Association of Fast Visual Field Loss With Risk of Falling in Patients With Glaucoma

Importance  Patients with glaucoma and a history of fast visual field loss might be at an increased risk for falls compared with those with a history of slow visual field loss, but, to date, this association has not been previously investigated in the literature.

Objective  To evaluate the association between self-reported falls and past rate of visual field loss in a cohort of patients with glaucoma followed up over time.

Design, Setting, and Participants  This observational cohort study included patients diagnosed as having glaucoma who had been followed up at the Visual Performance Laboratory, University of California, San Diego, at 6-month intervals for a mean (SD) of 7.5 (2.6) years from January 1, 2005, through December 31, 2015. Self-reported number of falls during the past year was obtained at the last follow-up visit. Integrated binocular fields were estimated from the monocular fields. Linear mixed models were used to calculate rates of change in binocular mean sensitivity over time. Poisson models were used to evaluate the association between the self-reported number of falls and rates of visual field loss. The models adjusted for the current level of visual field damage and other confounding variables.

Main Outcomes and Measures  Association between rates of binocular visual field loss and self-reported number of falls.

Results  The study included 116 patients with glaucoma with a mean (SD) age of 73.1 (10.7) years (55 women [47.4%], 84 white individuals [72.4%], and 32 black individuals [27.6%]). Of the 116 patients, 29 (25.0%) reported at least 1 fall in the previous year. The mean rate of change in binocular mean sensitivity was faster for patients who reported a history of falls vs those who did not (−0.36 vs −0.17 dB/y; mean difference, 0.20 dB/y; 95% CI, 0.09-0.31 dB/y; P < .001). History of fast visual field loss was significantly associated with falls (rate ratio, 2.28 per 0.5 dB/y faster; 95% CI, 1.15-4.52 db/y; P = .02), even after adjusting for confounding factors.

Conclusions and Relevance  The rate of visual field loss was associated with a self-reported history of falls in the past year even after taking into account the magnitude of visual field defect. However, although a positive association was found, further studies are necessary to establish whether a cause-and-effect relationship exists between rate of visual field loss and self-reported history of falls.

Falls are a major cause of morbidity and mortality and the leading cause of injury-related hospitalization in elderly populations.¹ It is estimated that up to one-third of individuals older than 65 years will fall each year. Of those who fall, up to 15% may have severe injuries, such as hip fractures and traumatic brain injury.²

Patients with glaucoma are reported to be at higher risk of falling compared with individuals without the disease.^3-9 Cross-sectional studies^6-9 have found that more severe visual field loss is associated with increased risk of falls and fear of falling. However, besides the actual severity of visual field defect, it is also possible that the velocity at which the patient developed these defects may be an important factor that contributes to the risk of falling. It is conceivable that individuals with a recent history of fast visual field deterioration may not have had time to develop compensatory mechanisms to adjust for the loss of vision, potentially leading to increased risk of falling. In contrast, individuals with a similar degree of visual field loss, but in whom the defects have been stable for a relatively long period, may have developed compensatory strategies that would put them at a decreased risk of falling compared with those with fast visual field loss. Despite the potential significance of this association, to our knowledge, no study has yet evaluated the association between velocity of visual field loss and risk of falling in patients with glaucoma. In the current study, we report the association between self-reported falls and rate of visual field loss in a cohort of patients with glaucoma followed up longitudinally.

Box Section Ref ID

Participants from this study were included in a prospective longitudinal study performed from January 1, 2005, through December 31, 2015, designed to evaluate functional impairment in glaucoma conducted at the Visual Performance Laboratory, Department of Ophthalmology, of the University of California, San Diego. The institutional review board at the University of California, San Diego, approved the methods, and written informed consent was obtained from all participants. The study adhered to the laws of the Health Insurance Portability and Accountability Act, and all study methods complied with the Declaration of Helsinki guidelines for human subject research. Data were deidentified.

At each visit during follow-up, participants underwent a comprehensive ophthalmologic examination, including review of medical history, slitlamp biomicroscopy, intraocular pressure measurement using Goldmann applanation tonometry, corneal pachymetry, gonioscopy, dilated ophthalmoscopy, stereoscopic optic disc photography, and standard automated perimetry (SAP) using the 24-2 Swedish Interactive Threshold Algorithm (Carl Zeiss Meditec Inc). Visual acuity was measured using the Early Treatment Diabetic Retinopathy Study chart. Only patients with open angles on gonioscopy were included. Patients with coexisting retinal disease, uveitis, or nonglaucomatous optic disc neuropathy were excluded from the study. Patients who had eye surgery during the study period were not excluded from this study.

