A Comment on the Selection of 'Healthy Controls' for Psychiatric Experiments

Abstract

To the Editor.— In a recent meeting on the design and analysis of multicenter clinical trials in psychiatric research sponsored by the MacArthur Foundation, Stanford, Calif, we were asked to present some statistical comments. Our first concern was the magnitude of center effects and potential center × diagnosis or center × treatment interactions. To obtain some insight into this problem we examined much of the world's data on cerebrospinal fluid biogenic amines (5-hydroxyindoleacetic acid [5-HIAA], homovanillic acid [HVA], and 3-methoxy-4-hydroxyphenylglycol). These data, for which summary statistics are presented in Table 1, came from the National Institute of Mental Health (NIMH) (Bethesda, Md) Intramural Study Center (access provided by William Potter, MD), the NIMH Biological Collaborative Study (Koslow et al,¹1983), The Nancy Pritzker Laboratory at Stanford (Calif) University (J. D. Barchas, MD, and K. F. Faull, PhD), the Karolinska Institute in Stockholm, Sweden (Marie Asberg, MD), the University of Uppsala,