Risk factors for poorer quality of life in patients with neovascular age-related macular degeneration: a longitudinal clinic-based study

Abstract

Background/Objectives

To examine the risk factors for poor vision-related and health-related quality of life (QoL) in patients with neovascular age-related macular degeneration (nAMD) who present for anti-vascular endothelial growth factor (anti-VEGF) therapy.

Methods

In a clinic-based cohort of 547 nAMD patients who presented for treatment, the National Eye Institute Visual Function Questionnaire-25 (NEI-VFQ25), Short-Form 36 (SF-36) and EuroQoL EQ-5D-5L questionnaires were administered to assess vision-related and health-related QoL. Of these, 83 participants were followed up one-year later to provide longitudinal data.

Results

Individuals with mild or moderate visual impairment or blindness at baseline had significantly lower NEI-VFQ-25 scores at follow-up. The presence of ≥3 chronic diseases was associated with lower SF-36 mental component scores (MCS) (p = 0.04) and EQ-VAS scores (p = 0.05). Depressive symptoms were associated with significantly lower MCS (p < 0.0001) and EQ-VAS scores (p = 0.02). Individuals with versus without impaired basic activities of daily living (ADLs) exhibited NEI-VFQ-25 and EQ-VAS scores that were 10.96 (p = 0.03) and 0.13 (p = 0.02) points lower. Those with impaired instrumental ADLs scored 11.62 (p = 0.02), 13.13 (p < 0.0001) and 15.8 (p = 0.0012) points lower in the NEI-VFQ-25, SF-36 physical component score and EQ-5D-5L summary score, respectively.

Conclusions

The QoL of nAMD patients is affected by visual acuity as well as patients’ medical history, mental health and functional status.

Introduction

Age-related macular degeneration (AMD) is the leading cause of visual impairment among older adults in developed countries [1]. The management of one of its subtypes, neovascular AMD (nAMD), has been revolutionised by the advent of anti-vascular endothelial growth factor (anti-VEGF) therapy, which limits disease progression. In Australia, aflibercept is the most costly drug under the Pharmaceutical Benefits Scheme (PBS) whilst ranibizumab is ranked eighth [2]. Given its heavy taxpayer burden plus the high prevalence of nAMD, research regarding the risk factors for poorer quality of life (QoL) among those undergoing anti-VEGF therapy is necessary. This will enable the development of evidence-based interventions that improve QoL whilst also maximising the value derived from government expenditure on anti-VEGF therapy.

Both AMD and nAMD specifically are associated with poorer QoL [3, 4]. Previous studies [3, 4] have assessed both health-related QoL (HRQoL) via instruments including the Short-Form 36 (SF-36) and EuroQoL (EQ-5D), and vision-related QoL (VRQoL) using the National Eye Institute Visual Function Questionnaire-25 (NEI-VFQ-25). Proxy measures such as anxiety and depression have also been analysed [5]. The impact of disease characteristics such as the development of bilateral nAMD [6] and visual acuity on QoL [7, 8] are well documented. Chatziralli et al. [3] study identified female gender, alcohol consumption and comorbidities including cardiovascular disease, diabetes and hypertension as risk factors for poorer QoL among Greek wet and dry AMD patients. Less data is available regarding the relationship between impaired activities of daily living (ADLs) and QoL. Matamoros et al. [9]. used the ‘purchase of visual aid equipment’ as a proxy measure for analysing ADL impairment in patients with bilateral nAMD. Meanwhile, Williams et al[10]. utilised validated ADL scales however this study predates anti-VEGF therapies.

We offer a comprehensive cross-sectional and temporal analysis of a large, clinic-based cohort of nAMD patients presenting for anti-VEGF therapy. Whilst previous studies have predominantly focused upon HRQoL or VRQoL, we will examine both outcome measures. Our analysis also extends beyond merely analysing the effect of disease characteristics. That is, we utilise a biopsychosocial approach that also considers how patients’ other comorbidities, functional status and mental health may impact their QoL. This approach analyses nAMD patients as whole individuals rather than simply a pair of eyes affected by a disease. Thus, our study aims to determine the risk factors associated with poor vision-related and health-related QoL in nAMD patients presenting for anti-VEGF therapy.

