Functional neuroimaging of autistic disorders.

Functional neuroimaging methods hold promise for elucidating the neurobiology of autistic disorders, yet they present difficult practical and scientific challenges when applied to these complex and heterogeneous syndromes. Single-state studies of brain metabolism and blood flow thus far have failed to yield consistent findings, but suggest considerable variability in regional patterns of cerebral synaptic activity. Patients with idiopathic autism are less likely to show abnormalities than are patients with comorbid illness or epilepsy. Activation studies have begun to suggest alterations in brain organization for language and cognition. Neurotransmitter studies using positron emission tomography (PET) suggest abnormalities of serotonergic and dopaminergic function. Studies using magnetic resonance spectroscopy (MRS) have begun to document metabolic deficits in the frontal cortex and cerebellum. A single study using magnetoencephalography suggests a high incidence of epileptiform activity in children with autistic regression. Research needs include well-controlled developmental studies, particularly of young subjects and relatively homogeneous subgroups, which balance scientific rigor with ethical constraints. Investigations of the serotonergic and dopaminergic systems, limbic-based memory and emotional systems, and the role of epileptiform activity in autism represent priorities for future research.