Rates of Visual Field Change in Patients With Glaucoma and Healthy Individuals: Findings From a Median 25-Year Follow-up

Key Points

Question  What are the rates of visual field change in patients with glaucoma and healthy individuals over an extensive follow-up period?

Findings  In this cohort study of 40 patients receiving treatment for glaucoma and 29 healthy individuals, the visual field of 1 eye per participant was followed up for a median of 25.65 years and 19.56 years, respectively. Patients with glaucoma had a −0.032 dB/y faster mean rate of mean sensitivity change compared with healthy participants, but this difference was not statistically significant.

Meaning  This study suggests that over a median follow-up of more than 25 years, the rate of visual field change in most patients receiving treatment for glaucoma was comparable with that of healthy individuals.

Importance  Estimating the rate of glaucomatous visual field change provides practical assessment of disease progression and has implications for management decisions.

Objective  To assess the rates of visual field change in patients receiving treatment for glaucoma compared with healthy individuals over an extensive follow-up period and to quantify the impact of important covariates for these rates.

Design, Setting, and Participants  This prospective longitudinal cohort study was conducted in a hospital-based setting from January 1991 to February 2020. The study included 40 patients receiving treatment for open-angle glaucoma and 29 healthy participants. One eye of each participant was randomly selected as the study eye.

Exposures  Patients with glaucoma and healthy participants received testing with standard automated perimetry every 6 months. Individual rates of mean sensitivity change were computed using ordinary least-squares regression analysis, and linear mixed-effects modeling was used to estimate the mean rates of mean sensitivity change in the 2 groups and the impact of baseline mean sensitivity, baseline age, and follow-up intraocular pressure for rate estimates.

Main Outcomes and Measures  Rate of mean sensitivity change in patients with glaucoma and healthy participants.

Results  A total of 40 patients with glaucoma (median age, 53.07 years [IQR, 48.34-57.97 years]; 21 men [52%]) and 29 healthy participants (median age, 48.80 years [IQR, 40.40-59.07 years], 17 women [59%]) were followed up for a median of 25.65 years (IQR, 22.49-27.02 years) and 19.56 years (IQR, 16.19-26.21 years), respectively. Most participants (65 individuals [94%]) self-identified as White, with the exception of 2 patients with glaucoma (1 self-identified as Black and 1 as South Asian) and 2 healthy participants (both self-identified as South Asian). The mean follow-up intraocular pressure of patients with glaucoma (median, 15.83 mm Hg [IQR, 13.05-17.33 mm Hg]) was similar to that of healthy participants (median, 14.94 mm Hg [IQR, 13.28-16.01 mm Hg]; P = .25). In an ordinary least-squares regression analysis, 31 patients (78%) with glaucoma had rates of mean sensitivity change within the range of healthy participants (ie, between −0.20 dB/y and 0.15 dB/y). Linear mixed-effects modeling revealed that the mean (SE) rate of mean sensitivity change in healthy participants was 0.003 (0.033) dB/y (95% CI, −0.062 to 0.068; P = .93). In comparison, patients with glaucoma had a mean (SE) rate of mean sensitivity change that was −0.032 (0.052) dB/y faster, but this difference was not statistically significant (95% CI, −0.134 to 0.070; P = .53). Among covariates, only baseline mean sensitivity was associated with the rate of mean sensitivity change (mean [SE], 0.021 [0.010] dB/y/dB; 95% CI, 0.002-0.041; P = .03).

Conclusions and Relevance  The results of this cohort study suggest that over a median follow-up of more than 25 years, the rate of visual field change in patients receiving treatment for glaucoma was comparable to that of healthy individuals. These findings could guide practitioners in making management decisions.

Assessing the extent of progressive visual field damage is one the most challenging aspects of glaucoma care because it substantially impacts treatment and monitoring decisions. Accurately estimating the rate of visual field change is important because the same rate has different consequences depending on the patient’s age, life expectancy, and severity of visual field damage at diagnosis. For example, a younger patient with moderate damage is at higher risk of lifetime visual impairment compared with an older patient with the same level of damage and rate of change.