Glaucoma was defined by the presence of 2 or more consecutive abnormal SAP test results at baseline, defined as a pattern SD with P < .05 and/or glaucoma hemifield test results outside normal limits, and evidence of glaucomatous optic neuropathy based on masked assessment of stereophotographs. A patient was considered to have glaucoma if damage was present in at least 1 eye. During the follow-up period, visual field tests were performed at 6-month intervals. The study included a total of 1354 SAP visits, with a mean (SD) of 11.7 (4.6) visits per patient.

Monocular SAP was performed using the 24-2 Swedish Interactive Threshold Algorithm standard test. Only reliable tests (≤33% fixation losses and ≤15% false-positive results) were included. To evaluate binocular visual field loss, sensitivities of the monocular SAP examinations of the right and left eyes were used to calculate an integrated binocular visual field, according to a previously described method.¹⁰ According to this model, the binocular sensitivity can be estimated using the following formula:

with Sr and Sl indicating the monocular threshold sensitivities for corresponding visual field locations of the right and left eyes, respectively. To calculate the binocular sensitivity from this formula, light sensitivity had to be converted to a linear scale (apostilbs) and then converted back to logarithmic scale (decibels).

Evaluation of rates of visual field change was performed using the mean sensitivity (MS) of the binocular visual field. We also calculated the MS values for the inferior and superior hemifields of the integrated binocular field. In addition, we report rates of change according to the mean deviation (MD) values of the better and worse eyes, as defined from baseline visual fields.

History of falls was obtained using the Falls Screening and Referral Algorithm (FSRA)¹¹ administered at the last follow-up visit. The FSRA consists of the Elderly Fall Screening Test and the Multi-factor Falls Questionnaire. The questionnaires were administered in a printed version and asked about the number of falls the patient had during the past year. The single question used for this study was, “Have you had any fall in the past 12 months?” A fall was considered when the individual found himself or herself suddenly on the ground, without intending to get there from a sitting or standing position.

The level of physical activity was investigated as a potentially confounding variable using the Physical Activity Scale for the Elderly (PASE) questionnaire, where higher scores indicate a greater level of physical activity.^12,13 Personal and anthropomorphic information collected included age, sex, race, height, and weight. Body mass index (BMI) was calculated as the weight in kilograms divided by height in meters squared. Data on the presence of the following comorbidities were also collected: diabetes, arthritis, high blood pressure, heart disease, chronic pulmonary obstructive disease, stroke, and cancers. A simple summation score was used to create a comorbidity index.¹⁴

The main outcome variable was the rate of falls as obtained by the FSRA questionnaires administered during the last follow-up visit. We then analyzed the association between rate of falls and the current status of visual field loss, as indicated by the integrated binocular MS of the visual field test closest to the date of questionnaire administration. In addition, we calculated the rates of visual field loss using visual field data from the period that preceded the administration of the falls questionnaire. All reliable visual field tests performed during the study period were used to calculate the rates of loss. This approach allowed us to investigate whether the number of self-reported falls was associated with history of fast visual field loss, adjusting for the current level of visual field damage, according to our hypothesis.

Rates of visual field loss were obtained by linear mixed models and reported using the MS of the integrated binocular visual field data. Ability to predict history of falls was investigated with Poisson regression models, where the number of falls during the previous year was used as the dependent variable and visual field results as the independent variables. Age, sex, BMI, PASE score, comorbidity index, and visual acuity were considered potentially confounding variables in multivariable analyses. Results of the Poisson models were given as the effect of the variables on the rate ratio (RR).

All statistical analyses were performed with STATA software, version 13 (StataCorp).

The study included 116 patients with glaucoma with a mean (SD) age of 73.1 (10.7) years at the date of administration of the FSRA questionnaires (55 women [47.4%], 84 white individuals [72.4%], and 32 black individuals [27.6%]). Of the 116 patients, 29 (25.0%) reported at least 1 fall in the previous year. Of the 29 patients who reported falling, 15 (51.7%) had 1 fall, 8 (27.6%) had 2 falls, 4 (13.8%) had 3 falls, and 2 (6.9%) had 4 falls. The eTable in the Supplement gives the demographic and clinical variables for patients with a history of falls vs those without at the time of administration of the questionnaire. Patients with a history of falls were older than those without a history of falls (mean [SD] age, 77.8 [7.1] vs 71.5 [11.3] years; P = .006). In addition, patients with a history of falls had a lower level of physical activity as indicated by the PASE score (mean [SD], 123.2 [65.8] vs 168.6 [96.2]; P = .02).

Quiz Ref IDPatients who reported a history of falls had worse visual field test results at the time of questionnaire administration (eTable in the Supplement). The mean MD in the better eye was −4.6 and −2.8 dB in those with and without a history of falls, respectively (P = .02). Corresponding numbers for MD in the worse eye were −9.1 vs −6.3 dB (P = .04), respectively. For binocular MS, corresponding numbers were 26.1 and 27.9 dB (P = .01), respectively. For binocular inferior hemifield MS, the differences were 26.7 and 28.7 dB, and for the binocular superior hemifield MS, the differences were 25.4 and 27.1 dB, respectively, in those with and without a history of falls (eTable in the Supplement).

