In the story of America’s opioid crisis a recent tripling in prescriptions of the painkillers is generally portrayed as the villain. Researchers and policy makers have paid far less attention to how social losses—including stagnating wages and fraying ties among people—can increase physical and emotional pain to help drive the current drug epidemic.

But a growing body of work suggests this area needs to be explored more deeply if communities want to address the opioid problem. One study published earlier this year found that for every 1 percent increase in unemployment in the U.S., opioid overdose death rates rose by nearly 4 percent.

Another recent study from researchers at Harvard University and Baylor College of Medicine reported U.S. counties with the lowest levels of “social capital”—a measure of connection and support that incorporates factors including people’s trust in one another and participation in civic matters such as voting—had the highest rates of overdose deaths. That review of the entire U.S. mined data from 1999 through 2014 and showed counties with the highest social capital were 83 percent less likely to be among those with high levels of overdose. Areas with low social capital, in contrast, were the most likely to have high levels of such “deaths of despair,” with overdose alone killing at least 16 people per 100,000

Overdose is now the nation’s leading cause of death for people in the prime of life. And suicide- and alcohol-related deaths have also risen—most dramatically in regions with the highest levels of economic distress. “It will be hard to address the addiction and overdose crisis without better understanding and addressing the neurobiology linking opioids, pain and social connectedness," says Sarah Wakeman, medical director of the Substance Use Disorder Initiative at Massachusetts General Hospital and an assistant professor of medicine at Harvard Medical School.

Connecting opioid use to social stress is not a new idea. Forty years ago the late neuroscience pioneer Jaak Panksepp first proposed the now widely accepted hypothesis that our body’s naturally produced opioids—endorphins and closely related enkephalins—are critical to the nurturing bonds that develop between parents and offspring and also between monogamous mates in mammals. Panksepp’s work and that of others showed that blocking one opioid system in the brain—which relies on the mu-opioid receptor—increased the distress calls of infants separated from their mothers in species as varied as dogs, rats, birds and monkeys. Giving an opioid drug (in doses too low to produce sedation) reduced such cries.

Panksepp also observed similarities between maternal love and heroin addiction. In each situation animals would persist in a behavior, despite negative consequences, in order to gain access to solace from the partner—or the drug. But, as Panksepp (who died in April) said in an interview several years ago, major journals rejected his paper in the 1970s because editors said the idea that motherly love was similar to heroin addiction was “too hot to handle.”

Since then, however, data supporting the link between opioids and bonding has only grown. It has been expanded on by researchers including Thomas Insel, former head of the National Institute on Mental Health; Robin Dunbar at the University of Oxford; and Larry Young, professor of psychiatry at Emory University.

Young showed that oxytocin, a hormone previously linked mostly with labor and nursing, is crucial to the formation of pair bonds as well as bonds between parents and infants. “The feelings that infants or adults feel when being nurtured—warmth, calmness and peacefulness—come from a combination of opioids and oxytocin,” he says. “These are the same feelings that people who take opioids report: a feeling of warmth and being nurtured or loved.” When a social bond is formed, oxytocin reconfigures the mu-opioid system so that a loved one’s presence relieves stress and pain—and that person’s absence, or a threat to the relationship, increases distress. For example, Young’s research shows normally monogamous prairie voles do not develop pair bonds with their mates if their mu-opioid system is blocked; other studies have found that mice genetically engineered to have no mu-opioid receptors do not prefer their mothers to other mice the way normal baby mice do.

A paper Young published this month, in collaboration with researchers at the University of Tsukuba in Japan, found prairie voles that have bonded with a mate not only experience more anxiety when separated from their partners—they also experience more physical pain during the separation, by various measures including response to a painful injection and pain from heat. “Bonding somehow changes your pain threshold—so if you lose that bond, then your pain reduction and natural analgesia is lost,” he says. This paper did not measure mu-opioid receptor binding, but other work with humans strongly suggests it is involved.

Recent human studies have specifically found that a partner’s presence can reduce pain, and supportive touching such as hugging is linked to activation of mu-opioid receptors in the brain. In addition, a study published last year found that administering an opioid blocker decreased people’s feelings of social connectedness—both when they were in the lab receiving e-mails of support from close friends or relatives and when they were at home during the four days they took the drug—compared with when they took a placebo. And, whereas the drug reduced overall levels of positive emotion, it had a larger effect on positive emotions related to feeling connected and loved.

All of this suggests that recognizing the connections between bonding, stress and pain could be critical to effectively addressing the opioid crisis. “Understanding the biology and commonalities between trusting social relationships and the opioid system can change the way we think about treatment,” Young says, noting that neither the punitive approach of the criminal justice system nor harsh treatment tactics are likely to increase connectedness. In essence, if we want to have less opioid use, we may have to figure out how to have more love.