Open Access
Review  |   May 2016
The Effect of Statins on Intraocular Pressure and on the Incidence and Progression of Glaucoma: A Systematic Review and Meta-Analysis
Author Affiliations & Notes
  • Paul McCann
    Centre for Experimental Medicine Queens University Belfast, Belfast, United Kingdom
  • Ruth E. Hogg
    Centre for Experimental Medicine Queens University Belfast, Belfast, United Kingdom
  • Richard Fallis
    Medical Library, Queen's University Belfast, Belfast, United Kingdom
  • Augusto Azuara-Blanco
    Centre for Experimental Medicine Queens University Belfast, Belfast, United Kingdom
  • Correspondence: Paul McCann, Centre for Experimental Medicine, School of Medicine, Dentistry & Biomedical Sciences, Queen's University Belfast, Institute of Clinical Science, Block A, Grosvenor Road, Belfast BT12 6BA, Northern Ireland, UK; pmccann45@qub.ac.uk
Investigative Ophthalmology & Visual Science May 2016, Vol.57, 2729-2748. doi:10.1167/iovs.15-18595
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Paul McCann, Ruth E. Hogg, Richard Fallis, Augusto Azuara-Blanco; The Effect of Statins on Intraocular Pressure and on the Incidence and Progression of Glaucoma: A Systematic Review and Meta-Analysis. Invest. Ophthalmol. Vis. Sci. 2016;57(6):2729-2748. doi: 10.1167/iovs.15-18595.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: We conducted a systematic review and meta-analysis of observational studies to evaluate the effect of oral statins on intraocular pressure (IOP) and the incidence and progression of glaucoma.

Methods: This was a systematic review of the literature and meta-analysis. Searches of PubMed/Medline and Embase were conducted to include all types of studies. Gray literature abstracts were also considered for inclusion. Last search date was February 2016. Risk of bias was assessed using the Newcastle-Ottawa scale independently by two reviewers. Odds ratios (OR) or hazard ratios (HR) and 95% confidence intervals (CI) were extracted from each study. Pooled ORs for incidence of glaucoma were calculated using a random-effects model.

Results: We identified seven cohort studies, three case–control studies, and one cross-sectional study with a total number of 583,615 participants. No randomized controlled trials were retrieved. Pooled ORs demonstrated a statistically significant association between short-term statin use (≤2 years) and reduced incidence of glaucoma (OR 0.96, 95%CI 0.94, 0.99). Pooled ORs of long-term statin use (>2 years) did not demonstrate statistically significant reduction in incidence of glaucoma (OR 0.70, 95%CI 0.46, 1.06). There was inconsistent evidence for the protective effect of statins against the progression of glaucoma, although there was no standard definition for progression across studies. There was no significant difference in IOP associated with statin use.

Conclusions: Short-term statin use is associated with a reduced incidence of glaucoma. The effect of statins on glaucoma progression and IOP is uncertain.

Glaucoma is a progressive optic neuropathy characterized by structural optic nerve head changes and visual field loss. The leading cause of irreversible blindness worldwide, glaucoma affects 64.3 million people, and this is expected to increase to 111.8 million by 2040.1 The global prevalence of open-angle glaucoma (OAG) between 40 and 80 years is estimated at 3.54% worldwide.1,2 
Major risk factors for OAG include age and intraocular pressure (IOP).3 Intraocular pressure is currently the only modifiable major risk factor for OAG development and progression.4 Medical and surgical therapies have been successfully introduced that lower IOP by reducing aqueous production and increasing outflow; however, these therapies are not without adverse effects. Furthermore, it is not uncommon for the disease to progress despite successful IOP reduction.5 Therefore demand continues for the discovery of novel therapeutic agents that offer patients protection from the onset and progression of glaucomatous visual loss. 
During development pipelines, 90% of drug candidates fail at some point, leaving only 10% as a marketable product.6 Failure late in clinical development results in greater amounts of time, money, and effort invested with little or no return. Drug repurposing is a process of finding new uses for drugs outside the scope of the original indication.7 This benefits from reduced risk and costs because the drug candidates have either already been approved for clinical use or been through several stages of clinical development with known safety and pharmacokinetic profiles.8 
Statins are inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, which is a rate-limiting enzyme necessary for the production of the intermediate product L-mevalonate in the biosynthetic pathway of cholesterol.9 Statins are a relatively well tolerated class of cholesterol-lowering medication commonly prescribed in patients with dyslipidemia for the primary and secondary prevention of cerebrovascular and cardiovascular disease. Clinical and scientific evidence suggests that statins are capable of reducing the risk of cerebrovascular and cardiovascular disease independent of their effect on cholesterol levels.10,11 The so-called pleiotropic properties of statins such as inhibition of isoprenylation of Rho-GTPase12 and immunomodulation13 have been proposed to protect retinal ganglion cells (RGCs) against glaucomatous damage.14 Thus there has been increasing interest in the potential role of statins in glaucoma pathologic mechanisms and therapeutics.15 
The purpose of this literature review is to examine the current clinical and epidemiologic evidence investigating the strength and consistency of the association between clinical statin use and the incidence or progression of glaucoma and its effects on IOP. 
Methods
We followed the MOOSE guidelines16 and registered our review at PROSPERO International Prospective Register of Systematic Reviews (http://www.crd.york.ac.uk/PROSPERO, registration no: CRD42015014875). This article adheres to the PRISMA statement17 checklist for the preferred reporting of systematic reviews and meta-analysis. 
Eligibility Criteria for Considering Studies for This Review
This systematic review focused on studies that investigated the association between statin use and glaucoma incidence or progression and the effect on IOP. Included studies were limited to primary research; however, they were not limited by design, sample size, participants, follow-up, or primary outcome measures. 
Search Methods for Identifying Studies
Medline (1946-February week 3 2016) and Embase (1980-2016 week 8) were searched on 24 February, 2016 (Fig. 1). The search strategy used subject headings in both databases: MeSH terms in Medline and Emtree terms in Embase (Appendix 1). PubMed and Google Scholar searches were also conducted using the search terms “glaucoma” and “statins” to pick up any articles that had not been added to Medline yet. The search strategies were limited to human studies and English language. Included studies were limited to published studies to the exclusion of editorials, commentaries, and article summaries. Reference lists of articles were also interrogated for additional relevant papers. Abstracts were also included. 
Figure 1
 
