Experts split on link between stress, disease

USA Today

MONTEREY, Calif. - Can psychological factors such as stress

influence the course of a disease?

That was the question debated by medical experts during a recent meeting of the American Psychosomatic Society.

In the "yes" corner sat Redford Williams, Duke University cardiologist and pioneer in studying the link between hostility and heart disease. Also of this persuasion: University of Miami AIDS researcher Neil Schneiderman.

Arguing "no" were Arnold Relman, editor of the New England Journal of Medicine from 1977 to 1991, and Marcia Angell, also a formal editor of the journal.

In the middle sat George Lundberg, a free-lance "provocateur." He was fired as editor of the Journal of the American Medical Association during the President Clinton impeachment controversy for publishing survey results suggesting the majority of college students don't consider oral sex as having "had sex." He's now editor in chief of Medscape, a medical Web site.

Williams kicked off the debate by saying evidence abounds that hostility, anxiety and depression promote heart disease.

Although some studies have been small, others are long-term, involve 1,000 or more people and take into account key factors, such as other physical conditions, that could explain the findings.

The personality-disease link can lead to treatments, too. Post-heart-attack patients who attend counseling groups to curb destructive qualities - for example, Type A behavior - are less likely to have repeat coronaries than are patients not in the groups. Stress management treatment also can cut repeat heart attacks, studies show.

There's no mystery here, Schneiderman said. "Biologically plausible" reasons explain why stress or anxiety can raise blood pressure and impose strain on the heart.

In AIDS and cancer research, cortisol, a stress hormone, hampers the ability of natural killer cells to fight infection or tumors. Attending stress relief and support groups can help HIV-positive patients cope, and even improve their immune function, Schneiderman said.

Still, he said, we lack evidence from large, long-term clinical trials that randomize patients into "treatment" or "no treatment" groups, with investigators who don't know which patients are getting treatment. "Blind" investigators monitor patient progress and can take into account all likely factors that could explain the results.

"But there is enough evidence now to justify doing these large-scale trials," Schneiderman said.

Relman thinks not. Even the best research in the mind-body field is riddled with serious flaws, he said. "None (of the studies) is good strong evidence. ... At best, a few are only suggestive."

One error often made: interpreting one-time checks of blood pressure, stress hormones or the immune system as durable effects. "So many things can affect these readings," Relman said.

In other studies, scientists have excluded certain patients or failed to locate so many at follow-up that the findings are not reliable.

Many, such as an often-cited study that found group counseling led to longer survival of women with advanced breast cancer, don't consider treatment or lifestyle factors that might account for longer lives, Relman said.

Most of the research shows "weak effects" for psychological factors on health, and the studies also are small. That, Angell said, inspires doubt.

There's lots of potential for mislabeling something as a "cause."

For example, in a study of hostility and heart disease, men were split into quartiles by their hostility levels, but the percentage of smokers in the four groups ranged from 8 percent to 39 percent. So differences in smoking behavior could have explained the heart disease findings, instead of hostility, Angell said.

"There's a universal human desire that we have some control over illness, but wishes are no substitute for evidence," she said.

Added Relman: "Show us the beef. Where are the numbers?"

But Williams pointed to often-tiny studies published in reputable medical journals in the 1970s showing modest gains from now-accepted beta blockers. "We didn't hear you go around then, saying, 'Where are the numbers?' We need to have a level playing field" for behavioral research.

Replied Relman: "I'm offended by the implication we were unfair. We don't use a double standard."

Because publication of evidence in peer-reviewed journals is the gold standard for medical credibility, the question of editorial fairness is a key one, Lundberg said.

"I assure you, editors are full of biases," he said. "There are gender biases, race biases. They're biased by discipline, by culture, age ... even by the school you went to."

At the Journal of the American Medical Association, about half of submissions "are thrown out in-house before they even get to peer review, and that's where the bias really comes in," he said.

Still, there's enough published evidence on psychological and social factors possibly influencing health "to merit doing larger clinical trials," said debate observer Esther Sternberg, who directs National Institutes of Health research on how brain hormones affect diseases. "It took the last decade to get this field over the hump. It had carried such baggage of quackery and flakiness." NIH is designing and financing many new studies in the mind-body area, she said. "They're harder to do than single-drug studies, but our evidence base on this will be expanding greatly."