Sociodemographic risk factors for autism in a US metropolitan area.
Family money and mom’s education only look like autism risks when clinics, not communities, do the counting.
01Research in Context
What this study did
Bhasin et al. (2007) looked at family income and mom’s education for kids with autism in one large US city.
They split the kids into two groups: autism only, and autism plus mental retardation.
Records came from schools, clinics, and a research registry.
What they found
Higher income and more schooling predicted autism without mental retardation.
The same link did not show up for kids who also had mental retardation.
Where the case came from changed the numbers, so method matters.
How this fits with other research
Cao et al. (2023) found the opposite in China: toddlers whose moms left school early were twice as likely to get an autism label.
The clash is mostly about how kids are found. In China, rural screening catches poor kids who were missed before; in the US city, richer families reach clinics sooner.
Reiss et al. (1982) saw no class effect when Iowa gave free autism services to every child, proving that equal access can wipe the pattern away.
Singh et al. (2024) in Malaysia and Safer-Lichtenstein et al. (2021) across US states both show the same thing: family resources and local systems decide who gets counted, not who has autism.
Why it matters
If you screen only at private clinics, you will over-diagnose high-SES kids and miss others.
Always ask, "Where did this case come from?" before you trust a risk factor.
Push for universal screening in schools and public clinics so income does not decide who gets help.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Add a low-SES community site to your screening rotation this month.
02At a glance
03Original abstract
The present study examined the association between autism and sociodemographic factors, overall and in subgroups of children with autism with and without mental retardation (Autism/MR and Autism/No MR, respectively); the association was further examined in subanalyses by child's source of ascertainment to assess the presence of ascertainment bias. In the main analyses, one marker of higher social class (higher median family income) was significantly associated with autism overall. Both markers of higher social class (higher maternal education and higher median family income) were significantly associated with autism/no MR, but not associated with autism/MR. In the subanalyses, associations with social class varied by ascertainment source. Future studies should consider phenotypic subgroups of children with autism and must consider potential ascertainment bias.
Journal of autism and developmental disorders, 2007 · doi:10.1007/s10803-006-0194-y