Search strategy flow diagram.
Figure 1
 
Search strategy flow diagram.
Study Selection
Two authors (REH and PM) screened all titles and abstracts generated from the searches to find studies that contained information on the topic of interest. Full articles were retrieved for detailed assessment by two authors, and papers that did not meet the inclusion criteria were excluded. 
Data Collection
Each study was characterized by extracting methodological details onto a predesigned form by two independent reviewers.18 Relevant outcomes and results were extracted into another form and were screened for comparability. When there were inconsistencies between reviewers' opinions, there were further discussions until consensus was reached. In studies in which more than one estimate of effect was presented, agreement was reached about the most appropriate “adjusted” estimate to include. Attempts were made to contact authors by e-mail when papers presented insufficient data. 
Risk of Bias Assessment
Risk of bias in the nonrandomized observational studies was assessed using Newcastle-Ottawa Quality Assessment Scale (NOS) for cohort and case–control studies as outlined in The Cochrane Handbook of Systematic Reviews.18 The NOS includes a star system in which a study is judged on three domains (Appendix 2); representativeness of study group selection (four items), comparability of groups (two items), and ascertainment of either the exposure or outcome in case–control studies or cohort studies (three items). Studies score a star for each item addressed with a score ranging from 0 to 9. Those studies scoring greater than 7 were distinguished from scores ≤7 as having a lower risk of bias. The cutoff of 7 was used as it had been adopted by a previous review.19 Two independent reviewers repeated this process, and inconsistencies were discussed until consensus was reached. When insufficient information was available to ascertain the NOS score, attempts were made to contact authors for further details. 
Data Synthesis and Analysis
Statistical analysis and meta-analysis were performed using RevMan 5.3 software (The Cochrane Collaboration, Copenhagen, Denmark). We combined the results of different study designs in the meta-analysis because we used the “rare disease assumption” that odds ratios (OR) and risk ratios can be considered equivalent when the disease has a prevalence less than 5%.20,21 The χ2 test of between-study heterogeneity was used to test the null hypothesis that the underlying treatment effect of statins is identical in all studies. The test statistic, Q, follows a χ2 distribution with the degrees of freedom equal to the number of studies minus 1. The I2 statistic measured the degree of inconsistency in the observed treatment effect of statins by measuring the percentage of total variation across the studies that is due to heterogeneity rather than chance. Forest plots were used to graphically represent the investigation of heterogeneity. Within the forest plots, estimates were stratified into subgroups on the basis of length of exposure to statins (≤2 and >2 years) because the primary studies made these stratifications. Overall effect size was then determined for each of the subgroups. A further meta-analysis was performed on estimates that were not subgrouped by length of exposure to statins. We used the most conservative of the two effects models, random effects, to estimate the pooled effect size. This takes into account extra variations when assuming that the studies are estimating different underlying treatment effects. Publication bias was checked for using funnel plots. Sensitivity analysis was performed to examine the impact of poor-quality studies upon the meta-analysis. For the purposes of the description of the results, study outcomes were classified into the three domains relevant to glaucoma: incidence, progression, and IOP. 
Results
The initial searches identified 307 records after the removal of duplicates (Fig. 1). Following screening of these 307 records, 293 were excluded due to being either irrelevant or nonepidemiologic studies. Full texts of 14 potentially relevant manuscripts were retrieved. Three were excluded due to being editorials, commentaries, or summaries of other included studies. The remaining 11 studies explored the association between primary OAG and statin use and were included. Of these 11 studies, 9 were full studies and 2 were abstracts. No randomized controlled trials were retrieved. Four studies investigated glaucoma incidence, one study investigated both glaucoma incidence and progression, and five other studies investigated glaucoma progression. The effect of statin therapy on IOP was reported in three studies. The publication dates for all the studies ranged between 2004 and 2015. 
Descriptions of populations, sample sizes, and outcome measures are outlined in Table 1 and the design for each study is defined in Table 2. The definition of the glaucoma-related outcome measure, the method of ascertainment of statin exposure, and the estimated effects of statins on incidence, progression, and IOP for each included study are presented in Tables 3 to 5, respectively. 
Table 1
 
List of Features of All Included Studies
Table 1
 
List of Features of All Included Studies
Table 1
 
Extended
Table 1
 
Extended
Table 1
 
Extended
Table 1
 
Extended
Table 2
 
List of Study Design Features
Table 2
 
List of Study Design Features
Table 3
 
Features and Results of Studies Investigating Association Between Statin Use and Glaucoma Incidence
Table 3
 
