Vision‐related quality of life among 70‐year‐olds diagnosed with glaucoma

To evaluate the effect of glaucoma on vision‐related quality of life (VRQoL), lifestyle, physical factors and socioeconomic status in a 70‐year‐old population in Gothenburg, Sweden.

Others show that certain types of training may reduce IOP, but that the positive effects wear off with time (Hamilton-Maxwell & Feeney, 2012;Najmanová et al., 2020;Zhu et al., 2018).There are also research results that report the opposite, stating that physical training does not protect against ophthalmic diseases, with the exception of diabetic retinopathy (Wang et al., 2019).Regarding other lifestyle habits, some studies about smoking have asserted that being an active smoker constitutes a high risk for glaucoma while previous smoking is not a risk factor (Jain et al., 2017;Pérez-De-Arcelus et al., 2017).A German study has shown that high alcohol consumption is associated with glaucoma (Lamparter et al., 2018), but an American study reports the opposite, stating that alcohol intake does not influence the risk for glaucoma (Kang et al., 2007).Conversely, looking at the effect of glaucoma on lifestyle habits and quality of life (QoL), it has been reported that persons with glaucoma have poorer balance control compared with people without ocular diseases, which may lead to difficulties in mobility and lower daily physical activity (Kotecha et al., 2012;Shabana et al., 2005;Zwierko et al., 2021).Older age and low body mass index (BMI) are associated with faster rate of progression (Lin et al., 2018).
The aim of this study was to evaluate the effect of glaucoma on VRQoL, lifestyle habits, physical factors and socioeconomic status in a 70-year-old population in the Gothenburg region of Sweden.

| M AT ER I A L S A N D M ET HOD S
The population was part of the 2014-2016 H70 birth cohort study conducted at the University of Gothenburg, which examined ageing and health in 1203 70-year-olds born in 1944 and has been described in detail elsewhere (Rydberg Sterner et al., 2019).The study was performed in accordance with the tenets of the Declaration of Helsinki as well as the General Data Protection Regulation (GDPR).The Regional Ethical Review Board in Gothenburg, Sweden, gave approval to the study .All participants gave written and informed consent.
To measure VRQoL, the Swedish 2000 version of the validated National Eye Institute Visual Functioning Questionnaire −25 (NEI VFQ-25) was used (Hyman et al., 2005;Mangione et al., 1998).Participants answered the NEI VFQ-25 before an ophthalmic examination (Figure 1).This vision-targeted survey assesses the impact of visual impairment on health-related quality of life.The conventional manual for this multi-item scale contains 11 vision-targeted subscales and one general health question (Mangione et al., 2001).The mean subscale scores are used to create a composite score.A cut-off for good/poor VRQoL was set at the mean VRQoL score for the entire examined cohort.
The NEI VFQ-25 can also be analysed using Rasch methodology (Pesudovs et al., 2010;Petrillo et al., 2017).This is a new methodology for analysing categorical data which takes into account that response options may not be psychologically equally spaced in the conventional manual.Instead of presenting 12 subscales, as in the conventional manual, the results are presented in two domains.The scoring manual identifies two areas of focus of VRQoL: the Activity Limitation domain and the Socio-Emotional Functioning domain.This modern psychometric method provides, according to Petrillo et al. (2017), a more robust examination of validity and interpretability compared with traditional psychometric methods.On the other hand, the original manual shows all the areas of QoL that can be affected when living with ophthalmic diseases.The conventional format is still widely used in both ophthalmological clinics and research settings.To be able to compare research studies it would be valuable to present both conventional and modern analyses of the NEI VFQ-25 (Wan et al., 2021).
The extensive examination procedure, including semi-structured interviews, and psychological and physical tests, was performed ahead of an ophthalmic examination.Self-rating questionnaires included questions on alcohol and tobacco consumption, physical activity and socioeconomic status.The Short-Form Health Survey (SF-36) was used to measure general QoL, namely physical, mental and social health (Sullivan et al., 1995).Participants also self-reported ophthalmic diseases, and one of the questions was specifically about glaucoma: "Are you currently being, or have you previously been, treated for glaucoma?",see Figure 1.The results of the self-reported ophthalmic diseases have been reported previously (Havstam Johansson et al., 2020).
One of the physical activity tests was a balance test.It consisted of a one-leg static test, performed without shoes and with the eyes open.Three attempts were made, each with a maximum duration of 30 s, to stand on one leg without swaying or losing the balance (Heitmann et al., 1989).The best result was registered.
About half of the participants (N = 560) underwent an ophthalmic examination at Sahlgrenska University Hospital.The randomization was based on the date of birth of participants: those born on a day containing the number 0 or 5 were invited to attend the examination.We have previously that we found a total of 27 individuals with glaucoma in the population: 15 persons (2.7%) with previously undiagnosed glaucoma and 12 (2.1%) with known glaucoma as confirmed at the examination (Havstam Johansson et al., 2023).There were six participants (1.1%) who self-reported glaucoma even though they had ocular hypertension (OHT).The diagnosis of open-angle glaucoma (OAG) was based on the appearance of optic disc rim and/or retinal nerve fibre layer (RNFL) loss, accompanied by corresponding VFD.The study protocol has been described elsewhere (Havstam Johansson et al., 2020).The flowchart in Figure 2 illustrates the selfreported as well as newly diagnosed glaucoma cases.