The accuracy and precision of rate estimates depends mainly on the degree of damage, test variability, and frequency and spacing of tests.^1,2 The United Kingdom Glaucoma Treatment Study³ found that frequent tests over a short period can yield enough precision to discern differences among 2 treatment groups. Although this approach is relevant for clinical trials in which outcomes are assessed over a short period, it is not appropriate for a typical scenario in clinical practice in which testing is considerably less frequent and generally much lower than recommended guidelines.^4-9

The mean rate of progression, expressed in mean deviation (MD) in observational studies and clinical trials, ranges between −0.80 dB/y and 0.03 dB/y in patients who have received treatment^3,10-13 and between −1.08 dB/y and −0.29 dB/y in patients who have not.^3,14 Variability in reported rates is likely associated with population differences, baseline damage, and treatment protocols. Most studies have had a follow-up time of 5 years or less, with only a few extending to approximately 10 years.^10,11 However, these follow-up times are likely unrepresentative of the entire course of the disease for a given patient. Although shorter follow-up periods have identified linear, exponential, or episodic progression,^15-17 a linear model is the most parsimonious with which to explain long-term visual field behavior in most patients¹⁸ and can provide more realistic assessments of long-term progression and visual prognosis compared with shorter follow-up periods. Only 2 studies^19,20 have had a follow-up period of up to 20 years, but these studies either used infrequent testing to examine the incidence of visual field loss in the general population or used a grading scheme rather than an assessment of the rate of change.

The purpose of this study was to measure the rates of visual field change in patients with glaucoma and healthy participants enrolled in prospective studies with a median follow-up of 25 years. In addition, we evaluated the impact of factors associated with progression.

This longitudinal prospective cohort study was conducted in a hospital-based setting from January 1991 to February 2020. The study was approved by the research ethics board of the Nova Scotia Health Authority. All participants provided written informed consent and were offered nominal compensation for transportation and parking. The study was conducted in accordance with the Declaration of Helsinki²¹ and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

Patients with glaucoma and healthy participants were selected from a longitudinal prospective study initiated in 1991 that examined the value of new perimetric techniques for detecting the earliest progression of open-angle glaucoma.^22,23 Some individuals transitioned to other prospective studies but had been followed up with standard automated perimetry since initial recruitment.

Patients with glaucoma were included in the initial study according to the following criteria: (1) clinical diagnosis of open-angle glaucoma, including pseudoexfoliation glaucoma and pigmentary glaucoma, based on evidence of optic disc and visual field changes; (2) visual field MD between −10 dB and −2 dB using standard automated perimetry (30-2 test and full-threshold strategy of the Humphrey Field Analyzer; Carl Zeiss Meditec, Inc); and (3) open iridocorneal angle based on gonioscopic examination. Healthy individuals were included if they had (1) a normal ocular examination, (2) intraocular pressure (IOP) lower than 22 mm Hg, (3) no family history of glaucoma, and (4) no history of undergoing an ocular surgical procedure.

Exclusion criteria for both patients with glaucoma and healthy individuals were (1) best-corrected visual acuity lower than 20/40, (2) concomitant chronic ocular disease (with the exception of glaucoma in the glaucoma group), (3) systemic disease or medication known to produce changes in the visual field, and (4) refractive error exceeding ±5 D sphere or ±3 D astigmatism. One eye of each participant was randomly selected as the study eye.

All individuals received testing with standard automated perimetry every 6 months using the appropriate near correction for testing distance. Measurements of IOP were performed using the Goldmann applanation tonometer on the same day as the visual field tests.

From the whole sample, we selected all patients with glaucoma and healthy participants with at least 15 years of follow-up. Because the study began before introduction of the Swedish interactive thresholding algorithm (SITA) standard strategy, the visual field series contained a mixture of tests using the 30-2 full-threshold and 24-2 SITA standard strategies. For each individual, we extracted all reliable visual field tests (defined as false-positive and false-negative errors <15% and fixation losses <20%) as well as notes from the perimetrist.

