A few years ago, Dr. Mark Zimmerman, a psychiatrist at Rhode Island Hospital, started noticing that many patients were coming to his practice seeking treatment for bipolar disorder. They’d received the diagnosis elsewhere, and they “were invested in it,” he said.

But as Zimmerman and his colleagues talked to these patients, it seemed to them that a significant proportion didn’t meet the definition of bipolar disorder, a mental illness characterized by dramatic mood swings.

So Zimmerman, director of outpatient psychiatry at Rhode Island Hospital and a psychiatry professor at the Warren Alpert Medical School of Brown University, decided to put his observations to a rigorous test. He conducted a study — and found that fewer than half those who were told they had bipolar disorder actually did.

His findings, if replicated in other studies, could upend the traditional notion that bipolar disorder is under-diagnosed and show that, in fact, it’s widely over-diagnosed.

Such errors can put patients at risk for the side effects of drugs that won’t help them, and deprive them of the treatment they truly need.

Zimmerman’s group assessed 700 patients between May 2001 and March 2005 with a structured three-hour interview that is considered the gold standard in psychiatric research. Those 700 patients were also asked whether they had received a diagnosis of bipolar disorder. Of them, 145 said they had. But fewer than half of the 145 — 43.4 percent — actually met the criteria for bipolar disorder.

To verify his findings, Zimmerman also looked at family history, because bipolar disorder is known to have a genetic component. The researchers compared three groups of patients: those who had never had a diagnosis of bipolar disorder and whom the researchers also didn’t think had it; patients who had been diagnosed with bipolar disorder but whose diagnosis was not confirmed by the structured interview; and patients diagnosed as bipolar whose illness was confirmed in the interview.

The family histories of the first two groups showed a low incidence of bipolar disorder in relatives. But those whose diagnosis of bipolar was confirmed by the interview also had significantly more relatives with the same disorder, suggesting that the third group had been accurately diagnosed.

Zimmerman’s study was published online last week in the Journal of Clinical Psychiatry, and he presented it Wednesday at the annual meeting of the American Psychiatric Association.

Believing that one has bipolar disorder when one doesn’t can have serious consequences, Zimmerman said. The drugs given to treat it can have harmful side effects, including damage to the kidneys, liver and immune or endocrine systems.

Additionally, he said, some patients “are very much invested in their diagnosis and disorder and live a lifestyle that is consistent with that. They stigmatize themselves. They view themselves as not being able to do certain things.” Some patients are “looking for a magic pill that will cure all ills” when they really need to do the hard work of psychotherapy.

Zimmerman speculates that the over-diagnosis occurs because doctors “prefer to treat something they think is more likely to get better,” and medications are available to treat bipolar disorder. “Also, there’s been a lot of p.r. in both the professional and public press about bipolar disorder,” including celebrities announcing they have it.

Asked about Zimmerman’s study, Dr. Michael E. Thase, professor of psychiatry at the University of Pittsburgh Medical Center and the Western Psychiatric Institute and Clinic, said that he, too, has seen people diagnosed with bipolar disorder who don’t meet the criteria. “I’m not surprised or shocked by these findings,” Thase said of Zimmerman’s study. After many years of hearing that bipolar is under-diagnosed, he said, “the pendulum has swung the other way.”

Dr. Gary S. Sachs, founder and director of the Bipolar Clinic and Research Program at Massachusetts General Hospital, in Boston, says that Zimmerman’s study goes to the heart of “a serious issue for our field”: inaccurate diagnoses, arrived at through casual impressions rather than the careful application of formal criteria.

“This is the sacred duty of a caretaker — to make sure they have the diagnosis right,” he said.

Sachs urged patients who receive a psychiatric diagnosis to ask their doctor how many criteria for the illness the patient fulfills. “If the doctor can rattle that off, they have done a formal assessment,” Sachs said. The assessment should also include medical records and conversations with family members, because people often don’t accurately perceive their own moods and behaviors.

Donna Howard, who heads the local chapter of the Depression and Bipolar Support Alliance, was unconvinced by Zimmerman’s study. “I know it’s just the opposite,” she said, asserting that for 10 years, psychiatry has been biased against bipolar disorder.

Zimmerman’s methods, she said, “ignore the nature of bipolar,” which she said can vary in form and intensity over time. “One could be diagnosed and then six months or a year later present to a different clinician … a different set of symptoms and not meet the very narrow criteria he’s using,” she said.

ffreyer@projo.com