The patients included in the study had been followed up for a mean (SD) of 7.5 (2.6) years (range, 2.0-10.0 years) before assessment of falls. Table 1 gives the rates of visual field loss for patients who reported a history of falls vs those who did not. Mean rate of change in binocular MS was faster for patients who reported a history of falls vs those who did not (−0.36 vs −0.17 dB/y; mean difference, 0.20 dB/y; 95% CI of difference, 0.09-0.31 dB/y; P < .001). Figure 1 illustrates the time course of changes in binocular MS for patients with and without a history of falls. It can be seen that more rapid visual field loss occurred in those who reported a recent history of falls compared with those who did not. Differences in mean rates of change were also found for the other visual field parameters evaluated.

Quiz Ref IDTable 2 gives the results of univariable Poisson regression models investigating factors associated with the rate of falls. The severity of visual field defect as measured by the binocular MS at the time of questionnaire administration was associated with falls (RR, 1.67 per 5 dB lower; 95% CI, 1.18-2.36; P = .004). The rate of change in binocular MS was also associated with falls. Each 0.5-dB/y faster rate of visual field loss was associated with a more than 2 times higher rate of falls (RR, 2.07 per 0.5-dB/y faster rate of visual field loss; 95% CI, 1.45-2.96; P < .001). Table 2 also gives the results of the multivariable Poisson models that investigated factors associated with falls after adjusting for confounding variables. The association between rate of change and history of falls persisted after adjustment for disease severity at the time of questionnaire assessment and the other potentially confounding variables (RR, 2.28 per 0.5-dB/y faster rate of visual field loss; 95% CI, 1.15-4.52; P = .02). Figure 2 illustrates the association between the probability of falling and rates of change in MS.

When rates of change were considered by hemifield, rates of change in the binocular inferior hemifield had a stronger association with a history of falls (RR, 2.47 per 0.5-dB/y faster rate of visual field loss; 95% CI, 1.26-4.84; P = .009) than those from the binocular superior hemifield (RR, 1.90 per 0.5-dB/y faster rate of visual field loss; 95% CI, 1.01-3.54; P = .04), after adjustment for confounding variables (P = .03 for comparison between RRs) (Table 2). Figure 3 illustrates the RRs for association with falls for rates of visual field loss for the different parameters investigated.

In the current study, we found that history of fast visual field loss was significantly associated with risk of falling in patients with glaucoma. Such an association was present even after adjustment for the degree of disease severity closest to the date of assessment of falls. This finding suggests that not only the amount of visual field damage but also the rate at which it has occurred may have an effect on explaining the risk of falling in patients with glaucoma. These findings may have implications for the understanding of risk factors for falls in patients with glaucoma. In addition, they underscore the importance of evaluating rates of visual field change in the functional rehabilitation of patients with glaucoma.

Patients who reported falling had rates of visual field loss that were a mean of more than 2 times faster than in those who did not report a history of falls. As Figure 2 shows, in patients with a history of fast visual field loss, the risk of falls would be much higher than in those with slower loss. For example, a patient with a rate of change of −1.0 dB/y would have a probability of falling of approximately 70% compared with less than 20% for a patient with a rate of change of −0.1 dB/y. Notably, a rate of loss of 1 dB/y in the binocular field corresponds to a devastating loss, leading to substantial functional damage because the binocular field is in general dominated by the better eye of the patient.¹⁰

Quiz Ref IDThe association between rates of visual field change and falls was even stronger when the inferior hemifield of the integrated binocular field was considered. This finding is not surprising considering the presumed greater importance of the inferior field in navigation and obstacle avoidance.^5,15-17 A study by Black et al⁶ found that the severity of inferior visual field loss in the integrated binocular field had greater association with risk of falls than measurements from the superior field. They found that a 5-dB lower sensitivity of the inferior integrated binocular field was associated with a RR of 1.56. In fact, our RR was very similar to the one reported by Black et al,⁶ with an RR of 1.65 for a 5-dB lower MS of the inferior binocular field. However, the study by Black et al⁶ used only cross-sectional visual field assessment. Our study found that in addition to the severity of visual field loss at the time of falls, the velocity that the loss occurred was a major determinant of the risk of falling. For the inferior binocular MS, each 0.5-dB/y faster rate of visual field loss was associated with an RR of falling of 2.47. For the superior binocular MS, the association was weaker, with each 0.5-dB/y faster rate of visual field loss associated with an RR of 1.90.

