A DISTURBING number of the experts who help write psychiatry's most influential diagnostic manual have financial ties to drug companies, raising concerns about the independence of diagnostic advice in the manual.

While such possible conflicts of interest are not uncommon, psychiatry is of particular concern because diagnosis is so tricky. Physicians rely heavily on the Diagnostic and Statistical Manual of Mental Disorders, or DSM, which categorises psychiatric illnesses and their diagnostic criteria. "The existence of disease categories validates the need for drugs," says Mildred Cho, a bioethicist at Stanford University in California. "Companies have an incentive to influence those creating the categories."

Lisa Cosgrove, a clinical psychologist at the University of Massachusetts in Boston, began to worry about such conflicts when she discovered that a majority of the members of a panel formed to consider whether to include "premenstrual dysphoric disorder" in the manual had received money from Eli Lilly. In 2000, Lilly won approval from the US Food and Drug Administration to market Prozac, rebranded as Sarafem, to treat the condition.

Together with Sheldon Krimsky of Tufts University in Medford, Massachusetts, Cosgrove looked at whether members of other DSM panels had financial ties to drug firms. Such ties included receipt of funding for research, acting as a consultant and being paid for speaking. Overall, 56 per cent of panel members had such links, and all members of the panels for "schizophrenia and other psychotic disorders" and "mood disorders" had such links (Psychotherapy and Psychosomatics, vol 75, p 154). Drugs are used in both areas, and in the case of mood disorders drug companies have been accused of "disease mongering" (New Scientist, 15 April, p 38).

Even subtle changes to the DSM can have a big effect on patterns of prescribing. This is a worry for conditions such as attention deficit hyperactivity disorder (ADHD). "There has been a gradual broadening of the diagnostic criteria," says James Swanson of the University of California, Irvine. Cosgrove and Krimsky found that 62 per cent of the DSM panel dealing with disorders such as ADHD had links to drug firms.

The American Psychiatric Association (APA), which publishes the DSM, says its experts are not influenced by their financial ties. However, those recruited for the next edition, to appear in 2011, will be required to declare such interests. "This information will appear in the publication," says the APA's Darrel Regier.

Krimsky argues that the APA should ensure that no DSM panel has a majority of members with ties to drug companies. "It is time that the profession of psychiatry takes a serious look at itself from an ethical standpoint," he says.