Materials and methods

Study population

The present study utilises data from a clinic-based cohort of AMD patients that presented for anti-VEGF therapy at ophthalmology clinics at either Westmead Hospital or Sydney West Retina, both based in Sydney, Australia. Data was collected from 621 participants between 2012 and 2015. Of these, 547 were diagnosed with nAMD by an ophthalmologist (PM). Due to resourcing issues, only 83 participants were followed up one year later to provide longitudinal data. It is noteworthy that this study examines a sample of patients presenting for anti-VEGF therapy rather than those commencing therapy. Thus, it includes both treatment-naïve patients and those that had already started anti-VEGF therapy prior to study recruitment. All participants provided informed consent. The study’s methods have been described elsewhere [11]. The Western Sydney Local Health District Human Research Ethics Committee approved this study. This study adheres to the tenets of the Declaration of Helsinki.

Assessment of AMD

Eye examinations were completed on all participants. This included bilateral 30^o or 40^o stereoscopic colour retinal photographs. Cirrus spectral-domain optical coherence tomography (OCT) was used to analyse macular lesions. The inclusion criteria was the presence of nAMD in at least one eye, as shown by retinal or subretinal haemorrhage, pigment epithelial or neurosensory detachment, subretinal fibrosis or old atrophic disciform or photocoagulation scars. The clinical features of nAMD that were analysed were: the presence of early or late AMD in the better and worse eye, the presence of unilateral or bilateral disease, the presence of subretinal fluid, intra-retinal fluid or pigment epithelial detachment in the better and worse eye and central macular thickness. The number of anti-VEGF injections received by each patient was also recorded and patients were divided into tertiles according to the total number of injections received at baseline and follow-up.

Assessment of QoL

Informed by the existing literature, participants’ QoL was measured via three different survey instruments – National Eye Institute Visual Function Questionnaire-25 (NEI-VFQ-25), Short Form-36 (SF-36) and EuroQoL EQ-5D-5L.

The NEI-VFQ-25 ref. [12] addresses VRQoL, or visual function, in the following areas: general vision, difficulty with near-vision and distance-vision activities, role limitations, limitations in social functioning, mental health, driving difficulties dependency on others, colour vision, ocular pain and general health [12]. Responses are coded to produce an overall score with higher scores indicating better vision-related QoL.

The SF-36 ref. [13] examines HRQoL across eight domains – limitations in physical activities, limitations in social activities, limitations in usual role activities due to physical problems and emotional problems, bodily pain, mental health, vitality and general health perceptions. Responses are coded to produce separate physical component scores (PCS) and mental component scores (MCS). Higher scores indicate better health states. The EuroQol 5D scale [14] (EQ-5D-5L) assesses HRQoL in five domains – mobility, self-care, usual activities, pain/discomfort and anxiety/depression. Responses in each category are scored between 1–3 and tallied to produce composite physical and mental scores. Participants also rate their health between 0 (worst imaginable health state) to 100 (best imaginable health state) on a visual analogue scale (EQ-VAS). In the absence of an Australian value set, the United Kingdom value set was applied [15].

Assessment of other covariates

Trained interviewers conducted face-to-face interviews with participants. Information was obtained regarding participants’ sociodemographic and functional characteristics, and quality of life. Patients’ clinical histories were obtained from their medical records. Activities of daily living (ADLs) were assessed using the Older Americans Resources and Services (OARS) questionnaire [16]. This questionnaire assesses both basic ADLs (BADLs) and instrumental ADLs (IADLS). BADLs refer to activities such as eating, dressing and showering whilst IADLs include cooking, cleaning and dressing. The presence of depressive symptoms was measured via the Centre for Epidemiological Studies Depression 10-point scale [17] (CESD-10) and the Mental Health Index (MHI), which is a component of the SF-36 ref. [13]. Finally, self-rated health (SRH) was assessed via the question: “would you say your overall health is: excellent, very good, good, fair or poor”. Responses in the two latter categories were classified as ‘poor SRH’.

As explained above, a number of different questionnaires were administered during this study. Questionnaires were provided to participants during their baseline and follow-up clinic visits. Participants were encouraged to complete the questionnaires during the clinic appointment. However, given the high number of questionnaires in the study, participants were also provided with the option to complete the questionnaires at home and return the results via mail.