The fact that the association between rate of field loss and falls persisted despite adjustment for the amount of field damage at the time of falls is interesting. In principle, one would think that the risk of falls would depend essentially on how severe the visual field defect is at a given moment. However, our results revealed that for 2 patients with the same disease severity, the patient who has been worsening faster would have higher risk of falling. This finding might be related to the mechanisms used by patients to compensate for the field losses. Patients in whom the disease has worsened slowly might be able to develop compensatory mechanisms to overcome the limitations caused by their field defects. On the other hand, those in whom the disease worsens fast, particularly affecting the better eye (and consequently the binocular field), might not have enough time to work out these compensatory mechanisms and would be at increased risk for falling. It is possible, for example, that postural adjustments in patients with rapid loss of the inferior field, such as turning the head down to use assistance from the superior field when walking, could cause loss of balance and predispose patients to falls. Although patients with inferior field loss who worsened slowly may also use head movements to compensate for the field defects, it is possible that other body adjustments may have taken place over time to compensate for the imbalance caused by head movement. The mechanisms that explain our findings, however, are still not clear and should benefit from further research. Of note, previous studies^14,18-20 have found that rapid worsening of visual fields is associated with decreases in quality of life and depression, as assessed by self-reported questionnaires. Patients with fast visual field loss report more difficulty with driving, climbing stairs, or seeing objects on the side, even after adjustment for the level of disease severity.^14,18,20 It is also possible that instead of developing compensatory mechanisms, patients with more gradual visual field loss had fewer falls because they were less mobile and had decreased their life space over time.²¹

Quiz Ref IDOur study has limitations. The primary outcome measure of the self-reported number of falls was obtained at only 1 time point, which may have underestimated the true incidence of falls. A previous study²² found that elderly persons often do not recall falls that occurred within this period. However, a retrospective recall of falls has high specificity compared with a prospective ongoing collection of falls data, despite a lower level of sensitivity.²³ It is therefore possible that a prospective collection of falls data may reveal an even stronger association between the number of falls and the rate of visual field loss, although future studies are required to examine this. We were careful to adjust our analyses to the current level of field defects and other confounding variables and found clinically relevant associations that should serve to motivate further prospective longitudinal studies investigating these issues. Another limitation of our study was that we did not measure field loss and rates of change in regions extending beyond the central 60º. However, although field defects that affect more peripheral field regions could also contribute to risk of falls, a previous study⁶ suggested that the integrated binocular field is as good as strategies that test more peripheral points in identifying patients with glaucoma at risk for falling. Our study did not collect information on depressive symptoms, which is a risk factor for falls.²⁴ Because a previous study²¹ found the rate of visual field loss was associated with the worsening of depressive symptoms in patients with glaucoma, future studies are required to elucidate whether both these factors contribute independently to the risk of falls. Our findings should be tempered by recognition that they are based on an association but do not necessarily imply a cause-and-effect relationship between rate of visual field loss and a recent self-reported history of falls.

Rate of visual field loss was associated with risk of falling in patients with glaucoma. A history of fast visual field loss in the inferior integrated binocular field had a major effect on risk of falls in our cohort. Given the potentially serious consequence of falls, our work suggests the need for careful assessment of rates of visual field loss that affect the inferior field in patients with glaucoma. Further studies should clarify the mechanisms potentially explaining the associations uncovered in the present work.

Submitted for Publication: January 7, 2016; final revision received April 14, 2016; accepted April 18, 2016.

Corresponding Author: Felipe A. Medeiros, MD, PhD, Visual Performance Laboratory, Department of Ophthalmology, University of California, San Diego, 9500 Gilman Dr, La Jolla, CA 92093-0946 (fmedeiros@ucsd.edu).

Published Online: June 9, 2016. doi:10.1001/jamaophthalmol.2016.1659.

Author Contributions: Dr Medeiros had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Diniz-Filho, Medeiros.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Baig, Diniz-Filho, Wu, Abe, Gracitelli, Cabezas, Medeiros.

Critical revision of the manuscript for important intellectual content: Baig, Diniz-Filho, Wu, Abe, Gracitelli, Medeiros.

Statistical analysis: Baig, Diniz-Filho, Wu, Abe, Gracitelli, Medeiros.

Obtained funding: Medeiros.

Administrative, technical, or material support: Diniz-Filho, Cabezas, Medeiros.

Study supervision: Baig, Diniz-Filho, Medeiros.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Funding/Support: This study was supported in part by grants EY021818 and EY025056 and core grant P30EY022589 from the National Institutes of Health/National Eye Institute (Dr Medeiros), an unrestricted grant from Research to Prevent Blindness to the Department of Ophthalmology, University of California, San Diego, fellowship grant 233829/2014-8 from the Brazilian National Council for Scientific and Technological Development (Dr Diniz-Filho), fellowship grant 1104985 from National Health and Medical Research Council (Dr Wu), and fellowship grant 12309-13-3 Brazilian National Research Council (Dr Gracitelli).

Role of the Funder/Sponsor: The funding sources 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.