Do you think that online social networks have no bearing on your life? Think again. Scientists who studied Facebook activity and mortality rates of registered California voters found that people who received many friend requests were far less likely to die over a two-year period than those who did not. Initiating friend requests, however, seemed to have no effect on death rates.

The findings, published in the Proceedings of the National Academy of Sciences, hint at deeper complexities in the relationship between humans’ health and their social networks, whether those networks are online or in person.

Senior author James Fowler, a social scientist at the University of California at San Diego, has spent years studying the link between human well-being and real-world social networks, including how happiness and even obesity may spread through them. But he and his colleagues wondered if perhaps online networks could also be connected to health.

“We’ve known for a long time — for decades now — that offline social networks, especially social integration, [were] related to longer life,” said lead author William Hobbs, who is now at Northeastern University. “But we didn’t know if that extended to online interactions, too.”

To try to get at this question with a large and reliable sample, the researchers took 12 million Facebook users and matched them to California vital statistics as well as the voter registration database. The data was anonymized, and the scientists checked how many had passed away over two years of follow-up. (All people in the study were born between 1945 and 1989, and all comparisons were made between people of about the same age and gender.)

The scientists found, to their surprise, that there was no correlation between how many friend requests people sent and their longevity. But there was a clear link between the number of friend requests they accepted and how long they lived. People who received and accepted the most friend requests were 34 percent less likely to have died than those who received and accepted the fewest friend requests.

Those who posted a lot of photos showing real-world interactions also had a lower risk of death — a sign that face-to-face interactions were linked to better health. Other activities showed complicated results: Writing wall posts and sending messages in moderation seemed to be linked to lower mortality, but writing very few or very many was not.

Scientists have long argued that having people expand their social networks might improve their health, because of the ­well-known connection between the strength of social connections and well-being, Fowler said. But studies of real-world interactions can’t easily tell which of two people initiated the friendship, and this appears to be key, he added.

The findings, then, seem to show that simply trying to make more friends (whether face to face or online) might not have much effect on your health.

To be clear, the scientists noted, this study shows only a correlation — there’s no way to tell what the cause may be. It may be that healthier people get stronger networks, not the other way around. There may also be an unknown factor that affects both health and social network strength.