SUBJECTS AND METHODS
Sixty-two patients (29 men and 33 women; mean age 52.9 ± 9.4 years) participated in the study. The inclusion criteria were: (a) a diagnosis of Major Affective Disorder (Major Depression, Recurrent, n = 32; Bipolar Disorder, n = 30) established on the basis of DSM-III criteria (1); (b) a past history of at least three major depressive episodes; (c) no history of CNS disease, head trauma with loss of consciousness, or alcohol or drug abuse; (d) no evidence of concurrent organic or neurological illnesses; (e) no steroid medication intake for at least 3 months prior to the study. The patients were classified as having a definitely anxious or nonanxious depression whether they had scored respectively more than 9 points or less than 6 points on the Covi anxiety scale (6) during at least three different depressive episodes (Sacchetti et al., this volume). According to this criterion, 29 patients (17 men, 12 women; mean age 55.6 ± 6.7 years) were classified as anxious and 33 (12 men, 21 women; mean age 50.5 ± 10.6 years) as nonanxious. The CT control group was composed of 49 healthy age- and sex-matched subjects who had undergone CT scans for minor accidental head trauma without loss of consciousness (most had been involved in automobile accidents). Inclusion criteria for the controls were the same as items (c), (d), and (e) described above plus no past history or present evidence of psychiatric disorder. Ten to 12 CT slices were obtained for each patient and control. Cerebral ventricular size was measured by manual planimetry on the CT slice that showed the lateral ventricles at their largest and was expressed as Ventricular Brain Ratio (VBR) (24). All the measurements were performed by two raters who were unaware of the nature of the study and the subjects’ diagnoses. Interrater reliability was determined and found to be high (r = 0.91).