Startle modulation studies in autism.

Castañe et al. (1993) wired up people with autism and neurotypical peers to loud clicks. They measured the eye-blink startle reflex to see if the groups jumped differently.

The lab kept noise and timing the same for everyone. Only tiny timing shifts were tracked.

Both groups blinked the same amount and at the same speed. The startle machine could not tell who had autism.

Only a few milliseconds of latency difference showed up, too small to matter for diagnosis.

Spates et al. (2013) looked at younger kids and a cousin reflex, the acoustic stapedial reflex. They found delayed timing and uneven ears in toddlers later diagnosed with autism. This extends the 1993 work by showing a brain-stem marker can flag risk, but only in very early childhood.

South et al. (2017) and Lemons et al. (2015) also saw flat physiological lines. Skin-conductance and pupil size stayed the same between groups, matching the null startle result. Together these papers say: most lab arousal measures look normal in autism.

Symons (2019) systematic review pulls all these null and mixed findings under one roof. It warns that no single biological signal yet predicts anxiety or autism, so keep your toolbox wide.

Do not spend session time on startle probes for diagnosis. Use the minutes for skill-building instead. If you need early red flags, ask about stapedial reflex screening in toddlers, not startle in older clients. Remember: flat lines on a machine do not rule out real anxiety or sensory issues—listen to parent reports and behavior.