Sensitivity values from the full-threshold and SITA strategies are not interchangeable (the full-threshold strategy yields lower sensitivity values compared with the SITA standard strategy).²⁴ Therefore, we converted values obtained using the full-threshold strategy to SITA standard equivalents using recently described nomograms.²⁵ In brief, for each visual field test, we converted the pointwise sensitivity values using the appropriate nomogram for either patients with glaucoma or healthy participants. We excluded test locations from the 30-2 test pattern that did not overlap with the 24-2 test pattern as well as the 2 blind spot locations. We computed mean sensitivity as the mean of pointwise sensitivity values based on the 24-2 test pattern (Figure 1). We analyzed mean sensitivity rather than the more conventional MD because the latter is age corrected, thereby not permitting longitudinal analyses of healthy participants.

We assessed a variety of follow-up variables, including follow-up time, number of visual field tests, interval between tests, and IOP. The Mann-Whitney test was used to compare baseline and follow-up variables between the 2 groups. For descriptive purposes, the rate of visual field change was computed using ordinary least-squares regression analysis.

A linear mixed-effects (LME) model with random slopes and intercepts for each individual was used to compute the rate of mean sensitivity change and estimate the group difference. The mean sensitivity for each visit was used as the dependent variable in the model. Diagnosis of glaucoma, baseline mean sensitivity, baseline age, IOP, and follow-up time were considered fixed effects. We included interaction terms between follow-up time and each of the following covariates: diagnosis of glaucoma, IOP, baseline mean sensitivity, and baseline age. Intercept, follow-up time, and IOP were considered random effects. Baseline mean sensitivity and age were centered on the median values of healthy participants for interpretation of the LME output. Further details on the LME model are available in eTable 1 in the Supplement.

Data were reported as means with SDs or medians with IQRs, as appropriate. Analyses were conducted using the lme4 package, version 1.1-27.1, for R software, version 3.6.0 (R Foundation for Statistical Computing).²⁶ The significance threshold was P = .05. All P values were 2-sided, with no adjustments for multiple analyses.

A total of 40 patients with glaucoma (21 men [52%] and 19 women [48%]) and 29 healthy participants (12 men [41%] and 17 women [59%]) were eligible for inclusion in the study. Most participants (65 individuals [94%]) self-identified as White, with the exception of 2 patients with glaucoma (1 self-identified as Black and 1 as South Asian) and 2 healthy participants (both self-identified as South Asian). Two patients (5%) had pseudoexfoliation glaucoma, and 3 patients (8%) had pigmentary glaucoma. At baseline, patients with glaucoma vs healthy participants had similar ages (median, 53.07 years [IQR, 48.34-57.97 years] vs 48.80 years [IQR, 40.40-59.07 years]), worse MD (median, −4.06 dB [IQR, −5.96 to −3.20 dB] vs −0.68 dB [IQR, −1.62 to −0.11 dB]; P < .001), worse pattern standard deviation (median, 3.61 dB [IQR, 2.89-7.18 dB] vs 2.01 dB [IQR, 1.78-2.50 dB]; P < .001), and worse mean sensitivity (median, 26.42 dB [IQR, 25.06-27.45 dB] vs 29.69 dB [IQR, 28.71-30.44 dB]; P < .001) (Table 1).

The median follow-up among patients with glaucoma was longer than that of healthy participants (25.65 years [IQR, 22.49-27.02 years] vs 19.56 years [IQR, 16.19-26.21 years], respectively), with 25 patients (62%) followed up for 25 years or more compared with 10 healthy participants (34%) (Table 1; Figure 2A). As a result, patients with glaucoma also received a higher median number of visual field tests during follow-up (median, 53.00 tests [IQR, 47.75-57.25 tests] vs 32.00 tests [IQR, 27.00-46.00 tests]; P < .001), whereas the interval between visual field tests was longer among healthy participants compared with patients with glaucoma (median, 7.26 months [IQR, 6.44-7.82 months] vs 5.86 months [IQR, 5.25-6.12 months]; P < .001) (Table 1; Figure 2B). A total of 28 healthy participants (97%) had 1 or more interval of 12 months or greater between consecutive tests compared with 27 patients (68%) with glaucoma.

