Topical Beta-Blockers and Cardiovascular Mortality: Systematic Review and Meta-Analysis with Data from the EPIC-Norfolk Cohort Study

Figures & data

Table 1. Summary of three published studies reporting associations between topical beta-blocker use and cardiovascular mortality.

Author	Study design	Participants	Exposure	Exposure assessment	Outcome	Outcome assessment	Covariables	Results
Lee et al.^2	Cohort study	N = 3654, 49–97 years, Blue Mountains, Australia	Glaucoma status, baseline medications	Detailed baseline eye examinations using Humphrey 30-2 visual field test for glaucoma. Face-to-face interviews for medication history.	All-cause mortality and cause of death	Matching participant demographic data to the Australian National Death Index using a probabilistic linkage package. Non-exact matches examined manually and corroborated with information from family members.	Age, sex, hypertension, diabetes, heart attack, angina, stroke, cancer, smoking, alcohol consumption, oral beta-blocker use, myopia, cataract	No association between all-cause mortality and glaucoma. Increased risk of cardiovascular mortality in those with glaucoma in <75 years cohort and previously diagnosed glaucoma (RR 2.78 95% CI 1.20–6.47). Further stratified analysis showed increased risk in those with previously diagnosed glaucoma (RR 1.85, 95% CI 1.12–3.04), more so in those using topical timolol as part of a trend analysis (RR 2.14, 95% CI 1.18–3.89). No association of increased risk of CVD shown in the 88 baseline users of timolol vs non-users (RR 1.47, 95% CI 0.90–2.41).
Muskens et al.^9	Cohort study	N = 3842, 55+ years, Rotterdam, Netherlands	Topical beta-blockers	All pharmacies in region using fully automated single network on use, date and type of beta-blocker. Matched to participants.	All-cause mortality and cause of death	From municipal registry, vital status checked on bi-weekly basis along with GP reported deaths. All deaths checked by specially trained personnel. CVD deaths according to relevant ICD-10 codes. Mortality data in 100% of cases. Certainty of cause in 84% of cases.	Age, sex, smoking, hypertension, diabetes, angina	No association between topical beta-blocker use and all-cause mortality (HR 0.94, 95% CI 0.71–1.25) or cardiovascular mortality (HR 1.02, 95% CI 0.56–1.86).
Wu et al.^3	Cohort study	N = 4092, 40–84 years, Barbados	Glaucoma status, baseline medications	Detailed baseline eye examination including Humphrey automated perimetry and detailed visual field assessment.	All-cause mortality and cause of death	Verified from death certificates held at Ministry of Health.	Age, sex, diabetes, hypertension, CVD, stroke	No association between OAG and increased mortality. In those with OAG at baseline, those treated with timolol had higher cardiovascular mortality (RR 1.91, 95% CI 1.04–3.50). However when comparing CVD mortality in timolol users to all non-users, no increased risk of cardiovascular mortality was observed (RR 1.58, 95% CI 0.97–2.58).

CI, confidence interval; CVD, cardiovascular disease; GP, general practitioner; HR, hazard ratio; ICD-10, international classification of diseases 10th revision; OAG, open-angle glaucoma; RR, relative risk.

Figure 1. Flowchart of the literature review on the association between topical beta-blockers and cardiovascular mortality illustrating the exclusion process of retrieved publications.

Table 2. Comparison of baseline characteristics between exposure groups in participants in the European Prospective Investigation into Cancer Norfolk Cohort Study (EPIC-Norfolk, 1993–2013).

Baseline characteristic	No topical beta-blocker use (n = 25,460)	Topical beta-blocker use (n = 179)	p-value for difference between exposure groups^a
Age, mean (SD), years	59.2 (9.3)	66.6 (7.6)	<0.001
Sex, % female	54.8	42.2	<0.001
Diabetes, %	2.3	5.6	0.003
Myocardial infarction, %	3.1	10.1	<0.001
Angina, %	5.1	11.7	<0.001
Self-reported hypertension, %	14.2	25.7	<0.001
Self-reported raised cholesterol level, %	8.2	14.0	0.004
Stroke, %	1.4	3.4	0.028
Glaucoma, %	1.4	89.9	<0.001
Systolic blood pressure, mean (SD) mmHg	135.0 (18.4)	141.0 (19.6)	<0.001
Diastolic blood pressure, mean (SD) mmHg	82.5 (11.3)	83.7 (11.8)	0.178
Smoking status, % current/ever	54.0	61.8	0.038
Social class, % non-manual	60.0	66.5	0.080
Education level, % higher	52.9	49.7	0.397

