VanderWeele said that other studies have suggested a similar link between service attendance and decreased mortality, but his team aimed to prove that service attendance actually causes the better health outcomes.
Because the nurses answered questionnaires periodically over a long time frame, he said, the researchers were able to look at whether a change in service attendance led to a change in health.
They found numerous benefits associated with attending services. Women who started going to services then became more likely to quit smoking and less likely to show signs of depression, for instance - even when the researchers controlled for a long list of other variables, from age and exercise habits to income and other non-religious social engagement.
The effect of religious attendance, they found, was stronger than that of any other form of participation in a social group like a book club or a volunteer organization.
"We were a bit surprised, initially, by the magnitude of the findings," VanderWeele said. He said they found a long list of positive effects: "Service attendance is increasing social support. Through social norms, it's also decreasing the likelihood of smoking. Perhaps through some of the messages of hope, it's decreasing depressive symptoms. Perhaps self-discipline, a sense of meaning or purpose in life - it's not just one pathway."
He continued, "I don't think it's one single reason that this effect is emerging. I think it's that service attendance affects so many different aspects of life."
Richard Sloan, a Columbia medical school professor who has expressed skepticism of similar studies in the past, said these latest results should not lead doctors to talk about faith in the examining room. "Physicians threaten to compromise the religious freedom of patients to make decisions about religious practice on their own," Sloan said. "It's perfectly reasonable for physicians to make medical recommendations which they expect patients to follow... When physicians stray from a medical agenda to some kind of social agenda, it's a violation of the patient's autonomy."
But VanderWeele said doctors should be aware of the apparent benefits of religious attendance.
One of the team's most striking findings was on breast cancer. Women who attended services were no more or less likely to contract breast cancer. But those who attended services were substantially less likely to die of it.
"We were quite struck by that," VanderWeele said. "Maybe it is a sense of hope or of faith, even in the face of illness and disease. A capacity to try to find meaning in the disease experience. Or feeling supported by a community even while struggling with illness. That would be my speculation as to those results. But I do agree it was surprising."
Daniel Hall, a University of Pittsburgh medical professor not involved in this study who trained as both a doctor and a minister, said that pious people might see this study as affirmation that there is a God listening to the prayers at those worship services, and others might see non-faith-based explanations. "Human beings are so religious in their behaviours. Quite apart from a truth clam in whether there is a God or not, it's just anthropologically one of the strongest ways human communities are held together," he said.
Hall said that just as doctors learned in recent decades to be less squeamish about asking about patients' sex lives, since the information can have medical value, physicians should keep patients' faith lives in mind.
"Talking about people's religious beliefs and practices is one of the last taboos," he said.
"Something's going on there that we ignore. You see something that has that much of an effect? In what way would it be appropriate to come alongside it and understand it - even if you can't manipulate it in the same way as blood pressure."
