Document resource
Objectives While the vast majority of autism research has been conducted in high-income countries (HIC), 95% of the world’s autistic children live in low- and middle-income countries (LMIC). A lack of autism expertise and culturally adapted/validated diagnostic tools in LMICs pose a serious challenge to the diagnosis of autism. Thus, there is a clear need to develop efficient, accurate, and scalable methods of autism diagnosis in LMICs. In previous HIC studies, eye-tracking has proven to be a feasible method for identifying toddlers and young children with autism. The objective of this study is to determine whether our eye-tracking biomarkers differentiate children with and without autism in western Kenya.Methods To date, eye-tracking biomarker data have been collected from sixty, 24- to 72-month-old autistic (n=21) and non-autistic (n=39) children in Kenya (mean: 3.55 years; 17 females). The eye-tracking battery (10- to 15-minute assessment) measures 6 independent metrics shown to predict autism outcomes in HICs – non-social preference, attentional disengagement, pupillary light reflex, and oculomotor metrics.Results Preliminary results from one of the six biomarkers (non-social preference; non-social looking / total looking time) demonstrate that our adapted measure significantly predicts autism outcomes in Kenyan children (p < 0.0001). Based on application of an existing threshold (40% non-social looking time), 60% of autistic and 100% of non-autistic children were correctly identified using this single metric.Conclusions Data collection and analysis are ongoing; however, preliminary findings are consistent with research in HIC and suggests that eye-tracking biomarkers will have diagnostic utility in LMICs.