Features and Results of Studies Investigating Association Between Statin Use and Glaucoma Incidence
Table 4
 
Features and Results of Studies Investigating Association Between Statin Use and Glaucoma Progression
Table 4
 
Features and Results of Studies Investigating Association Between Statin Use and Glaucoma Progression
Table 5
 
Features and Results of Studies Investigating Association Between Statin Use and IOP
Table 5
 
Features and Results of Studies Investigating Association Between Statin Use and IOP
Risk of Bias Assessment
Using the NOS, six cohort studies were judged to score ≥8 and the remaining two cohort studies were judged to score ≤7 in quality (Table 6). The lowest-scoring cohort studies were from the two gray literature abstracts with scores of 5 and 0 out of 9. One case–control study was judged to score ≥8 on the NOS, and the other two were judged to have scored ≤7 (Table 7). 
Table 6
 
Newcastle-Ottawa Scale: Cohort Studies
Table 6
 
Newcastle-Ottawa Scale: Cohort Studies
Table 7
 
Newcastle-Ottawa Scale: Case–Control Studies
Table 7
 
Newcastle-Ottawa Scale: Case–Control Studies
Statin Use and Incidence of Glaucoma
The association between statin use and incidence of glaucoma was examined in five studies: two nested case–control studies, one case–control study, one retrospective cohort study, and one prospective cohort study. The outcomes for each study were stratified by the length of exposure to statins as per the primary studies and were then outlined in forest plots (Figs. 2, 3). A further meta-analysis was performed on outcomes reported from studies that did not stratify by the length of exposure (Fig. 4). Overall estimates for incidence of glaucoma were presented in forest plots. For exposure to statins for ≤2 years, overall estimated OR was 0.96 (95%CI [confidence interval] 0.94, 0.99) and for >2 years, overall estimate OR was 0.70 (95%CI 0.46, 1.06). Meta-analysis of outcomes that were not stratified by length of exposure did not show a statistically significant reduction in the incidence of OAG (OR 0.94, 95%CI 0.83, 1.06). 
Figure 2
 
Forest plot of incidence of glaucoma and statin use ≤2 years versus controls. McGwin et al.,22 refers to exposure for <12 months (top) and 12 to 23 months (second from top). McGwin et al.,22 (second from top): Upper limit of 95%CI (1.38) is not exactly equivalent to upper limit of 95%CI in Table 3 (1.39) due to rounding in meta-analysis software.
Figure 2
 
Forest plot of incidence of glaucoma and statin use ≤2 years versus controls. McGwin et al.,22 refers to exposure for <12 months (top) and 12 to 23 months (second from top). McGwin et al.,22 (second from top): Upper limit of 95%CI (1.38) is not exactly equivalent to upper limit of 95%CI in Table 3 (1.39) due to rounding in meta-analysis software.
Figure 3
 
Forest plot of incidence of glaucoma and statin use >2 years versus control. Marcus et al.,25 upper limit of 95%CI (0.92) is not exactly equivalent to upper limit of 95%CI in Table 3 (0.94) due to rounding in meta-analysis software.
Figure 3
 
Forest plot of incidence of glaucoma and statin use >2 years versus control. Marcus et al.,25 upper limit of 95%CI (0.92) is not exactly equivalent to upper limit of 95%CI in Table 3 (0.94) due to rounding in meta-analysis software.
Figure 4
 
Forest plot of incidence of glaucoma and statin use from outcomes not stratified by length of exposure. Marcus et al.,25 upper limit of 95%CI (0.94) is not exactly equivalent to upper limit of 95%CI in Table 3 (0.96) due to rounding in meta-analysis software. Owen et al.,30 upper limit of 95%CI (1.07) is not exactly equivalent to upper limit of 95%CI in Table 3 (1.06) due to rounding in meta-analysis software.
Figure 4
 
