The population characteristics and prevalence of visual impairment in a Southern European population

Abstract

Purpose

To describe the population characteristics, prevalence, and causes of visual impairment in Maltese adults aged 50–80 years.

Methods

The Malta Eye Study invited 4,006 random adults aged 50–80 for eye exams and interviews at the Malta and Gozo state hospitals (September 2021–July 2024). Tests included visual acuity, refraction, anthropometry, clinical measurements, tonometry, slit-lamp and dilated fundus exams, retinal imaging, and cognitive screening. Interviews covered sociodemographic, medical, ocular, and medication histories.

Results

A representative sample of 1,794 individuals (44.8% turnout) were assessed. Visual impairment in either eye was found in 23.9% (95%CI 21.9%–25.9%), and bilateral impairment in 6.0% (95%CI 4.9%–7.2%). The most common causes in either eye were uncorrected/undercorrected refractive error (12.3%), amblyopia (5.0%), cataract (3.8%), pathological myopia (1.3%), diabetic retinopathy (0.8%), age-related macular degeneration (0.6%), and glaucoma (0.4%). Predictors of visual impairment included older age, lower education, diabetes requiring insulin and tablets, and dementia. Protective factors were sunglasses use and angiotensin receptor blocker therapy.

Conclusions

Although overall visual impairment rates are favourable, this study highlights the need to strengthen screening and treatment services for avoidable causes, particularly by improving public optometry access, among older adults, socioeconomically vulnerable groups, and those with poorly controlled diabetes.

Introduction

In 2020, global age-standardised prevalence of visual impairment (VI) in those aged ≥50 years was 1.85% for blindness, 11.18% for moderate-to-severe VI (MSVI), 7.73% for mild VI, and 22.33% for uncorrected presbyopia. ¹ Eastern Europe had lower blindness (0.88%) but similar MSVI (11.5%), mild VI (4.56%), and higher presbyopia (30.6%), ¹ with country-specific MSVI rates reported as 15.2% in Moldova ² and 5.6% in Hungary ³ . Central Europe showed lower blindness (0.58%), MSVI (6.78%) and mild VI (2.75%) with presbyopia close to global levels (23.5%). ¹ Western Europe reported low blindness (0.60%), MSVI (5.15%), mild VI (3.63%), and very low presbyopia (2.03%), and variation exists in MSVI rates between Northern (0.59% in Iceland, ⁴ 0.66% in the UK ⁵ ) and Southern Europe (3.1% in Italy ⁶ ). Higher VI in Eastern than in Western Europe likely reflects differences in healthcare access.

A 1989 Malta report found 0.9% blindness and 9.3% MSVI in adults aged ≥40^{. 7} Among adults ≥18, the 2010 European Health Examination Survey (EHES) reported 1.1% low vision, while the 2008 and 2019/20 European Health Interview Survey (EHIS) recorded 0.8% blindness and 2.9% and 10.4% non-blind VI, respectively^.8,9 

Globally, the main causes of blindness are cataract (45.5%) and glaucoma (11%), followed by uncorrected/undercorrected refractive error (URE) (6.6%), age-related macular degeneration (ARMD) (5.6%), and diabetic retinopathy (DR) (2.5%).^10–12 URE is the leading cause of MSVI (41%), followed by cataract (38.9%)^10,12–15. In high-income countries, glaucoma is the leading cause of blindness (28.2%) and URE the leading cause of MSVI (44.5%)^10–12. URE and cataract, being common and reversible, are classified by the World Health Assembly as “avoidable” causes of visual impairment. ¹⁶

Malta, a small island state in Southern Europe, has a densely growing and ageing population of 563,443. With a GDP per capita slightly above the EU average (109 vs 100 adjusted for purchasing power), ¹⁷ Malta is a high-income country. It has the highest prevalence of overweight and obesity (69.75%) ¹⁸ and one of the highest diabetes rates (10.31%) among EU states,^18,19 both rising over the past 30 years. High UV radiation and air pollution,^20,21 combined with age and diabetes ²² as key ocular risk factors, may influence VI prevalence.

Previous data on VI in Europe are over a decade old, while more recent data is scarce.^2,3 With changing risk profiles and scarce recent national and European data, updated VI prevalence estimates, especially from small islands like Malta, are crucial for public health planning.

The Malta Eye Study (TMES) is a cross-sectional population-based study estimating the prevalence of common eye diseases and VI in Maltese adults aged 50–80. This paper presents the population's characteristics, VI prevalence, causes and predictors.

Method

TMES methodology, including sample size derivation from expected VI prevalence using the one-proportion confidence interval (CI) formula, is detailed elsewhere^{. 23} The target of 1,800 was feasible and sufficient for all estimates. The study followed the STROBE checklist ( Supplementary Table 1 ).

Ethical clearance and permissions

Ethical approval (FRECMDS_1819_94) was obtained from the University of Malta Faculty Research Ethics Committee before data collection. The study adhered to the Declaration of Helsinki and GDPR, with hospital and Data Protection Officer approvals. Informed consent was obtained and data pseudonymised.

