Almost half of college-age adults had a psychiatric disorder over a one-year span, based on research criteria that ranged from bipolar disease to substance abuse including smoking. Few sought treatment, the study found.

About one in five students failed to fulfill an obligation, had a legal problem, did something dangerous or caused social problems by using alcohol, the study found. The next most common psychiatric problems were so-called personality disorders, including obsessive-compulsive behavior, at 18 percent, according to the report in the Archives of General Psychiatry.

The mental health of young adults, who often don't know to reach out for help, is a growing concern, said Mark Olfson, the study's author and a professor of clinical psychiatry at Columbia University in New York. This is the first research to examine the prevalence of mental problems in college students, the researchers said. Alcohol and drug abuse had the lowest rates of people seeking treatment in the study.

"Only a small proportion of those who meet criteria for disorders seek treatment," Olfson said. "I don't want to leave you with the impression that college is creating these problems."

About 44,000 people ages 19 to 25 were interviewed in 2001 and 2002 for the study. They were questioned about their behavior in the previous year, based on symptoms listed in the Diagnostic and Statistical Manual of Mental Disorders, a book used by doctors to diagnose mental illness. The interviewers were lay people, and students were characterized based on whether they met certain criteria.

Olfson and colleagues analyzed a subset of the larger survey group comprised of about 5,000 of those interviewed.

The study's findings that about half of those surveyed had a psychiatric disorder "seems extraordinarily high," Joseph Glenmullen, a psychiatrist at Harvard Medical School in Boston, said Monday in a telephone interview.

The results may reflect "a watering down of the diagnostic criteria such that they capture more people with milder symptoms," said Glenmullen, the author of "The Antidepressant Solution," and a critic of the overuse of psychiatric medications.

People ages 19 to 25 who didn't attend college were also studied. The researchers found few differences between those who attended college and those who didn't, suggesting the disorders were age-based, not the result of environmental factors. College students were less likely to smoke or be addicted to tobacco than those not attending college.

Previous research has shown that onset of alcohol-use disorders peaks at 19, and subsequently drops.

Risks for a psychiatric disorder included being male; having lost a steady relationship; being widowed or divorced; living in a rural setting; and living away from one's parents, the study found.

Whether a clinician would have diagnosed those interviewed the same way "is a separate matter" Olfson said. The study didn't see if the use disorders persisted, he said.