Forest plot of incidence of glaucoma and statin use from outcomes not stratified by length of exposure. Marcus et al.,25 upper limit of 95%CI (0.94) is not exactly equivalent to upper limit of 95%CI in Table 3 (0.96) due to rounding in meta-analysis software. Owen et al.,30 upper limit of 95%CI (1.07) is not exactly equivalent to upper limit of 95%CI in Table 3 (1.06) due to rounding in meta-analysis software.
Among studies evaluating the short-term use of statins, McGwin et al.,22 Owen et al.,23 and Stein et al.24 used diagnostic read codes to define glaucoma incidence, whereas Marcus et al.25 used a clinical diagnosis. McGwin et al.22 were unable to demonstrate a statistically significant effect of statin use for less than 12 months (OR 1.03, 95%CI 0.77, 1.39) or for 12 to 23 months (OR 0.75, 95%CI 0.46, 1.23). Consistent with this result, Owen et al.23 did not find a significant association between short-term statin use and glaucoma incidence when adjusted for a socioeconomic index, comorbidities, and other medications taken (OR 0.98, 95%CI 0.89, 1.08). Marcus et al.25 were unable to demonstrate a statistically significant protective effect of cumulative statin use for less than 2 years (hazard ratio [HR] 0.89, 95%CI 0.41, 1.94). However, Stein et al.24 found statistically significant protective effects of statin use for 1 year using two parameters of glaucoma incidence: OAG onset from no previous diagnosis (HR 0.960, 95%CI 0.933, 0.988) and incidence of medical treatment for OAG (HR 0.950, 95%CI 0.924, 0.976). 
Regarding the long-term use of statins, three studies reported an association with reduced OAG incidence. McGwin et al.22 and Stein et al.24 used diagnostic read codes to define OAG incidence, whereas Marcus et al.25 used a clinical diagnosis. McGwin et al.22 demonstrated a statistically significant association between incidence of glaucoma and statin use for greater than 23 months (OR 0.60, 95%CI 0.39, 0.92). In support of this, Marcus et al.25 demonstrated a statistically significant protective effect of cumulative statin use for more than 2 years (HR 0.46, 95%CI 0.23, 0.94). Finally, Stein et al.24 found statistically significant protective effects of statin use for 2 years using OAG onset from no previous diagnosis (HR 0.922, 95%CI 0.870, 0.976) and incidence of medical treatment for OAG (HR 0.902, 95%CI 0.854, 0.953). 
Our meta-analysis suggests that statin therapy for ≤2 years confers a 4% reduction in the incidence of OAG (Fig. 2) while statin therapy for >2 years did not confer a statistically significant reduction in the incidence of OAG (Fig. 3). Statin use not stratified by length of exposure to statins also did not confer a statistically significant reduction in the incidence of OAG (Fig. 4). 
Publication Bias
Funnel plots that plot the OR on the log scale (x-axis) against the standard error of the log odds (y-axis) were used to examine publication bias and the possibility of type 1 error. In Figure 5 there is no evidence of asymmetry in the funnel plot examining short-term statin use, and consequently no publication bias is apparent in these studies. Funnel plots were not conducted for longer-term statin use and statin use not stratified by length of exposure because too few studies were available. 
Figure 5
 
Funnel plot examining publication bias investigating short-term (≤2 years) statin use and incidence of glaucoma.
Figure 5
 
