santos (who used a lot of misinformation in his testimony to get the city to passe the ban)even noted that for 20 years, the animal control officers have received breed identification training to recognize the characteristics of the three main breeds.  You can read the comment here, on page 18.  Note, being able to identify three breeds means squat when there are well over 100 breeds, and without having any knowledge of those breeds, you cannot identify them, and thus, everything becomes one of the three breeds.

 

Rarely are dogs that enter shelters purebred dogs -- and seldom are people allowed to just call the dogs what they are -- mixed breed dogs. Thus, they are asked to identify the dog by what does it look mostly like.   This information ends up in a database, and then ends up getting used for dog bite studies and dog bite fatality studies.

 

However, with DNA testing now at least an option beyond strictly visual idenification, studies are showing that a dog's primary breed is reportedly different on DNA test vs visual identification in 87.5% of cases.

 

So in essence, DNA testing is starting to show that any dog bite studies based on visual identification or media reports based on visual identification could have an accuracy rate of about 12% based on true, actual, DNA information.

 

And while people who have for decades now have said that media reports have shown bias based on faulty breed identification, scientific evidence is starting to agree with them on a the subject -- and by a wide margin.

 

Interestingly, most 'pit bull' advocates were very much against DNA testing when it first came out because they feared that cities would use the DNA tests as a way to go out and kill certain breeds of dogs.  Now, it may be proving what they knew all along, which is that literally thousands of dogs each year are being called 'pit bulls' and really have not DNA related to any of the 'pit bull' breeds.

 

Meanwhile -- the pictures in this post represent almost exactly what the study talks about. The first picture is a picture of Jeffery.  A gorgeous, well behaved dog that, because of his wide head and short fur, would, in most juristictions be labeled as a "pit bull".  However, he has a known ancestory - -with one parent being a purebred French Bulldog (2nd pic) and the other being a purebred Labrador Retriever (3rd pic).  However, with any dog that doesn't have a known ancestory (which is pretty much any dog that was adopted, stray or otherwise came through a shelter or rescue), that ancestory would not be known, and Jeffery would be classified as a 'pit bull'. 

 

Which is why the whole argument of breed specific legislation is failed. Even if a certain breed of dog, because of its DNA, was more prone toward aggressive behavior, it would still be so costly to accurately enforce for cities using DNA testing that it would not be feasible for cities to implement it.  And if they decide that making the determination without DNA testing and based solely on how the dog looks is a better option (which almost all do at this point), 87.5% of the dogs that are affected by the ban would not actually be the desired breed that is affected at all.

 

And if breed identification based on looks is only 12.5% accurate, then the idea of a breed ban, based solely on looks, is so unconstitutinally vague that there is no way it would stand up in court.

 

Which would take us to the natural conclusion, the one that should have been reached long ago, ordinances that are based on owner behavior are FAR more fair and effective than anything that focuses on the look of a dog.

 

It is time to end the madness.

 

 

 

Comments

[themacinator]

i am not super familiar with DNA testing, but i've heard that i'ts not very reliable- that the database doesn't include all breeds, and that dogs of known parentage can come up with bizarre mixtures of breeds. have you heard any of this?

[Brent Toellner]

Mac,

There are two different types of tests -- one is completely usesless, the other has some promise.

The swab tests (that use saliva samples) have yielded completely innaccurate results -- and also has a very small list of breeds that it covers. I think the Mars Wisdom Panel test (which uses blood samples taken by veterinarians -- which is what this study is based on) has shown some promise --although it is fair to say that the final verdict on it is not out yet. It has over 100 breeds represented in their database -- so it's fairly robust.

It's a fair question though....and one that I'm sure a lot more research will be done on.

[Sarah]

i would disagree with the statemtent "Rarely are dogs that enter shelters purebred dogs" - It's not a rare occurence at all. I think the nationwide statistic is something close to 25% of dogs in shelters are pure-bred. At our shelter, it's closer to 50%.

[Brent Toellner]

Sarah,

By "purebred" in this case, I'm talking dogs that enter the shelter with known heritage. So unless the dogs come in with paperwork, then it is only assumed by shelter staff, based on its appearance, that the dog is purebred. I may very well be that the dog is purebred, but as the data is showing, is that even folks in the rescue business aren't so good at judging dog breeds based on appearance.

So when I refer to pure-bred dogs at shelters, I'm talking dogs with known lineage.

[MichelleD]

For arguments sake, let's say DNA testing isn't accurate. If you can't tell a breed with science how the hell are people getting away with visual ids standing up in a court of law!?!?

[Lisa]

This point is key to fighting BSL. It's easy to forget sometimes when you're in the middle of arguing the million different things that are wrong with breed specific laws, but people who support BSL overwhelmingly don't know anything about dogs to begin with.

Start at the beginning. When someone says they want to ban or otherwise restrict pit bulls, stop them at "What is a pit bull, and how do you identify one?" Most of the time, they haven't even thought about it that far into the equation. It's all predicated on the notion that breed is definitive and easily identified.

This is the argument that was thrown out in Denver's ongoing case, but only because the court found that the complainants didn't have grounds to sue because they moved out of Denver. This is the argument that is going to kill BSL, though, as soon as it's heard in court. Every other point--every other fudged study or ridiculous opinion is 100% dependent on the myth of accurate breed ID.

People need to be shown that not only do a lot of dogs end up being falsely IDed as 'pit bulls' based on arbitrary physical characteristics; but also that a lot of dogs that do have a fair amount of bully breed in them look nothing like what they expect.