^aDichotomous covariables; p-value for independence calculated using Chi-squared test. Continuous variables; p-value for difference calculated from unpaired Welch’s T-test.

SD, standard deviation.

Table 3. Associations between topical beta-blocker use and cardiovascular mortality in the European Prospective Investigation into Cancer Norfolk Cohort Study (EPIC-Norfolk, 1993–2013).

	Hazard Ratio (95% confidence interval)	p-value
Model 1: Cardiovascular mortality^a
Topical beta-blocker use^b	1.116 (0.802–1.553)	0.513
Age, per year	1.190 (1.182–1.198)	<0.001
Sex, female	0.491 (0.450–0.535)	<0.001
Model 2: Cardiovascular mortality^a
Topical beta-blocker use^+	0.931 (0.668–1.299)	0.677
Age, per year	1.178 (1.170–1.187)	<0.001
Sex, female	0.590 (0.537–0.648)	<0.001
Diabetes	2.118 (1.783–2.516)	<0.001
Myocardial infarction	2.287 (1.968–2.659)	<0.001
Angina	1.294 (1.128–1.486)	<0.001
Stroke	2.272 (1.869–2.761)	<0.001
Systolic hypertension	1.579 (1.432–1.740)	<0.001
Social class, non-manual	0.854 (0.782–0.933)	<0.001
Smoking status	1.339 (1.216–1.475)	<0.001

^a6525 deaths (25.4%); 2158 deaths due to CVD.

^b179 participants on topical beta-blockers, of whom 82 (45.8%) died; 36 from CVD.

Model 1 adjusted for age and sex (N = 25,639), model 2 adjusted for age, sex, diabetes, myocardial infarction, angina, stroke, hypertension, social class, and smoking status, using cox proportional hazards model (N = 24,724 due to missing covariables).

CVD, cardiovascular disease.

Table 4. Associations between topical beta-blocker use and cardiovascular mortality in 514 participants with self-reported glaucoma in the European Prospective Investigation into Cancer Norfolk Cohort Study (EPIC-Norfolk, 1993–2013).

	Hazard Ratio (95% confidence interval)	p-value
Model 1: Cardiovascularmortality in those withglaucoma
Topical beta-blocker use	0.757 (0.485–1.180)	0.219
Age, per year	1.203 (1.150–1.257)	<0.001
Sex, female	0.598 (0.394–0.907)	0.016
Model 2: Cardiovascularmortality in those withglaucoma
Topical beta-blocker use	0.885 (0.557–1.401)	0.605
Age, per year	1.195 (1.140–1.252)	<0.001
Sex, female	0.765 (0.484–1.210)	0.252
Diabetes	2.550 (1.376–4.729)	0.003
Myocardial infarction	1.119 (0.519–2.413)	0.775
Angina	1.735 (0.905–3.327)	0.097
Stroke	5.119 (2.316–11.316)	<0.001
Hypertension	1.015 (0.608–1.694)	0.956
Social class, non-manual	0.743 (0.465–1.189)	0.216
Smoking status	1.246 (0.756–2.054)	0.389

Model 1 adjusted for age and sex, model 2 adjusted for age, sex, diabetes, myocardial infarction, angina, stroke, hypertension, social class, and smoking status, using cox proportional hazards model.

Figure 2. Forest plot of random effects meta-analysis examining effect of topical beta-blocker use on cardiovascular mortality.

ES, effect size; CI, confidence interval; Blue Mountains, Blue Mountains Eye Study; Rotterdam, Rotterdam Study; Barbados, Barbados Eye Studies; EPIC-Norfolk, European Prospective Investigation into Cancer Norfolk Cohort Study.