| Statistical analysis
For conventional analysis of the NEI VFQ-25 responses, we used the NEI VFQ scoring algorithm.The scores are presented as mean and standard deviation (SD), with p-values calculated using Student's t-test.Bonferroni correction for multiple comparisons was used, with a significance level of p < 0.004.The NEI VFQ-28-R scoring manual, version 1, by Petrillo et al. (2017) was used for Rasch analysis of the NEI VFQ-25 responses.The scores are presented as mean and SD, with p-values calculated using Student's t-test.The Satterthwaite p-value was used to avoid inference errors.Comparison of VRQoL was done using the NEI VFQ-25 questionnaire for the entire cohort and subsequently for the group of participants who underwent an ophthalmic examination.
To compare differences in lifestyle-related responses between participants with and participants without selfreported glaucoma, the independent samples t-test was conducted and reported as mean and SD for continuous variables.Some parameters were dichotomised and hence treated as categorical variables, using Chi-square test for statistical analysis.
Logistic regression was performed to assess the impact of a number of factors on the likelihood that respondents would report an effect on VRQoL.The dependent variable was the low value of the NEI VFQ-25 composite score, with a cut-off at the mean score of the cohort.A score above the mean implied better VRQoL compared with a lower value.The model contained of eight independent variables: low household economy (<27 000 Swedish krona (SEK) after tax per month), self-reported glaucoma, previous smoking, low body balance (<20), low physical activity (exercise once per month or less), high alcohol consumption (≥4 times per week), low education (no university education) and high BMI (>30).For odds ratio, 95% confidence intervals (CIs) were given.
The SF-36 scores were calculated using Student's t-test and presented as mean and SD; the O'Brien permutation test was implemented to avoid statistical interference.The scoring manual for the RAND 36-Item Health Survey, version 1.0, was used to calculate item scores.The score range is from 0 to 100, with a higher score representing a more favourable health state (Ware & Sherbourne, 1992).

