An Investigation of Eye Color and Autism: The Distribution of Iris Coloration Among Children with ASD in the United States

Background: The etiology of autism still lacks definitive genetic, neurochemical, or neuroanatomical markers to enable reliable diagnosis and improved early intervention.  However, investigations of eye color have received increased interest among epidemiologists.  Iris coloration has been associated with the phenotype of multiple genetic disorders, including albinism, schizophrenia, and Waardenburg, Angelman, and Prader-Willi Syndromes.  It is also related to risk of macular degeneration, glaucoma, and melanoma.  Melanin-based coloration is implicated in physiological, morphological, and behavioral traits.  Researchers have identified significant correlations between eye color and behavior in animals and humans, including significant differences in physiological reactivity between light-eyed and dark eyed persons, suggesting a possible role of neuromelanin.  Although eye color genes are located on chromosomes of interest in autism, and researchers have determined that melanin may play a role in Ascending Reticular Activation System functions related to arousal among individuals with ASD, little research has investigated possible relationships between eye color and autism.

Objectives: The purpose of this study was to investigate the distribution of iris coloration within a national sample of children with ASD in the United States, and determine relationships between eye color in ASD and the general population.  The relationship between eye color and severity levels in the core deficits of autism (communication, socialization, and restricted/repetitive behaviors) was also examined. 

Methods: Direct observations of the eyes of 162 children with ASD were conducted at autism centers in Hawaii, California, Colorado, and Texas using a systematic observation protocol.  Eye colors were categorized into seven categories ranging from blue to black.  Data collectors were required to complete training and demonstrate mastery of eye color ratings before conducting observations.  Intervention fidelity checklists were completed for all observations, and interobserver reliability observations were conducted for approximately 80% of participants (overall agreement level of 87%).  Clinical supervisors rated the severity levels of autism symptoms for each participant, and parents were asked if they have observed unusual characteristics related to their child’s eyes (e.g., eyes change color during periods of extreme behaviors).

Results: Chi-square tests for goodness of fit were conducted to determine if there were statistically significant differences between the eye color distribution of the sample of children with ASD, and the general distribution of eye color in the United States as reported by Soni and Neuhoff (1985) and to check if autism symptom severity levels were related to eye color categories.  There was statistical significance for iris coloration (p<.05).  Dark eyes (black, brown, hazel) were over-represented in the ASD sample, and there were significantly fewer blue and green eyed persons with autism compared to the national population.  There was no trend in symptom severity by eye color category, with non-significant results in communication, socialization, and behavior. 

Conclusions: The possible significant over-representation of dark eyed individuals with ASD in a national U.S. sample warrants further investigation.  Additional analyses of eye color data, including expected levels by ethnicity, are underway.  Additional studies of the role of melanin could result in increased understanding about iris coloration as a possible genetic marker or correlate in ASD.