Neurobiology of autism: An update

Consideration of available studies suggests that many cases of autism are caused by a neurodevelopmental disorder. In other cases, a known disease entity is found, either during life or at autopsy. A common factor linking primary idiopathic autism with double syndrome cases having autistic behavior may be malfunction in specific neuroanatomic systems, defects in which give rise to the clinically defined autistic symptomatology. The available evidence suggests that the limbic system is abnormal in most cases of autism, and that the hippocampus, basal forebrain, cingulate and orbitofrontal cortices are specifically involved. Evidence for neocortical involvement is less strong; additional investigations will be necessary to define the role of neocortical pathology observed in some, but not all, individuals with autism. Similarly, the role of the thalamus and hypothalamus and their subdivisions needs to be better defined. The role of the cerebellum in the causation of autistic symptoms is controversial. However, the consistency of the findings of the Boston group suggest that additional study, especially studies focused on the connections of the cerebellum to the diencephalic and telencephalic structures, is warranted. The evidence, then, suggests that autism is a disorder of connectivity, often but not exclusively arising during the gestational period and ongoing degeneration of involved neural systems may occur in some individuals. Since different investigators, who study different populations of autistic individuals, have found involvement of multiple neuroanatomic sites, neural network(s) may be involved in pathogenesis of this complex behavior. A defect at any point in the network could produce autistic behavior, and differences in the specific network defect between individuals might account for observed differences in clinical phenotype. The recent identification of abnormalities in serotonin synthesis in autistic individuals suggests that serotonergic systems are likely involved (23). However, the complexity of the brain's circuitry, especially in tine limbic system (56), and the presence of multiple neurotransmitters in any given anatomic site in tine brain, suggests that investigations of additional neurotransmitter systems might be useful as well. While autism is now accepted to be an intrinsic disorder of the brain, much additional work needs to be done to elucidate tine precise biochemical and physiologic defects that lead to the observed pathologic changes. Application of basic neuroscience methods to clinical material will hopefully elucidate the pathogenesis of this disorder and lead to effective therapy.