Funnel plot examining publication bias investigating short-term (≤2 years) statin use and incidence of glaucoma.
Sensitivity Analysis
In the sensitivity analysis, the overall heterogeneity and effect size was calculated following exclusion of the studies scoring ≤7 in the NOS (n = 2). When McGwin et al.22 was removed from the analysis there was no change in the pooled OR comparing statin use for ≤2 years versus controls (OR 0.96, 95%CI 0.94, 0.99). There was a change in the pooled OR comparing statin use for >2 years versus controls when McGwin et al.22 was removed but it did not affect the statistical significance of the result (OR 0.71, 95%CI 0.37, 1.38). When McGwin et al.22 and Chen et al.26 were removed from the analysis of pooled ORs that were not stratified by length of exposure, there was no effect on the statistical significance of the result (OR 0.77, 95%CI 0.44, 1.35). 
Statin Use and Progression of Glaucoma
The association between statin use and progression of glaucoma was reported in four full studies and two abstracts (Table 4). Among these there were five retrospective cohort studies and one prospective cohort study. There were different definitions of glaucoma progression across all of the studies, which meant that meta-analysis could not be performed. There were conflicting results across studies regarding association between statin use and progression. De and coauthors (De M, et al. IOVS 2006;47:ARVO E-Abstract 3398) defined progression as the average change in mean deviation of the visual field test per year. They found no statistically significant difference in the average change in mean deviation per year or pattern standard deviation per year between controls and users of statin for greater than 23 months. De Castro et al.27 defined OAG progression using various clinical parameters. They found no statistical difference among the number of patients who progressed to “outside normal limits” on glaucoma hemifield visual field test in the statin group compared to controls. However, they did find significant differences in the progression of multiple confocal scanning laser ophthalmoscopy parameters per year including rim volume, retinal nerve fiber layer cross-sectional area, and mean global retinal nerve fiber layer thickness, which favored the statins group when adjusted for multiple systemic and ocular factors. An abstract by Tong28 in 2008 found that univariate analysis of statin use was correlated with stable disease. However, descriptions of the study population, method of assessment, and adjustment for confounders were not reported. The study scored 0 on NOS. Iskedjian et al.29 used read code data for the addition of adjunctive medical therapy in those taking prostaglandin analogues for glaucoma as a surrogate marker for progression. They found that the proportion of patients initiating adjunctive medical therapy for glaucoma in the statin group was less than in those not taking any systemic medication, although this did not reach statistical significance. In a prospective cohort study of normal-tension glaucoma, Leung et al.30 found that the proportion of patients who took statins in the group that remained stable was significantly higher than the proportion of patients who took statins in the group who progressed. A logistic regression model adjusting for a history of disc hemorrhages, cerebrovascular disease, and age at baseline showed that simvastatin use conferred a significant protective effect against visual field progression. In a retrospective cohort study, Stein et al.24 used read code changes from “suspect OAG to OAG diagnosis” and “surgical treatment for OAG” as proxies for progression. Those who took statins for 1 or 2 years had decreased hazard of progressing to OAG from OAG suspect compared to those who did not receive statins (Table 4). However, hazard of an individual with OAG later requiring laser or incisional glaucoma surgery was not significantly reduced with statin exposure. 
Statin Use and IOP
The association between statin use and IOP was presented in three studies (Table 5). Leung et al.30 and Marcus et al.25 reported no significant changes in IOP associated with statin use. Khawaja et al.31 reported a significant reduction in IOP among statin users compared to non-statin users when adjusted for age and sex (β −0.31, 95%CI −0.51, −0.12 P = 0.002). However, when adjusted for beta-blocker therapy, the association was no longer significant. 
Discussion
To date this is the only systematic review that evaluates the association between statin use and glaucoma. Our search yielded no randomized controlled trials but 11 observational and case–control studies with sample sizes ranging from 76 to over 500,000 participants. Meta-analysis of the effect of short-term statin therapy on the incidence of glaucoma demonstrated a 4% reduced risk of glaucoma; however, long-term therapy did not demonstrate a statistically significant effect. Similarly, we did not find any significant association between statin use and incidence of glaucoma when outcomes were not stratified according to length of exposure to statin therapy. A previous meta-analysis by Macedo et al.19 evaluating the unintended effects of statins identified only three studies investigating the association with glaucoma and statin use, whereas we have identified a more complete set of evidence. Furthermore those authors did not report on short- versus long-term exposure to statin therapy. Macedo et al.19 found an overall pooled OR estimate of 0.86 (95%CI 0.69, 1.08). 
Read codes are a system by which diagnostic codes are allocated to patients within databases based on the clinical diagnosis as entered in the system, but not necessarily independently validated. The use of read code to classify glaucoma incidence and progression in several studies2224,26,29 poses the risk of misclassification bias. Caution must therefore be employed when interpreting these studies. By far the largest identified study was conducted by Stein et al.24 with a study population of over 500,000 individuals. The sample was identified by the individuals' hyperlipidemia status. Hence the generalizability of these results may be limited to the population with hyperlipidemia. As the largest study identified, the study by Stein et al.24 carries most weight; however, it is retrospective and uses read code data to define glaucoma, and therefore the quality of evidence from this study is relatively poor and the results need to be interpreted with caution. 
The use of nonstatin cholesterol-lowering drugs (NSCLDs), a possible confounding factor, was reported by McGwin et al.,22 Stein et al.,24 Marcus et al.,25 and Chen et al.26 McGwin et al.22 found that NSCLD use for less than 12 months was associated with reduced incidence of OAG (OR 0.38, 95%CI 0.18, 0.79), and Stein et al.24 found that persons who took NSCLD for 2 years had a 14% decreased risk of being prescribed a glaucoma medication (adjusted HR 0.862, 95%CI 0.785, 0.946). In contrast, Marcus et al.25 and Chen et al.26 did not demonstrate statistically significant protective effects of NSCLDs in glaucoma. In these studies NSCLDs were defined as a heterogeneous group of medications encompassing various classes of drugs. Certain classes of NSCLDs such as peroxisome proliferator–activated receptor alpha (PPARα) agonists (fibrates) have been shown to exhibit immunomodulatory pleiotropic effects independent of their lipid-lowering properties32 and have been shown to work synergistically with statins.3335 Statins may induce IOP lowering by increasing aqueous outflow.36 The confounding effect of systemic beta-blocker therapy on the effect of statins on IOP lowering was reported by Khawaja et al.31 They reported that the observed IOP-lowering effect of statins was no longer significant following adjustment for systemic beta-blocker therapy. Thus the confounding effects of NSCLDs and systemic beta-blockers should be considered in the design and analysis of future interventional studies. 
From our study we cannot rule out confounding by indication,37 and we must ask if it is the hyperlipidemia that might be protective or the statin use. A study by Newman-Casey et al.38 showed that hyperlipidemia was associated with a decreased risk in developing OAG; however, they could not determine whether it was the treatment for hyperlipidemia that reduced the risk or the hyperlipidemia itself. A study by Wang et al.39 showed that dyslipidemia was not significantly associated with the prevalence of glaucoma; however, they showed that dyslipidemia was associated with higher IOP and beta zone of parapapillary atrophy in a Chinese population. Chen et al.26 demonstrated that higher dosages of statins are associated with increased risk of OAG (OR 1.24, 95%CI 1.03, 1.49). They proposed that higher dosages of statins were an indication of poorer lipid control that was the cause of the increased risk of OAG. 
