Modeling the effects of aging on retinal ganglion cell density and nerve fiber layer thickness

Purpose The effects of aging on retinal nerve fiber layer (RNFL) thickness should reflect the age-related losses in retinal ganglion cells (RGCs), but published data suggest that the relative rate of thinning of RNFL thickness with age is less than predicted by age-related losses of RGCs. Therefore, the present study was undertaken to reconcile the differences in age-dependency on measures of RGCs and axons that are incorporated in normative clinical data. Methods Normative data for RNFL thickness and visual field sensitivities were obtained from the printouts of standard optical coherence tomography (OCT) and standard automated perimetry (SAP) for patients aged between 25 and 95 years, in decade steps. These data were used in models to estimate the number of RGCs underlying each measure. Results The age-related losses of RGCs derived from normative perimetry data agreed closely with published histologic data, without an age-dependent variable in the model. In contrast, the age-related losses of RGCs derived from normative total RNFL thickness data required an age-dependent decrease of 0.007 axons/mu m(2)/year in axon density in the RNFL to account for the relatively slower rate of RNFL thinning than RGC loss. Conclusions The analysis of normative data suggests a model of age-related thinning of RNFL in which the relationship between RNFL thickness and the density of RGC axons varies with the number of neurons that are lost through normal aging. This model posits that the OCT measurement of total RNFL thickness of a normal retina represents two components: 1) an age-dependent population of RNFL axons, and 2) a non-neural component that partially compensates for the age-related decrease in axons in the nerve fiber layer.