Study population

A random sample of 4006 individuals aged 50–80 years, stratified by age, sex, and locality from the electoral register, was invited for ophthalmic assessment between September 2021 and December 2023. Invitations, sent 2–4 weeks prior in Maltese and English, included appointment details, contact information, and instructions. Reminders were made via phone or SMS, and participants could reschedule by phone or email. Participation was promoted through free hospital parking, flexible scheduling including late sessions, additional Gozo sessions, social media campaigns, television features, second recruitment calls, follow-up explanations of the study's importance and small incentives such as gifts and free refraction prescriptions. Hospital transport was offered to participants with limited mobility or access, though uptake varied due to waiting times when drivers collected multiple passengers.

Ophthalmic assessment

Examinations took place at Mater Dei (September 2021–May 2024) and Gozo General Hospital (June–July 2024). Vision testing included monocular presenting visual acuity (MPVA) via electronic ETDRS chart with distance glasses if available, expressed in logMAR; pinhole was performed if MPVA > 0.3. Best-corrected visual acuity followed autorefraction (Visionix^®) and subjective refraction. Cognitive impairment prompted guidance, such as pointing to optotypes or examiner interpretation. Slit-lamp ocular exams included tonometry and dilated fundoscopy (90D lens). Imaging comprised colour fundus photography and swept-source optical coherence tomography (OCT, 3D wide, macula and optic disc). Anthropometrics included height and weight; clinical measures included capillary glucose and blood pressure. The questionnaire recorded sociodemographic, medical, ocular, and drug histories, plus cognitive screening (Quick Mild Cognitive Impairment Score).

Response rate and missing data

Of 1,794 participants (44.8% response), age–gender representativity was checked against national census data ²⁴ using binomial and chi-squared tests; weights (population proportion/sample proportion) corrected minor under-representation (Supplementary Table 6). The questionnaire was completed by 1,786 participants. Missing data were minimal: 0.4% for most items, 0.2% for VI classification, 8.1% for education (post-survey calls), 4.8% for autorefraction, and 2.3% for anthropometry.

Definitions used

VI was defined per ICD-11 using monocular presenting visual acuity (MPVA): any VI if MPVA > 0.3 in ≥1 eye; bilateral VI (VIOU) if MPVA > 0.3 in the better eye; unilateral VI if MPVA > 0.3 in one eye and ≤0.3 in the fellow eye. Severity (better eye or affected eye for unilateral VI) was mild (MPVA > 0.3–0.5), moderate-to-severe (MSVI; MPVA > 0.5–1.3), or blindness (MPVA > 1.3). Near VI was presenting near vision worse than N6 at 40 cm. Other definitions are in Table 1's legend. Causative pathologies were examiner-identified; URE was classified regardless of lens status and counted if pinhole improved MPVA to ≤0.3.

Table 1.

Crude descriptive statistics for demographic, behavioural, and medical characteristics of the population.