Statistical analysis

Statistical analyses were completed using SAS software 9.4 (SAS Institute, North Carolina, USA). Baseline and longitudinal data were initially assessed using t-tests and chi-square tests. Associations between variables were then analysed via general linear models (GLM). Analyses of covariance (ANCOVA) were utilised for the multivariable models. This approach was justified as the outcomes were continuous variables. With regards to the assumptions of the tests, there was some negative skew and imbalances of variances in the data. However, as Barton et al[18]. explain, ANCOVA is considered robust to some deviations from normality and imbalance of variances. In this approach, the assumption of normality is most crucial when the sample size is small whilst the assumption of variance is important when the size of the cells is small (<30). Our study involves a large sample size and a high number of observations in the cells. Thus, our use of general linear models is justified. Models were initially adjusted for age and sex. Based on both our findings and the existing literature, the multivariable models were adjusted for age, sex, visual acuity, number of anti-VEGF injections received, hospitalisation within the past 12 months, the presence of three or more chronic diseases, the presence of depressive symptoms on the CESD or MHI scales and basic and instrumental ADL impairment.

Results

Data was collected from 547 individuals, who were included in the final analysis. These participants provided information regarding their sociodemographic, functional and clinical characteristics whilst also completing surveys regarding their quality of life. Participants with 1-year follow-up data (n = 83) were less likely to be current smokers (p = 0.04) but more likely to self-report poor health (p = 0.01) than those that were not followed up.

Cross-sectional data analysis

Worsening visual acuity (VA) is associated with progressively poorer NEI-VFQ-25 scores (Table 1). Those that had received a lower number of injections exhibited the poorest SF-36 scores (p = 0.0012). Hospitalisation within the past 12 months was associated with PCS and EQ-5D-5L scores that were 2.18 points (p < 0.0001) and 4.00 points (p = 0.04) lower than those that had not been hospitalised. Individuals with three or more chronic diseases scored 4.78 (p < 0.0001), 0.04 (p = 0.02) and 4.96 points (p = 0.01) lower in the PCS, EQ-VAS and EQ-5D-5L summary score respectively. Presence of depressive symptoms as measured by the CESD-10 was associated with poorer VRQoL and HRQoL. Using the MHI, the presence of depressive symptoms was associated with EQ-VAS and EQ-5D-5L summary scores that were 0.09 (p = 0.0004) and 5.92 points (p = 0.01) lower respectively than those without this risk factor.

Table 1.

Multivariable-adjusted cross-sectional associations with vision-related and health-related quality of life scores for individuals with nAMD.

