Correlations Between Subjective Evaluation of Quality of Life, Visual Field Loss, and Performance in Simulated Activities of Daily Living in Glaucoma Patients

Supplemental Digital Content is available in the text. Precis: Glaucoma patients displayed alterations in their quality of life (QoL) and their ability to perform activities of daily living. The visual field (VF) of the worse eye might serve as a good marker for QoL evaluation. Purpose: The purpose of this study was to explore the correlations between VF defects, performance in simulated activities of daily living, and subjective evaluation of QoL in glaucoma patients. Methods: Thirty-two patients with glaucoma and 10 age-matched control subjects were included. All participants answered a QoL questionnaire and underwent an assessment of visual function including monocular and binocular best-corrected visual acuity, binocular contrast sensitivity test (LogCS), and monocular and binocular VF. All subjects also carried out a series of simulated activities of daily living in a controlled environment. Results: Glaucoma patients had lower QoL scores compared with controls for the composite score, near and distance activities, social functioning, mental health, role difficulties, dependency, and color vision. With regard to performance in the simulated mobility task, the number of mobility incidents was higher for glaucoma patients than for control subjects. For the reaching and grasping tasks, the overall movement duration for small objects was significantly longer in glaucoma patients compared with controls. The VF mean deviation of the worse eye was correlated with most of the QoL subscores. Mobility incidents as well as the reaching and grasping task parameters were not significantly correlated with QoL scores. Conclusions: Glaucoma patients showed an alteration of performance in simulated daily living activities, associated with a decreased QoL. There was no clear correlation between alterations in QoL and ability to perform activities of daily living. The QoL related to vision was mostly correlated to the visual function of the worse eye.

G laucoma is one of the leading causes of irreversible blindness; it may affect up to 111.8 million people worldwide in 2040. 1 It is generally characterized by a progressive loss of retinal ganglion cell axons, associated with a reduction of visual field (VF). 2,3 The progressive visual function loss, as well as the medical or surgical treatment, or even the diagnosis of glaucoma itself, can affect the daily life of glaucoma patients and, consequently, their quality of life (QoL). 4 QoL is a complex and subjective concept that involves the individual perception of emotional, physical, material, and social factors. As the goal of glaucoma treatment is to preserve the patient's QoL, 5 direct or indirect evaluation of QoL is of central importance in the management of the glaucoma patient.
In contrast to visual impairment and stage of glaucoma, which can be measured precisely by objective tests such as visual acuity (VA) measurement, contrast sensitivity, and VF testing, evaluation of the impact of glaucoma on daily activities and QoL remains a challenge. 5 The monocular VF is the gold standard used in clinical practice to evaluate glaucoma patients' visual function and glaucoma progression, so as to adjust treatment accordingly. 6 The better eye VF mean deviation (MD) has been directly correlated with visual function and is used in QoL studies in glaucoma. [7][8][9] As glaucoma patients use one eye to compensate for the other, binocular VFs, or systems integrating both monocular VF tests to reproduce a binocular VF, are also used. [9][10][11][12] However, using a concomitant evaluation of 4 different activities of daily living, our group showed, consistently with several other studies, that neither monocular nor bilateral VFs can thoroughly describe the effect of glaucoma on a patient's everyday activities of daily living. 9,[13][14][15][16] Although visual impairment due to glaucoma has a significant negative impact on patients' ability to perform activities of daily living, the exact correlation between objective visual changes and their consequences on QoL are not fully understood and remain difficult to evaluate. Simply stated, patients with similar visual function might experience different performance in activities of daily living and/or rate their QoL differently. Thus, the purpose of the present study was to evaluate correlations between subjective QoL evaluation and objective evaluation of visual function and performance in simulated activities of daily living, to better understand patients' perception of their disease and disabilities.