| R E SU LT S
The NEI VFQ-25 questionnaire was answered by 1129 participants before the ophthalmic examination, 54% of whom were female and 46% male.
Participants with self-reported glaucoma had lower VRQoL compared with participants without selfreported glaucoma.Both analyses, conventional and Rasch analysis, demonstrated that VRQoL was affected in patients with glaucoma.For the conventional method, see Table 1: participants who self-reported glaucoma had lower scores in many of the subscales, although these differences were not significant after Bonferroni correction.However, the composite score and the Mental Health subscale were significantly lower in patients with glaucoma compared with participants without glaucoma.The mean composite score of the cohort was 91.The subscale of ocular pain almost reached significance (p = 0.004).Results of the Rasch analysis of the NEI VFQ-25 scores are presented in Table 2. Significant areas were the Activity Limitation domain as well as the total score.Examples of the Activity Limitation domain include self-reported near and distance vision problems, problems with walking stairs, noticing people's reactions and difficulties in participating in social events.The Socio-Emotional Functioning domain, including the parameters of frustration, decreased control and feelings of limitations due to vision, was borderline significant.
The VRQoL of participants who underwent an ophthalmic examination is shown in Table 3.The group with previously known glaucoma, the group with glaucoma diagnosed at the examination, and participants with OHT had lower scores compared with the control group consisting of individuals without neither glaucoma nor OHT.However, none of the scores differed significantly between groups.
The mean monthly income of the study population was 27 924 SEK; €2549 (date of the exchange rate from SEK to €, 11 October 2017).There were no differences in lifestyle habits, physical factors or socioeconomic status between participants with and participants without selfreported glaucoma (N = 915); see Table 4. Body balance and BMI were not associated with glaucoma.Low household was significantly associated with poorer VRQoL 5); odds ratio (OR) 1.63 (95% confidence interval (CI) 1.14-2.33,p = 0.01).We could not find any significant associations with other included covariates.Therefore, self-reported glaucoma, smoking, body balance, physical activity, alcohol consumption, educational level and obesity did not seem to affect VRQoL.
The SF-36 questionnaire was answered by 1137 participants (Table 6).Those self-reporting glaucoma, had significantly lower general health compared with the individuals without glaucoma.Otherwise, the QoL was not affected since the glaucoma population did not differ in physical, emotional or social functioning, nor did they differ in energy level, pain or mental health index score.

| DI SC US SION
In the present study, glaucoma was found to affect VRQoL but not general QoL in 70-year-olds.Lifestyle habits, physical factors and socioeconomic status did not differ between participants with and participants without self-reported glaucoma.
Studying the whole population (N = 1129) with the NEI VFQ-25 instrument we found significantly lower VRQoL in the group of participants self-reporting glaucoma.T A B L E 3 Vision-related quality of life in participants who underwent an ophthalmic examination (N = 560): patients with glaucoma and/or ocular hypertension compared with participants without these conditions.

NEI VFQ-25 subscales
Non-glaucoma a (N = 527) Both conventional Rasch analysis showed a general in this population, which was reflected in the score in conventional analysis and in the total score of the Rasch analysis.Rasch-based scoring has been reported to improve the psychometric performance compared with the original scoring of the NEI VFQ-25 questionnaire (Petrillo et al., 2017).On the other hand, using Bonferroni correction together with the conventional method yielded almost the same results as obtained using Rasch analysis.There were differences, however: the conventional analysis method showed that mental health was affected but according to the Rasch analysis, socio-emotional functioning was not affected.Using different scoring systems gave us two different answers.Does the low mental health score indicate a fear of blindness, or is this result unreliable?The participants with glaucoma had lower VRQoL even though they had normal best corrected visual acuity (BCVA) and normal contrast sensitivity and were in the early stage of the disease (Havstam Johansson et al., 2020).
The conventional method of analysing the NEI VFQ-25 has been criticized for measuring visual function rather than QoL (Pesudovs et al., 2010).This Australian study refers to the method as having validity issues linked to multidimensionality and claims that Rasch analysis is preferable.Perhaps, the truth is somewhere in between.In an American study, the NEI VFQ-25 was shown to correspond to the SF-36 (Naik et al., 2013).Therefore, it was of interest to use the SF-36 in the present study.b <27 000 Swedish krona (SEK) after tax per month.
c "Yes" to the question: "Are you currently being, or have you previously been, treated for glaucoma?" d Tested using one-leg static standing.
e Physical activity once a month or less.
f Drinking at least 4 times per week.
g No university degree.
**Logistic regression.A p-value of <0.05 was considered statistically significant.
The participants with reported glaucoma did not describe problems in general QoL compared with the non-group.The SF-36 subscales were the same either you had glaucoma or not.One significant difference was shown; the self-reported glaucoma participants reported lower general health in the SF-36.The question, then, is, whether glaucoma gives people experience of a lower state of health and also, whether the population is more affected by other diseases, such as heart diseases, cancer, and neurological diseases in the glaucoma compared with a non-glaucoma population?
It is well known that glaucoma is not easy to recognize by the affected individual before late in the progress because VFDs are harder to interpret than low visual acuity (Yanagisawa et al., 2012).None of 27 participants with a confirmed diagnosis of glaucoma, who underwent an ophthalmic examination, had a VFI of <50% (Havstam Johansson et al., 2023).However, the Los Angeles Latino Eye Study (LALES) concluded that patients may experience measurable loss of VRQoL early in the disease (Kim & Varma, 2010).They found that people with no knowledge about glaucoma, and with manifest but previously undiagnosed glaucoma, exhibited lower VRQoL compared with people without the disease.The LALES group also found that VRQoL differed depending on the location of visual field loss (VFL), with larger impact on VRQoL for participants with central VFL.In our study, analysis of the VRQoL of the 560 participants who underwent ophthalmic examination did not show significant differences (Table 3).This may be due to the small sample size of the glaucoma population in the general population of 70-year-olds.The trends are obvious, however: VRQoL is lower in persons with both known and unknown glaucoma.Also, the participants with OHT showed a trend of lower VRQoL.The latter did not have any VFD at all, but still reported lower VRQoL.It may be that fear of glaucoma made them believe they had reduced VRQoL.A Turkish study has reported that a glaucoma diagnosis is associated with increased anxiety, stress and depression (Tastan et al., 2010).
When we analysed what had contributed to lower VRQoL we found that the socioeconomic factor of household economy had an important impact on VRQoL.We had no knowledge of whether glaucoma would lead to a lower VRQoL, but the adjusted OR indicated a trend towards lower VRQoL in individuals with self-reported glaucoma.No other socioeconomic factor, such as educational level, was significant.
In summary, we asked persons with self-reported glaucoma about their vision issues (NEI VFQ-25).They informed us they had deteriorating vision and that this affected their daily life to a greater extent compared with people without glaucoma.However, when we looked at predictors for low VRQoL, glaucoma was not the answer.Indeed, this was logical as our glaucoma population had normal vision, that is, normal BCVA and contrast sensitivity plus the disease was in an early state (visual field index >50%).The important predictor for lower VRQoL in the population was, instead, low household economy.Poverty can thus lead to an inability to buy glasses or pursue other options in life.The fact that glaucoma does not interfere with general QoL is further supported by the SF-36.People with glaucoma are not outstanding in any other categories, other than experiencing lower general health status.

