Study Finds Confounding Factors in ASD Prevalence
By Stacy W. Kish on November 14, 2012
Background: Can where you live increase a child’s risk of autism? Previous studies have tried to link spatially distributed environmental factors, like air pollution or agriculture pesticides, to autism risk. However, the spatial distribution of confounding factors, such maternal education level, have made it difficult to narrow environmental factors that could place a child at risk for autism.
What’s new: In a new study published in Environmental Health, researchers examined the geographic distribution of ASD patients in North Carolina. They linked birth records to surveillance data obtained from the Autism and Developmental Disabilities Monitoring (ADDM) Network. The researchers used a generalized additive model (GAM) to estimate ASD prevalence in the area. Using the model, the researchers found that geographic variability of ASD prevalence declined when the model accounted for predictive and diagnostic factors—those factors that increase ASD risk or make a parent more likely to recognize ASD symptoms—such as year of birth, race/ethnicity, level of the mother’s education, smoking habits during pregnancy, plurality, and maternal age. The diminished variability limited the researcher’s ability to identify an environmental factor of concern.
Why it’s important: Unlike previous work, this study used GAM, which better adjusts individual risk factors when modeling ASD prevalence in the area. This study underscores the importance of understanding confounding factors to interpret geographic patterns properly when examining the role of the environment in ASD risk.
Help me understand :