Eighteen patients (45%) with glaucoma and 6 healthy participants (21%) underwent cataract surgery during the follow-up. In addition to medical therapy, 26 patients (65%) with glaucoma required further IOP-lowering interventions (single, multiple, or combined). At least once during follow-up, 16 patients (40%) with glaucoma received argon laser trabeculoplasty, 4 patients (10%) received peripheral laser iridotomy, 13 patients (32%) received trabeculectomy, 4 patients (10%) received phacotrabeculectomy, and 1 patient (5%) received a tube implant (eTable 2 in the Supplement).

The mean follow-up IOP was similar between the 2 groups, with a median of 15.83 mm Hg (IQR, 13.05-17.33 mm Hg) among patients with glaucoma vs 14.94 mm Hg (IQR, 13.28-16.01 mm Hg; P = .25) among healthy participants.

Among patients with glaucoma, the mean (SD) rate of mean sensitivity change derived using ordinary least-squares regression analysis was −0.13 (0.21) dB/y, and the median rate was −0.03 dB/y (IQR, −0.19 to 0.01 dB/y) (Figure 3). Among healthy participants, the mean (SD) rate of mean sensitivity change was −0.01 (0.06) dB/y, and the median rate was 0 dB/y (IQR, −0.04 to 0.02 dB/y). Thirty-one patients (78%) with glaucoma, including those with pseudoexfoliation glaucoma and pigmentary glaucoma, had rates of mean sensitivity change within the range of those in healthy participants (ie, between −0.20 dB/y and 0.15 dB/y). Among patients with glaucoma, the mean (SD) rate of MD change was −0.07 (0.21) dB/y, and the median rate of MD change was 0.01 dB/y (IQR, −0.15 to 0.07 dB/y). Among healthy participants, the mean (SD) rate of MD change was 0.05 (0.07) dB/y, and the median rate of MD change was 0.06 dB/y (IQR, 0.01-0.09 dB/y).

Results from the LME model and the impact of the covariates are shown in Table 2. Among healthy participants, the mean (SE) rate of mean sensitivity change was 0.003 (0.033) dB/y (95% CI, −0.062 to 0.068; P = .93). After adjusting for baseline mean sensitivity, baseline age, and IOP, the mean (SE) rate of mean sensitivity change was −0.032 (0.052) dB/y faster among patients with glaucoma vs healthy participants; however, this difference was not statistically significant (95% CI, −0.134 to 0.070; P = .53). Among the covariates evaluated in the model, IOP (mean [SE] estimated rate of change, 0 [0.001] dB/y/mm Hg; 95% CI, −0.002 to 0.001; P = .63) and baseline age (mean [SE] estimated rate of change, −0.003 [0.002] dB/y/y; 95% CI, −0.007 to 0.001; P = .11) were not associated with the rate of mean sensitivity change. Baseline mean sensitivity was the only covariate that altered the rate of mean sensitivity change, with higher mean sensitivity at baseline associated with slower rate of change (mean [SE] estimated rate of change, 0.021 [0.010] dB/y/dB; 95% CI, 0.002-0.041; P = .03).

In this cohort study, we estimated the rate of visual field change in patients receiving treatment for glaucoma who were followed up prospectively for a median of more than 25 years, with 2 visual field tests per year. To contextualize this rate, we also followed up a parallel group of healthy participants. The rate of visual field change in patients receiving treatment for glaucoma was not different from that of healthy participants, with 31 patients (78%) having values within the range of healthy participants. To our knowledge, no other published prospective studies on visual field change have been conducted with such an extended follow-up time.