Participants
A total of 32 patients with glaucoma, aged 18 to 80 years, and 10 age-matched control subjects were included. All participants were informed of the purpose of the study, and their signed consent was obtained before inclusion. The study was approved by the Saint-Antoine Ethics Committee (CPP IDF VI P16-03) and the National Agency for the Safety of Medicines and Health Products (ANSM) (2016-A01371-50).
Glaucoma patients were followed-up regularly at the Quinze-Vingts National Ophthalmology Hospital, Paris, France. They were required to have stable glaucoma, confirmed with at least 3 VF tests over the past 3 years, and monocular VA in both eyes of at least 0.6 [best-corrected visual acuity (BCVA) <0.2 logarithm of the minimum angle of resolution (LogMAR)]. They were required to be autonomous without motor or cognitive problems that could interfere with the patient's full understanding of testing instructions. The noninclusion criteria were pregnancy or breastfeeding, inability to personally give informed consent, systemic disease, or medications that might cause visual or cognitive impairment.
Glaucoma patients were divided into 2 groups according to the Hodapp, Parish, and Anderson (HPA) classification. 6,17 Three principal criteria are considered for this classification: the overall extent of damage using both the MD index and proximity of the defects to fixation, and the number of defective points on the Humphrey Statpac-2 pattern deviation probability map of the 24-2 SITA-STANDARD test (Humphrey Visual Field). According to this classification, 26 patients had early-stage glaucoma (EG group, stage 0 to 1), and 6 had advanced glaucoma (AG group, stage 2 to 4).
The control group (CO) had no systemic or ocular disease in terms of VF impairment or VA loss. Monocular VA in both eyes was at least 20/25 (BCVA < 0.1 LogMAR).

Procedure
All participants in the study underwent a self-administered QoL questionnaire, a complete visual function assessment, and an objective evaluation of simulated activities of daily living.
All patients underwent an evaluation of monocular and binocular BCVA, converted to the LogMAR, and a binocular contrast sensitivity test (LogCS). For all patients, monocular and binocular VFs were also recorded. For monocular tests, patients underwent a Humphrey perimeter 24-2 threshold test with the SITA-Standard program on the Humphrey Visual Field Analyzer (HFA) (Carl Zeiss Meditec, Dublin, CA). The MD index of the better (MD-BE) and worse (MD-WE) eyes was recorded. For the binocular VF, patients performed an Esterman binocular VF using the HFA. The Esterman score was reported as points seen out of the 120 points evaluated in a VF extending over 140 degrees horizontally and 110 degrees vertically.
The National Eye Institute Visual Function Questionnaire-25 (NEI VFQ-25) was used as a self-administered QoL questionnaire. The NEI VFQ-25 consists of 25 questions and 14 additional optional questions (total of 39 questions) that address 12 aspects of daily living: general health (GH), general vision (GV), ocular pain (OP), near activities (NA), distance activities (DA), social functioning (SF), mental health (MH), role difficulties (RD), dependency (Dp), color vision (CV), peripheral vision (PV), and driving (D). In this study, we evaluated the 39 questions except for the 2 questions about D, as most of the patients did not drive. We used a French translated version validated for glaucoma and ocular hypertension. 18 Patients answered the questionnaires themselves or, when necessary, with the assistance of a family member. Each question has a score between 0 and 100, where "100" represents a better QoL. Results were distributed according to 12 subscores covering each of the 12 individual domains and a composite score (CS) representing the mean score of all subscores, except for GH, which is not directly related to the visual condition, 19 and D, which was not analyzed in the present study.
Two different simulated activities of daily living were performed in the present study: a mobility task and a reaching and grasping task. Both tasks were performed on the StreetLab platforms of the Vision Institute (IHU FOReSIGHT, Paris, France) according to previously validated protocols. 9 These platforms provide a controlled environment with an adjustable light system and reproducible experimental conditions. Briefly, the mobility task consisted of a 15 m indoor course with controlled illumination intensity (250 lux, 4350 K). Subjects were instructed to walk at their preferred walking speed, following an indoor route with various obstacles (chairs, desks, or tables). Four mobility courses with a similar level of complexity and number of obstacles were assessed in the same order for all subjects. The time to travel the path (TP) and the number of mobility incidents (MI), such as bumps, stumbling, or stops, were recorded during the trial. For the reaching and grasping tasks, the subject was seated in front of a counter with 3 small objects (S) and 3 large objects (L). They randomly performed 5 reachand-grasp tasks for each target object, located on the right (near and far), on the left (near and far), or in the middle of the counter, in a randomized order. The movement onset (MO), representing visual search and movement initiation time, and the overall movement duration (OMD), corresponding to the reaching and grasping time, were recorded.

