Autism and Classical Conditioning

Plaques and tangles build up in certain structures of the brains of Alzheimer patients. Interestingly, people with autism have pathologies in some of the same structures: fewer cells in the cerebellum and increased cell density in the hippocampus, among other abnormalities.

Joseph Steinmetz, professor of psychology at Indiana University Bloomington, is studying the effect of these brain abnormalities on simple learning tasks. Steinmetz devised an experiment to see whether people with autism acquired classically conditioned eye-blink responses differently from control subjects. Eye-blink conditioning is a well-established experimental model--first used, incidentally, by former IU professor Isidore Gormezano in the early 1960s--whereby subjects hear a tone and then receive a puff of air in an eye. After a while, they begin blinking when the tone sounds, without any intervening air puff. (The Russian physiologist Ivan Pavlov introduced the world to classical conditioning almost a century ago when he showed that dogs could be trained to salivate at the flashing of a light before food was placed in their mouths.)

Steinmetz found some significant differences between people with autism and control subjects. People with autism actually acquired the conditioned response faster than control subjects, but they mistimed their blinks. "They blink too early," he explained. "It's not a very adaptive response because the learned eye blink occurs before the air puff is delivered."

This experiment on people with autism was Steinmetz's first use of human subjects after more than a decade of basic research on animals. "My major interest when I entered graduate school," he says, "was to understand how the brain was involved in learning and memory. I discovered that at the level of the kinds of questions I wanted to ask--which included where in the brain were learning and memory occurring and exactly what cellular processes were going on--human subjects weren't appropriate."

So Steinmetz studied the brains of rats and rabbits, pinpointing the crucial role of the cerebellum and hippocampus in conditioned responses and the effects of certain pathologies on those responses. "Then you actually can use findings from this basic animal research," he says, "to make hypotheses about human disorders."

And this is exactly what happened. A graduate student named Lonnie Sears, who was interested in studying the human brain's role in learning disabilities, pointed out to Steinmetz that the very brain structures he had been studying for their involvement in classical conditioning, the cerebellum and hippocampus, were implicated in autism in humans. The result was the eye-blink experiment described above, the first in a planned series of experiments on people with autism that, Steinmetz and Sears hope, will shed light on the causes of the impairment and perhaps lead to treatments. "It's a wonderful example," he says, "of how you can mix basic science research with ongoing observations from clinical research to advance scientific understanding of the human condition."

--Mark Buechler