Descriptive epidemiology of autism in a California population: who is at risk?

Williams et al. (2002) tracked every baby born in California from 1989 to 1994. They checked state disability records to see who later received an autism label.

The team then asked: Do boys, twins, older moms, highly educated moms, or Black moms show up more often in the autism files?

All five groups did show up more. Male sex, multiple births, Black maternal race, older maternal age, and higher mom education each raised the chance a child was recorded with autism.

The study did not give new counts; it simply flagged these patterns in the huge California file.

A second paper by the same authors that year, Williams et al. (2002), looked at the same birth file but counted cases. They saw autism labels rise while mental-retardation-without-autism labels fell, hinting that doctors were swapping codes rather than finding brand-new autism.

Wong et al. (2009) later audited the same California agency and found no proof that workers were quietly moving kids from the mental-retardation pile to the autism pile. Together the three California studies say: risk factors are real, yet some of the surge is still likely better detection, not epidemic.

Coo et al. (2008) in British Columbia schools added numbers: one-third of their autism jump came from kids who used to carry a different special-ed label. The swap story repeats outside California.

When you see a spike in new autism cases on your caseload, remember the California picture. Part of the rise traces to who gets counted—boys, twins, kids of older or highly educated moms, Black families—not necessarily more actual autism. Ask intake teams if the child carried another label last year; the answer helps you gauge true incidence versus diagnostic shift and plan realistic staffing.