Although research has shown that people who have greater ties to family, friends, church and community are more likely to ward off the common cold--and may even recover from illness more quickly--health professionals have largely ignored how the quality of those relationships influences a person's sickness or health, believes Sheldon Cohen, PhD, a psychology professor at Carnegie Mellon University known for his research on the association between social relationships and immune function.

"The correlational research in this area has very clearly shown that when you have strong social support from your friends and family--from your natural social network--that you are protected from stress-induced illness," said Cohen, in a presidential-track talk at APA's 2004 Annual Convention in Honolulu that doubled as his 2004 APA Distinguished Scientific Contributions Award address. "Interestingly, research on interventions has ignored this literature and has focused almost entirely on whether providing support from strangers improves our health."

In designing health interventions, Cohen maintained that psychologists and others should take advantage of people's "natural networks" of support--their friends and family, their co-workers, people they volunteer and go to church with--rather than relying solely on the support of strangers in artificially constructed support groups.

In fact, he said, interventions should focus on building people's existing relationships with friends and family, developing and strengthening ties with their communities and reducing conflicts in their social relationships. And more research is needed to identify the types of interventions that work best for people with different social environments, Cohen added.

How social relationships get under the skin

Cohen, the Robert E. Doherty Professor of Psychology at Carnegie Mellon, has spent much of the past 25 years unraveling the impact that stress, social relationships and social characteristics have on people's immune system and their susceptibility to infectious illness.

For example, in his well-known Pittsburgh Common Cold Study, Cohen surveyed 273 Pittsburgh-area participants ages 18 to 55 about their active social roles, such as close friend, neighbor, church member, student, teacher, wife, father and husband, and found that more socially isolated people--those with only one to three active roles--were four times more likely to get sick when exposed to an experimentally administered cold virus than people who reported a high level of social integration and relationship participation, or had six or more active social roles.

However, additional studies by Cohen and his colleagues have revealed that our social environment isn't always good medicine. In one such study, Cohen looked at how people's levels of chronic social conflict--or having a stress-invoking, demanding relationship with a spouse, friend or family member affected their susceptibility to the cold virus. He found that those who had experienced a month or more of such conflict were 2.5 times more likely to get sick when exposed to the cold virus than those who hadn't.

Cohen found a similar negative effect on health when he looked at loneliness or when people's actual social networks didn't meet their expectations. In that study, he found that people who scored high on measures of loneliness produced substantially less protective antibody in the weeks following a flu vaccine compared with less lonely people.

Toward better interventions

These findings and others have led Cohen to conclude that multiple aspects of social environments must be considered in health-intervention design.

In fact, Cohen collaborated on a study with colleagues Vicki Helgeson, PhD, and Richard Schulz, PhD, in which they randomly assigned breast cancer patients to one of four groups: a group that provided emotional support for one another, a group that received information about cancer through lectures, a group that received both information and emotional support and a no-treatment control group. Interestingly, those in the peer-support group fared worse than all of the other groups on measures of physical and mental health. However, he noted, a closer look at the data revealed that some participants did benefit from the peer support--those with previous low levels of social support and high levels of conflict with families and friends.

"Those with strong levels of social support were actually harmed by peer support," he explained. "They started comparing their family, their natural support group, to the artificial support group."

"It's a mistake to use just a social support kind of approach," since it's just one of the social factors that influences health, he noted. "We need to increase social integration and decrease negative interactions, and we need to identify those who benefit most and least from various types of interventions."