Cover Story

Psychologists are lead players in a booming field that only 15 years ago was considered on the fringe of science: psychoneuroimmunology, or PNI, the examination of the relationship between stress, the immune system and health outcomes.

"The sheer weight of the studies in PNI--both animal and human--has really been very convincing," says Janice Kiecolt-Glaser, PhD, of The Ohio State University, who along with her husband Ronald Glaser, PhD, is one of the field's pioneers. "There's much more acceptance among basic scientists and clinicians that the links between stress, the immune system and disease really do mean something."

PNI researchers began their studies in the 1960s making broad-brush associations between stress and immune functioning. Today's investigators are increasingly sophisticated and specific in their work. They're examining how psychosocial factors such as optimism and social support moderate stress responses; they're mapping the biological and cellular mechanisms by which stress affects the immune system; and they're testing new theories, such as the idea that the immune system acts as a "sixth sense" that gives the brain valuable information about a person's health status.

And the field is becoming truly interdisciplinary, as basic scientists join forces with more socially oriented researchers.

"PNI is no longer this little area off by itself," comments University of California, Los Angeles (UCLA) researcher Shelley E. Taylor, PhD, who is widely known for her optimism research and a new theory of the female stress response. "It's become a big area, and lots of people are working in the field who at one time in their careers might never have gone near it."

While researchers still have further to go to show definitive links between PNI variables and disease and to devise targeted interventions based on these findings (see articles on the following pages), they have turned up promising discoveries on the mind-body connection and its potential to fight AIDS, cancer and the common cold. Below are some of the field's recent findings.

A 'bi-directional' model

One of the newest PNI theories proposes a flow of information that hasn't been posited before: a data link from the immune system to the brain rather than the other way around, says Steven Maier, PhD, director of the Center for Neuroscience at the University of Colorado at Boulder.

"It says that things that go on in the body can have potent effects on the brain's activity," he explains.

According to the theory, the immune system is itself a messenger that alerts the brain to infection or injury by releasing proteins called proinflammatory cytokines. These cytokines, research is finding, travel to the central nervous system and brain to communicate information about the body's distress. In turn, the brain releases its own cytokines that signal the central nervous system to initiate a cascade of responses such as fever and listlessness that, theorists believe, help the body adapt by reducing energy output. (PNI researchers are, in fact, increasingly interested in cytokine activity because it represents an immune response that can go awry.)

The work also raises intriguing questions about the origins of psychological and cognitive states such as depression and memory dysfunction. Researchers posit that these states may have arisen as part of the sickness response to conserve energy during times of infection. Maier and others are investigating that idea through animal studies.

So far, they have found that mice exposed to tail shocks show the same types of plasma-level changes that infection produces, and that these changes fail to occur if cytokine activity is blocked in the brain. In addition, they've shown that brain-based cytokines interfere with the consolidation of memory and induce depressed behaviors, providing further evidence that infection and stress responses are linked, Maier notes.

A new theory on stress and its impact

PNI researcher Margaret Kemeny, PhD, of the University of California, San Francisco (UCSF), and colleagues at UCLA are testing a theoretical model that's challenging Hans Selye's well-known general theory of stress.

Selye's theory maintains that no matter what the stressor, people respond with a similar set of hormonal and immune changes. The California team posits instead that there are two stress reactions: the classic "fight-or-flight" response and a withdrawing reaction where a person pulls back to conserve energy, says Kemeny, who directs UCSF's health psychology and behavioral neuroscience program. These reactions have been noted in animals and Kemeny's team is now extending them to humans.

She and colleagues have been examining a cognitive angle to this theory as well. They're finding that people respond to the same stressful event in different physiological ways depending on how they appraise it. Take the example of a loved one's death. Depending on how they perceive the loss--as sad but normative or as a major threat--they will either withdraw or become agitated, explains Kemeny. "The body is responding to what's going on in the brain, not to what's going on in the environment," she explains.

The researchers also have been investigating which types of cognitive appraisals most affect people's health. In a study conducted on HIV-positive men before the advent of protease inhibitors, Geoffrey M. Reed, PhD, and colleagues including Kemeny and UCLA's Taylor found that asymptomatic HIV-positive men who held negative though realistic beliefs about their disease progression, and who had lost a close friend or partner to AIDS within the last year, developed AIDS symptoms more quickly than HIV-positive men with a more optimistic outlook on their prognosis or men who had not been bereaved.

In addition, Kemeny is finding that negative self-appraisal--blaming one's self for negative events--is linked to poorer health outcomes, such as faster rates of CD4 T-cell decline in patients with HIV. These findings are discussed in a summary article by Taylor, Kemeny, Reed and colleagues in the American Psychologist (Vol. 55, No. 1).

