To the Editor: We thank Dr. Gepner for his letter concerning our article and the letter of Drs. Kemner and van Engeland discussing their findings (1) relative to ours. A word of accuracy, however, needs to precede our reply. Drs. Kemner and van Engeland were not reacting to our review in the Journal (that included a single case illustration) but to our case-control series, which appeared in the Archives of General Psychiatry(2). And it was our suggestion (3), not that of Drs. Kemner and van Engeland, that the discrepancy in results between the two studies could be due to the type of stimuli used in the two studies: static, i.e., pictures, by Dr. Kemner and her colleagues (1) versus dynamic, i.e., videotaped social situations by us (2).

Dr. Gepner’s hypothesis of a rapid visual-motion integration deficit in autism is interesting, but we must take issue with his explanations of our data. First, Dr. Gepner hypothesizes that some individuals with autism may avoid rapid physical and biological movements (considered as aversive stimuli), which would, developmentally, disrupt social interaction. In our clinical experience, young children with autism may in fact be fascinated with rapid movements, particularly if these are repetitive or create unusual sensory sensations (e.g., shining reflections as in spin-top or repetitive patterns like in a computer screensaver). As in many areas of perceptual research in autism in which a basic process was proposed to underlie more global visual attention to social stimuli, we feel that it is important to study the given process in both nonsocial and social paradigms so as to ensure that the deficit is general rather than a phenomenon that occurs only in social situations.

Second, Dr. Gepner hypothesizes that our finding of increased visual fixation on the mouth region (rather than the eyes) of people in our videotaped stimuli reflects the participants’ attempt to capture information that is difficult to process accurately and efficiently in naturalistic social situations. In other words, individuals with autism would focus on the mouth because they were overwhelmed by the rapid movement nature of naturalistic social phenomena. While this is a possibility, we doubt that this is the only factor accounting for our results. For example, in some of our eye-tracking illustrations (e.g., Figure 3 in our article, p. 900), there is hardly any movement taking place. In addition, our eye-tracking studies of toddlers with autism (4) showed that young children with autism also focus a great deal on mouths, and in our video stimuli of a parent’s approach to a toddler, the mouth region is where most of the movement is taking place (because the caregivers are continuously talking).

Third, Dr. Gepner’s own work (his references 3–5) suggests that lower-functioning individuals with autism (i.e., those with a degree of mental retardation) are impaired in the processing of physical environmental movement, particularly rapid movement, while higher functioning individuals with autism (i.e., those without accompanying mental retardation) are less impaired in such tasks. In other words, low-functioning children do worse than high-functioning children on visual tasks involving rapid movement. Applying this notion to the discussion of our results and those of van der Geest and colleagues, Dr. Gepner hypothesizes that the discrepancy may be due to the severity of individuals with autism included in the two studies, with our participants being more cognitively disabled than those included in the study by van der Geest and colleagues. Inspection of subject characterization data on the two studies does not support this hypothesis. The participants in our study were both older and more cognitively able than the group of participants in the study by van der Geest and colleagues. In fact, the viewer with autism whose individual eye-tracking data were described and illustrated in detail in our review article was a 38-year-old man who has been followed-up in our center for most of his life and who has a very impressive list of accomplishments, including college graduation in a competitive institution and two master’s-level degrees. He was, nevertheless, quite socially disabled and appeared to have a typical manifestation of autism in a cognitively able individual.

In sum, we agree with Dr. Gepner about the importance of future research on this topic but do not believe that his suggestions account for the results obtained in our eye-tracking studies to date.