On "The Science of Prevention"

We feel a painful duty to criticize the article by Coie et al. (October 1993) that purported to provide a conceptual framework and some directions for a national research program based on the "science of prevention." Our pain is occasioned by our belief that people committed to the prevention of mental disorders should support one another. There are too few of us willing to defend the value of prevention programs in a professional culture that focuses the lion's share of effort and resources on individual treatment and that largely ignores the role of social stresses in favor of organic and genetic factors.

However, the Coie et al. (1993) article appears to us to be telling funding agencies what they want to hear rather than what they should hear. Political conservatives strongly defend a prevention model that ignores social injustice, that insists on finding specific physical causes of "specific clinical disorders" as essential first steps in efforts at prevention.

The Coie et al. (1993) article also insisted that prevention programs must "sell," specifically in schools, hospitals, playgrounds, homes, clinics, industries, and community agencies nationwide. In our experience, many sound prevention programs are not popular with all of these groups. Sex education, amniocentesis, toxic waste disposal, parental leave, and so forth, are often vigorously opposed by the "religious right" and business-oriented groups.

Numerous examples in Coie et al. (1993) discussed "disease processes," such as schizophrenia, in contexts that have little to do with primary prevention but that, for example, aim at affording "some protection against schizophrenic episodes" (p. 1004), clearly an example of secondary prevention. And the same is true of interventions with dysthymic children to prevent more severe forms of later depression (p. 1018).

Coie et al. (1993) argued that "effective prevention requires a developmental theory" (p. 1017). This is a highly restrictive prevention strategy. Many demonstrably effective primary prevention programs do not consider developmental factors. Nor do many programs use control groups, or long-term follow-up of samples, both deemed essential in the article.

The specified essential "scientific criteria" demand adequate sampling, careful measurement, and appropriate statistics. We are reminded of the comment of Maccoby and Alexander (1979) :

You can either work on only the most important problems you can handle with precision, or you can work on the most important problems with the best of inadequate research methods. (pp. 99—100, italics added)

For an article that purports to suggest directions for a "scientific" national agenda, there are curious omissions and unusual inclusions in the literature cited. A number of major "scientific" reviews are not cited, for example, Cowen's (1982) special issue of the American Journal of Community Psychology, containing carefully chosen programs; 14 Ounces of Prevention: A Casebook for Practitioners, edited by Price, Cowen, Lorion, and Ramos-McKay (1988) and published by the American Psychological Association; and Prevention of Mental/Emotional Disorders, which resulted from a major effort by a commission of the National Association for Mental Health ( NAMH; 1986 ) using the testimony of experts who reviewed successful prevention programs across the life span. Also absent is any mention of the NAMH's annual Lela Rowland Prevention Award, which applies a rigorous screening process and clearly deserves the designation scientific. Absent is any reference to the Report of the Task Panel on Prevention of the President's Commission on Mental Health (1986) or to the annotated bibliography of 1,000 references on prevention by Buckner, Trickett, and Course (1985) . Any reference to the carefully designed studies of the Michigan Department of Mental Health ( Tableman, 1987 ) is also missing.

Our most serious reservations concern the near total absence of prevention program proposals that strike at the social injustices that play a major role in the appearance of mental and emotional problems. The major sources of stress involve economic status. For example, on the basis of a review of several large-scale social surveys, Mirowski and Ross (1989) found that 85% of the psychological stress reported in these surveys was from individuals whose economic status was below the national median income. Furthermore, involuntary unemployment, premature pregnancy, and the risk of low birth weight are associated with poverty and poor bonding. Absent is any mention of homeless people, of the unavailability of low-cost housing, of the stressful life in the inner cities, of the proximity of toxic dumps to the concentrations of people with darker skin, including, especially, Native Americans on reservations. Absent is any concern with sexism, racism, and other forms of discrimination so clearly associated with the incidence of mental and emotional disorders.

Conservative political philosophy argues, of course, that it is not the business of mental health professionals to meddle into social problems. But their view depends on the acceptance of an individual disease model and the denial of the power of social environment as causal. And where is a concern for poor diet, for iron deficiency, lead poisoning, accidents, and other causes of mental retardation and brain damage? Do we need control groups to recommend immunization or to enrich the diet of pregnant women and children?

Prevention science, without epidemiology, can hardly be justified as science, and there is precious little epidemiology in Coie et al. (1993) . Over the past 60 years, psychiatric epidemiologists have repeatedly identified the relationship of social forces to mental disorders. Social class is the major variable in both morbidity and mortality as well as in rates of mental disorders in humans. The National Institute of Mental Health (NIMH) Epidemiologic Catchment Area Study found significant association between poverty and increased rates of several mental disorders, including major depression. High rates of depressive symptoms, especially among minority populations, are found among people living in poverty.

Two further problems with the Coie et al. (1993) article trouble us. One is the almost exclusive focus on voluntary behavior change. There is little mention of mandated changes, an important component of many public health prevention programs. Laws requiring parental leave, premarital screening for syphilis, protective infant seats and seat belts in cars, compulsory inspection of foods, pasteurization of milk, and phenylketonuria (PKU) infant testing are a few examples of mandated prevention programs.

Another problem with Coie et al. (1993) is the failure to mention ethical issues. To identify specific groups as "high-risk" can lead to self-fulfilling prophecies ( Bond & Albee, 1990 ). Are we to intervene individually with abused children, with children from large families, with the organically handicapped? These are microlevel approaches, and they raise serious ethical problems.

Clearly, we do need a national agenda for research into the primary prevention of mental and emotional problems. However, the agenda should be developed independently of a model already committed to individual defect over social injustice, biogenetic causes over social learning, microover macrointerventions, and traditional processes over positive outcomes.