Is it all in our heads? Or is none of it?

The New England Journal of Medicine has re-ignited a centuries-old controversy over whether what you think affects your health.

This is no mere rhetorical struggle between the medical establishment and the burgeoning unorthodox fringe, either.

Dr. Marcia Angell, pathologist and deputy editor of the influential New England Journal, in an editorial in the June 13 issue, dismisses virtually all recent studies purporting to show a link between state of mind and illness.

"In short," she wrote, "the literature contains very few scientifically sound studies of the relation, if there is one, between mental state and disease."

"I think," says Dr. David Rubinow, a unit chief in the psychobiological branch of the National Institute of Mental Health, "that the editorial discards with one fell swoop all the research that has been done on attitude and health as well as the body of information that suggests there is a profound and mutually interactive relationship between the central nervous system and the immune system."

Angell's editorial was written as a companion to a study published in the same journal issue on two groups of seriously ill cancer patients with poor prognoses, either for survival or recurrence. The study found that psychosocial factors "are not useful clinical predictors of the length of survival in newly diagnosed patients with advanced metastatic disease" nor do they "predict the time to recurrence of disease in patients with high-risk primary melanoma or stage II breast cancer."

The study was conducted at the University of Pennsylvania Cancer Center in Philadelphia by a team headed by Dr. Barrie R. Cassileth.

The Philadelphia researchers conceded in their discussion that "this study did not address the possibility that psychosocial factors or events might influence either the cause of disease or the outcome for patients with more favorable cancer diagnoses."

However, Angell used the negative conclusion of the study to suggest "that we have been too ready to accept the venerable belief that mental state is an important factor in the cause and cure of disease."

In general, mainstream research scientists like NIMH's Rubinow would not disagree with her conclusions when she refers to those who promote attitudinal changes as clinical treatments. These treatments range from author Norman Cousins' own purely anecdotal tale of laughing himself well, to the imaging techniques of Carl and Stephanie Simonton, who teach cancer patients to picture strong white cells eating up the malignancies.

Indeed, says Rubinow: "If one assumes that life style or positive outlook or happiness may influence a disease process, and then jumps to the conclusion that that's the treatment of choice and avoids biological treatments, well I think that is a horrible mistake." But, "you can certainly say that there are several routes that are reasonably well established by which personality factors or emotional responses could influence immune system functions, and by exploring those routes, we might get an idea of how personality and emotion may in fact ultimately influence the development or expression of disease. Instead of looking at one end product, saying it doesn't, and discarding the whole field, you ought to take a look at the evidence."

Because the Philadelphia study found no relationship between the mind and cancer, Rubinow says that when Angell concludes that attitude is irrelevant and therefore shouldn't be considered in treatment, "she is making the same sort of mistake as those who ignore biological treatments. In a way she is as guilty of overgeneralizing as the people she accuses. Obviously both positions are wrong."

Dr. Sandra Levy of the University of Pittsburgh psychiatry department and former chief of the National Cancer Institute's program on behavioral medicine says she agrees "with 95 percent of the editorial."

Levy is a prominent researcher on the relationship of attitude and cancer. Her widely quoted findings suggest that the psychosocial responses of breast cancer victims to their illness correlates with the activity of certain immune system cells that act against cancer activity.

A recent study by Levy found that passive, resigned patients who perceived a lack of social support had a lower activity of the immune cells needed to fight cancer. "What is suggested here," says the study, "is that the pathways of influence of brain chemicals and stress hormones on immune system cells extend to higher cortical function, overt behavioral response and the perceived social context."

However, Levy quite agrees with Angell's conclusions on what Levy terms "the touchy-feely, blame-the-victim sort of thing," but, she said, "the conclusion it left the reader with was that this is the definitive study and there are no further questions to ask."

In a brief telephone interview, Angell said she had been unable to find the Levy study in the Harvard Medical Library. Levy has written a response, both to the Philadelphia study and to portions of the editorial, but Angell said she hadn't seen that either, that it "would be looked at in about six weeks," in its turn.

Dr. Marvin Stein, chief of psychiatry at the Mount Sinai Medical School in New York, is coauthor of a study on bereavement mentioned by Angell as "overinterpreted." Stein believes Angell, by attacking the notion that there is a cause and effect relationship between mood and illness, has created a straw man. "I thought we were beyond that," he said last week.

For example, Stein says in his own bereavement study, two main facts emerged: One was epidemiological evidence that suggests bereavement, especially loss of a spouse, is associated with increased illness and death. The other was the discovery of impaired immune functions in the subjects, a group of men whose wives had died of breast cancer.

"There has been," he said, "a tendency to lump these two facts together, and it is important to emphasize that biological findings associated with bereavement do not adequately explain the epidemiological findings. We haven't been able to relate these two facts yet, and I agree that there has been a tendancy to suggest that this is the cause of illness. No one would say that the stress causes the cancer, but it is a factor that may play a role."

Stein, Levy and Rubinow are among a growing number of scientists who are exploring the mechanisms by which mood and behavior might influence health. They have now amassed evidence pointing to brain chemicals and other hormones as possible mediators between life-stresses and illnesses, including cancer.

They agree with Angell that unwarranted conclusions can be drawn by the naive and the unscrupulous, but they are unanimous that she is "premature in closing the scientific door," as Levy puts it.

Angell denied that she is closing any doors. "My point is," she said, "that in science, if you are asserting something, the onus is on you to prove it. My point is that the evidence is not there. My guess is when adequate studies are done, we will find that any effects are small effects.

"It strikes me as egocentric to think that nature revolves around human wishes, like making rain by doing a rain dance. We shouldn't allow the wish to be father to the thought."