Fluctuation of intraocular pressure and glaucoma progression in the early manifest glaucoma trial.

PURPOSE To investigate whether increased fluctuation of intraocular pressure (IOP) is an independent factor for glaucoma progression. DESIGN A cohort of patients was followed up in a randomized clinical trial. PARTICIPANTS Two hundred fifty-five glaucoma patients from the Early Manifest Glaucoma Trial (EMGT; 129 treated and 126 control patients). METHODS Study visits, conducted every 3 months, included ophthalmologic examinations, IOP measurements, and standard automated perimetry, with fundus photography every 6 months. Intraocular pressure values were included only until the time of progression in those eyes that showed such progression. Individual mean follow-up IOP and IOP fluctuation, calculated as the standard deviation of IOP at applicable visits, were the variables of main interest. Cox regression with time-dependent variables was used to evaluate the association between IOP fluctuation and time to progression, both with and without IOP mean in the models. These analyses also controlled for other significant variables. MAIN OUTCOME MEASURES Glaucoma progression, as defined by a predetermined visual field criterion, worsening of the disk, assessed by an independent disc reading center, or both. RESULTS Median follow-up time was 8 years (range, 0.1-11.1 years). Sixty-eight percent of the patients progressed. When considering mean follow-up IOP and IOP fluctuation in the same time-dependent model, mean IOP was a significant risk factor for progression. The hazard ratio (HR) was 1.11 (95% confidence interval [CI], 1.06-1.17; P<0.0001). Intraocular pressure fluctuation was not related to progression, with an HR of 1.00 (95% CI, 0.81-1.24; P = 0.999). CONCLUSIONS These results confirm our earlier finding that elevated IOP is a strong factor for glaucoma progression, with the HR increasing by 11% for every 1 mmHg of higher IOP. Intraocular pressure fluctuation was not an independent factor in our analyses, a finding that conflicts with some earlier reports. One explanation for the discrepancy is that our analyses did not include postprogression IOP values, which would be biased toward larger fluctuations because of more intensive treatment. In contrast, in this EMGT report, no changes in patient management occurred during the period analyzed.