Everybody knows that. Or do they?

Since the 1980s, research into the link between healthy children and having a pet has supported the common wisdom that pets are good for kids, both emotionally and physically.

For instance a 1995 study published in the Journal of Applied Developmental Psychology indicates that living with pets can have a positive impact on a child’s self-esteem. A 2002 study published in the Journal of the American Medical Association concluded that exposure to two or more cats or dogs in early childhood may reduce the risk of allergic sensitivity to multiple allergens during childhood.

And in 2005, the journal BMJ published a literature review that concluded it’s likely “pet ownership itself is the likely cause of the reported benefits,” because “no studies have found significant social or economic differences between people who do or do not have pets that would adequately explain [these] differences in health.”

Until now.

Because a recent study by the RAND corporation turns the truism that having pets positively impacts kids’ health on its head

It’s a study that poleaxed even the people involved.

“Everyone on the research team was surprised,” said Layla Parast, a statistician at RAND and a coauthor of the study. "Given our personal relationship with our current and past pets, we all truly believed that the results of our study would simply demonstrate the obvious—that pets are good for kids.” But, “we could not find evidence that children from families with dogs or cats are better off either in terms of their mental wellbeing or their physical health.”

The big difference between the RAND study, published in August in Anthrozoös, a multidisciplinary journal of the interactions of people and animals, and earlier studies is a matter of depth: the RAND study dug deeper.

Researchers factored in more than 100 variables when constructing their model of pet ownership and children’s health, including household income, language skills, and type of family housing. The study analyzed information from over 5,000 households collected as part of the 2003 California Health Interview.

And while the initial results supported the earlier studies—children in pet-owning homes were significantly healthier than children in non-pet-owning homes in areas like better general health and higher levels of activity—the results changed when all the variables were factored in: the effects were smaller and, more importantly, “no longer statistically significant.”

The study pins the difference on “confounding factors.” In statistics, a confounding factor is a variable that influences other variables, causing a “spurious association.”

Which can lead to false assumptions. Like, owning a pet is good for kids’ health.

And when you factor in other small but potential health risks associated with pet ownership—800,000 people a year are treated for dog bites in the US, at least half of which are children—the health benefits to kids get even murkier.

But before you start worrying that everything you think you know about kids and pets is wrong, consider this: the study concedes that the research has some limitations. The data were collected from a geographically limited area, despite the relatively large sample size. And the survey did not assess things like the length of pet ownership or the level of interaction with the pet, both of which could significantly impact the results.

And as coauthor Parast herself noted, “I would not say that our results completely disprove [the link between pet ownership and kids’ health], but they show that the benefits are likely to be quite small compared to other factors in a child’s life and perhaps be longer term.”

For example, households with pets also tended to have other characteristics associated with better physical and mental health, such as higher income and home ownership.

Parast said that when researchers looked at their own households, the dissonance disappeared. “We realized that we fit these characteristics almost exactly.”