Characteristics			Valid Responses	No. (%)	95% CI of %
Demographic Factors	Sex	Male	1786	879 (49.2)	46.9 - 51.6
		Female	1786	907 (50.8)	48.4 - 53.1
	Age Group	50–54	1786	223 (12.5)	11.0 - 14.1
		55–59	1786	294 (16.5)	14.8 - 18.3
		60–64	1786	353 (19.8)	17.9 - 21.7
		65–69	1786	356 (19.9)	18.1 - 21.9
		70–74	1786	314 (17.6)	15.8 - 19.4
		75–80	1786	246 (13.8)	12.2 - 15.5
	District	Northern Harbour	1786	474 (26.5)	24.5 - 28.7
		South Eastern	1786	266 (14.9)	13.3 - 16.6
		Western	1786	256 (14.3)	12.7 - 16.0
		Southern Harbour	1786	328 (18.4)	16.6 - 20.2
		Northern	1786	322 (18.0)	16.3 - 19.9
		Gozo and Comino	1786	140 (7.8)	6.6 - 9.2
	Education Level	Primary Level	1651	535 (32.4)	30.1 - 34.7
		Secondary Level	1651	582 (35.3)	32.9 - 37.6
		Tertiary Level	1651	534 (32.3)	30.1 - 34.7
	Employment History	Employed	1786	674 (37.7)	35.5 - 40.0
		Retired	1786	879 (49.2)	46.9 - 51.6
		Domestic Work	1786	233 (13.0)	11.5 - 14.7
	Nationality	Maltese	1794	1732 (96.5)	95.6 - 97.3
		Foreign	1794	62 (3.5)	2.7 - 4.4
		Maltese parents / grandparents	1776	1613 (90.8)	89.4 - 92.1
		Foreign parents / grandparents	1776	163 (9.2)	7.9 - 10.6
		Either parent foreign	1786	121 (6.8)	5.7 - 8.0
		Both parents foreign	1786	86 (4.8)	3.9 - 5.9
		One parent foreign	1786	35 (2.0)	1.4 - 2.7
Behavioural Factors	Tobacco Use *	Never Smoked	1777	1267 (71.3)	69.1 - 73.4
		Light Smoker	1777	199 (11.2)	9.8 - 12.8
		Moderate Smoker	1777	140 (7.9)	6.7 - 9.2
		Heavy Smoker	1777	171 (9.6)	8.3 - 11.1
	Alcohol Consumption **	No Drinking	1786	1282 (71.8)	69.6 - 73.9
		Normal Drinking	1786	334 (18.7)	16.9 - 20.6
		Heavy Drinking	1786	170 (9.5)	8.2 - 11.0
	Recreational Drug Use	Any	1794	76 (4.2)	3.4 - 5.3
		Marijuana	1794	46 (2.6)	1.9 - 3.4
		Cocaine	1794	5 (0.3)	0.1 - 0.6
		MDMA (ecstasy)	1794	3 (0.2)	0.0 - 0.5
		Heroin	1794	2 (0.1)	0.0 - 0.4
		LSD	1794	3 (0.2)	0.0 - 0.5
	Hat Wearing in the sun		1780	1032 (58.0)	55.6 - 60.3
	Sunglasses Wearing in the sun		1781	1317 (73.9)	71.8 - 76.0
Clinical Assessments	Body Mass Index (BMI) ***	Overweight	1756	651 (37.2)	34.9 - 39.5
		Obese	1756	698 (39.9)	37.6–42.2
	Random Glucose Level****	Normal	1776	1461 (82.3)	80.4 - 84.0
		Relatively High	1776	218 (12.3)	10.8 - 13.9
		Hyperglycaemia	1776	97 (5.5)	4.5 - 6.6
	Blood Pressure	Hypertensive	1773	957 (54.0)	51.6 - 56.3
	Cognitive Impairment *****	Mild Cognitive Impairment	1782	683 (38.3)	36.1 - 40.6
		Dementia	1782	242 (13.6)	12.0 - 15.3
Medical History	Self-Reported Diabetes Mellitus	Insulin only	1786	12 (0.7)	0.3 - 1.2
		Oral Hypoglycaemic Agents only	1786	283 (15.8)	14.2 - 17.6
		Both Insulin and Oral Hypoglycaemic Agents	1786	31 (1.7)	1.2 - 2.5
		No Treatment	1786	27 (1.5)	1.0 - 2.2
	Self-Reported Other Medical Conditions	Hypertension	1786	787 (44.1)	41.7 - 46.4
		Stroke	1786	35 (2.0)	1.4 - 2.7
		Ischaemic heart disease	1786	89 (5.0)	4.0 - 6.1
		Hypercholesterolaemia	1786	760 (42.6)	40.2 - 44.9
		Malignancy	1786	157 (8.8)	7.5 - 10.2
		Thyroid	1786	250 (14.0)	12.4 - 15.7
		Kidney	1786	9 (0.5)	0.2 - 1.0
		Asthma	1786	37 (2.1)	1.5 - 2.8
		Psychiatric	1786	40 (2.2)	1.6 - 3.0
		Orthopaedic	1786	47 (2.6)	1.9 - 3.5
		Neurological	1786	31 (1.7)	1.2 - 2.5
		Vascular	1786	15 (0.8)	0.5 - 1.4
		Gastrointestinal	1786	47 (2.6)	1.9 - 3.5
		Dermatological	1786	13 (0.7)	0.4 - 1.2
		Inflammatory	1786	19 (1.1)	0.6 - 1.7
		ENT	1786	10 (0.6)	0.3 - 1.0
		Haematological	1786	7 (0.4)	0.2 - 0.8
		Respiratory	1786	55 (3.1)	2.3 - 4.0
		Urological	1786	30 (1.7)	1.1 - 2.4
		Gynaecological	1786	2 (0.1)	0.0 - 0.4
		Fibromyalgia	1786	11 (0.6)	0.3 - 1.1
	Medication Use	History of Regular Medication	1785	1346 (75.4)	73.3 - 77.4
		Antihypertensives	1786	773 (43.3)	41.0 - 45.6
		Antiplatelets	1786	124 (6.9)	5.8 - 8.2
		Anticoagulants	1786	51 (2.9)	2.1 - 3.7
		Lipid Lowering Agents	1786	665 (37.2)	35.0 - 39.5
		Psychiatric medications	1786	136 (7.6)	6.4 - 8.9
		Anti-convulsant	1786	25 (1.4)	0.9 - 2.1
		Antihistamines	1786	14 (0.8)	0.4 - 1.3
		Anti-Parkinsonian Rx	1786	8 (0.4)	0.2 - 0.9
		Proton Pump Inhibitors	1786	122 (6.8)	5.7 - 8.1
		Non Steroidal Anti-Inflammatories	1786	5 (0.3)	0.1 - 0.7
		Long-Acting Beta-Agonists	1786	22 (1.2)	0.8 - 1.9
		Long-Acting Muscarinic Antagonists	1786	8 (0.4)	0.2 - 0.9
		Systemic Steroid	1786	5 (0.3)	0.1 - 0.7
		Steroid Inhalers	1786	43 (2.4)	1.7 - 3.2