The findings are pioneering in two ways, notes Reed, now the APA Practice Directorate's assistant executive director for professional development. For one, they imply that cognition itself may play a role in health outcomes, challenging the traditional view that emotions mediate between psychological factors and immunity. "We're suggesting that there may be something about cognitive processes that has a direct effect on the immune system," he says.

They also call into question a view commonly touted by mental health practitioners: that a realistic acceptance of one's situation is synonymous with mental health.

"Our data show that people who have a life-threatening illness can say these unrealistic, positive things about their future and still behaviorally do things to take care of their health," Reed says. "Being overly optimistic may actually be healthy." Future research could test some of these findings in interventions, he suggests.

Personality variables in HIV progression

UCLA's Steve Cole, PhD, is using the same data set as the UCSF and UCLA researchers to study personality factors in disease progression. In particular, he's examining why some gay men may hide their sexual orientation in the tolerant city of Los Angeles, and why those same men who have HIV might become ill more quickly than HIV-positive peers who are out of the closet.

Studies by Cole, Kemeny, Taylor and others show that the men who stay closeted share characteristics of the nervous system that render them especially sensitive to social rejection. And the physiological correlates of this introverted personality type--in particular a greater tendency to go into "fight-or-flight" mode when facing stress--may make them more vulnerable to attack by HIV, Cole contends.

Cole has supported his theories through studies on the men's stress hormones, immune-cell functioning and psychological characteristics. He's also thinking about interventions that might help reduce these men's automatic stress responses. Scientists know, for example, that enhancing brain levels of serotonin can mitigate people's tendencies toward shyness. So, one intervention, Cole speculates, is medications that increase serotonin activity; another is stress-reducing behavioral interventions like meditation or biofeedback.

Tend and befriend

In still another theory that challenges old views of stress, UCLA's Taylor has proposed a model of the female stress response that she calls "tend and befriend," in contrast to the "fight-or-flight" model. Taylor has found that in response to stress, females tend to engage in nurturing activities that protect themselves and their offspring. She believes these patterns have hormonal underpinnings, and she's particularly interested in the role of a pituitary hormone, oxytocin. While the hormone is found both in males and females, it's strongly linked to maternal behavior in females, and may foster the tend-and-befriend pattern. For example, oxytocin's calming effects are enhanced by estrogen, but antagonized by androgen, she notes. Taylor first delineated her theory in Psychological Review (Vol. 107, No. 3) that's since been updated in a book by Taylor released this March, "The Tending Response" (Holt).

Now she is exploring the links between estrogen and oxytocin by looking at postmenopausal women on estrogen-replacement therapy compared with those who are not. Preliminary evidence finds that those on the treatment are better able to manage stress, in keeping with the theory's contention that estrogen augments the effects of oxytocin.

Stress and marriage

Another recent finding by PNI researchers is exploring the impact of marriage on health.

A study in press at APA's Journal of Clinical and Consulting Psychology by The Ohio State University's Kiecolt-Glaser, Cynthia Bane, PhD, Ronald Glaser, PhD, and William B. Malarkey, MD, reports on a longitudinal study in which the team took blood samples of married couples to test stress factors at the beginning of the study, and then examined them in relation to marital status 10 years later. They found that couples who divorced at 10 years had already displayed a 34 percent higher rate of the stress hormone norephinerphine at the beginning of the study than couples who stayed married.

In addition, stress hormones were higher in the conflicted couples at the beginning of the study even when they weren't arguing, suggesting a chronic pattern of stress in these marriages that started early, the study finds.

"These findings show us in a microcosm how close personal relationships can get translated into health outcomes," comments Kiecolt-Glaser. "They show that marital quality is as good a predictor of subsequent health as any of the usual major prognostic indicators."

Susceptibility to colds

Sheldon Cohen, PhD, of Carnegie Mellon University, was the first to show a link between stress and upper-respiratory infections. Now he's plying that research in two new directions, one physiological and the other social.

In an in-press study in Psychosomatic Medicine, Cohen and colleagues report that people in the lab who displayed high levels of the stress-related hormone cortisol in response to a task that provoked a defensive reaction were more likely to get sick than those who didn't react as strongly. The finding supports the theory of a "reactive" personality style that seems to be linked to health problems, including heart disease, he notes.

Another study in press in Health Psychology by Natalie Hamrick, PhD, of Carnegie Mellon, Cohen and colleagues looks at how real-world variables play into Cohen's earlier lab findings that social support provides greater immune protection from colds. The new study suggests that a well-rounded social life may bear a cost, too: Well-connected people who reported a lot of stress became sick more often than anyone else, the study finds. In keeping with Cohen's earlier results, however, nonstressed people with diverse networks got sick less often than their less-connected counterparts who didn't report stress. The findings support the hypothesis that people with more diverse social networks may be more vulnerable to illness than others when under stress because they're exposed to a wider range of pathogens.

