The popular drugs known as atypical antipsychotics, prescribed for an array of conditions, including schizophrenia, autism and dementia, double patients' risk of dying from sudden heart failure, a study has found.

The finding is the latest in a succession of recent reports contradicting the long-held assumption that the new drugs, which include Risperdal, Zyprexa and Seroquel, are safer than the older and much less expensive medications that they replaced.

The risk of death from the drugs is not high, on average about 3 percent in a person being treated at least 10 years, according to the study, published today in The New England Journal of Medicine. Nor was the risk different from that of the older antipsychotic drugs.

But it was significant enough that an accompanying editorial urged doctors to sharply reduce their prescribing of antipsychotic drugs, especially to children and elderly patients, who can be highly susceptible to the drugs' side effects, including rapid weight gain.

In recent years, the newer drugs, which account for about 90 percent of the market, have become increasingly controversial, as prescription rates to children and elderly people have soared. Doctors use the drugs to settle outbursts related to a host of psychiatric disorders, including attention deficit disorder and Alzheimer's disease. Most are not approved for such use; doctors prescribe them "off label."

In recent years, after an analysis of study data, the Food and Drug Administration required that all antipsychotics' labels contain a warning that the drugs were associated with a heightened risk of heart failure in elderly patients. The new study, an analysis of more than 250,000 Medicaid records, is the first to rigorously document that risk for the newer drugs in adults over 30 without previous heart problems.

The study finding "doesn't come as complete surprise; we did know much of what they found," said Dr. Jeffrey Lieberman, chief of psychiatry at Columbia University Medical Center and the New York State Psychiatric Institute, who was not part of the study. But the study, he added, "puts one more nail in the coffin of this notion of the tremendous superiority of the newer class of medications."

In the study, researchers at Vanderbilt University and the Nashville Veterans Affairs Medical Center analyzed Tennessee Medicaid records for 276,907 people ages 30 to 74. About a third of them began taking an antipsychotic medication in the period studied, from 1990 to 2005, either a newer atypical or an older drug. Two-thirds made up a control group, people of the same age and health who did not take an antipsychotic.

The researchers excluded patients with heart disease or other problems that might put them at higher risk of cardiac failure. Antipsychotic drugs, sometimes called major tranquilizers, can affect heart rhythm in some vulnerable people.