*Tobacco Use Definition ⁴⁷ :

Never Smokers: 0.0 Pack Years

Light Smokers: 0.1–20.0 Pack Years

Moderate Smokers: 20.1–40.0 Pack Years

Heavy Smokers: > 40.0 Pack Years

**Alcohol Use Definition ⁴⁸ :

No drinking: Answers “never” to question related to alcoholic drinks

Normal drinking: Males: < 5 drinks on any day or <15 per week; Females: < 4 drinks on any day or <8 per week

Heavy Drinking: Males: ≥ 5 drinks on any day or ≥15 per week; Females: ≥ 4 drinks on any day or ≥8 per week

*** Overweight and Obesity

Overweight: BMI ≥25.0 kg/m² < 30.0 kg/m²

Obesity: BMI ≥30.0 kg/m²

****Random Glucose Level Normal <7.9 mmol/L Relatively High 7.8-11.0 mmol/L Hyperglycaemia ≥11.0 mmol/L *****Cognitive Impairment by the Quick Mild Cognitive Impairment Score ⁴⁹

No Cognitive Impairment: ≥ 62

Mild Cognitive Impairment: ≥ 45, < 62

Dementia: < 45

Distance effective refractive error coverage (eREC) was the proportion achieving MPVA ≤ 0.3 among those with refractive error; near eREC was analogous for N6. ²⁵ Effective cataract surgical coverage (eCSC) was the proportion of cataract-operated eyes achieving postoperative MPVA ≤ 0.5 among cataract-related VI cases. ²⁶

Statistical analysis

Analyses used IBM SPSS^® v23. Demographics were compared with census data. Descriptive statistics provided prevalence estimates with 95% CIs; crude prevalences described population characteristics. Age- and gender-adjusted prevalences were calculated via direct standardisation. Categorical VI predictors were tested with Chi-square or Fisher's exact tests (Supplementary Tables 2–5). Significant univariate predictors (p < 0.05) entered backward stepwise binary logistic regression (Wald test), with absence of VI as reference; final steps were presented. Minimal missing data, unrelated to VI, were handled by listwise deletion without imputation.

Results

Baseline population characteristics

Participants completing the full ophthalmic examination had a mean age of 64.8 years (SD 8.0). Compared with the national 50–80-year census population (mean 63.8, SD 8.4), the sample was representative for age, gender, and district (all p > 0.05) overall but underestimated males (p < 0.001) and females (p = 0.009) aged 50–59 in the Northern Harbour, and females 50–59 in Gozo (p = 0.004); weighting corrected these discrepancies (Supplementary Data Table 6).

Overweight and obesity affected 698 individuals (77.0%; 95%CI 75.0–79.0%), 97 persons (5.5%; 95%CI 4.5–6.6%) were hyperglycaemic, and 957 persons (54.0%; 95%CI 51.6–56.3%) were hypertensive. Cognitive impairment was present in 925 individuals (51.9%; 95% CI 49.6–54.3%) (Table 1).

Hypertension (44.1%, 95%CI 41.7%-46.4%), hypercholesterolemia (42.6%, 95%CI 40.2%-44.9%) and diabetes (19.8%, 95%CI 17.9%-21.7%) were the commonest self-reported medical conditions while antihypertensives (43.3%, 95%CI 41.0%-45.6%) and lipid lowering agents (37.2%, 95%CI 35.0%-39.5%) were the commonest reported used medications.

Prevalence of visual impairment

Adjusted population data showed that 23.9% (95%CI 21.9%–25.9%) of adults aged 50–80 had VI in either eye, with 6.0% (95%CI 4.9%–7.2%) having VIOU and 17.9% (95%CI 16.1%–19.7%) unilateral VI. By severity, mild VIOU was prevalent in 4.5% (95%CI 3.6%–5.5%), MSVI OU in 1.3% (95%CI 0.9%–2.0%), and blindness in 0.2% (95%CI 0.1%–0.6%) of individuals. Mild unilateral VI was the most frequent subtype (10.3%, 95%CI 9.0%–11.8%), followed by unilateral MSVI (4.9%, 95%CI 4.0%–6.0%) and unilateral blindness (2.6%, 95%CI 1.9%–3.5%). Near VI occurred in 10.2% (95%CI 8.1%–12.7%) individuals. VI indices varied mainly by age rather than sex (Figure 1; Supplementary Figures 1–2).

Figure 1.

The prevalence of visual impairment (vi) laterality by age group and sex in the adjusted TMES population.

Predictors of visual impairment

Multivariate models were adjusted for age, education, employment, nationality (self and parent), sunglasses use, alcohol intake, diabetes type, cerebrovascular disease, cognitive status and the use of calcium channel blockers, angiotensin converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB) and lipid-lowering agents (Table 2).

