Purchase this article with an account.
J.M. Miller, D.L. Sherrill, V. Dobson, E.M. Harvey; Stability of Spherical Equivalent Refraction in Native American Preschool Children . Invest. Ophthalmol. Vis. Sci. 2003;44(13):4774.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose. To examine the stability of mean refractive error in a preschool population in which there is a high prevalence of astigmatism. Methods. Subjects were 208 participants in the Tohono O'Odham Head Start program who were >3 and <5 years of age at the first test session. Refractive error was measured using cycloplegic (cyclopentolate 2%, followed by cyclopentolate 1%) autorefraction (Nikon Retinomax K+) on 3 to 6 occasions (mean = 3.7), over an average period of 1.4 years (minimum=0.6 year, maximum=2.5 years. Test-retest variability in autorefractor results were calculated based on 99 test-retest comparisons from 73 children tested twice in at least one test session. These variability results were used in Monte Carlo simulations to generate a 95% confidence interval on the slope describing the change in diopters of right eye spherical equivalent (SEQ) per year for each subject, in order to determine whether this slope differed significantly from zero. Results: Mean SEQ was 0.98D (SD=0.99, median=0.88D) at the first refraction, and 1.15D (SD=0.98, median=1.13D) at the last. Most subjects (121/208, 58%) showed stable SEQ (SSEQ) over time (Monte Carlo confidence interval estimate including 0). A significant myopic shift (SMS) was observed in 24 subjects (12%), with average change of -0.55D/yr (range -1.48 to -0.15). In 63 subjects (30%), a significant hyperopic shift (SHS) was seen, averaging 0.53D/yr (range 0.16 to 1.21). Average baseline SEQ was 1.07D (SD=0.98) for SSEQ, 1.41D (SD=0.79) for SMS, and 0.64 (SD=1.01) for SHS subjects. Multinomial logistic regression (comparing SMS and SHS to SSEQ) did not show any significant differences between groups attributable to gender, age at enrollment, or duration of follow-up. Baseline SEQ <1.0 D was predictive of SHS (p=0.006), but unassociated with SMS, when compared to SSEQ in a multnomial logistic regression. However, neither baseline astigmatism >1.50D or anisometropia (vector dioptric difference >0.75D) was predictive. Conclusions: Among preschool Native American children with a high prevalence of astigmatism, the SEQ refractive error is approximately three times as likely to demonstrate a hyperopic shift than a myopic shift, should a shift occur. These shifts in SEQ were independent of baseline levels of astigmatism or anisometropia, but a significant hyperopic shift over time was observed more frequently in children having SEQ <1.0 D at baseline.
This PDF is available to Subscribers Only