	NEI-VFQ-25 composite score	p-value	SF-36 physical component score	p-value	SF-36 mental component score	p-value	EQ-5D-5L visual analogue scale (VAS)	p-value	EQ-5D-5L summary score	p-value
Age (each 1-yr increase)	−0.15 ± 0.09	0.07	−0.05 ± 0.06	0.40	−0.01 ± 0.06	0.82	−0.00 ± 0.00	0.58	0.05 ± 0.11	0.67
Gender		0.36		0.37		0.13		0.55		0.34
Female	55.82 ± 1.16		37.79 ± 0.77		47.47 ± 0.80		0.63 ± 0.02		66.88 ± 1.42
Male	54.51 ± 1.27		38.64 ± 0.84		45.90 ± 0.89		0.62 ± 0.02		65.21 ± 1.58
Visual acuity (VA)a
Normal	69.29 ± 1.20	Ref.	38.88 ± 0.80	Ref.	44.91 ± 0.83	Ref.	0.64 ± 0.02	Ref.	66.99 ± 1.46	Ref.
Mild VI	57.65 ± 1.51	<0.0001	37.03 ± 1.01	0.12	46.63 ± 1.07	0.18	0.64 ± 0.02	0.89	66.68 ± 1.89	0.89
Moderate VI	48.78 ± 1.78	<0.0001	38.08 ± 1.19	0.56	47.07 ± 1.27	0.14	0.62 ± 0.02	0.41	64.35 ± 2.21	0.30
Blind	44.92 ± 2.11	<0.0001	38.88 ± 1.39	1.00	48.11 ± 1.48	0.05	0.61 ± 0.03	0.26	66.17 ± 2.61	0.78
Number of injectionsb
Nil (0) [n = 123/547]	54.45 ± 1.59	Ref.	40.70 ± 1.06	Ref.	46.48 ± 1.12	Ref.	0.62 ± 0.02	Ref.	66.44 ± 1.94	Ref.
Tertile 1 (1–5) [n = 141/547]	56.49 ± 1.58	0.31	36.34 ± 1.05	0.0012	46.16 ± 1.12	0.82	0.62 ± 0.02	0.97	65.91 ± 1.96	0.83
Tertile 2 (6–22) [n = 142/547]	55.83 ± 1.48	0.48	38.02 ± 0.98	0.04	47.82 ± 1.03	0.34	0.64 ± 0.02	0.42	66.70 ± 1.84	0.92
Tertile 3 (23–79) [n = 141/547]	53.87 ± 1.58	0.77	37.81 ± 1.04	0.03	46.28 ± 1.09	0.89	0.62 ± 0.02	0.94	65.14 ± 1.93	0.60
Hospitalisation within the past 12 months		0.46		0.04		0.12		0.11		0.04
Yes	54.59 ± 1.38		37.13 ± 0.91		45.81 ± 0.97		0.61 ± 0.02		64.05 ± 1.71
No	55.73 ± 1.11		39.31 ± 0.75		47.55 ± 0.76		0.65 ± 0.01		68.05 ± 1.38
Presence of ≥3 chronic diseases		0.12		<0.0001		0.22		0.02		0.01
Yes	54.05 ± 1.25		35.83 ± 0.83		46.06 ± 0.88		0.61 ± 0.02		63.57 ± 1.55
No	56.27 ± 1.18		40.61 ± 0.78		47.31 ± 0.80		0.65 ± 0.02		68.53 ± 1.46
Presence of depressive symptoms (CESD-10)		<0.0001		0.05		<0.0001		<0.0001		<0.0001
Yes	49.86 ± 1.29		37.03 ± 0.86		38.92 ± 0.92		0.57 ± 0.02		60.88 ± 1.63
No	60.46 ± 1.36		39.40 ± 0.92		54.44 ± 0.80		0.69 ± 0.02		71.22 ± 1.68
Presence of depressive symptoms (MHI)		0.56		0.09		N/Ac		0.0004		0.01
Yes	54.61 ± 1.54		39.31 ± 1.04				0.58 ± 0.02		63.09 ± 1.91
No	55.71 ± 1.16		37.12 ± 0.77				0.67 ±0.02		69.01 ± 1.44
Impaired BADLs		0.16		<0.0001		0.09		<0.0001		0.03
Yes	54.00 ± 1.57		36.00 ± 0.91		45.68 ± 0.96		0.55 ± 0.02		63.79 ± 1.69
No	56.32 ± 1.18		40.44 ± 0.79		47.68 ± 0.82		0.71 ± 0.02		68.31 ± 1.47
Impaired IADLs		<0.0001		<0.0001		0.15		<0.0001		<0.0001
Yes	49.77 ± 0.93		33.66 ± 0.62		45.81 ± 0.65		0.57 ± 0.01		60.68 ± 1.15
No	60.56 ± 1.59		42.78 ± 1.07		47.56 ± 1.11		0.69 ± 0.02		71.42 ± 1.99

Model adjusted for age, sex, number of injections, hospitalisation within the past 12 months, presence of ≥3 chronic diseases, presence of depressive symptoms as per the CESD or MHI scales, impaired BADLs, impaired IADLs and visual acuity.

All figures presented in age category are estimate values/regression co-efficients (±standard error) as age is a continuous variable. The figures presented for all other variables in the table are least square means (±standard error) as these are categorical variables.

NEI-VFQ-25 National Eye Institute Visual Function Questionnaire-25, SF-36 Short-Form 36, EQ-5D EuroQoL, CESD-10 Centre for Epidemiological Studies Depression-10 scale, MHI Mental Health Index, BADLs basic activities of daily living, IADLs instrumental activities of daily living, VA visual acuity, VI visual impairment.

Bold values represent significant findings (p < 0.05).

^aVisual acuity ranges: normal = VA > 20/40, mild VI = ≤20/40 to >20/80; moderate VI = ≤20/80 to >20/200; legally blind = ≤20/200.

^bNumber of injections listed as a range in the round brackets for each tertile.

^cNot calculated as the MHI is a component of the SF-36.

