Laurel Alexandra Fisha b✉, Pär Nyströmc d e, Teodora Gligaf g, Anna Guih i, Jannath Begum Alih i, Luke Masonh i, Shruti Gargj k l, Jonathan Greenj k l, Mark Henry Johnson
Papers open inside SciRah from the original publisher or repository; SciRah never hosts them.
Background Although autism spectrum disorder (ASD) is heritable, the mechanisms through which genes contribute to symptom emergence remain unclear. Investigating candidate intermediate phenotypes such as the pupillary light reflex (PLR) prospectively from early in development could bridge genotype and behavioural phenotype. Methods Using eye tracking, we longitudinally measured the PLR at 9, 14 and 24 months in a sample of infants ( N = 264) enriched for a family history of ASD; 27 infants received an ASD diagnosis at 3 years. We examined the 9‐ to 24‐month developmental trajectories of PLR constriction latency (onset; ms) and amplitude (%) and explored their relation to categorical 3‐year ASD outcome, polygenic liability for ASD and dimensional 3‐year social affect (SA) and repetitive/restrictive behaviour (RRB) traits. Polygenic scores for ASD (PGS ASD ) were calculated for 190 infants. Results While infants showed a decrease in latency between 9 and 14 months, higher PGS ASD was associated with a smaller decrease in latency in the first year (β = −.16, 95% CI = −0.31, −0.002); infants with later ASD showed a significantly steeper decrease in latency (a putative ‘catch‐up’) between 14 and 24 months relative to those with other outcomes (typical: β = .54, 95% CI = 0.08, 0.99; other: β = .53, 95% CI = 0.02, 1.04). Latency development did not associate with later dimensional variation in ASD‐related traits. In contrast, change in amplitude was not related to categorical ASD or genetics, but decreasing 9‐ to 14‐month amplitude was associated with higher SA (β = .08, 95% CI = 0.01, 0.14) and RRB (β = .05, 95% CI = 0.004, 0.11) traits. Conclusions These findings corroborate PLR development as possible intermediate phenotypes being linked to both genetic liability and phenotypic outcomes. Future work should incorporate alternative measures (e.g. functionally informed structural and genetic measures) to test whether distinct neural mechanisms underpin PLR alterations.