Statistical Analyses
R, v.3.4.2 Development Core Team (2008) was used for statistical analysis. Descriptive statistics were used to analyze demographic data. Qualitative variables, such as sex or errors in the various performance tasks, were compared between the CO group and the glaucoma groups with a χ 2 test. Quantitative variables were compared between groups using the nonparametric Kruskal-Wallis H test; multiple pairwise-comparisons between groups were calculated using the post hoc Dunn test (false discovery rate) with Benjamini-Hochberg adjustment for significant differences. Subsequently, in the glaucoma population, the nonparametric Spearman partial correlation was used, controlling the age factor between functional scores, QoL scores, and the performance of objective tasks. The level of statistical significance was defined as 0.05. and AG groups. The BCVA-WE was significantly higher in the CO group compared with the EG group (P = 0.0147). The Esterman score was significantly higher in the AG group than in either the CO or EG groups (P = 0.0012 and 0.003, respectively). With regard to the MD-BE and MD-WE, the AG group had the lowest values compared with both the CO and EG groups (P < 0.001 for all comparisons). Similarly, the MD-BE and MD-WE were lower in the EG group compared with the CO group (P = 0.0115 and 0.001, respectively). There was no difference in binocular contrast sensitivity between the 3 groups. Demographic and visual function data are summarized in Supplemental Table 1 (Supplemental Digital Content 1, http://links.lww.com/ IJG/A417).
For QoL, the GH score was not different between the 3 groups. The CS was significantly lower in the glaucoma groups compared with controls (P = 0.008 and 0.0015 for EG and AG, respectively). There was no difference between the EG and AG groups for the CS. The GV score was significantly lower in the AG group than in either the EG or CO groups (P = 0.0006 and 0.0188, respectively). Similarly, the GV score was lower in the EG compared with the CO group (P = 0.0135). With regard to the QoL subtype analyses, glaucoma patients showed significant differences compared with controls for NA, DA, MH, RD, and PV. Moreover, the AG group had significantly further decreased scores compared with normal subjects and the EG group for SF, Dp, and CV QoL subtypes. Results of the QoL analyses are presented in Supplemental With regard to performance in the simulated mobility task, no significant difference was found between groups for TP, but the number of MI was higher for the AG group than in the other 2 groups (P = 0.0126 and 0.0281, for CO and EG, respectively) and for the EG group compared with CO. For the reaching and grasping tasks, the MO.L was longer in the AG group compared with the CO and the EG groups (P = 0.0207 and 0.0431, for CO and EG, respectively). The OMD.S was significantly longer in the glaucoma groups compared with CO (P = 0.0246 and 0.0193, for EG and AG, respectively). For the reaching and grasping tasks, errors were not analyzed, because they concerned only 2 patients within the glaucoma groups (representing only 0.62% of all trials). Results of simulated activities of daily living are presented in Supplemental

Correlation Between QoL Scores and Performance in Simulated Activities of Daily Living
Within the mobility task, TP was only significantly correlated with DA (r = 0.39; P = 0.022). MI and the reaching and grasping task parameters were not significantly correlated with QoL scores. Within the simulated tasks, there were significant correlations between TP and MO.L (r = 0.43; P = 0.011), between MO.S and L (r = 0.66; P < 0.001), and between OMD.S and L (r = 0.88; P < 0.001). Correlation results are presented in Supplemental

