Correlations between familial refractive error and children's non-cycloplegic refractions

We examined the spherical equivalent refractions of 237 subjects who had been seen four or more times in a longitudinal study of refractive development in normal children. We employed both photorefraction and autorefraction using a Canon R1 autorefractor. We performed an analysis of variance (F-test for Lack of Fit) to determine the significance of a linear regression in fitting these refractions against three different measures of familial refractive status. One measure included only the number of myopic parents, a second took into account the number of hyperopic parents, and a third included the refractive states of extended genetic relatives. We found no significant correlation between photorefractive data and familial refractions; however, we did find significant, albeit weak, correlations between all measures of familial refractive status and infrared autorefractions of their children. A linear relationship between autorefractive data and overall familial refractive state was found to be most significant. Linear regression of children's autorefractions against a measure of parental myopia and hyperopia was less significant, while a linear model fitting only a measure of parental myopia was least significant, in addition to showing other non-linear trends. We attribute the failure to find a significant correlation between parent and offspring refractions using children's photorefractive data to vignetting by the apparatus and to the short distance of the fixation target. We believe the significant correlation patterns found with the autorefractor reflect the inheritance patterns of parental and familial refractive states.