The prevalence of ADHD in childhood has led to questions about the validity of the diagnosis and the long-term effects of the illness and its treatment. Articles in this issue indicate that the need for treatment lasts into adulthood and point to the increasing evidence for brain abnormalities in the illness. Persistence of ADHD and its association with mental illnesses in adulthood are well known in boys. Biederman et al. (CME, p. 409) extend this finding to girls. Over 40% of the girls were still receiving treatment in early adulthood, and there was increased incidence of antisocial personality disorder, mood and anxiety disorders, and other psychopathology (figure left). Ivanov et al. (p. 397) report that high-resolution images of the thalamus show deficits in children with both hyperactive and inattentive subtypes. Children who were treated with stimulant medication had smaller deficits than children who did not receive consistent treatment. In an editorial, Dr. Phillip Shaw (p. 363) points out that several studies have now shown that children who were treated for ADHD with psychostimulants have less brain volume loss than those who were untreated.

A common variation in a gene that has the potential to promote cancer is less common in patients with schizophrenia than in healthy subjects. This protective variant of the MET proto-oncogene also influences neurodevelopment, and within the healthy subjects it was associated with higher cognitive ability. The study by Burdick et al. (p. 436) of variable DNA sequences in this gene stems from the lower incidence of cancer in schizophrenia patients than in the general population. Another region of the MET gene has previously been linked to autism.

A large study demonstrates that neurocognitive abilities in African Americans are inherited and are affected by schizophrenia, as they are in Caucasians. Calkins et al. (CME, p. 459) analyzed functioning in 10 neurocognitive domains for 610 African American patients with schizophrenia or schizoaffective disorder, their relatives, and community comparison subjects. The patients were slower and less accurate in most domains than their relatives—both with and without other psychiatric disorders—and the relatives were more impaired than the comparison subjects. These abilities may therefore provide quantifiable markers of schizophrenia, or endophenotypes, that could be used to identify individuals for genetics research.