Thursday, May 20, 2010: 11:45 AM
Grand Ballroom AB Level 5 (Philadelphia Marriott Downtown)
10:00 AM
Background: In utero tobacco exposure has been associated with neurodevelopmental deficits such as behavior problems and intellectual disabilities and may therefore exert neurotoxic effects that contribute to the development of autism. Several European cohort studies have found an elevated prevalence of autism in children whose mothers smoked during pregnancy.
Objectives: We estimated the association between maternal smoking during pregnancy and prevalence of autism spectrum disorders (ASD) at age 8, using a large, population-based case-cohort design.
Methods: The study cohort included all children who resided at the time of birth in a county subsequently under surveillance for autism through the Autism and Developmental Disabilities Monitoring (ADDM) network. This included children born in 1992, 1994, 1996, and 1998 from select counties in Alabama, Arkansas, Colorado, Georgia, Illinois, Maryland, Missouri, New Jersey, North Carolina, Pennsylvania, South Carolina, and Wisconsin. Variables for this base population were obtained from publicly available natality files from the National Center for Health Statistics. All children from less populous counties (< 100,000) were excluded because vital statistics records for these counties lack geographic identifiers needed to select the cohort. Cases were defined as children from the base population who were identified by the ADDM surveillance network as having ASD in these same counties at age 8 years. ADDM is a multiple-source, active, records-based surveillance network that relies on documented behavioral symptoms in developmental records to determine whether a child meets standardized criteria for ASD based on the DSM-IV-TR. Information on maternal smoking during pregnancy was obtained from birth certificates. We estimated prevalence ratios of maternal smoking and ASD using logistic regression, adjusting for maternal age, race, education, and marital status.
Results: Of 591,639 total birth records meeting inclusion criteria with complete covariate data, 2663 were later identified as meeting ADDM criteria for ASD. About 13% of the total sample and 11% of ASD cases had a report of maternal smoking in pregnancy. We estimated an adjusted prevalence ratio of 0.9 with 95% confidence limits of 0.8-1.0.
Conclusions: Unlike previous reports, we did not find an association between maternal smoking in pregnancy and subsequent ASD. Strengths of this study include its large sample size using a population-based birth cohort, use of standard ASD diagnostic criteria, and ascertainment of maternal smoking prior to ASD diagnosis. Limitations include under ascertainment of maternal smoking due to reliance on information from birth certificates, potential residual confounding due to social class, and inability to remove from the cohort denominator individuals who moved out of the study area after birth. Such children could not be identified as cases, introducing a potential selection bias if residential mobility is related to tobacco use. Further research is needed to determine the extent to which these limitations have contributed to the estimated lack of association between maternal smoking and ASD observed in this study or alternatively, whether maternal smoking during pregnancy does not contribute to ASD.
Objectives: We estimated the association between maternal smoking during pregnancy and prevalence of autism spectrum disorders (ASD) at age 8, using a large, population-based case-cohort design.
Methods: The study cohort included all children who resided at the time of birth in a county subsequently under surveillance for autism through the Autism and Developmental Disabilities Monitoring (ADDM) network. This included children born in 1992, 1994, 1996, and 1998 from select counties in Alabama, Arkansas, Colorado, Georgia, Illinois, Maryland, Missouri, New Jersey, North Carolina, Pennsylvania, South Carolina, and Wisconsin. Variables for this base population were obtained from publicly available natality files from the National Center for Health Statistics. All children from less populous counties (< 100,000) were excluded because vital statistics records for these counties lack geographic identifiers needed to select the cohort. Cases were defined as children from the base population who were identified by the ADDM surveillance network as having ASD in these same counties at age 8 years. ADDM is a multiple-source, active, records-based surveillance network that relies on documented behavioral symptoms in developmental records to determine whether a child meets standardized criteria for ASD based on the DSM-IV-TR. Information on maternal smoking during pregnancy was obtained from birth certificates. We estimated prevalence ratios of maternal smoking and ASD using logistic regression, adjusting for maternal age, race, education, and marital status.
Results: Of 591,639 total birth records meeting inclusion criteria with complete covariate data, 2663 were later identified as meeting ADDM criteria for ASD. About 13% of the total sample and 11% of ASD cases had a report of maternal smoking in pregnancy. We estimated an adjusted prevalence ratio of 0.9 with 95% confidence limits of 0.8-1.0.
Conclusions: Unlike previous reports, we did not find an association between maternal smoking in pregnancy and subsequent ASD. Strengths of this study include its large sample size using a population-based birth cohort, use of standard ASD diagnostic criteria, and ascertainment of maternal smoking prior to ASD diagnosis. Limitations include under ascertainment of maternal smoking due to reliance on information from birth certificates, potential residual confounding due to social class, and inability to remove from the cohort denominator individuals who moved out of the study area after birth. Such children could not be identified as cases, introducing a potential selection bias if residential mobility is related to tobacco use. Further research is needed to determine the extent to which these limitations have contributed to the estimated lack of association between maternal smoking and ASD observed in this study or alternatively, whether maternal smoking during pregnancy does not contribute to ASD.