Factors That Influence Age of Identification of Children with Autism and Pervasive Developmental Disorder-Not Otherwise Specified

Background: Early diagnosis of autism spectrum disorders (ASD) is crucial because early identification can lead to early intervention, which has been shown to improve the outcome of affected children. 

Objectives: This study explored the current mean age of diagnosis of Autism and Pervasive Developmental Disorder, NOS in the United States, whether average age of diagnosis of children with Autism and Pervasive Developmental Disorder NOS is declining and whether average age of diagnosis differs in the four regions in the United States.  A linear regression analysis was performed to determine which socio-demographic factors, clinical factors and pediatrician practices predict age of diagnosis.

Methods: Potential participants were recruited with the assistance of the Interactive Autism Network (IAN) Research Database at the Kennedy Krieger Institute and Johns Hopkins Medicine in Baltimore, sponsored by the Autism Speaks Foundation.   Participants were selected from a pool of participants who have registered with IAN, who have agreed to participate in autism research, and whose children met the study criteria.  The potential participants were sent a recruitment letter via email by IAN research personnel, which included a link to the on-line survey.  The study description and a link to the survey were also posted on the IAN Community Research Opportunities Bulletin Board.
Results: A sample size of 654 participants was used for data analysis.  Current age of diagnosis in months of autistic disorder and PDD-NOS in the United States was 37.78, and SD=16.269.  Results of a one-way ANOVA indicated differences in age of diagnosis among the four regions in the United States F (3, 650) = 7.618, p=.01).   Tukey’s post hoc comparisons of the groups indicated that the mean age of diagnosis in months in the Midwest (M=42.1) was significantly later than in the Northeast (M=35.3, p=.000) and South (M=35.5, p=.000).  Results of a Pearson Correlation indicated a significant negative correlation (r=-.409, n=654, p=.000) between date of birth and age of diagnosis, with later birthdays being associated with lower age of diagnosis.   Regression results were statistically significant, F(25, 620)=7.549, p<.001, R²=.233.  Having autistic older sibling/s, being referred to ECI, and having a pediatrician who performed an in-depth screening in response to parent concerns, were negatively correlated with the logarithm of age of diagnosis, while having first symptoms of non-autism specific behavioral difficulties, being given a non-ASD diagnosis prior to ASD diagnosis and switching pediatricians during first five years of life, were positively correlated with the logarithm of age of diagnosis. 
Conclusions: Although the results of this study demonstrate that average age of diagnosis is still within the range reported almost ten years ago, the results are also promising, as younger children in the study were likely to be diagnosed at younger ages.  Results of this study suggest that more targeted efforts need to take place in the Midwest, where age of diagnosis was found to be significantly later than in the Northeast and South regions.  This study also identifies some factors associated with age identification of ASD, which may useful when developing programs for early identification of ASD.