There were a number of strengths in this review. The sensitivity of our search strategy was maximized by restricting the exclusion criteria during the screening stage. However, the observational studies included are susceptible to various systematic biases depending on whether they are case–control or cohort designs. Case–control studies are generally prone to selection bias and require strict case definition to prevent misclassification bias. Cohort studies are considered methodologically superior to case–control studies; however, they are expensive and must be well conducted to prevent loss to follow-up. Cross-sectional studies are useful to estimate prevalence but are of limited value when investigating incidence. For each study we addressed the risk of bias using a range of tools recommended in The Cochrane Handbook of Systematic Reviews.18 In addition, the comprehensive approach adopted to ascertain confounding factors in each study added to the strength of the review. Potential confounding factors identified and controlled for in each study are outlined in Table 1
Weaknesses of the study include the exclusion of literature in languages other than English. To reach a wider audience, significant results tend to be published in English; therefore a degree of publication bias may be introduced by language restriction. Our investigation of publication bias did not reveal type 1 error in the results of studies investigating the short-term effects of statin use and incidence of glaucoma. A limitation in our study is that we had too few studies to investigate possible publication bias in studies investigating long-term statin exposure and those not stratified by length of exposure to statins. Although abstracts were identified and included, a formal search of gray literature databases was not performed, which may have contributed to publication bias. Another limitation in the reporting of our results is defining glaucoma as “commencing glaucoma medications” because some people may have ocular hypertension and not glaucoma. However, we addressed this by not including these estimates in the meta-analysis. 
In conclusion, the results of our meta-analysis provide evidence for the association between the short-term use of statin therapy and a reduced incidence of glaucoma. However, the observational design of the studies in the meta-analysis limits the ability to make inferences about whether or not exposure to statins causes reduced incidence of glaucoma. There was inconsistent evidence for the IOP-lowering effect of statins and the effect of statins on the progression of OAG. The associations observed in this review warrant a prospective interventional randomized controlled study with short- and long-term follow-up to provide further insight into the role of statin therapy in the prevention of onset or progression of glaucoma and its effects on IOP. 
Acknowledgments
Disclosure: P. McCann, None; R.E. Hogg, None; R. Fallis, None; A. Azuara-Blanco, None 
References
Tham YC, Li X, Wong TY, et al. Global prevalence of glaucoma and projections of glaucoma burden through 2040: a systematic review and meta-analysis. Ophthalmology. 2014; 121: 2081–2090.
Burr JM, Mowatt G, Hernandez R, et al. The clinical effectiveness and cost-effectiveness of screening for open angle glaucoma: a systematic review and economic evaluation. Health Technol Assess. 2007; 11: iii–iv, ix–x, 1–190.
Rudnicka AR, Mt-Isa S, Owen CG, et al. Variations in primary open-angle glaucoma prevalence by age, gender, and race: a bayesian meta-analysis. Invest Ophthalmol Vis Sci. 2006; 47: 4254–4261.
Miglior S, Bertuzzi F. Relationship between intraocular pressure and glaucoma onset and progression. Curr Opin Pharmacol. 2013; 13: 32–35.
Chang EE, Goldberg JL. Glaucoma 2.0: neuroprotection, neuroregeneration, neuroenhancement. Ophthalmology. 2012; 119: 979–986.
Kola I, Landis J. Can the pharmaceutical industry reduce attrition rates? Nat Rev Drug Discov. 2004; 3: 711–715.
Bradley D. Why big pharma needs to learn the three ‘R's. Nat Rev Drug Discov. 2005; 4: 446.
Pritchard JF, Jurima-Romet M, Reimer ML, et al. Making better drugs: decision gates in non-clinical drug development. Nat Rev Drug Discov. 2003; 2: 542–553.
van der Most PJ, Dolga AM, Nijholt IM, et al. Statins: mechanisms of neuroprotection. Prog Neurobiol. 2009; 88: 64–75.
Downs JR, Clearfield M, Weis S, et al. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/TexCAPS. Air Force/Texas Coronary Atherosclerosis Prevention Study. JAMA. 1998; 279: 1615–1622.
Sillesen H, Amarenco P, Hennerici MG, et al. Atorvastatin reduces the risk of cardiovascular events in patients with carotid atherosclerosis: a secondary analysis of the stroke prevention by aggressive reduction in cholesterol levels (SPARCL) trial. Stroke. 2008; 39: 3297–3302.
Rikitake Y, Liao JK. Rho GTPases, statins, and nitric oxide. Circ Res. 2005; 97: 1232–1235.
Vohra R, Tsai JC, Kolko M. The role of inflammation in the pathogenesis of glaucoma. Surv Ophthalmol. 2013; 58: 311–320.
Schmeer C, Kretz A, Isenmann S. Therapeutic potential of 3-hydroxy-3-methylglutaryl coenzyme a reductase inhibitors for the treatment of retinal and eye diseases. CNS Neurol Disord Drug Targets. 2007; 6: 282–287.
Pokrovskaya O, Wallace D, O' Brien C. The emerging role of statins in glaucoma pathological mechanisms and therapeutics. Open J Ophthalmol. 2014; 4: 124–138.
Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000; 283: 2008–2012.
Panic N, Leoncini E, de Belvis G, et al. Evaluation of the endorsement of the preferred reporting items for systematic reviews and meta-analysis (PRISMA) statement on the quality of published systematic review and meta-analyses. PLoS One. 2013; 8: e83138.
Reeves BC, Deeks JJ, Higgins JPT, Wells GA. Including non-randomised studies. In: Higgins JPT, Green S, eds. The Cochrane Handbook of Systematic Reviews of Interventions. The Cochrane Collaboration; 2008: 391–432.
Macedo AF, Taylor FC, Casas JP, et al. Unintended effects of statins from observational studies in the general population: systematic review and meta-analysis. BMC Med. 2014; 12: 51.
Vandenbroucke JP, Pearce N. Case-control studies: basic concepts. Int J Epidemiol. 2012; 41: 1480–1489.
Magnus M. Applied Analytic Approaches. Intermediate Epidemiology Methods That Matter. Burlington MA: Jones and Bartlett Learning; 2016: 176.
McGwin G,Jr, McNeal S, Owsley C, et al. Statins and other cholesterol-lowering medications and the presence of glaucoma. Arch Ophthalmol. 2004; 122: 822–826.
Owen CG, Carey IM, Shah S, et al. Hypotensive medication, statins, and the risk of glaucoma. Invest Ophthalmol Vis Sci. 2010; 51: 3524–3530.
Stein JD, Newman-Casey PA, Talwar N, et al. The relationship between statin use and open-angle glaucoma. Ophthalmology. 2012; 119: 2074–2081.
Marcus MW, Muskens RP, Ramdas WD, et al. Cholesterol-lowering drugs and incident open-angle glaucoma: a population-based cohort study. PLoS One. 2012; 7: e29724.
Chen HY, Hsu SY, Chang YC, et al. Association between statin use and open-angle glaucoma in hyperlipidaemia patients: a Taiwanese population-based case-control study. Medicine. 2015; 94: e2018.
De Castro DK, Punjabi OS, Bostrom AG, et al. Effect of statin drugs and aspirin on progression in open-angle glaucoma suspects using confocal scanning laser ophthalmoscopy. Clin Experiment Ophthalmol. 2007; 35: 506–513.
Tong MK. Correlation of statins and disease progression in patients with normal tension glaucoma. Clin Experiment Ophthalmol. 2008; 36 (supp 1): A264.
Iskedjian M, Walker JH, Desjardins O, et al. Effect of selected antihypertensives, antidiabetics, statins and diuretics on adjunctive medical treatment of glaucoma: a population based study. Curr Med Res Opin. 2009; 25: 1879–1888.
Leung DY, Li FC, Kwong YY, et al. Simvastatin and disease stabilization in normal tension glaucoma: a cohort study. Ophthalmology. 2010; 117: 471–476.
Khawaja AP, Chan MP, Broadway DC, et al. Systemic medication and intraocular pressure in a British population: the EPIC-Norfolk Eye Study. Ophthalmology. 2014; 121: 1501–1507.
Chen Y, Hu Y, Lin M, et al. Therapeutic effects of PPARalpha agonists on diabetic retinopathy in type 1 diabetes models. Diabetes. 2013; 62: 261–272.
ACCORD Study Group, ACCORD Eye Study Group; Chew EY, et al. Effects of medical therapies on retinopathy progression in type 2 diabetes. N Engl J Med. 2010; 363: 233–244.
Inoue I, Itoh F, Aoyagi S, et al. Fibrate and statin synergistically increase the transcriptional activities of PPARalpha/RXRalpha and decrease the transactivation of NFkappaB. Biochem Biophys Res Commun. 2002; 290: 131–139.
Martin G, Duez H, Blanquart C, et al. Statin-induced inhibition of the rho-signaling pathway activates PPARalpha and induces HDL apoA-I. J Clin Invest. 2001; 107: 1423–1432.
Song J, Deng PF, Stinnett SS, et al. Effects of cholesterol-lowering statins on the aqueous humor outflow pathway. Invest Ophthalmol Vis Sci. 2005; 46: 2424–2432.
Salas M, Hofman A, Stricker BHC. Confounding by indication: an example of variation in the use of epidemiologic terminology. Am J Epidemiol. 1999; 149: 981–983.
Newman-Casey PA, Talwar N, Nan B, Musch DC, Stein JD. The relationship between components of metabolic syndrome and open-angle glaucoma. Ophthalmology. 2011; 118: 1318–1326.
Wang S, Xu L, Jonas JB, Wang YX, You QS, Dyslipidemia Yang H. and eye diseases in the adult Chinese population: the Beijing eye study. PLoS One. 2012; 7: e26871.
Appendix 1
Table A1
 