Both BADL and IADL impairment were associated with poorer QoL. BADL impairment was associated with PCS, EQ-VAS and EQ-5D-5L summary scores that were 4.44 (p < 0.0001), 0.16 (p < 0.0001) and 4.52 points (p = 0.03) lower. Individuals with IADL impairment demonstrated significantly lower SF-36 PCS, EQ-VAS and EQ-5D-5L scores (Table 1). In addition, IADL impairment was also associated with NEI-VFQ-25 scores that were 10.79 points (p < 0.0001) lower.

Longitudinal data analysis

Table 2 shows the multivariable model outlining the relationship between baseline risk factors and QoL scores at 1-year follow up. Worsening visual acuity was associated with poorer visual function 12 months later. Interestingly, mild VI was associated with MCS that were 6.39 points higher (p = 0.03). With regards to anti-VEGF therapy, individuals that had received up to five injections at baseline exhibited EQ-5D-5L summary scores that were 17.07 points higher (p = 0.03) than those that had received no injections.

Table 2.

Multivariable-adjusted longitudinal (12-month) associations with vision-related and health-related quality of life scores for individuals with nAMD.

	NEI-VFQ-25 composite score	p-value	SF-36 physical component score	p-value	SF-36 mental component score	p-value	EQ-5D-5L visual analogue scale (VAS)	p-value	EQ-5D-5L summary score	p-value
Age (each 1-yr increase)	0.29 ± 0.30	0.34	−0.06 ± 0.18	0.76	−0.09 ± 0.18	0.64	−0.01 ± 0.00	0.20	−0.07 ± 0.28	0.79
Gender		0.49		0.50		0.49		0.61		0.67
Female	52.60 ± 3.75		38.96 ± 2.38		48.35 ± 2.27		0.58 ± 0.05		66.52 ± 3.60
Male	49.84 ± 3.71		40.59 ± 2.31		46.70 ± 2.20		0.61 ± 0.05		64.89 ± 3.53
Visual acuity (VA)a
Normal	67.67 ± 3.12	Ref.	38.23 ± 1.92	Ref.	43.48 ± 1.89	Ref.	0.56 ± 0.04	Ref.	62.15 ± 2.96	Ref.
Mild VI	50.77 ± 4.13	0.0010	39.78 ± 2.55	0.61	49.87 ± 2.49	0.03	0.61 ± 0.05	0.46	60.31 ± 3.96	0.70
Moderate VI	47.33 ± 5.69	0.0019	42.55 ± 3.58	0.27	48.43 ± 3.32	0.17	0.55 ± 0.07	0.94	63.11 ± 5.43	0.87
Blind	39.12 ± 7.15	0.0005	38.54 ± 4.56	0.95	48.32 ± 4.33	0.30	0.66 ± 0.09	0.29	77.26 ± 7.02	0.05
Number of injectionsb
Nil (0) [n = 9/83]	45.87 ± 6.49	Ref.	36.69 ± 3.92	Ref.	47.78 ± 3.84	Ref.	0.49 ± 0.08	Ref.	57.65 ± 6.16	Ref.
Tertile 1 (1–5) [n = 11/83]	53.85 ± 5.68	0.31	42.08 ± 3.55	0.27	50.16 ± 3.52	0.62	0.63 ± 0.07	0.17	74.72 ± 5.39	0.03
Tertile 2 (6–22) [n = 27/83]	53.18 ± 3.49	0.28	40.10 ± 2.22	0.39	45.07 ± 2.07	0.49	0.66 ± 0.04	0.04	60.86 ± 3.35	0.61
Tertile 3 (23–79) [n = 36/83]	51.98 ± 3.95	0.37	40.22 ± 2.53	0.39	47.10 ± 2.27	0.87	0.61 ± 0.05	0.16	69.59 ± 3.77	0.07
Hospitalisation within the past 12 months		0.31		0.98		0.68		0.90		0.72
Yes	53.55 ± 4.35		39.80 ±2.64		46.97 ± 2.58		0.59 ± 0.05		64.92 ± 4.15
No	48.89 ± 3.39		39.74 ± 2.22		48.08 ± 2.05		0.60 ± 0.04		66.49 ± 3.23
Presence of ≥ 3 chronic diseases		0.49		0.23		0.04		0.05		0.45
Yes	49.78 ± 3.79		38.25 ± 2.40		44.91± 2.29		0.54 ± 0.05		64.22 ± 3.62
No	52.66 ± 3.72		41.30 ± 2.34		50.14 ± 2.22		0.65 ± 0.05		67.19 ± 3.57
Presence of depressive symptoms (CESD-10)		0.42		0.86		<0.0001		0.25		0.07
Yes	48.75 ± 4.24		40.11 ± 2.61		41.75 ± 2.47		0.55 ± 0.05		60.19 ± 4.14
No	53.69 ± 4.54		39.44 ± 2.82		53.31 ± 2.16		0.64 ± 0.06		71.23 ± 4.34
Presence of depressive symptoms (MHI)		0.75		0.86		N/Ac		0.02		0.47
Yes	50.18 ± 5.01		40.12 ± 3.22				0.50 ± 0.06		63.41 ± 4.78
No	52.26 ± 3.96		39.43 ±2.36				0.69 ± 0.05		68.00 ± 3.88
Impaired BADLs		0.03		0.22		0.78		0.02		0.07
Yes	45.74 ± 4.40		37.95 ± 2.77		47.12 ± 2.69		0.53 ± 0.05		61.39 ± 4.19
No	56.70 ± 3.52		41.60 ± 2.19		47.93 ± 2.02		0.66 ± 0.04		70.03 ± 3.39
Impaired IADLs		0.02		<0.0001		0.23		0.28		0.0012
Yes	45.41 ± 3.13		33.21 ± 1.94		45.85 ± 1.91		0.56 ± 0.04		57.81 ± 3.01
No	57.03 ± 4.68		46.34 ± 2.95		49.22 ± 2.71		0.63 ± 0.06		73.61 ± 4.47

