Extraordinarily high rates of untreated mental illness reported by community studies are false

Mar 9 2006

According to widely reported community-based research, almost half the U.S. population suffers from depression. But research by two sociologists indicates that percentage is greatly exaggerated or is a misrepresentation.

The extraordinarily high rates of untreated mental illness reported by community studies are false, say Allan V. Horwitz, a sociology professor in the Institute of Health at Rutgers University, and Jerome Wakefield, a professor in the School of Social Work at New York University. Community studies rely on standard, closed-format questions about symptoms with no context provided to differentiate between reactions to normal life stress (i.e., a death, a romantic break up, work or school stress) and pathological conditions that indicate clinical mental illness.

"These numbers are largely a product of survey methodologies that, by nature, overstate the number of people with mental illness." Reporting the findings in Contexts magazine (Winter 2006), published by the American Sociological Association, the authors state, "Moreover, because people experiencing normal reactions to stressful events are less likely than the truly disordered to seek medical attention, such questions are bound to inflate estimates of the rate of untreated disorder."

In the past, diagnoses relied on treatment studies, but it became apparent that the number of treated patients understated the problem for a variety of reasons such as lack of access to treatment and reluctance to seek appropriate help. Today tightly structured questions are used in community studies to allow researchers to better diagnose a population.

The problem is that the criteria used in the community surveys are not necessarily valid for diagnosing mental disorders. One reason for this is that people self-select when seeking treatment and use their judgment to decide if their feelings exceed normal responses to stressful events. Second, clinicians make contextual judgments when they diagnose patients because some depressive symptoms might occur as a normal response to a loss of a job or a marriage unraveling. In surveys, interviewers are forbidden to judge the validity of a response or discuss the intent of a question. In addition, the duration criteria of community surveys only require that symptoms last at least two weeks, causing transient and self-correcting problems to be counted as disorders.

"In contrast to clinical settings," say Horwitz and Wakefield, "symptom-based diagnoses in community studies consider everyone who reports enough symptoms as having the mental disorder of depression. Symptoms that would not require treatment may nevertheless qualify as a disorder in a community survey."

"Community surveys could more adequately separate normal responses to stressful situations from mental disorders by including questions about the context in which symptoms develop and persist," say Horwitz and Wakefield. The decision not to include contextual criteria in community surveys, they speculate, might have to do not only with efficiency and practicality but also with resistance to change by groups that benefit from high rates of mental health problems.

There are a number of reasons that these high rates are perpetuated. One is that political support is more likely for an agency devoted to preventing and treating a widespread disease such as the National Institute of Mental Health. Another reason is that pharmaceutical companies capitalize on these survey findings to broaden their markets. The explosive growth in sales of antidepressants may indicate its effectiveness. Lastly, advocacy groups lay claim to the prevalence of mental disorders. They equate the millions of people that surveys identify with disorders with the serious mental disorders in order to reduce the social distance between the mentally disordered and others, thereby lowering the stigma. This may only hinder the truly disabled by shifting resources from where it is truly needed.

http://www.asanet.org