The changeable mood and impulsivity typical of borderline personality disorder suggest that it could be part of a bipolar spectrum. Gunderson et al. (p. 1173) found only modest connections to bipolar disorder among 196 patients with borderline personality disorder. They had a higher rate of co-occurring bipolar disorder than patients with other personality disorders, 19% versus 8%. Over the next 4 years, bipolar disorder developed in 8% of the patients with borderline personality disorder who were not bipolar at baseline and in 3% of those with other personality disorders. Despite these differences, the rates of bipolar disorder in the borderline patients were low, indicating that the two disorders do not commonly coexist. Many borderline patients nevertheless receive only a diagnosis of bipolar disorder, which can divert their treatment away from appropriate, psychosocial interventions for borderline personality disorder. An editorial on the joint presentation and treatment of these disorders by Dr. Michael Stone appears on p. 1126.

In clinical practice, a sequence of treatments may be needed before major depression remits. Guidance in this process is provided by the NIMH collaborative study of the treatment of depression, STAR*D (Sequenced Treatment Alternatives to Relieve Depression). Fava et al. (p. 1161) describe Level 3, in which 235 patients who had not benefited adequately from two previous antidepressants were randomly assigned to a third, either mirtazapine or nortriptyline. Both have pharmacological actions different from those of the antidepressants used in Levels 1 and 2. Both produced remission rates of 10%-20%, lower than in previous studies of carefully chosen patients with uncomplicated conditions. An editorial on the clinical implications of the STAR*D studies by Dr. Matthew Menza appears on p. 1123.

During the 1990s, the proportions of the U.S. population who consulted psychiatrists and other physicians for psychiatric disorders more than doubled. Among the people who received mental health services, treatment shifted away from combined medication and psychosocial therapy—either by a psychiatrist or by multiple service providers—toward treatment by a nonpsychiatrist physician only. Wang et al. (p. 1187) compared the services reported in the National Comorbidity Survey in 1990-1992 with those found in the repeat survey a decade later. The changes also included a substantial decline in treatment by nonphysician mental health professionals, e.g., psychotherapy, among people receiving services. The shift away from combined medication and psychotherapeutic interventions may signal a decline in the intensity of care for mentally ill patients.

In children and adolescents, bipolar disorder is often particularly severe and disabling. There are no randomized, controlled trials of most treatments for these bipolar children. Wagner et al. (p. 1179) report a multicenter study of oxcarbazepine for 110 children and adolescents with bipolar disorder. The oxcarbazepine and placebo groups did not differ significantly on any outcome measure, and mania ratings fell by about one-third in both groups. Within children ages 7-12 years, however, a 50% reduction in manic symptoms occurred in 41% of those taking oxcarbazepine and 17% of those taking placebo; the rates among adolescents were 43% and 40%. Adverse physical effects among the oxcarbazepine patients were similar to those for patients with epilepsy and led to treatment discontinuation for 19%. An editorial by Dr. Ellen Leibenluft on the treatment of bipolar children appears on p. 1129.

The metabolic syndrome, a combination of disorders that collectively increases the risk of diabetes and cardiovascular disease, is an increasingly problematic effect of several second-generation antipsychotics. The prevalence of its various features in clozapine-treated patients has not been previously established. Of 93 patients receiving clozapine for schizophrenia or schizoaffective disorder, 54% had the metabolic syndrome, compared to 21% of a matched sample of the U.S. population. Lamberti et al. (p. 1273) based these findings on measurements of waist circumference, triglycerides, HDL cholesterol, blood pressure, and glucose; the metabolic syndrome is defined as high values for three or more of these. An editorial on the clinical presentation of metabolic syndrome by Dr. Gary Remington appears on p. 1132.