Model adjusted for age, sex, number of injections, hospitalisation within the past 12 months, presence of ≥3 chronic diseases, presence of depressive symptoms as per the CESD or MHI scales, impaired BADLs, impaired IADLs and visual acuity.

All figures presented in age category are estimate values/regression co-efficients (± standard error) as age is a continuous variable. The figures presented for all other variables in the table are least square means (± standard error) as these are categorical variables.

NEI-VFQ-25 National Eye Institute Visual Function Questionnaire-25, SF-36 Short-Form 36, EQ-5D EuroQoL, CESD-10 Centre for Epidemiological Studies Depression-10 scale, MHI Mental Health Index, BADLs basic activities of daily living, IADLs instrumental activities of daily living, VA visual acuity, VI visual impairment.

Bold values represent significant findings (p < 0.05).

^aVisual acuity ranges: normal = VA > 20/40, mild VI = ≤20/40 to >20/80; moderate VI = ≤20/80 to >20/200; legally blind = ≤20/200.

^bNumber of injections listed as a range in the round brackets for each tertile.

^cNot calculated as the MHI is a component of the SF-36.

Those with ≥3 chronic diseases had mean MCS and EQ-VAS scores that were 5.23 (p = 0.04) and 0.11 points (p = 0.05) lower than those without. Those with depressive symptoms at baseline as per the CESD-10 had mean MCS that were 11.56 points (p < 0.0001) lower than those without. Using the MHI, those with depressive symptoms at baseline exhibited lower EQ-VAS scores (p = 0.02).

Individuals with versus without BADL impairment at baseline displayed lower NEI-VFQ-25 (p = 0.03) and EQ-VAS scores (p = 0.02) at follow-up (Table 2). Individuals with IADL impairment at baseline had mean NEI-VFQ-25 scores at follow-up that were 11.62 points lower (p = 0.02) than those without. On the SF-36 PCS and EQ-5D-5L summary scores, those with versus without baseline IADL impairment scored 13.13 (p < 0.0001) and 15.80 points (p = 0.0012) lower, respectively.

Discussion

This clinic-based study provides novel epidemiological data showing that visual acuity, physical and mental health, and functional status are all factors that are significantly and independently associated with the longer-term vision-related and health-related QoL of nAMD patients presenting for anti-VEGF therapy.

Our findings show that VA independently influences the VRQoL of nAMD patients after 12 months. The longitudinal data suggests a causal relationship between VA and visual function; a finding that has extensively been described in the literature [19, 20]. Interestingly, mild VI was associated with a higher MCS; a finding which is difficult to explain. Future studies involving a higher number of participants at follow-up may better clarify this finding.