"These are important findings, because in the 1960s, everyone was studying what was bad about social relationships, and in the 1970s, they were saying that social relationship are always good for you," Cohen says. "Our results show that both are true."

Language, connection and health

At the University of Texas at Austin, James Pennebaker, PhD, is conducting a new series of studies showing that the way people use language--in particular pronouns like "we," "I" and "you"--is highly related both to personality styles and to health variables such as depression, suicide-proneness and health-center visits.

"I never in a million years thought I'd be talking to anybody about pronouns," he chuckles. "But these pronouns are very important because they give us a sense of how socially connected people are, the degree to which they're self- or other-focused."

A study by Sherlock Campbell of the University of Texas at Austin and Pennebaker, in press at Psychological Science, demonstrates that people who are more flexible in their use of pronouns when writing expressively about traumatic events in their lives show greater health improvements, as measured by number of physician visits, than those who consistently used the same patterns of pronouns.

Likewise, a study by Shannon Wiltsey Stirman of the University of Pennsylvania and Pennebaker that's reported in Psychosomatic Medicine (Vol. 63, No. 4), shows that famous suicidal poets made far few references to other people late in their careers than nonsuicidal poets.

Stress and older adults

PNI researcher Susan Lutgendorf, PhD, has been conducting a number of studies on how healthy older adults cope with moderate life stresses. In one recent study, reported in Psychology & Aging (Vol. 14, No. 4), she looked at how well elderly people's immune systems functioned a month before they were to move from their homes of many years to group-living facilities.

She found that the event did have a short-term effect: Movers showed about a 30 percent decline in natural killer-cell activity during this stressful time compared to nonmoving controls.

In addition, Lutgendorf has found that older women with severe life stresses, such as caring for a partner with Alzheimer's disease, show greater disregulation of interleukin-6, a type of proinflammatory cytokine implicated in many age-related diseases. As compared with older women with moderate or minimal life stress, this finding suggests there may be a "dose-response" effect of life stress to immune decline among the elderly above and beyond that of normal aging. This work was reported in the Journal of Gerontology: Medical Sciences (Vol. 54A, No. 9).

Lutgendorf is also interested in what may buffer stress among these older adults. She used a measure called "sense of coherence," which assesses people's ability to find meaning in their lives and in stressful events. She found that older people with the highest sense of coherence had natural killer-cell levels in the same range as controls a month before they moved. By contrast, movers with a low sense of coherence showed the poorest immune response. The findings, also in the Psychology & Aging article, suggest people could benefit from interventions that help them perceive meaning in stressful activities, she says.

Intervention directions

As their research continues, PNI researchers are also exploring ways to apply their findings.

"Our challenge is to take basic science and bring it into the clinic in a logical and straightforward way and maybe make people's lives a lot less miserable," says Christopher Coe, PhD, professor of psychology at the University of Michigan and immediate past-president of the Psychoneuroimmnology Research Society. "A lot of PNI is basic science, but in the next few years we'll see more direct applications of it," he predicts.

Researchers have already begun to identify which intervention characteristics may bolster the immune response. For instance, a recent meta-analysis of 85 psychological interventions found that more successful interventions included classical conditioning procedures drawn from animal models, hypnosis and interventions such as those advocated by Pennebaker that encourage people to write about distressing experiences. The study, by Gregory Miller, PhD, of Washington University and Cohen of Carnegie Mellon, is reported in Health Psychology (Vol. 20, No. 1).

To the authors' surprise, however, stress-management interventions--usually group-therapy sessions that teach patients about their illness and how to manage their stress--showed weaker effects than other interventions examined in the meta-analysis, Miller says. Two exceptions are a series of studies on HIV infection by Michael Antoni, PhD, and Neil Schneiderman, PhD, of the University of Miami, and a study of melanoma patients by Fawzy Fawzy, MD, and Kemeny, both of UCLA. Both trials showed positive changes in immune responses--and in Fawzy's study, longevity--following the interventions, perhaps because the authors designed them for specific stages in the patients' illnesses, Miller speculates.

APA's Reed emphasizes, however, that stress-management interventions can show health benefits besides immune changes. A recent study by Duke University psychologist James Blumenthal, PhD, underscores the point: Male heart patients who participated in a stress-management intervention had fewer cardiac events over a five-year follow up than those in an exercise or a standard-care intervention (see the March Monitor).

"While stress-management interventions may not be associated as strongly with immunological changes as other kinds of interventions, it doesn't mean those interventions wouldn't be effective in other ways for other populations," Reed says.

Miller notes that the main strength of his and Cohen's meta-analysis lies in suggesting specific intervention characteristics worthy of future examination.

"Naively, we all expected in the beginning that any intervention we threw at people would influence the immune system," he says. "It's turning out to be more subtle than that, but we seem to be finding the way."

Tori DeAngelis is a writer in Syracuse, N.Y.