Table 2.

Multivariate analyses of predictors of visual impairment (vi) in at least 1 eye, uncorrected / undercorrected refractive error (URE) vi, amblyopia vi and cataract vi

	Predictor	Reference Category	Category	Odds Ratio (95%CI)	p value
VI in either eye vs no VI	Age Group	50–59	60–69	1.30 (0.95–1.77)	0.103	0.003
			70–80	1.80 (1.28–2.53)	0.001
	Education Level	Primary	Secondary	0.74 (0.54–1.00)	0.051	0.003
			Tertiary	0.57 (0.41–0.78)	0.001
	Nationality		Maltese	3.15 (0.90-11.03)	0.072	N/A
	One parent Foreign	−	+	2.10 (0.99–4.46)	0.052	N/A
	Sunglasses Wearing	−	+	0.70 (0.54–0.91)	0.007	N/A
	Diabetes type	No Diabetes	On Insulin Only	2.98 (0.89–9.98)	0.078	0.026
			On OHAs Only	0.97 (0.69–1.36)	0.850
			On Both Insulin and OHAS	2.88 (1.22–6.81)	0.016
			On no medical treatment	2.03 (0.81–5.11)	0.131
	Angiotensin Receptor Blockers	−	+	0.60 (0.41–0.89)	0.011	N/A
	Cognitive Impairment Grade	No Cognitive Impairment	Mild CI	1.16 (0.89–1.52)	0.279	0.001
			Dementia	2.00 (1.40–2.87)	<0.001
URE VI vs No VI	Education level	Primary	Secondary	0.82 (0.56–1.19)	0.292	0 . 049
			Tertiary	0.59 (0.38–0.90)	0.014
	One parent Foreign	−	+	3.28 (1.44–7.44)	0.005	N/A
	Sunglasses Wearing	−	+	0.72 (0.51–1.01)	0.057	N/A
	Cognitive Impairment Grade	No Cognitive Impairment	Mild	1.14 (0.79–1.63)	0.485	0.002
			Dementia	2.25 (1.42–3.56)	0.001
	Malignancy	−	+	0.51 (0.26–1.02)	0.055	N/A
Amblyopia VI vs no VI	Age Group	50–59	60–69	2.15 (1.14–4.03)	0.018	0.058
			70–80	1.79 (0.89–3.60)	0.101
	Sunglasses Wearing	−	+	0.60 (0.35–1.02)	0.058	N/A
Cataract VI vs no VI	Age Group	50–59	60–69	2.06 (0.59–7.19)	0.257	<0.001
			70–80	9.88 (3.23–30.24)	0.000
	Cognitive Impairment Grade	No Cognitive Impairment	Mild	1.17 (0.55–2.48)	0.692	0.008
			Dementia	3.12 (1.43–6.84)	0.004
	Diabetes type	No Diabetes	On Insulin Only	5.43 (0.41–72.74)	0.202	0.010
			On OHAs Only	1.28 (0.59–2.76)	0.532
			On Both Insulin and OHAS	8.95 (2.38–33.66)	0.001
			On no medical treatment	3.46 (0.67–17.81)	0.138

Compared with the 50–59-year group, those aged 70–79 had higher odds of VI in either eye (OR 1.80). Tertiary education was protective versus primary education (OR 0.57). Diabetes treated with both insulin and OHAs was associated with greater odds of VI (OR 2.88), as was dementia (OR 2.00). ARB use was linked to lower odds (OR 0.60).

Sunglasses wear was also protective (OR 0.70) and associated with reduced URE VI (11.3% vs 14.8%; p = 0.046) and cataract VI (3.0% vs 6.0%; p = 0.003). These remained significant after mutual adjustment (URE VI OR 0.72, 95%CI 0.53–0.98, p = 0.036; cataract VI OR 0.47, 95%CI 0.28–0.76, p = 0.002). No association was observed between sunglasses and ARMD, while cortical opacities were linked to lower sunglasses use (data not shown).

Visual impairment by causes

Prevalence of visual impairment by causes

The commonest cause of distance VI of any severity or laterality was URE (12.3%, 95%CI 10.8%–13.9%), followed by amblyopia (5.0%, 95%CI 4.1%–6.1%) and cataract (3.8%, 95%CI 3.0%–4.8%). Other age-related causes accounted for 7.1% (95%CI 5.9%–8.4%), including pathological myopia (1.3%, 95%CI 0.8%–1.9%), DR (0.8%, 95%CI 0.5%–1.4%), ARMD (0.6%, 95%CI 0.3%–1.0%), and glaucoma (0.4%, 95%CI 0.2%–0.9%). Avoidable VI (URE and/or cataract) was present in 15.8% (95%CI 14.1%–17.5%) of individuals and comprised most cases, while non-avoidable causes represented 8.1% (95%CI 6.9%–9.5%) of the population.

