In the 1980 study of heart attack survivors, researchers found that 28 percent of pet owners had survived to the one-year follow-up, compared to only six percent of non-owners. This finding, widely cited by the popular press at that time, continues to inform the public's impression of the benefits of pets. Here is a result we didn't hear about. The researchers had analyzed pet ownership along with seven other social and physiological variables. Pet ownership was the least important. For the statistically minded, this amounted to a discriminant function coefficient of only 0.12. Critics have pointed out that pet ownership makes a relatively trivial contribution compared to other factors included in the study such as age, severity of disease, place of residence. Subsequent research found that pet owners were more likely than non-owners were to die or be readmitted to a hospital within a year of heart attack. But you won't read about that study in the media. You won't hear about studies that found no evidence that pets are associated with cardiovascular benefits. Some studies found that pet owners had higher diastolic blood pressure, higher body mass index measures, and higher rates of cigarette smoking. Although pet owners were younger compared to non-owners, they were also slightly more overweight and engaged in less exercise. They were more likely to have diabetes and to use medication for cardiac issues. Nearly every study that finds an association between pet ownership and better health, can be matched with one finding no differences in the health of pet owners compared to non-owners. For example, whereas one study found an association between dog ownership and perceived physical quality of life, another found that dog owners didn't differ from non-owners and perceived health. One study of older adults of both sexes found that pet ownership didn't contribute significantly to explain variance in health and well-being. But another found more positive health-related characteristics among dog owners in particular. The impact of pets on psychological health and well-being is also conflicting. Ever since the initial study that found that giving older adults a pet bird to care for improve their well-being, one large study of older adults found more symptoms of depression among pet owners. Consider too, that pets are strongly associated with emergency room visits for injuries from falls, the leading cause of nonfatal injuries in the United States. Although pets present a fall hazard for people of all ages, those over 75 have higher injury rates and more injuries occur among women than men. Seven times more fall injuries are associated with dogs compared with cats. In the United States, a person is 100 times more likely to be seriously injured or killed by a dog than by a venomous snake. Further, people can contract a number of diseases from companion animals including roundworm, E. coli, salmonella, giardia, ringworm and cat scratch fever. Most of us are delighted to learn that having pets keeps us healthy and happy. Before we consider why we didn't hear about the studies that find little or no benefits from having pets, let's look first at why the results are so often inconsistent. Conflicting results in the research occur largely because the methodologies used differ substantially. Varying research designs compromise the methodological rigor required to understand the direct benefits of pets. Results don't always reach statistical significance, which indicates the likelihood that a relationship between two or more variables is caused by something other than chance. Statistical studies are often underpowered, having sample sizes too small to detect even a moderate effect. Many use cross-sectional designs which may establish associations but can't establish causality. Some reveal few details about samples and provide no measure of the size of the effect they claim to find. Very few studies manipulate pet ownership using a quasi-experimental design. Although it would be impossible to blind a person to the presence of a pet, increased rigor could still come from other methodological strategies, including randomization and longitudinal designs. Many studies rely on self-reports by pet owners who volunteer to participate, raising questions about confirmation bias or the tendency report information that confirms what we'd like to be true. Researchers too can bring threats to validity by having a stake in the outcome of their studies. As pet lovers, some might come to the research already confident of the positive influence of pets on health. Finally, funding for some studies comes from pet related sources which have an investment in finding the benefits of pet ownership. Now, why don't we hear about studies that find the pet spring no benefits to human health and happiness? Psychologist Hal Herzog attributes this to what's called the file drawer problem, which is found in many areas of research. This is the tendency for studies with negative results, meaning those finding no statistical significance or causal relationship to remain unpublished. Scientific journals prefer to publish positive results. Studies with negative results end up in the researcher's filing cabinet or on a thumb drive in current terms, rather than in print. They're never seen by others and they aren't picked up by the media. This problem has consequences for researchers and for the public. When a researcher reviews the existing literature, the publications available suggests stronger effects than actually exist. Because insignificant results have been consigned to the file drawer, the body of knowledge itself is biased. For members of the public who learn about research mainly through the media, we've been getting only part of the story. We hear only about the healing powers of animals. People have enjoyed the company of pets for a very long time, and there are many good reasons to bring a dog or a cat into your life. Their healing powers just may not be one of them.