Search Strategies
Table A1
 
Search Strategies
Appendix 2 Newcastle-Ottawa Quality Assessment Scales
Case–Control Studies
Note: A study can be awarded a maximum of one star (*) for each numbered item within the Selection and Exposure categories. A maximum of two stars can be given for Comparability. 
Selection
  •  
    (1) Is the case definition adequate?
  •  
    •  
      (a) Yes, with independent validation*
    •  
      (b) Yes, for example, record linkage or based on self-reports
    •  
      (c) No description
  •  
    (2) Representativeness of the cases
  •  
    •  
      (a) Consecutive or obviously representative series of cases*
    •  
      (b) Potential for selection biases or not stated
  •  
    (3) Selection of controls
  •  
    •  
      (a) Community controls*
    •  
      (b) Hospital controls
    •  
      (c) No description
  •  
    (4) Definition of controls
  •  
    •  
      (a) No history of disease (endpoint)*
    •  
      (b) No description of source
Comparability
  •  
    (1) Comparability of cases and controls on the basis of the design or analysis
  •  
    •  
      (a) Study controls for _______________ (select the most important factor)*
    •  
      (b) Study controls for any additional factor* (this criterion could be modified to indicate specific control for a second important factor)
Exposure
  •  
    (1) Ascertainment of exposure
  •  
    •  
      (a) Secure record (e.g., surgical records)*
    •  
      (b) Structured interview where blind to case/control status*
    •  
      (c) Interview not blinded to case/control status
    •  
      (d) Written self-report or medical record only
    •  
      (e) No description
  •  
    (2) Same method of ascertainment for cases and controls
  •  
    •  
      (a) Yes*
    •  
      (b) No
  •  
    (3) Nonresponse rate
  •  
    •  
      (a) Same rate for both groups*
    •  
      (b) Nonrespondents described
    •  
      (c) Rate different and no designation
Cohort Studies
Note: A study can be awarded a maximum of one star for each numbered item within the Selection and Outcome categories. A maximum of two stars can be given for Comparability. 
Selection
  •  
    (1) Representativeness of the exposed cohort
  •  
    •  
      (a) Truly representative of the average _______________ (describe) in the community*
    •  
      (b) Somewhat representative of the average ______________ in the community*
    •  
      (c) Selected group of users, for example, nurses, volunteers
    •  
      (d) No description of the derivation of the cohort
  •  
    (2) Selection of the nonexposed cohort
  •  
    •  
      (a) Drawn from the same community as the exposed cohort*
    •  
      (b) Drawn from a different source
    •  
      (c) No description of the derivation of the nonexposed cohort
  •  
    (3) Ascertainment of exposure
  •  
    •  
      (a) Secure record (e.g., surgical records)*
    •  
      (b) Structured interview*
    •  
      (c) Written self-report
    •  
      (d) No description
  •  
    (4) Demonstration that outcome of interest was not present at start of study
  •  
    •  
      (a) Yes*
    •  
      (b) No
Comparability
  •  
    (1) Comparability of cohorts on the basis of the design or analysis
  •  
    •  
      (a) Study controls for _____________ (select the most important factor)*
    •  
      (b) Study controls for any additional factor* (this criterion could be modified to indicate specific control for a second important factor)
Outcome
  •  
    (1) Assessment of outcome
  •  
    •  
      (a) Independent blind assessment*
    •  
      (b) Record linkage*
    •  
      (c) Self-report
    •  
      (d) No description
  •  
    (2) Was follow-up long enough for outcomes to occur?
  •  
    •  
      (a) Yes (select an adequate follow-up period for outcome of interest)*
    •  
      (b) No
  •  
    (3) Adequacy of follow-up of cohorts
  •  
    •  
      (a) Complete follow-up—all subjects accounted for*
    •  
      (b) Subjects lost to follow-up unlikely to introduce bias—small number lost: ≤20%, or description provided of those lost)*
    •  
      (c) Follow-up rate <80% and no description of those lost
    •  
      (d) No statement
Figure 1
 