DISCUSSION
Consistently with previous studies, we showed an alteration in objective visual function and performance in activities of daily living, and decreased QoL in glaucoma patients compared with a control group of normal subjects. 10,20 This result was obtained even though the patients included in our study presented with VA of at least 0.6 in each eye. Patients with AG also showed a significant decrease in QoL compared with patients with EG, particularly in terms of CV, Dp, SF, and GV. Although some studies have reported poor correlations between disease-specific QoL measures, such as the NEI VFQ-25 and VF changes, 15,21,22 our results showed that glaucoma has a direct impact on patients' QoL even at an early stage and that QoL decreases in accordance with the severity of glaucoma, as measured by VF testing.
Considering simulated activities of daily living, although the time to travel the mobility course (TP) was not different between groups, patients with AG showed poorer performance than either the EG patients or CO subjects, with more mobility incidents. These results are in accordance with previous studies showing almost no difference between controls and patients with early to moderate glaucoma, but differences appearing in more advanced stages of glaucoma. 12, 23 Turano et al, 24 in a study enrolling a broad distribution of glaucoma stages, showed no difference in mobility incidents, but a decrease in velocity in glaucoma patients compared with controls. Contrarily, Popescu et al 25 included more AG patients and observed a decrease in performance in glaucoma patients during locomotion and balance tasks compared with healthy subjects. In a previous study, we showed that the mobility task was more closely related to changes observed on binocular testing such as the Esterman VF than on monocular VF testing. 9 Early glaucoma patients often showed no change in binocular VF, allowing for effective and unchanged mobility. As glaucoma progresses, both eyes may develop severe VF loss, leading to possible differences in performance on mobility tasks. Interestingly, in the present study, there was a correlation between the Esterman binocular VF score and the patients' QoL, particularly in terms of social life (SF and Dp). This emphasizes the threshold effect on performance of severe binocular alterations compared with severe monocular alterations. Whereas the mobility task was only altered in AG patients, the reaching and grasping tasks were altered even in EG patients. The OMD.S was significantly higher in both EG and AG glaucoma patients. Similarly, Sippel et al 13 showed that glaucoma patients took a longer time to find items in a supermarket than controls, and those who found objects had more glances toward the area of the VF defect. In a study by Smith et al, 26 glaucoma patients also took significantly longer than controls for a visual search task on a computer and showed fewer saccades per second.
Some authors have found that the impaired performance on visual search tasks may be attributed to the lack of peripheral information available to patients with peripheral VF loss. 26,27 As PV provides critical wide-field information about the environment and brings targets into the fovea for high-resolution inspection, peripheral VF defects negatively impact visual search for objects 28 and central near vision. 29,30 This might explain our findings of deficient visual search tasks in both EG and AG patients.
Intuitively, as the binocular VF is determined mainly by the VF of the better eye, one might suspect that QoL in glaucoma patients would be correlated with visual function of the better eye. However, in the present study, the correlation analysis between QoL and visual function tests showed a positive correlation between most of the QoL subscores and the MD-WE. Murata et al 31 previously created a QoL prediction system to correlate VF test points with vision-related QoL. These authors also found that the WE (BCVA-WE) was the most important parameter linked to all activities of daily living related to vision (reading, walking, going out, and dining). Moreover, van Geste et al 32 found that binocular MD and MD-BE might have a similar impact on QoL, but the MD-WE had an independent effect on QoL. The relationship between QoL and visual function remains controversial between different studies. In the Early Manifest Glaucoma Trial (EMGT), although patients showed good results in terms of QoL (not different between the treated and untreated groups), QoL scores were correlated with low VA in the eye with the better MD and with the eye with the worse MD. 10 In patients with suspect or EG, Alqudah et al 7 found correlations between visual function (VA and MD) of the better eye and the QoL CS. Takahashi et al 33 found significant correlations between bilateral objective visual functions and 5 QoL subscores (GV, NA, DA, D, and PV) in glaucoma patients, implying compensation of the better eye as a determinant of QoL. As previously mentioned, the relationship between QoL and glaucoma is complex and not simply related to visual function. Most studies have found only a modest correlation between NEI VFQ-25 scores and VF status in patients with glaucoma. 10,32,[34][35][36] Moreover, similarly to objective performance, a threshold effect might exist in the relationship between VF alteration and QoL impairment, as a report from the EMGT showed that many patients with VF loss of <50% in the better eye rated their vision-related QoL at a level similar to that reported by patients with no VF loss in the better eye. 37 Meanwhile, we found no correlation between performance of simulated activities of daily living and QoL scores in this study. Both the NEI VFQ-25 and simulated activities of daily living have been shown to be valid measures of performance 5 : the latter tests the patient's ability to perform an activity, while the former interprets the patient's perception of his or her abilities. Some investigators have also evaluated the correlation between an objective evaluation of vision-specific ability to perform activities of daily living, clinical visual function tests, and QoL questionnaires. 31,38,39 Interestingly, the results of binocular visual function tests have shown higher correlations with objective performance tests than with questionnaire scores. 31,39 Our results, showing that QoL was related more to the visual function of the worse eye than the binocular visual function, might explain these results. During activities of daily living, the alteration in visual function of the worse eye is compensated for by the better eye, giving the patient more effective and less altered binocular vision. Consequently, there was no correlation between the ability to perform tasks and the QoL evaluation. These results also suggest that the Humphrey VF MD of the worse eye might serve as a good marker for QoL evaluation in glaucoma patients.
While this study has some limitations, including the small number of AG patients and the inclusion of patients with preserved VA in both eyes, our results show that EG patients display alterations in their QoL and/or their ability to perform activities of daily living. However, there is no clear correlation between alterations in QoL and ability to perform activities of daily living. These results clearly demonstrate that the impairment in QoL in glaucoma involves far more complex factors than simply altered visual function.