Abnormal transient pupillary light reflex in individuals with autism spectrum disorders.

Fan et al. (2009) flashed a brief light into the eyes of kids with and without autism. They filmed the pupil with an infrared camera and measured how fast and how far it shrank.

The team used simple math to turn the video into three numbers: latency, size of shrink, and speed. They asked if these numbers could tell the groups apart.

Kids with autism took longer to start shrinking, shrank less, and did it slower. A computer rule using these three scores put each child in the right group 92 times out of 100.

No one had to talk or follow instructions; the light reflex happened on its own.

Bradford et al. (2018) saw the same slower start in teens, but they used a pen-light instead of a flash and reached 72% accuracy. The direction matches; the flash just gives a clearer split.

Ring et al. (2020) moved from reflex to memory. They found adults with autism did not show the typical pupil size jump when seeing old photos. The pupil still marks autism, but now the marker sits in the thinking system, not the basic reflex.

Emerson et al. (2025) looked at pupil size during a mental rotation game and saw no group difference in overall dilation. This seems to clash, but they timed peak load minutes after the trial started; the reflex studies catch the first quarter-second, so both can be true.

You can run the 30-second flash test while the child sits on the bus, watches a tablet, or waits for therapy. If latency is slow, flag for further screening; if it is typical, keep watching. No language, no cooperation needed—just a light and a phone camera with an app.