Saturday, May 22, 2010
Franklin Hall B Level 4 (Philadelphia Marriott Downtown)
10:00 AM
Background: There is much interest in the theory that environmental factors following a seasonal pattern, (e.g., temperature, viral infection, availability or consumption of certain foods at specific times of the year), and acting during the prenatal or perinatal period, could be related to autism spectrum disorders (ASD). An environmental risk factor is hypothesized to act either directly or through interaction with other environmental or genetic factors, and adversely influence infant brain development. Findings from previous studies investigating this theory have been inconsistent. Methodological limitations of studies of season of birth in ASD, such as the use of clinical rather than population-based samples or small sample sizes may explain some inconsistencies in the findings. Additional factors contributing to variable findings include differences in (1) types of comparison groups (sibling controls vs. population-based controls) used; (2) diagnostic criteria applied; (3) definition of seasonality (e.g. months included in “winter”) used; (4) hemispheric region study in which the study was conducted (5) statistical analysis.
Objectives: The objective of this study was to address many of the limitations of previous studies by using a large, national, population-based sample to investigate the association between seasonality and all subtypes of ASD. Birth month of individuals diagnosed with each ASD diagnostic subtype (Infantile Autism, Asperger’s Syndrome and Pervasive Developmental Disorder), rather than “season”, which may be defined differently in various regions, was examined for deviation from the monthly pattern of birth of children in the general population without autism.
Methods: All ASD cases born in 1990-2002 were identified from the Finnish Hospital Discharge Register (FHDR). Information concerning total births in Finland in 1990-2002 was obtained from the Finnish Medical Birth Registry (FMBR). Descriptive Chi-Squared “goodness-of-fit” tests, adjusted for year of birth, were conducted to examine how well observed frequencies of births of children with ASD differed from expected frequencies. Females and males were analyzed separately in order to examine any differences by gender. Poisson regression analyses will be used to test the data against a sinusoidal function.
Objectives: The objective of this study was to address many of the limitations of previous studies by using a large, national, population-based sample to investigate the association between seasonality and all subtypes of ASD. Birth month of individuals diagnosed with each ASD diagnostic subtype (Infantile Autism, Asperger’s Syndrome and Pervasive Developmental Disorder), rather than “season”, which may be defined differently in various regions, was examined for deviation from the monthly pattern of birth of children in the general population without autism.
Methods: All ASD cases born in 1990-2002 were identified from the Finnish Hospital Discharge Register (FHDR). Information concerning total births in Finland in 1990-2002 was obtained from the Finnish Medical Birth Registry (FMBR). Descriptive Chi-Squared “goodness-of-fit” tests, adjusted for year of birth, were conducted to examine how well observed frequencies of births of children with ASD differed from expected frequencies. Females and males were analyzed separately in order to examine any differences by gender. Poisson regression analyses will be used to test the data against a sinusoidal function.
Results: Preliminary results examining all births revealed a deviation in birth pattern in the months of June (p=.01) and October (p=.06) for all three subtypes of ASD. There was an association between a decrease in births of females diagnosed with all ASD subtypes (p=.04) in the month of June. There was an increase in males diagnosed with infantile autism born in October (p=.05), and the month of December approached significance (p=.08).
Conclusions: These results suggest that environmental exposures may contribute to increased risk of ASD in males. The finding that females born in June are at decreased risk of being diagnosed with ASD warrants further exploration.