Are controls in schizophrenia research "normal"?

The psychiatric assessment by structured interview and family history of mental disorder in normal volunteers recruited by advertisement for a study of brain structure and function in psychosis is described. Nine of 51 volunteers (17.6%) who passed a phone screen were excluded after a structured interview for major psychopathology. Of 35 completers, 10 (28.6%) had subthreshold mood or substance use but were included in the study. Only 16 subjects (45%) had a negative family history by FH-RDC. Diagnoses in family members included substance abuse (31%), mood disorder (11%), psychosis (9%), and other/undiagnosed (14%). Ventricular enlargement was evaluated by magnetic resonance imaging in two planes. Ventricular size was bimodally distributed in the males, and the group with larger ventricles was more educated and had higher scores on the 8 (Schizophrenia) scale of the MMPI (F = 5.44, p = .0099). Our results suggest that 'normal' volunteers for psychiatric research have personal or family psychopathology which motivates them to participate. As the sensitivity of biological instrumentation increases, the characteristics of the control group must be anticipated in the design and recruitment.