Quantifying Change: The Significant Increase of Autism Spectrum Disorder Prevalence in a Rural Population

Background:    The prevalence of Autism Spectrum Disorder (ASD) is now 1 in 88, a 78% increase over the past six years.  Efforts are being made to quantify this change.   

Previous studies have found a significant urban-rural difference in the prevalence of psychological and other mental disorders.  Yet little data on geographic differences has been collected on ASD.  A recent study in the UK found that the rate of ASD was higher in the rural population (Kiani), but a study from Taiwan showed that the incidence of ASD was higher in urban areas (Lai).  Due to these conflicting reports and different locations, it is imperative that we determine the relationship between ASD and geographical classification so that we may better allocate resources for diagnosis and follow-up care. 

Objectives:   Using the Bureau of the Census definition of a rural area (any incorporated place or census designated place with fewer than 2500 inhabitants that is located outside an urban area), we will determine the rate of change for children meeting criteria for ASD from the South Carolina Autism and Developmental Disabilities Monitoring Program (SC ADDM) in an urban versus rural setting. 

Methods:   Data for this study came from SC ADDM (43% rural in 2000 and 37% rural in 2008).   ADDM is a collaborative agreement between 14 sites and the Centers for Disease Control to conduct ASD surveillance in the United States.  Data includes all 8-year-old children with an ASD who lived in one of 21 counties in the Coastal and Pee Dee regions of South Carolina in 2000 or 2008 and who were designated as having ASD based on record review.  ASD status was determined through extensive screening and records abstraction at multiple educational and clinical sites, followed by expert clinician review and final case determination. This multi-site population-based approach has been used since 2000 and has collected data on 5 cohorts of data (children born in 1992, 1994, 1996, 1998, and 2000).  Prevalence will be compared for rural and urban areas and assessed over time to determine whether an increase in rural prevalence is responsible for any of the increase in prevalence seen over the past 6 years.   Chi-square tests for differences in two independent proportions will be used to determine significance of changes in the proportion of children with ASD in rural and urban areas.

Results:    In urban SC, prevalence increased 57% from 7 per 1,000 in 2000 to 11 per 1,000 in 2008 (x²=14.6, p=0.0001).  In rural SC, prevalence increased 137% from 5.8 per 1,000 in 2000 to 13.7 per 1,000 (x²=32.5, p<0.0001).

Conclusions: The current increases in ASD prevalence in South Carolina can partly be attributed to increases within rural areas.  Awareness and outreach throughout the state are possible reasons.  Expansion of the current analysis to more ADDM states will focus on quantification of the amount of prevalence increase due to rural prevalence increases.