For VIOU by cause and severity, URE mild VI was most frequent (3.1%, 95%CI 2.4%–4.1%), followed by cataract mild VI (1.2%, 95%CI 0.7%–1.8%) and URE MSVI (0.7%, 95%CI 0.3%–1.2%).

Uncorrected presbyopia caused near VI in 4.5% (95%CI 3.1%–6.3%) and this is significantly higher than that in Western Europe (2.03%, 95%CI 1.32%, 2.88%) ²⁷ and high-income countries (2.37%, 95%CI 1.59%, 3.31%). ²⁸

Table 3 shows adjusted prevalences of causes of VI by eye and severity, while Supplementary Figures 3–9 present adjusted rates stratified by age and sex.

Table 3.

Prevalence rates (with 95% confidence intervals) of the main visually impairing causes in either eye and in the better eye (stratified by severity) in the adjusted TMES population

			Frequency	Percentage (%)	L95% CI	U95% CI
URE	Either Eye		221	12.3%	10.8%	13.9%
	Better Eye	mild VI	56	3.1%	2.4%	4.1%
		MSVI	12	0.7%	0.3%	1.2%
		Blindness	0	0.0%	0.0%	0.2%
		Total	68	3.8%	3.0%	4.8%
Amblyopia	Either Eye		90	5.0%	4.1%	6.1%
	Better Eye	mild VI	4	0.2%	0.1%	0.6%
		MSVI	5	0.3%	0.1%	0.7%
		Blindness	0	0.0%	0.0%	0.2%
		Total	9	0.5%	0.2%	1.0%
Cataract	Either Eye		68	3.8%	3.0%	4.8%
	Better Eye	mild VI	21	1.2%	0.7%	1.8%
		MSVI	4	0.2%	0.1%	0.6%
		Blindness	2	0.1%	0.0%	0.4%
		Total	27	1.5%	1.0%	2.2%
Pathological Myopia	Either Eye		23	1.3%	0.8%	1.9%
	Better Eye	mild VI	2	0.1%	0.0%	0.4%
		MSVI	3	0.2%	0.0%	0.5%
		Blindness	0	0.0%	0.0%	0.2%
		Total	5	0.3%	0.1%	0.7%
DR	Either Eye		15	0.8%	0.5%	1.4%
	Better Eye	mild VI	6	0.3%	0.1%	0.7%
		MSVI	2	0.1%	0.0%	0.4%
		Blindness	1	0.1%	0.0%	0.3%
		Total	9	0.5%	0.2%	1.0%
ARMD	Either Eye		10	0.6%	0.3%	1.0%
	Better Eye	mild VI	2	0.1%	0.0%	0.4%
		MSVI	1	0.1%	0.0%	0.3%
		Blindness	1	0.1%	0.0%	0.3%
		Total	4	0.2%	0.1%	0.6%
Glaucoma	Either Eye		8	0.4%	0.2%	0.9%
	Better Eye	mild VI	1	0.1%	0.0%	0.3%
		MSVI	0	0.0%	0.0%	0.2%
		Blindness	1	0.1%	0.0%	0.3%
		Total	2	0.1%	0.0%	0.4%

URE: Uncorrected/Undercorrected Refractive Error; DR: Diabetic Retinopathy; ARMD: Age-Related Macular Degeneration; MSVI: Moderate-Severe Visual Impairment; VI: Visual Impairment.

Effective refractive error coverage and effective cataract surgical coverage

TMES reported better rates for distance eREC than high income countries (92.8%, 95%CI 91.0%-94.4% vs 79.1%, 95%CI 72.4%-85.0%; p < 0.001) ²⁵ and eCSC than Hungary (94.5%, 95%CI 88.5%-98.0% vs 70.3%, 95%CI 65.8%-74.9%; p < 0.001). ²⁶ The near eREC in TMES was 93.7% (95%CI 91.2%-95.6%).

Proportion of causes of visual impairment

Figure 2 and Supplementary Figures 10–16 show proportions of non-mutually exclusive VI causes by laterality and severity. URE was the leading cause of any VI (51.6%, 95%CI 46.8%–56.5%), unilateral VI (47.8%, 95%CI 42.2%–53.4%), and VIOU (63.0%, 95%CI 53.1%–72.2%). Amblyopia ranked second for VI in either eye (21.0%, 95%CI 17.3%–25.2%) and unilateral VI (25.3%, 95%CI 20.6%–30.4%), but less so for VIOU (8.4%, 95%CI 3.9%–15.4%). Cataract was third overall (15.9%, 95%CI 12.6%–19.7%), second for VIOU (24.3%, 95%CI 16.5%–33.5%), and third for unilateral VI (12.8%, 95%CI 9.4%–17.0%). Pathological myopia was fourth for VI in either eye (5.1%, 95%CI 3.2%–7.7%) and unilateral VI (5.3%, 95%CI 3.1%–8.4%). DR ranked third for VIOU (9.3%, 95%CI 4.5%–16.4%) and fifth overall (3.5%, 95%CI 2.0%–5.7%), but was less frequent in unilateral VI (1.6%, 95%CI 0.5%–3.6%).