Minimizing the risk of visual impairment among patients with glaucoma relies on making optimal clinical decisions, including the mode and intensity of therapeutic interventions. These decisions frequently depend on an accurate assessment of the rate of visual field change. However, the precision of such estimates is limited, by either short follow-up duration or infrequent testing. The rate of MD change in large clinical populations of patients with glaucoma has been reported to range from −0.62 dB/y to −0.05 dB/y,^12,27-29 whereas the rate in smaller clinical studies has ranged from −0.21 dB/y to 0.03 dB/y.^3,10,11 In our study, the mean rate of mean sensitivity change derived using an LME model was approximately −0.03 dB/y, which is slower compared with previous reports and may be associated with our study design. Patients were tested every 6 months, and visit adherence was high. Although patients were not required to be excluded from follow-up if their visual field damage or rate of change exceeded a certain value, there could have been selective retention of those with lower rates of mean sensitivity change. Furthermore, because patients in clinical studies might be more adherent to follow-up visits, they could also presumably be more adherent to therapy and thus progress at slower rates compared with patients from large clinical populations.³⁰ As a counterpoint to these arguments, a previous study²⁷ examined rates of visual field change among patients from a large clinical population who were not enrolled in research studies compared with rates among patients in the Canadian Glaucoma Study,^31,32 matched pairwise for baseline visual field damage. No difference in visual field change was observed among these pairs of patients, which was notable given that the Canadian Glaucoma Study mandated a standardized treatment intervention based on visual field progression.³¹ In addition, the distributions of the rates of visual field change in patients with different lengths of follow-up were highly similar.²⁷ Therefore, the results of the present study likely have relevance to clinical populations.

Our estimate of visual field rates of change could have been impacted by our methods of analyses. First, we chose to use all visual field data in the follow-up analysis, including those derived using the older full-threshold strategy. We applied nomograms to convert pointwise sensitivity values to SITA-equivalent values to increase the number of tests available for analysis. However, if the nomograms underestimated the SITA-equivalent sensitivity values, the net result when combining these values with the actual SITA sensitivity values obtained later in follow-up could have produced slopes with values that were less negative. To rule out this hypothesis, we performed a second analysis comparing the ordinary least-squares regression rate of visual field change derived exclusively using the full-threshold strategy with the rate derived using the SITA-equivalent sensitivity values. The median of individual rates derived using the full-threshold strategy was not substantially different from that derived using the SITA-equivalent sensitivity values (−0.07 dB/y vs −0.05 dB/y, respectively) and was similar to rates exclusively derived using actual SITA sensitivity values (−0.09 dB/y). Second, we used mean sensitivity to estimate the rate of visual field change because we wanted to contrast our results to those from the parallel cohort of healthy participants. Although MD is more typically used, it accounts for age-related loss and was therefore unsuitable for the present study.

In a study of a large clinical population, Heijl et al¹² reported a median MD rate of −0.62 dB/y, which is considerably faster than our estimate. However, the mean follow-up IOP in that study ranged between 18.1 mm Hg and 20.2 mm Hg,¹² whereas in the present study, patients with glaucoma had a median follow-up IOP of 15.83 mm Hg, which was not different from that of healthy participants (Table 1). The dissimilarity in follow-up IOP between these 2 studies is substantial and may explain the differences in visual field outcomes. Our study was not designed to assess the consequences of IOP lowering for visual field change because patients were not treated according to a standardized protocol but by clinical judgment and a target IOP that was set based on the clinical appearance of the optic nerve head and visual field. A range of treatment modalities was used, with 65% of patients with glaucoma receiving either laser or surgical treatment in addition to medical therapy to achieve target IOP.

Our findings were consistent with those of studies reporting that patients with worse visual fields progressed faster compared with patients with earlier damage.^3,28,33-35 Because of the study design, the median baseline MD of patients with glaucoma was −4.06 dB; hence, it is plausible that inclusion of patients with more advanced damage may have produced faster rates of visual field change, limiting the generalizability of our findings. Factors in addition to baseline visual field damage could have yielded a lower rate of visual field change, suggesting other potential sources of selection bias toward patients with lower risk of progression. However, patients were recruited consecutively; therefore, at baseline, the risk of progression among enrolled patients was likely similar to that of eligible patients who declined to participate or those who were unavailable for follow-up.

