Abstract
Purpose: :
Previously we have shown cross-sectional age changes in mean ocular refraction and astigmatism across the human lifespan (Irving et al ARVO 2009). The purpose of the current study was to determine if the same pattern of refractive change can be observed longitudinally.
Methods: :
Patient age and best corrected refraction (sphere, cylinder and axis) were abstracted from 88 files from the School of Optometry clinic, at the University of Waterloo using all available assessment dates. Patient visits occurred between 1968 and 2010. Files selected were from patients who had been seen for a minimum of 27 years yielding a total of 1419 file dates and an average of 16 (SD +/- 5) visits per patient. Mixed modeling analysis (using R and SPSS 18) was conducted to establish growth curves and changes in mean ocular and cylinder refraction over the lifespan. Two specific hypotheses were tested examining how growth curves vary. The first was that (random effect) growth curve slope variance was zero and the second was that there was no association between each subject’s growth curve intercept and slope.
Results: :
The hypothesis that (random effect) growth curve slope variance was zero, was rejected by the mixed model analysis (Wald test p<0.001). The second hypothesis, that there was no association between each subject’s growth curve intercept and slope, was also rejected. The associations between these 2 parameters for both mean spherical and cylindrical ametropia were significant (both p<0.001). The association was such that growth curves with more positive intercepts had increasingly negative slopes while those with negative intercepts had positive slopes.
Conclusions: :
When observed longitudinally both mean ocular refraction and astigmatism vary systematically with age similar to cross-sectional data.
Keywords: refractive error development • refraction • aging