Figure 2.

Proportions (with 95% ci bars) of non-mutually exclusive causes of visual impairment in at least one eye. ARMD: Age-Related Macular Degeneration; VI: Visual Impairment.

Avoidable causes explained 66.1% (95%CI 61.4%–70.6%) of VI in either eye and 82.2% (95%CI 73.7%–89.0%) of VI in the better eye. Multiple causes occurred in 9.8% (95%CI 7.2%–13.0%) of individuals with VI.

Predictors of causes of visual impairment

Predictor models for the main VI causes are shown in Table 2. URE VI was positively associated with dementia (OR 2.25) and inversely with higher education (tertiary vs primary OR 0.59). Amblyopia VI was associated with age 60–69 (OR 2.15 vs 50–59), while cataract VI was linked to age 70–80 (OR 9.88), dementia (OR 3.12), and diabetes treated with both insulin and oral agents (OR 8.95). Uncorrected presbyopia near VI was less frequent with higher education (tertiary vs primary OR 0.18, 95%CI 0.05–0.68, p = 0.011) and more frequent with dementia (OR 3.54, 95%CI 1.37–9.14, p = 0.009).

Discussion

The population characteristics

Obesity and diabetes rates in Maltese adults align with previous estimates.^18,29 High cognitive impairment rates reflect global post-pandemic trends ³⁰ and may be influenced by the older sample age, the more sensitive screening tool versus the standardised mini-mental state examination, ³¹ and potential cultural or educational effects on performance.

The prevalence of visual impairment

In TMES, one in four adults aged 50–80 had VI in at least one eye, of whom one-fourth had VIOU. A prior local study reported MSVI/blindness (better eye) at 10.11% (95%CI 8.89%–11.43%) in those aged 40+, ⁷ whereas TMES showed a lower prevalence with the same criteria (1.39%, 95%CI 0.89%–2.05%), despite excluding the typically lower-risk 40–49 age group.

Although Malta's population is ageing^{, 24} VI prevalence has declined, likely reflecting expanded eye care services. In 1988, with a population of 349,014 and only seven public service ophthalmologists, ⁷ access was limited. Today their number has more than quadrupled (n = 29), alongside major advances in screening and management of refractive error, cataract, retinal disease, and glaucoma.

Despite age group differences (TMES: 50–80 years; EHES: ≥ 18; EHIS: ≥ 15), TMES reported a comparable prevalence of low vision to EHES 2010 (1.1%) when using the same presenting visual acuity definition (Snellen 6/25–6/320), at 1.2% (95%CI 0.7%–1.8%). When grouping self-reported visual difficulty responses to match EHIS, TMES found 17.3% (95%CI 15.5%–19.6%) reported non-blind VI, broadly aligning with EHIS 2008 (2.9%) and 2019/20 (10.4%). Differences in age structures, survey tools, and data collection modes may explain these variations.

TMES data were age-standardised to the 2019 Global Burden of Disease standard population for comparison with the Vision Loss Expert Group meta-analysis. TMES distance VI and presbyopia rates differed significantly from Western Europe (p < 0.001; Table 4). When compared with Central and Eastern European estimates, TMES rates remained lower, placing Malta at the lower end of the European spectrum despite its Southern location. This likely reflects differences in eye care access, with lower eREC in Central and Eastern Europe contributing to higher VI. ²⁵ The upper age limit of 80 years in TMES may partly explain the lower prevalence compared with studies without such limits.

Table 4.

Comparison of age-standardised prevalence of vi grades between TMES and the Western Europe component of the global burden of disease (GBD) study. ⁵⁰

	TMES (age-adjusted)		GBD - Western Europe
	n (%)	95% CI (%)	n in thousands (%)	95% CI (%)	p
Mild VI	73 (4.09)	3.22–5.12	7970 (3.63)	3.10–4.13	<0.001
MSVI	19 (1.07)	0.64–1.66	11100 (5.15)	4.52–5.79
Blindness	3 (0.17)	0.03–0.49	1350 (0.60)	0.51–0.70
Near VI	69 (9.72)	7.64– 12.14	N/A	N/A	N/A
Uncorrected Presbyopia VI	33 (4.6%)	3.22–6.47	4190 (2.03)	1.32–2.88	<0.001

MSVI: Moderate Severe Visual Impairment; VI: Visual Impairment.

Malta's lower distance VI rates likely reflect short clinic travel distances and strong eye care access. With 85 ophthalmologists per million and a refractionist ratio of 1:9,237, Malta exceeds both high-income and WHO workforce benchmarks^32–34.