Eighteen patients (45%) with glaucoma and 6 healthy participants (21%) received cataract surgery during the follow-up period. In addition, 4 patients (10%) with glaucoma received a combined cataract and glaucoma procedure. A previous study³⁶ reported a significant increase in visual field sensitivity after cataract surgery among patients with glaucoma, whereas another study³⁷ found only minor change, likely because indications for cataract surgery vary among centers. A previous study³⁸ involving the same population of patients with glaucoma as the present study found that cataract surgery had a negligible impact for the visual field; thus, we chose not to censor data or account for cataract surgery in our current analysis. Nevertheless, to validate our approach, we repeated ordinary least-squares regression analysis using only data from before cataract surgery and found a median rate of mean sensitivity change of −0.07 dB/y (IQR, −0.30 to 0.02 dB/y) in patients with glaucoma and 0.01 dB/y (IQR, −0.05 to 0.02 dB/y) in healthy participants, values that were similar to those obtained with uncensored data.

Spry and Johnson³⁹ reported that the age-related decrease in visual field sensitivity was more pronounced with older age. Several studies have indicated that normal age-related decrease in visual field sensitivity^40-42 and structural measurements with optical coherence tomography^43,44 can explain a portion of the change associated with glaucomatous progression. Patients with glaucoma in our study were relatively young, with a median age of 53 years at baseline. If patients who were older at baseline had been included, the estimated rate of mean sensitivity change might have been faster. Although there was no statistically significant difference in the rate estimate between patients with glaucoma and healthy participants, the point estimate was 10 times faster in patients with glaucoma. The lack of statistical significance can be explained by a combination of the inherently variable rates of visual field change among individuals with rates close to 0 dB/y and the relatively small samples included in the study, which reduced the precision of the estimates. However, it is important to note that our estimates were not substantially different from those obtained among large clinical populations.^27-29

This cohort study found that the mean rate of visual field change among patients with early glaucoma who were followed up for more than 25 years was comparable to that of healthy individuals. However, consistent with previous studies,^27-29 a small proportion of patients exhibited faster rates of change and were at higher risk of visual impairment. Although inferential statements about the degree of IOP reduction cannot be made based on these findings, it is notable that the mean follow-up IOP was not different between patients with glaucoma and healthy participants. Although some selection biases may have underestimated the extent of visual field change, the findings of this study are relevant to a large portion of the population with glaucoma and highlight the fact that most patients who receive treatment for glaucoma have rates of visual field that are modest and within the range of healthy individuals.

Accepted for Publication: February 15, 2022.

Published Online: April 7, 2022. doi:10.1001/jamaophthalmol.2022.0671

Corresponding Author: Balwantray C. Chauhan, PhD, Department of Ophthalmology and Visual Sciences, Dalhousie University, 1276 South Park St, Victoria Building, Room 2035, Halifax, NS B3H 2Y9, Canada (bal@dal.ca).

Author Contributions: Drs Giammaria and Chauhan had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Giammaria, Rafuse, LeBlanc, Chauhan.

Acquisition, analysis, or interpretation of data: Giammaria, Hutchison, Shuba, Nicolela, Chauhan.

Drafting of the manuscript: Giammaria, Hutchison, Chauhan.

Critical revision of the manuscript for important intellectual content: Giammaria, Rafuse, Shuba, LeBlanc, Nicolela, Chauhan.

Statistical analysis: Giammaria, Chauhan.

Obtained funding: Chauhan.

Administrative, technical, or material support: Hutchison, Nicolela, Chauhan.

Supervision: Rafuse, Shuba, Nicolela, Chauhan.

Conflict of Interest Disclosures: Dr Nicolela reported receiving grants from Allergan and personal fees from Bausch + Lomb Canada and Labtician Théa outside the submitted work. Dr Chauhan reported receiving grants from the Canadian Institutes of Health Research during the conduct of the study. No other disclosures were reported.

Funding/Support: This study was supported by grants MOP11357 (Dr Chauhan), MOP200309 (Dr Nicolela), and PJT159564 (Dr Chauhan) from the Canadian Institutes of Health Research.

Role of the Funder/Sponsor: The funding institution had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.