Go to the Mars Wisdom Panel gallery, where they show pictures of dogs they've tested, and sort the results by "American Staffordshire Terrier" or one of the other targeted breeds. I'd say a good 25% of them would never be visually identified as a 'pit bull.' And chances are pretty good that some of those dogs will look enough like your BSL defenders' dogs that they might have to reconsider their ill-informed positions.

[Victor]

In regards to your comment about the Blood draw method being better than the cheek swab testing. Isn't DNA just DNA? I have heard of a test for people who have a dog pooping in their yard, and they want ot find out what neighbothood dog it is...you send in a "sample" and they'll tell you the breed of the "offender". My understanding of DNA collection for Canine testing (even human too), is that DNA is DNA is DNA. And I do believe that the cheek swab is not collecting saliva, but rather, skin cells.

[Brent]

Victor, I think you're probably right about the skin cells vs the saliva...

As for the DNA -- yes, DNA is DNA. However, it's my understanding (and I'm not a doctor, or an expert in this in any way) that there are some ways to get DNA that lead to partial samples and others that get more complete samples -- and that a blood sample would be more complete than the swab method. Also, the swab test does not have nearly as many breeds represented in their back end database which would force the matching system to find the "best" match, even if it's not accurate. The more breeds in the system, the more closely something can get matched.

It's not that the DNA is different per se, it's that it is more complete with a fuller database under the blood sample.

That's my understanding of it anyway...

[KC KS Kills Dogs]

AMEN! AMEN! KANSAS CITY, KANSAS ANIMAL CONTROL, MAYOR REARDON, AND THE ENTIRE CITY COUNCIL THE WHOLE WORLD REMEMBERS PIT BULL AMNESTY SUMMER 2006. YOU STOLE AND KILLED 250++ DOGS IN TWO WEEKS, THAT YOU THE CITY CLAIMED WERE PIT BULLS.

WE WILL NOT FORGET THAT YOU BROKE THE HEARTS OF INNOCENT DOG OWNERS AND CAUSED THE DEATH OF MRS. MC CONNELL. KANSAS CITY, KANSAS WHO ALWAYS HAD A LOUSY AND CRIMINAL AC DIVISION, BLAMED THEIR LACK OF PROVIDING A BASIC PUBLIC SERVICE - PUBLIC SAFETY - ON A SUPPOSED DOG BREED.

[Victor]

So really, what you're saying is that the blood draw test just has more breeds available than the cheek swab test...it's not that blood is better. I have been interested in having my dog tested with one of these tests. The little bit of research I have done, it is my understanding that the blood draw test is done by a candy company, and the biggest of the cheek swab tests (that I can find) is done by an actual genetics company. I'm leaning towards the cheek swab test, just for the fact that I would feel a little more comfortable knowing that people that do actual genetic testing on a day-to-day basis, will hopefully provide better results, even though they may have a few less breeds in their database (plus, my dog hates to go to the Vet!). So we'll see!

[Selma]

Fantastic post, Brent.

"For arguments sake, let's say DNA testing isn't accurate. If you can't tell a breed with science how the hell are people getting away with visual ids standing up in a court of law!?!?"

Bingo, Michelle.

It's because the people who write these laws really don't care about 'pit bulls' - or dogs.

What they are doing is banning all dogs when you drill down and really analyze the language. This may be because they have been fed the animalib line or because they are animalibbers themselves. Either way, it's what they are doing but they use 'pit bulls' to avoid the inevitable backlash if the truth gets out.

That's why they go vague - the vaguer the better because the net is cast wide. While we lambaste them for capturing too many dogs that aren't 'pit bull' types, that is actually the intent of the legislation - to cut a huge swath through the dog population and to associate dog ownership - one of the best things around - with fear and uncertainty. And it's working.

[Lisa]

At the top levels of the BSL movement, the animal liberation movement is definitely very present. Not so much at the local levels, though. I've talked to a lot of regular schmos and low-level politicians who are behind local BSL initiatives, and almost to a one, they're just plain ignorant. They don't know what animal liberation is, and most of them have dogs that they believe are 'safe breeds.' In fact, I've had a few of them call me a PETA member as an insult.

At least among the small time BSL proponents I've met, they honestly believe that 'pit bulls' are some easily identifiable, discrete type of dog, and that they have evil supernatural powers. One guy actually told me that he defined pit bulls as mean looking dogs over 60#.

It's shocking to me that so many people have such strong opinions about something they can't even articulate, but people like that are very frequently the driving forces behind local BSL.

[PoochesForPeace]

Thank you for such useful information. And i love the pictures in this post!

[Betty]

I recently came across your blog and have been reading along. I thought I would leave my first comment. I don't know what to say except that I have enjoyed reading. Nice blog. I will keep visiting this blog very often.

Betty

http://smallpet.info

[Karin]

It isn't that blood is better than cheek cells. The blood draw test is only better because the company has more breeds in their database: 167 vs. 100 or even 58 in a couple of other company's tests. The Wisdom Panel just chose to use blood. Maybe it makes them look more legitimate. I don't know, but the DNA in the other tests isn't inferior. It's their database that's inferior.

The advertising, and the information the other companies provide is simply false. For any breed that isn't in their database, they claim their test will show an "earlier breed in the ancestry," which it can't because the ancestral breeds aren't in their database. For example, Mexican Hairless dogs are not derived from any extant breed, so their test will give you a result, but it will be 100% wrong. Send them wolf DNA and you'll also get a result, but it won't be "wolf" and it won't be "breeds farther back in the ancestry" of a wolf because wolf is as far back as dog ancestry goes. I might trust the Wisdom Panel because the database is so much larger, but the others are worse than a waste of money -- they disseminate false information that people believe because they think "DNA never lies." DNA doesn't lie, but people do.