Disorder decline: U.S. mental ills take controversial dip. (Science news: this week).

Far fewer people suffer from mental disorders requiting treatment than was initially indicated by two national surveys, according to a reanalysis of them. However, some researchers argue that the revision understates the reach of serious mental illnesses in the U.S. population at large.

"Establishing the clinical significance of mental disorders in the community is crucial for estimating treatment need," say psychiatric epidemiologist William E. Narrow of the American Psychiatric Institute for Research and Education in Washington, D.C., and his colleagues. The outcome of this scientific debate is almost certain to influence political efforts aimed at expanding insurance coverage for mental illness.

Narrow and his coworkers probed the responses of 20,861 individuals surveyed from 1980 to 1985 and 8,098 volunteers interviewed between 1990 and 1992 (SN: 1/22/94, p. 55). Rates of mental disorders in these government-sponsored surveys were based solely on the presence of symptoms that met criteria for psychiatric diagnoses.

Narrow's team rated mental disorders as "clinically significant" if survey participants had reported that their symptoms led them to thoughts of suicide, made them seek mental-health treatment, or interfered markedly with their daily activities. Based on these criteria, the rates of any mental ailment for 18-to-54-year-olds in the year before each survey declined from 30 percent to 25 percent in one study and from 30 percent to 21 percent in the other.

These revisions represent a decrease of 13.3 million and 13.9 million people, with at least one mental illness, respectively, the researchers report in the February Archives of General Psychiatry. The adjusted rate for both data sets falls to 18.5 percent for all adults, they add, since mental disorders affect a smaller proportion of people older than 54.

Reasons for discrepancies in the rates of certain ailments between the two surveys remain unclear. For instance, the 1-year prevalence of major depression was 6.5 percent in the 1980s survey and 10.1 percent in the 1990s one. The reanalysis brought these rates closer together, to 5.2 percent and 6.4 percent, respectively.

The new study takes a necessary but still inadequate step toward assessing treatment needs of people with mental disorders, contend psychologist Jerome C. Wakefield of Rutgers University in New Brunswick, N.J., and psychiatrist Robert L. Spitzer of Columbia University, in a commentary accompanying the new report. Its clinical-significance ratings rest on self-reports, which likely understate the personal havoc wreaked by various symptoms, Wakefield and Spitzer note.

Moreover, they say, Narrow's group did not distinguish between clinically significant symptoms caused by biological disturbances and those that arose in response to stressful events.

The director of the 1990s survey, sociologist Ronald C. Kessler of Harvard Medical School in Boston, says the new report greatly underestimates the prevalence of serious mental disorders. Preliminary data from a new survey that he's conducting, which probes for symptom-related impairments more extensively than Narrow's group was able to, largely confirms the higher prevalence rates in his previous survey, Kessler says.