Pii: S1388-2457(00)00525-3

Objective: This work aims to evaluate whether glaucomatous visual ®eld defects could be related to an impaired retinal function, to a delayed neural conduction in postretinal visual pathways, or both. Methods: Visual ®eld by Humphrey perimeter (central 24-2 threshold test) and simultaneous recordings of visual evoked potential (VEP) and pattern electroretinogram (PERG) were assessed in 21 subjects with open angle glaucoma (POAG) and in 15 age-matched controls (C). Results: VEP: in POAG eyes we found P100 latency signi®cantly (P , 0:01) delayed when compared with controls and correlated with mean deviation (index of global visual ®eld damage, MD) of 24-2 Humphrey perimetry (P , 0:001); the P100 amplitudes were signi®cantly (P , 0:01) lower in POAG eyes than in control eyes and correlated with MD (P , 0:001). PERG: POAG eyes showed P50 latency signi®cantly (P , 0:01) delayed when compared with controls and correlated with MD (P ˆ 0:002); the P50 and N95 amplitudes were signi®cantly (P , 0:01) lower in POAG than in control eyes and correlated with MD (P50: P ˆ 0:006; N95: P ˆ 0:002). Retinocortical time (RCT: difference between VEP P100 and PERG P50 latencies) and latency window (LW: difference between VEP N75 and PERG P50 latencies) were signi®cantly (P , 0:01) longer in POAG eyes than in control eyes and correlated with MD (RCT: P , 0:001; LW: P , 0:001). No signi®cant correlations (P . 0:05) were found between electrophysiological parameters and the corrected pattern standard deviation (index of localized visual ®eld damage) of 24-2 Humphrey perimetry Conclusion: In patients with open angle glaucoma the reduction of the index of global visual ®eld damage (MD) could be ascribed to two sources of functional impairment: one retinal (impaired PERG) and one postretinal (delayed RCT and LW). In the postretinal impairment, a postsynaptic degeneration at the level of the lateral geniculate nucleus could be suggested. q 2001 Elsevier Science Ireland Ltd. All rights reserved.