Abstract
Purpose: :
Change in refractive error between three and nine months of age is linearly related to initial refractive errors between +1.00D and +5.00D; initial refractive errors beyond +5.00 tend to show deficient emmetropization and a non-linear relationship with refractive change (Mutti et al., IOVS, 2005). We investigated whether these associations between initial refractive error and change in refractive error were more related to: 1) defocus from accommodative error; or 2) the amount of the accommodative response.
Methods: :
Subjects were 222 infants seen at both three months and nine months of age. Near defocus error was measured by dynamic retinoscopy (DRET) at 57cm (lens neutralization), distance defocus error by modified Mohindra retinoscopy (MRET), and distance refractive error was measured by cycloplegic retinoscopy (WSEQ, cyclopentolate 1%). Accommodative responses (AR) were also calculated: AR distance = WSEQ - MRET; AR near = WSEQ + 1.75D - DRET. Residuals from the orthogonal regression between initial WSEQ and change in WSEQ provided estimates of departures from the linear effects of WSEQ on emmetropization.
Results: :
More hyperopic values of DRET and MRET at three months were related to a more hyperopic initial WSEQ (r = 0.21 and p = 0.0013 for DRET; r = 0.22 and p = 0.0003 for MRET), but neither DRET nor MRET was related to the change in WSEQ between three and nine months (r = -0.08 and p = 0.27 for DRET; r = -0.06 and p = 0.36 for MRET). Larger AR at three months at both distance and near was related to greater change in WSEQ between three and nine months (r = -0.36 for distance; r = -0.41 for near, p<0.0001 for each). One-way ANOVA indicated that the poorest emmetropization (highest departures from the linear effects of WSEQ) was found in infants with the most hyperopic values of WSEQ, DRET, and MRET (p values from <0.0001 to 0.043).
Conclusions: :
Hyperopic defocus at distance and near did not account for the linear effects of initial WSEQ on emmetropization and did not produce effective emmetropization. The amount of accommodative response was related to emmetropization, accounting for both the linear effects of WSEQ for moderate initial refractive errors and the non-linear effects at higher refractive errors. These findings are more consistent with accommodation rather than defocus as a visual signal for emmetropization.
Keywords: emmetropization • infant vision • accomodation