Our longitudinal data shows no association between the number of anti-VEGF injections received and VRQoL or SF-36 scores. Finger et al[19]. reported that anti-VEGF therapy has a positive impact on VRQoL. However, research regarding the effect of the number of injections received, on QoL is scarce [21]. Senra et al. [5] cross-sectional study reported that clinical levels of depression were highest in the earlier stages of treatment, particularly among individuals that had received up to three anti-VEGF injections. By contrast, we found no association between the number of injections received and NEI-VFQ-25 or SF-36 MCS scores. Most notably, the association between number of injections and SF-36 scores observed at baseline was not sustained at follow-up. In our study, 123 of 547 of patients at baseline were treatment naïve. At follow-up, 9 of the 83 patients were previously treatment naïve one year prior and had subsequently commenced anti-VEGF therapy. Thus, our findings must be interpreted with caution due to the low numbers at follow-up. Future studies may wish to use a larger longitudinal study to revisit this relationship between the number of injections received and QoL.

The present study also found significant associations between broader health characteristics and HRQoL. The presence of ≥3 chronic diseases at baseline correlated with poorer HRQoL at one-year follow-up. Chatziralli et al. [3] cross-sectional study of wet and dry AMD patients found that comorbidities including hypertension, diabetes and cardiovascular disease were each associated with lower MCS, PCS and EQ-5D scores. Our longitudinal study builds upon this finding by demonstrating that the presence of multiple comorbidities results in lower SF-36 MCS and EQ-5D-5L VAS scores in nAMD patients. Patients with multiple chronic diseases may perceive themselves as a burden upon their families whilst the management of their conditions may also cause financial distress [22]. These factors may result in poorer mental health thereby explaining the findings with the SF-36 MCS. Furthermore, patients may be burdened by the symptoms and treatment side effects of their multiple illnesses [22] thereby explaining their poorer HRQoL, as illustrated by the EQ-5D-5L VAS scores. Where nAMD affects older adults, a patient group that is more at risk of having multiple illnesses, it is necessary to consider how the overall health of nAMD patients may be optimised or more effectively managed. This is a task that will likely require ophthalmologists to collaborate with other healthcare providers including primary care physicians who may be best placed to manage patients’ multiple comorbidities and refer to other specialists as required.

Our study also explored the relationship between depressive symptoms and QoL. Chatzirilli et al. [3] cross-sectional study identified diagnoses of anxiety or depression to be associated with poorer QoL. We have built upon this research to show that the presence of depressive symptoms at baseline is associated with poorer HRQoL at one-year follow up. This illustrates how psychological wellbeing impacts the overall HRQoL of nAMD patients, as shown by the EQ-5D-5L VAS scores. Furthermore, depressive symptoms may cause actual or perceived limitations in function [23] such that an individual may become more greatly impacted by their nAMD. Given the high prevalence of depressive symptoms among nAMD patients [24], further research into this area is necessary to enable the implementation of evidence-based interventions, such as cognitive behavioural therapy, to support the mental health of nAMD patients and improve their HRQoL.

We also report a relationship between ADL impairment and poorer QoL. Matamoros et al[9]. found that the purchase of visual aid equipment and having ‘specific facilities at home’ were associated with lower NEI-VFQ scores. Our research builds upon this study through its longitudinal design and the use of a validated scale to assess ADL impairment. We are able to report that the presence of BADL and IADL impairment at baseline is independently associated with poorer VRQoL and HRQoL at follow-up as indicated by the NEI-VFQ-25 and EQ-5D-5L scores, respectively. IADL impairment also impacts physical wellbeing as measured by the SF-36 PCS. Interestingly, neither BADL nor IADL impairment correlate with lower MCS. Thus, it is possible that the presence of ADL impairment adversely affects the QoL of nAMD patients through its impact on patients’ physical wellbeing and functioning rather than their mental health. This is plausible as functional impairments limit patients’ abilities to adequately care for themselves, which may in turn affect their QoL. Gopinath et al[25]. reported that the presence of any AMD increases the risk of impaired IADLs and total ADLs. Thus, given our findings regarding the association between ADL impairment and poorer QoL, there is a need for further research into rehabilitative interventions and/or support services that may reduce the impact of ADL impairment among nAMD patients and thus improve their QoL.

