Early language profiles in infants at high-risk for autism spectrum disorders.

Hudry et al. (2014) tracked babies who have an older sibling with autism. These infants are 'high-risk' because autism runs in families.

The team checked each baby's understanding of words at 14 months and again at 24 months. They wanted to see which children would later be diagnosed with autism and which would not.

Babies who later got autism started with the same word understanding as other high-risk babies. By 14 months their growth slowed down.

The gap stayed open at 24 months. High-risk babies who did not develop autism caught up to typical levels. The autism group did not catch up.

Matson et al. (2013) looked at the same high-risk babies and found that higher thinking scores only slightly lowered autism signs. Kristelle's team shows language gives a clearer early red flag than overall IQ.

Reyes et al. (2019) followed similar children into preschool. They found temperament keeps changing in autism, so one-time checks are not enough. Kristelle's data say the same about language: check more than once.

Dembo et al. (2023) used the same math tool—latent profiles—to sort babies with Down syndrome. Both studies prove that babies with the same diagnosis still follow different paths.

You can spot likely autism before the second birthday by watching word comprehension, not just how many words a child says. If a high-risk baby loses his early edge by 14 months, start intervention right away instead of waiting for expressive delays to show. Repeat the test at 24 months; do not assume once is enough.