In the unadjusted model for monocular accommodative responses, risk group had a highly significant predictor effect (F = 42.04; P < 0.01) for static AR ( Fig. 4 a), with an age-related trend across risk groups (i.e., older children likely to demonstrate greater accommodation) (F = 2.78; P = 0.10).
The indirect effect was marginally significant (z = 1.84; P = 0.06), suggesting a possible mediating role for AR in the pathway from refractive error to myopia development, although this did not reach conventional levels of statistical significance.
When analyzed within risk groups, the negative correlation remained similar in magnitude, but did not reach statistical significance, likely owing to the smaller sample (LR, r = −0.27; P = 0.06; HR, r = −0.28; P = 0.07) ( Fig. 3 ).
In contrast, because accommodation lag accounts for both demand and response, the observed lags were more similar between risk groups, with borderline significant differences.