It is noteworthy that in our initial age and sex-adjusted models, few clinical features of nAMD other than visual acuity were associated with VRQoL at baseline or follow up whilst almost none were associated with HRQoL. Features such as early versus late AMD, fluid type, central macular thickness and bilateral disease were not significant in these models. Thus, these clinical features were not included in the final, parsimonious multivariable model. These findings are similar to the results of Chatziralli et al. [3] study. This suggests that whilst ophthalmologists may predominantly use clinical signs as indicators of disease severity, these may constitute poor indicators of patients’ VRQoL and HRQoL and thus their overall experiences with nAMD. As such, a more holistic or biopsychosocial approach that considers both VA and patients’ broader physical and mental health indices may be necessary to optimise the QoL of nAMD patients.

Finally, we found that the results of the SF-36 and EQ-5D-5L scales did not always correlate. Rather, variables that were associated with SF-36 scores were not necessarily associated with EQ-5D-5L scores and vice versa. Previous studies have predominantly used either SF-36 or EQ-5D-5L. We identified only one other study [3], which utilised both scales. This study also reported mismatches in findings with SF-36 and EQ-5D-5L scores. Where these scales utilise different domains to assess QoL, the difference in SF-36 and EQ-5D-5L scores is unsurprising.

A major strength of our study is that it collected a broad range of health and clinical factors that may impact on patients’ QoL. Whilst most existing studies focused upon the clinical characteristics of nAMD, we were able to assess whether other variables including patients’ physical and mental health, and functional status influenced their QoL. We also examined both VRQoL and HRQoL to capture a broader snapshot of patient wellbeing. Our analysis is also strengthened by the use of validated questionnaires to examine both QoL and independent variables such as ADL impairment and the presence of depressive symptoms. Furthermore, the majority of existing studies were cross-sectional in nature. Through a longitudinal study design, we were able to report novel relationships between baseline variables and outcomes at 12-month follow-up.

A key weakness of our study is the low number of participants at one-year follow-up, which decreases the statistical power of our study. Future studies involving a larger number of participants at follow-up may be able to confirm or challenge the findings that we have reported to determine whether our results may be generalised to all nAMD patients presenting for treatment. A further limitation of our study is the high number of lengthy surveys administered. This may result in survey fatigue that reduces the reliability of our results. In our study, information regarding the total duration of anti-VEGF therapy received by participants was not collected. Future studies may wish to include such data to assess whether the QoL of patients is impacted by lengthy treatment regimes. Finally, given the older demographic of nAMD patients, our study could also be improved by the use of a validated scale such as the Mini-Mental State Examination (MMSE) to screen for dementia, as cognitive impairment could have influenced observed associations with QoL.

Overall, we found that the 12-month vision-related and health-related QoL of nAMD patients is affected by not only visual acuity but also the broader physical and mental health indices and patients’ functional status. Future research may wish to examine a larger sample of both wet and dry AMD patients to determine whether the factors identified in our study are relevant to all AMD patients. Overall, our findings are likely to inform clinical practice by emphasising the need to consider the whole person, rather than simply treating a patient’s macular morphology. Thus, greater collaboration between ophthalmologists and other medical and allied healthcare professionals is necessary to improve both the visual function and health-related QoL of AMD patients.

Summary

What was known before

• That vision-related quality of life in patients with neovascular age-related macular degeneration (nAMD) is affected by visual acuity.

• That the prevalence of depressive symptoms among nAMD patients is high.

What this study adds

• A more comprehensive analysis of the risk factors for not only poor vision-related quality of life but also health-related quality of life.

• A biopsychosocial approach that analyses the impact of not only clinical features of disease but also other aspects of physical and mental health on the quality of life of patients.

Author contributions

Research design: KVV, PM, BG; Research execution: HD, KVV, PM; Data interpretation and analysis: KVV, BG, GB; Manuscript preparation: KVV, HD, GL, GB, PM, BG.

Funding

This study received funding from the Macular Disease Foundation Australia (MDFA).

Data availability

Data are available upon reasonable request. Deidentified participant data will be available upon request made to the corresponding author.

Competing interests

The authors declare no competing interests.