Enrollment and Participant Characteristics of SPARK, a National, Web-Based Cohort of Individuals with ASD and Their Family Members

Background: The heterogeneity of autism spectrum disorders (ASD) and the burden of large-scale phenotyping and genotyping [2] have historically been significant obstacles to the study of this group of neurodevelopmental disorders. Initiatives such as the Interactive Autism Network’s (IAN) pioneered efforts in creating large web-based autism registries and have successfully demonstrated the validity of parent-reported ASD diagnosis [3] and the phenotypic data provided [1]. To date, however, no studies of this scale have combined large scale parent-reported data with genetic data to rapidly advance ASD research.

Objectives: The aim of the current study is to describe SPARK and to summarize preliminary demographic and clinical characteristics of SPARK participants <18 years.

Methods:  Enrollment in SPARK occurs entirely online at SPARKforAutism.org. Participants are asked to indicate their ASD status and relationship to the individual in the family with ASD and to enter basic demographic information. After consenting to share medical, and optional genetic data, participants are then asked to provide the following additional information on the individual(s) with ASD: ASD diagnosis type, age at diagnosis, individual who diagnosed ASD, language ability, ever/never ASD-related services used, ever/never individualized educational plan for ASD, and ever/never intellectual disability diagnosis. Data collected through SPARKforAutism.org also includes a brief medical history and the Social Communication Questionnaire – Lifetime (SCQ) for children age 2 – 18 years. Participants who consent to share genetic data are then asked to provide a shipping address for saliva collection and are also given the option of inviting the other biological family members to participate.

Results: From the launch of SPARK’s pilot in December 22, 2015 through October 13, 2016, a total of 8,650 children with ASD enrolled in SPARK. Among child participants, the mean age at enrollment was 8.6 years (SD 4.1), and 80% were male. Mean age at diagnosis was 49.5 months (SD 30.8), and the most commonly assigned diagnosis was ASD (68%). Children were most commonly diagnosed by a clinical psychologist (24%), and the most common language level was use of longer sentences (57%). The majority of parents reported that children received ASD services (89%) and had an IEP (83%). 17% of participants received a diagnosis of ID, and the mean SCQ score was 22.5 (7.1 SD).

Conclusions: Initial enrollment supports the feasibility of implementing a large-scale, web-based registry for ASD that collects both phenotypic and genetic information. Characteristics of the participants <18 years recruited to SPARK are consistent with those described in both CDC’s Autism and Developmental Disabilities Monitoring Network study cohorts and IAN samples. Future research on the SPARK cohort will provide more in-depth examination of adults with ASD, evaluate the effectiveness of saliva collection for DNA analysis and examine the associated genetic findings.