Prevalence of Regressive Autism from 2000 to 2010: Findings from a Population-Based System

Background:  Estimates of the prevalence of regressive autism have varied widely, from 15% to 50%. This variation may be due to varying definitions of regression and to absence of information from comprehensive population-based studies.  

Objectives:  To determine the prevalence of regressive autism in 8-year-old children, in the same region, over time, using a modern epidemiologic method and to identify and trends in the prevalence or expression of regressive autism.  

Methods: Autism Spectrum Disorder (ASD) case-specific data, from an ongoing multi-year (2000, 2002, 2006, 2010) surveillance system in metropolitan New Jersey (NJ), including information on regression, were analyzed.  Regressive autism was defined as loss of skill(s) in areas of speech/language and/or social skill, documented in a professional evaluation, of a child satisfying the ASD case criteria, according to the Centers for Disease Control and Prevention (CDC) multiple source ascertainment method.  Demographic information was from source records.  Children were 8 years old in each of the study years.  Prevalence estimates were based on the numerator (number of cases confirmed for ASD and regression) and denominator data (post censual) for each study year.  Statistical analysis was performed using Chi-square tests and ANOVA. 

Results: Regressive autism affected 13%-19% of all ASD children, during the period.  The median age of regression was 18 to 24 months.   Male: female ratio for regressive autism was 4:1.  For study year (SY) 2000, the prevalence of regressive autism was 1.4 per 1,000; for SY2002, prevalence was 2.0 per 1,000; for SY2006 prevalence was 2.4 per 1,000; for SY2010 prevalence was 3.3 per 1,000, indicating a 74% increase from 2000 to 2010.  The prevalence of regressive autism among girls did not change during the period, but increased significantly among boys.  For SY2000, the prevalence of regressive autism among males was 1.9 per 1,000.  In SY2010, the rate of regressive autism among males was 5.4 per 1,000, an increase of 180%.  Regressive ASD prevalence rates increased among White, non-Hispanic children, and African-American, non-Hispanic boys, from 2000 to 2010; however, there was no increase for Hispanic children.  White, non-Hispanic children had regressive autism at a rate of 1.8 per 1,000 in 2000 and in SY2010 prevalence was 3.2 per 1,000 (78% increase).   Among African-American, non-Hispanic children, the rate of regressive ASD was 1.4 per 1,000 in SY2000 and 4.2 per 1,000 in SY2010 (200% increase).

Conclusions: Regressive autism accounts for a small but significant proportion of all ASD cases.  The proportion of children with regressive autism increased significantly during the study period.  The increase in regressive autism prevalence was driven by males especially by African-American, non-Hispanic boys. The reasons for the increase in regressive autism and for the differential increase among African-American, non-Hispanic boys are unknown.  More studies are needed to understand the etiology of regressive autism and to define the similarities and differences between regressive autism and non-regressive forms of ASD.