Macular and retinal nerve fiber layer thickness: which is more helpful in the diagnosis of glaucoma?

PURPOSE To explore factors affecting the glaucoma diagnostic capability of macular thickness and circumpapillary retinal nerve fiber layer (cRNFL) thickness as determined by spectral domain-optical coherence tomography. METHODS Area under the curve (AUC) of the receiver operating characteristics (ROC) discriminating healthy from glaucomatous eyes was determined using 10 macular thickness and 16 cRNFL thickness parameters. Glaucomatous eyes were categorized into two groups using four criteria according to whether cRNFL thickness or macular thickness measurement was better for glaucoma detection (cRNFL better group [RBG], macula better group [MBG], respectively). The clinical characteristics were compared between RBG and MBG. ROC regression analysis was conducted to identify variables affecting diagnostic capability using either macular thickness or cRNFL thickness measurements. RESULTS Four hundred twenty-four glaucomatous patients and 297 healthy subjects were analyzed. Of all cRNFL parameters, average thickness showed the largest AUC (0.958). Of macular parameters, the inferior outer sector showed the largest AUC (0.880). More eyes were placed into the RBG than the MBG, making use of all four criteria (90 vs. 24, 143 vs. 46, 76 vs. 18, and 103 vs. 36, respectively). RBG patients had a smaller optic disc area than did MBG patients in 3 of the 4 criteria. Signal strength affected the diagnostic performance of cRNFL thickness measurement (P = 0.043), whereas that of macular thickness was not affected by any covariate analyzed. CONCLUSIONS Overall, cRNFL thickness measurements were generally superior to those of macular thickness when used to diagnose glaucoma. Macular thickness parameters were of greater value in eyes with larger optic discs.