When I first graduated from vet school, I came into our profession armed with all of the most up to date knowledge at the time. However, I also immediately recognized that we didn’t have the same caliber of science guiding our practices as the human medical world. This was one of several reasons I pursued a Ph.D. – I wanted to contribute more to the science in our field, and I wanted to do it well.
After finishing my Ph.D. in immunology last year at a well-respected biomedical institute, I turned my attention back to the latest controversies in veterinary medicine. One area that immediately caught my attention was the research into the long-term health effects of age at the time of spays and neuters. Veterinarians cannot agree on an ideal age for sterilization based on the current literature, and I believe it is because we are lacking high quality studies.
Some of the most recent studies have heavily impacted veterinarians’ recommendations to clients, and yet to my eye they are far from definitive. We all want to guide our clients to make the best decisions for their families and their lifestyles, but more importantly we want to take our patients’ long-term health into account. Of course any studies showing an impact of early spays and neuters in dogs are going to affect our recommendations, but I would like to urge my colleagues to evaluate these reports with a critical eye, because they haven’t given us the data we really need to make informed recommendations. Here are 3 reasons I would urge caution when using these studies as a basis for our recommendations:
They are retrospective.
The strength of retrospective studies is that large amounts of data can be collected in a relatively short period of time because they are based upon previous clinical cases. However, medical records notoriously lack important details and frequently have incomplete information.
Prospective studies, where case criteria and data collection standards are defined prior to data collection, are the gold standard for establishing true clinical correlations, and comparisons in the human medical literature have shown that prospective data are inherently more accurate than retrospective data (Clark M, 2008 ).;
Correlation does not imply causation.
Both the incidence of diabetes and the number of people practicing yoga in the United States have increased in the past 10 years, but that does not mean yoga causes diabetes. Likewise, an increased incidence of cancer or orthopedic injury in neutered dogs does not prove a causative link.
It is fair to use these studies as a basis to explore the matter further, but veterinarians have started making clinical recommendations based upon a HYPOTHESIS. That’s right – even though it is published, it is still only a hypothesis that large breed dogs suffer an increase in CCL rupture and hip dysplasia due to early spaying and neutering. There are reasonable underpinnings to the hypothesis – the growth plates close long after 6 months of age in large breed dogs, and sex hormones can have an effect upon growth plates. However, no one has definitively proven that the removal of sex hormones at 6 months or younger is actually a causative factor in these maladies. The evidence is still circumstantial at best.
The problem with correlations is that they frequently don’t account for all of the other variables that might have an effect upon the measured outcome. For example, while median ages and body condition scores are mentioned in the text of the recent papers from UC Davis, no statistical tests to evaluate if they are different between the sterilized vs. intact groups are performed. You can have two groups with the same median that are still skewed in completely different directions. I find this problematic since both of these variables are known risk factors for cancer and joint disorders.
There are questionable statistical methods for establishing a causal relationship.
Most of our current studies on this topic suffer from the problem of small sample sizes, which can lead to a statistically significant effect when there really isn’t one (Button et al, 2013 ). Additionally, the authors of the UC Davis studies use hazard ratios to calculate the risks of specific outcomes (i.e. cancer, CCL rupture, etc.). While this statistical test is used frequently for clinical trial data, hazard ratios are prone to their own built-in selection bias (Hernan, 2013 ). Therefore, they are not well-suited for establishing a causal relationship.
There are other criticisms one can make about these studies, such as whether the patient population at a referral center is representative of all dogs. However, since I don’t have any personal insight into what kind of bias that may present, I’m focusing on what I can glean from the methods in the studies themselves.
The bottom line is that our current evidence base on this topic underwhelms me. Sadly, many of the benefits we tout for spaying and neutering also suffer from the same issues I have outlined above (e.g. the often-cited association of age of sterilization with reduced mammary cancer is also a retrospective and correlative study on a small sample set). Thus, right now I choose to go back to basics when making recommendations to clients about the age of sterilization.
We do know that spaying and neutering are important for preventing pet overpopulation, and they tend to work best for this endeavor when performed before sexual maturity. There are also pathologic conditions that we know sterilization will prevent, such as pyometra, benign prostatic hypertrophy, and prostatic abscesses, all of which cause high morbidity and mortality.
Finally, we really don’t need a scientific study to know we experience fewer surgical complications when performing sterilization prior to sexual maturity. I know I’m not the only veterinarian that has sweated bullets while wading through bloody abdominal fat during the spay of an adult female Rottweiler, praying that my ligatures stay in place around massive ovarian and uterine arteries.
I am not suggesting that we should ignore long-term health ramifications for sterilization, but we need to explain to our clients that the studies we currently have are not well controlled or definitive. I am also not suggesting that the conclusions of these studies are necessarily wrong; further research could still validate the conclusions, but my emphasis here is that further studies are needed. We are now taking steps to collect some prospective data, such as that generated by the Golden Retriever Lifetime study, and my genuine hope is that the new data will provide us with some clearer answers.
The views and opinions expressed in this article are those of the author and do not necessarily reflect the position of the DrAndyRoark.com editorial team.
ABOUT THE AUTHOR
Dr. Sarah LaMere is originally from Austin, Texas. After graduating from Cornell University School of Veterinary Medicine, she decided one graduate degree just wasn’t enough and completed her PhD in immunology from the Scripps Research Institute in San Diego. While pursuing her passion for OneHealth in a research setting, Dr. LaMere also works in clinical practice as a hospice veterinarian with Paws into Grace in San Diego, California.