Senescent changes of the normal visual field: an age-old problem.

PURPOSE: To examine the effect of normal aging on visual field sensitivity and identify the best-fitting function for normal populations.
METHODS: Full Threshold standard automated perimetry data (program 24-2, Humphrey Field Analyzer) was collated retrospectively from clinically normal subjects who had previously been recruited for studies of normal visual function. One eye of each subject was randomly selected for analysis. Linear, bilinear, and simple nonlinear candidate functions of mean sensitivity vs. age (independent variable) were applied to these cross-sectional population data. Differences in the aging effect by eccentricity and hemifield were also considered.
RESULTS: Data from 562 normal eyes were available for analysis. A significant negative relationship existed between age and mean visual field sensitivity within the cross-sectional population data. Coefficients of determination for the linear, bilinear, and nonlinear functions were 0.21, 0.20, and 0.26, respectively, indicating that the nonlinear function provided best characterization of the effect of age on mean sensitivity. A small but significant increase in the aging effect was present both peripherally (p < 0.001) and superiorly (p < 0.001). DISCUSSION: The nonlinear function that provided the best fit to cross-sectional population data indicates that age exerts an increasing effect on perimetric sensitivity with age, which has implications for clinical devices that use linear age coefficients to correct for normal aging effects. A linear age correction will overestimate aging changes for younger subjects and, therefore, may miss early pathologic changes in visual sensitivity. Conversely, linear age correction will also underestimate aging changes for older subjects and, thus, may overcall pathological sensitivity loss in this group. The effect of age within subjects requires further investigation to provide reliable estimates of the effect of age on sensitivity.