F
Flowchart showing the study population of 70-yearolds born in 1944.Participants self-reported glaucoma by answering "Yes" to the question: "Are you currently being, or have you previously been, treated for glaucoma?"NEI VFQ-25 = national eye institute visual function questionnaire.

F
I G U R E 2 Flowchart showing the study population of 70-yearolds born in 1944 within the ophthalmic examination.Participants self-reported glaucoma by answering "Yes" to the question: "Are you currently being, or have you previously been, treated for glaucoma?"Six participants (1.1%) self-reported glaucoma even though they had ocular hypertension (OHT).The diagnosis of open-angle glaucoma (OAG) was based on appearance of optic disc rim and/or retinal nerve fibre layer (RNFL) loss, accompanied by corresponding VFD.
Vision-related quality of life in participants with and without self-reported glaucoma a (N = 1129).
Differences in lifestyle, physical and social factors between persons with and persons without self-reported glaucoma a (N = 915).
Vision-related quality of life in subjects with and without self-reported glaucoma a (N = 1129).
a "Yes" to the question: "Are you currently being, or have you previously been, treated for glaucoma?"**Independentt-test.A p-value of <0.004 was considered significant after Bonferroni correction for multiple comparisons.T A B L E 2Abbreviations: NEI VFQ, National Eye Institute Visual Functioning Questionnaire (Rasch-analysis); SD, standard deviation.a Yes to the question: "Are you currently being, or have you previously been, treated for glaucoma?";**Independent t-test (Satterthwaite).A p-value of <0.05 was considered significant.
Previous smoking as opposed to never smoker, no person stated current smoking.Physical activity at least once a week.One-leg static standing (continuous variable).Independent samples t-test; BMI (continuous variable).Education, no university degree; Household economy (continuous variable).*A p-value of <0.05 was considered statistically significant.Multivariate analysis of the impact of variables on low vision-related quality of life a (N = 919).Composite score of the National Eye Institute's Visual Functioning Questionnaire, with cut-off at the mean score of the cohort.
d e f g h *a