Delayed language onset as a predictor of clinical symptoms in pervasive developmental disorders.

Eisenmajer et al. (1998) tracked kids with autism who started talking late. They looked at whether that early delay still mattered once the kids reached grade-school age.

The team checked clinical records and gave language and autism tests. They wanted to know if late talking at two or three still predicted autism severity at eight or nine.

Early language delay did link to more autistic traits in preschool. By middle childhood the link was gone.

The late-talking group also had weaker motor and receptive language scores, but autism severity scores evened out across groups.

Barbaro et al. (2012) extends this view. They showed the receptive-expressive gap is already visible at 18–24 months, tightening the window for first flags.

Reichard et al. (2019) looked later, from 4 to 8 years. They found a small but steady receptive-vocabulary gap that never closed, even though the predictive power of early delay fades.

Warnes et al. (2005) conceptually replicate the null result. Parent-recall onset patterns did not predict severity at 3–4 years, matching the idea that early timing loses forecasting value fast.

Stop using “late talker” as a long-term severity marker. After age six it tells you little about autism traits. Shift your assessment energy to current language, motor, and adaptive skills. Plan intervention goals from today’s data, not the toddler timeline.