The Causes of Visual Impairment

The age-adjusted prevalence of MSVI due to URE was significantly lower in TMES (0.6%; 95% CI: 0.3%-1.1%) than in comparable high-income populations (1.94%; 95%CI 1.67%-2.25%, p < 0.001). ²⁸ A similar trend was observed for cataract-related MSVI (0.1%, 95%CI 0.0%-0.4% vs 1.46%, 95%CI 1.22%-1.71%, p < 0.001). ¹¹ High coverage for both refractive correction and cataract surgery, as measured by distance eREC and eCSC, likely contributes to the low prevalence of VI observed in TMES. These results align with patterns reported in other high-coverage populations, highlighting the importance of effective refractive and cataract services.

Avoidable VI, though less prevalent than in Western Europe, explained the main causes of VIOU, accounting for two-thirds and one-quarter of cases, respectively. This reflects the lower proportion of other causes, amplifying the relative impact of avoidable VI. Despite rising diabetes prevalence, DR-related VI is not higher than in other countries, likely due to diabetes care, screening programmes, and DR management. Further reductions may be supported by shorter waiting times and improved public-sector optometry access.

Amblyopia VI was highest in the 60–69 age group (6.2%, 95%CI 4.4–8.3%, 60–69 vs 50–59 OR 2.15, p = 0.018), reflecting childhood ophthalmic care between 1952 and 1971. Orthoptic services were first established in the early 1970s, meaning those now aged 60 + missed the critical treatment window before age seven^35–37. With childhood vision screening and paediatric ophthalmic and orthoptic services now in place, lower amblyopia prevalence is expected in younger and future older cohorts.

Relevance of predictors of visual impairment

Despite good eCSC, prioritising cataract surgery in older adults remains essential, given the ageing population and burden of reversible cataract VI. Since late 2024, evening surgical lists have been introduced to meet the patient charter ³⁸ and reduce waiting times below six months. ³⁹ Expanding surgical capacity improves access, but health literacy declines with age, particularly beyond 75 years, ⁴⁰ where most cataract VI occurs ( Supplementary Figure 5 ), and may limit uptake. Age-friendly outreach through primary care, community hubs, or local councils should raise awareness of cataract as a treatable condition.

Higher education reduces VI and URE VI through improved health literacy, lifestyle, and care access. ⁴¹ Bridging gaps requires public education, awareness of free services, investment in primary eye care, and expansion of the public optometry workforce, as only 4 of 30 registered optometrists are government-employed.

Having one foreign-born parent was associated with higher VI and URE VI. This may reflect interactions between genetic background, ⁴² health awareness, ⁴³ and eye care access. In Malta, school vision screening has only been in place for the past 30 years, which may partly explain these findings, underlining the value of early detection and equitable refractive care through school screening.

Insulin- and oral-treated diabetes was significantly linked to VI, particularly cataract-related. Given diabetes’ role in cataractogenesis,^44,45 TMES recommends prioritising cataract assessment and surgery for poorly controlled individuals.

Dementia and cerebrovascular disease were associated with VI, especially URE and cataract VI. ⁴⁶ VI may also indicate underlying cognitive decline, supporting reciprocal screening for VI and dementia. ARBs may protect against retinal stress, inflammation, ⁴⁷ and vascular damage, ⁴⁸ key in DR, ARMD, and glaucoma, ⁴⁹ with meta-analyses supporting ACEI/ARB use in DR regression. ⁵⁰

Strengths of the study

Despite the response rate, the sample represented 1% of the population and was broadly representative by age, gender, and district, with weighting applied for full representativity. Information bias was minimised through validated tools, and all examinations were performed by a single observer, ensuring consistency. ²³ The rigorous assessment enabled detection of symptomatic and asymptomatic conditions. This study adds to the limited literature on unilateral VI, recognising common coexisting pathologies and inter-eye differences. Non–mutually exclusive cause categories were used to capture clinical complexity, an approach useful for future studies.

Weaknesses of the study

Turnout matched similar local studies^{, 18} but the 44.8% response rate may introduce nonresponse bias. Post-stratification weighting corrected the underrepresented 50–59 age group. Nonrespondents may still have differed systematically from participants in unmeasured ways, which could influence prevalence estimates. Other underrepresented subgroups included prisoners, foreign residents without contact details, and individuals with disabilities. Hospital-based screening excluded those unable or unwilling to attend; domiciliary visits, could have complemented data collection but were not feasible for this study. Cognitive impairment sometimes hindered cooperation, but examiners adapted procedures to obtain accurate outcomes. Subgroup analysis of bilateral blindness (n = 4) was not feasible. VI was defined solely by visual acuity, potentially underestimating disability, with limited detection of cerebral or subtle neuro-ophthalmological causes. Comparisons with other studies were limited by TMES’ 50–80-year age range.

Conclusions

In TMES, adjusted prevalence of VI in either eye was 23.9% (95%CI 21.9%–25.9%) and blindness in the better eye 0.2% (95%CI 0.1%–0.6%). Despite diabetes and UV risks, VI rates, especially DR-related, compared well internationally. Reducing avoidable VI requires stronger public optometry services and targeted cataract screening, particularly in older adults and those with poorly controlled diabetes.