Of the 170 experts in all who contributed to the manual that defines disorders from personality problems to drug addiction, more than half had such ties, including 100 percent of the experts who served on work groups on mood disorders and psychotic disorders. The analysis did not reveal the extent of their relationships with industry or whether those ties preceded or followed their work on the manual.

"I don't think the public is aware of how egregious the financial ties are in the field of psychiatry," said Lisa Cosgrove, a clinical psychologist at the University of Massachusetts in Boston, who is publishing her analysis today in the peer-reviewed journal Psychotherapy and Psychosomatics.

The analysis comes at a time of growing debate over the rising use of medication as the primary or sole treatment for many psychiatric disorders, a trend driven in part by definitions of mental disorders in the psychiatric manual.

Cosgrove said she began her research after discovering that five of six panel members studying whether certain premenstrual problems are a psychiatric disorder had ties to Eli Lilly & Co., which was seeking to market its drug Prozac to treat those symptoms. The process of defining such disorders is far from scientific, Cosgrove added: "You would be dismayed at how political the process can be."

The American Psychiatric Association, which publishes the guidelines in its bible of disorders, the Diagnostic and Statistical Manual (DSM), said it is planning to require disclosure of the financial ties of experts who write the next edition of the manual -- due around 2011. The manual carries vast influence over the practice of psychiatry in the United States and around the world.

Darrel Regier, director of the association's division of research, said that concerns over disclosure are a relatively recent phenomenon, which may be why the last edition, published in 1994, did not note them. Regier and John Kane, an expert on schizophrenia who worked on the last edition, agreed with the need for transparency but said financial ties with industry should not undermine public confidence in the conclusions of its experts. Kane has been a consultant to drug companies including Abbott Laboratories, Eli Lilly, Janssen and Pfizer Inc.

"It shouldn't be assumed there is a true conflict of interest," said Kane, who said his panel's conclusions were driven only by science. "To me, a conflict of interest implies that someone's judgment is going to be influenced by this relationship, and that is not necessarily the case. . . ."

The DSM defines disorders in terms of constellations of symptoms. While neuroscience and genetics are revealing biological aspects to many disorders, there has been unease that psychiatry is ignoring social, psychological and cultural factors in its pursuit of biological explanations and treatments.

"As a profession, we have allowed the biopsychosocial model to become the bio-bio-bio model," Steven Sharfstein, president of the American Psychiatric Association, said in an essay last year to his colleagues. He later added, "If we are seen as mere pill pushers and employees of the pharmaceutical industry, our credibility as a profession is compromised."

He stressed that the association has strict guidelines to police the role of the pharmaceutical industry but said the profession as a whole needs to do a better job monitoring ethical conflicts.

Sharfstein added yesterday that the presence of experts with ties to companies on the manual's expert panels is understandable, given that many of the top experts in the field are involved in drug research.

"I am not surprised that the key people who participate have these kinds of relationships," he said. "They are the major researchers in the field, and are very much on the cutting edge, and will have some kind of relationship -- but there should be full disclosure."

At least one psychiatrist who worked on the current manual criticized the analysis. Nancy Andreasen of the University of Iowa, who headed the schizophrenia team, called the new analysis "very flawed" because it did not distinguish researchers who had ties to industry while serving on the panel from those who formed such ties afterward.

Two out of five researchers on her team had had substantial ties to industry, she said. Andreasen said she would have to check her tax statements to know whether she received money from companies at the time she worked on the panel, but said, "What I do know is that I do almost nothing with drug companies. . . . My area of research is neuroimaging, not psychopharmacology."

The analysis could not determine the extent or timing of the financial ties because it relied on disclosures in journal publications and other venues that do not mention many details, said Sheldon Krimsky, a science policy specialist at Tufts University who also was an author of the new study. Whether the researchers received money before, during or after their service on the panel did not remove the ethical concern, he said.

Krimsky, the author of the book "Science in the Private Interest," added that although more transparency is welcome, the psychiatric association should staff its panels with disinterested experts.

"When someone is establishing a clinical guideline for the bible of psychiatric diagnosis, I would argue they should have no affiliation with the drug companies in those areas where the companies could benefit from those decisions," he said.