Two Stanford University researchers have found that adults in the United States who believed that they were less active than their peers died younger than those who believed they were more active, even if their actual activity levels were similar.

Alia Crum, an assistant professor of psychology, and Octavia Zahrt, a doctoral candidate at the Graduate School of Business, analyzed surveys from more than 60,000 adults from three national data sets, including one during which participants wore an accelerometer to measure their activity over a week. The surveys documented participants' levels of physical activity, health and personal background, among other measures.

The two were interested in one question in particular: "Would you say that you are physically more active, less active, or about as active as other persons your age?"

They then viewed death records from 2011, which was 21 years after the first survey was conducted. Controlling for physical activity and using statistical models that accounted for age, body mass index, chronic illnesses and other factors, they found that individuals who believed that they were less active than others were up to 71 percent more likely to die in the follow-up period than individuals who believed that they were more active than their peers.

"Our findings fall in line with a growing body of research suggesting that our mindsets - in this case, beliefs about how much exercise we are getting relative to others - can play a crucial role in our health," Crum said.

Crum's prior research shows that the health benefits people get out of everyday activities depend in part on their mindsets - that is, whether or not they believe that they are getting good exercise. In her 2007 study, Crum made a group of hotel room attendants aware that the activity they got at work met recommended levels of physical activity. Through this shift in mindsets, the workers, many of whom had previously perceived themselves as inactive, experienced reductions in weight, body fat and blood pressure, among other positive outcomes.