Search strategy flow diagram.
Figure 1
 
Search strategy flow diagram.
Figure 2
 
Forest plot of incidence of glaucoma and statin use ≤2 years versus controls. McGwin et al.,22 refers to exposure for <12 months (top) and 12 to 23 months (second from top). McGwin et al.,22 (second from top): Upper limit of 95%CI (1.38) is not exactly equivalent to upper limit of 95%CI in Table 3 (1.39) due to rounding in meta-analysis software.
Figure 2
 
Forest plot of incidence of glaucoma and statin use ≤2 years versus controls. McGwin et al.,22 refers to exposure for <12 months (top) and 12 to 23 months (second from top). McGwin et al.,22 (second from top): Upper limit of 95%CI (1.38) is not exactly equivalent to upper limit of 95%CI in Table 3 (1.39) due to rounding in meta-analysis software.
Figure 3
 
Forest plot of incidence of glaucoma and statin use >2 years versus control. Marcus et al.,25 upper limit of 95%CI (0.92) is not exactly equivalent to upper limit of 95%CI in Table 3 (0.94) due to rounding in meta-analysis software.
Figure 3
 
Forest plot of incidence of glaucoma and statin use >2 years versus control. Marcus et al.,25 upper limit of 95%CI (0.92) is not exactly equivalent to upper limit of 95%CI in Table 3 (0.94) due to rounding in meta-analysis software.
Figure 4
 
Forest plot of incidence of glaucoma and statin use from outcomes not stratified by length of exposure. Marcus et al.,25 upper limit of 95%CI (0.94) is not exactly equivalent to upper limit of 95%CI in Table 3 (0.96) due to rounding in meta-analysis software. Owen et al.,30 upper limit of 95%CI (1.07) is not exactly equivalent to upper limit of 95%CI in Table 3 (1.06) due to rounding in meta-analysis software.
Figure 4
 
Forest plot of incidence of glaucoma and statin use from outcomes not stratified by length of exposure. Marcus et al.,25 upper limit of 95%CI (0.94) is not exactly equivalent to upper limit of 95%CI in Table 3 (0.96) due to rounding in meta-analysis software. Owen et al.,30 upper limit of 95%CI (1.07) is not exactly equivalent to upper limit of 95%CI in Table 3 (1.06) due to rounding in meta-analysis software.
Figure 5
 
Funnel plot examining publication bias investigating short-term (≤2 years) statin use and incidence of glaucoma.
Figure 5
 
Funnel plot examining publication bias investigating short-term (≤2 years) statin use and incidence of glaucoma.
Table 1
 
List of Features of All Included Studies
Table 1
 
List of Features of All Included Studies
Table 1
 
Extended
Table 1
 
Extended
Table 1
 
Extended
Table 1
 
Extended
Table 2
 
List of Study Design Features
Table 2
 
List of Study Design Features
Table 3
 
Features and Results of Studies Investigating Association Between Statin Use and Glaucoma Incidence
Table 3
 
Features and Results of Studies Investigating Association Between Statin Use and Glaucoma Incidence
Table 4
 
Features and Results of Studies Investigating Association Between Statin Use and Glaucoma Progression
Table 4
 
Features and Results of Studies Investigating Association Between Statin Use and Glaucoma Progression
Table 5
 
Features and Results of Studies Investigating Association Between Statin Use and IOP
Table 5
 
Features and Results of Studies Investigating Association Between Statin Use and IOP
Table 6
 
Newcastle-Ottawa Scale: Cohort Studies
Table 6
 
Newcastle-Ottawa Scale: Cohort Studies
Table 7
 
Newcastle-Ottawa Scale: Case–Control Studies
Table 7
 
Newcastle-Ottawa Scale: Case–Control Studies
Table A1
 
Search Strategies
Table A1
 
Search Strategies
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×