Autism prodrome 39 of 89
syndrome autism (children with Asperger's disorder as well as other PDD disorders
were not included in the study) and 3,497,870 live births as comparisons. The
following data were retrieved from the live birth certificate electronic files: sex of the
child, birth weight, plurality, birth order, maternal age at delivery, race/ethnicity, birth
place, and current level of education. The findings indicated that among child
characteristics there is an increased risk for autism in males, in multiple births (but
Piven et al. (1993), reported equal distribution for probands with autism and siblings),
and that relative to 1st born children, third and later-born children are at a reduced risk,
but no association was found with low birth weight. Among maternal characteristics
older age and higher educational level involved higher risk with women over the age
of 35 at a significantly higher risk than women under 20 years of age, and women
with a post-doctorate education at significantly higher risk than women with less than
high-school education. In addition, Afro-American women were found to be at a
higher risk to have a child with autism compared to other ethnic groups. Yet although
it is a total population study in that all live births were potentially included, as the
authors state, the study is not an epidemiological study because some children may be
undiagnosed whereas other children diagnosed or undiagnosed, may not have been
enrolled in the state operated services through which the participating children with
autism were identified. To the extent that some self selection processes involved in
receiving the diagnosis and/or services are operative, the data may be biased.
Nonetheless, this is a unique study due to the huge number of participants with
autism, use of total live births as a comparison group, data retrieved from medical
files rather than from possibly biased parental recall, and the fact that each
characteristic was estimated in association with other characteristics. Similar studies
conducted in Denmark, Australia, Sweden and Israel also suggest that maternal