Intraocular pressure fluctuation - A risk factor for visual field progression at low intraocular pressures in the advanced glaucoma intervention study

Purpose: To investigate the relationship of intraocular pressure (IOP) fluctuation and mean IOP to visual field (VF) progression in patients enrolled in the Advanced Glaucoma Intervention Study (AGIS). Design: Retrospective analysis of a prospective randomized clinical trial. Participants: Three hundred one eyes of 301 patients enrolled in the AGIS were included. Eyes with more than one surgical intervention were excluded. Methods: Worsening of the VF was detected with pointwise linear regression. Long-term IOP fluctuation was defined as the standard deviation of IOP (millimeters of mercury) at all visits after initial intervention until the time of VF worsening or end of follow-up, whichever came first. A multivariate linear regression model was performed to identify predictors of VF progression. Terciles of mean IOP were identified, and the average IOP fluctuation in each stratum was calculated. Terciles of long-term IOP fluctuation were similarly evaluated. The proportion of eyes showing VF progression in each stratum was determined and compared. Main Outcome Measure: Visual field progression. Results: Visual field progression was detected in 78 eyes (26%). There were statistically significant differences, between progressing and nonprogressing eyes, for mean IOP (P = 0.006), IOP fluctuation (P<0.001), mean length of follow-up (P = 0.013), mean number of VFs (P = 0.005), and mean number of medications (P = 0.006). Three variables were associated with a higher probability of VF progression: greater IOP fluctuation (P = 0.009), argon laser trabeculoplasty (P = 0.004), and older age (P = 0.05). In this model, mean IOP was of borderline statistical significance (P = 0.09). Within the lower and upper terciles of mean IOP, IOP fluctuation was associated with VF progression in the low mean IOP group (P = 0.002) but not in the high mean IOP group (P = 0.2). When subjects were stratified according to IOP fluctuation, there was a statistically significant difference between lower and upper terciles of IOP fluctuation with respect to progression (P = 0.05). There was a weak correlation between mean IOP and IOP fluctuation (r(2) = 0.025, P = 0.01). Conclusions: In the AGIS, long-term IOP fluctuation is associated with VF progression in patients with low mean IOP but not in patients with high mean IOP.