Paw Print Genetics offers tests that can be categorized into two types. Most tests offered directly test for a DNA change (or mutation) that causes a disease. For a small number of diseases, we test for a mutation that increases “the chance” that a dog will develop a disease. These have been termed risk variants. Recently, we have been getting a lot of questions about these risk variants and what a positive result means for your dog. One example is dermatomyositis (DMS) testing, which generates an associated risk (low, moderate, high, or unknown) for this skin condition. Chondrodystrophy (CDDY) is another example in which the test is for a mutation that causes abnormal cartilage formation and having the mutation may put a dog is at an increased risk for intervertebral disc disease (IVDD). If you have a Labrador that carries one or two copies of the ATP7B mutation for copper toxicosis, this mutation puts a dog at a greater risk of developing the disease compared to dogs without the mutation. This risk may be mitigated if the dog also has one or two copies of the ATP7A protective mutation for copper toxicosis, which may reduce the risk when also inheriting the ATP7B risk variant. The word “risk” is something we intuitively understand, but at the same time generates thoughtful questions about what the risk is for the dog. The answer to these questions is the information we need to make actionable decisions about our dogs. In this blog, I will dissect what risk means in canine genetic testing and how you can apply this to the test results that only provide a risk assessment. In this first blog, I define risk and explain it in statistical significance. In a later blog, I further expound the use of risk as it applies to specific genetic testing.
Risk in simple terms is the likelihood or probability that an event will occur. Risk can be further divided into two concepts: absolute risk and relative risk. The absolute risk is the probability of an event within a population of interest. The relative risk is the probability of the event in an experimental group relative to a control group. In canine genetic testing, risk often translates to whether a dog will develop a disease. The absolute risk is for any dog to develop a specific disease, whereas the relative risk is for dogs with a specific disease-causing mutation will develop the disease as compared to dogs clear of the same mutation.
Many times, veterinary research cannot determine absolute risk because researchers are working with a limited sample size and cannot access every member of a particular breed. Therefore, it is easier, and sometimes only possible, to generate a relative risk for dogs carrying a specific disease-associated mutation. The downside is that interpreting this relative risk in terms of what that means for your dog is problematic.
Take the following optical illusion for example:
Which inner circle is bigger? The one on the left looks bigger, but this is just an illusion as the circles in the middle are the same size. The small circles on the left make the middle circle look relatively big, while the large circles on the right make the middle circle look relatively small. The point being is information needs to have perspective. Relative risk interpreted without appropriate perspective can be misleading.
Drug manufacturers love relative risk because the results tend to be more dramatic. If I told you a drug was 30% better at lowering the risk of disease, that seems significant. However, if the absolute risk was really 9% and adding the medication lowers that risk to 6%, does it still seem so remarkable? Many researchers believe that a relative risk, without the context of absolute risk is incomplete. However, due to logistical limitations of veterinary research, relative risk is sometimes all we get.
When we apply these concepts to genetic tests it helps impart perspective towards the results. With this perspective, you can judge what the implied risk is, and what that means for the dogs within a breeding program. One of the fundamental uses of genetic testing is helping decide which dogs to breed together. This is straightforward when the test result is for a disease-causing mutation. Breeding two dogs that carry a recessive mutation that causes disease should be avoided to prevent producing affected dogs. Knowing a dog has one copy of a dominant disease-causing mutation, tells you that dog may develop the disease and breeding that dog to a dog clear of the mutation still carries a 50% chance of producing an affected dog. However, what do we do with a test result that effectively says, your dog carries a mutation which places them at greater risk for a disease? Can we breed this dog? What are the risks for that individual dog? In the next part of this blog on risk I want to break down specific individual tests that are risk assessments and possibly answer these questions.
Woloshin S, Schwartz LM, Welch HG. Know Your Chances: Understanding Health Statistics. Berkeley (CA): University of California Press; 2008. Chapter 2, Putting Risk in Perspective. Available from: https://www.ncbi.nlm.nih.gov/books/NBK126167/