Statin use and the risk of cataracts: A systematic review and meta-analysis

Methodological quality of the review: High confidence

Author: Yu S, Chu Y, Li G, Zhang Q, Wu L.

Region: United States of America, Europe, Asia and North America.

Sector: Cataract

Sub-sector: Prevention

Type of cataract: Age-related cataract

Equity focus: None specified

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background:

Cataracts are the main cause of poor vision and blindness worldwide. The effects of statin administration on cataracts remain debated, as findings from different studies conducted reported mixed findings. Some studies have reported no association between statin use and cataracts, whereas others have found that statin use is protective against the incidence of cataracts, or that it is associated with an increased risk of cataracts.

Research objectives:

Investigate the association between statin use and cataracts.

Main findings:

In total, authors included 17 studies consisting of 12 observational studies and five randomized-controlled trials (RCTs) in the review. In the meta-analysis, authors included five RCTs and six case-control studies. Among the cohort studies, three were conducted in North America, and the remaining three were performed in Europe, Asia and Australia. Among the case-control studies, four were conducted in North America and two were conducted in Europe. Based on the methodological quality assessment, the means score of the cohort studies included in the analysis was 7. Four studies were of high quality (Newcastle-Ottawa Scale (NOS) ≥7), and 3 studies were of low quality (NOS <7). The mean score of the 6 case–control studies was 6.5. Three studies were of high quality (NOS ≥7), and 3 studies were of low quality (NOS <7).

Of the RCTs included in the review, two were conducted in the USA and three in Europe. The mean score of the RCTs was 5.4.

The pooled Risk Ratio (RR) of the cohort studies showed that the use of statins was associated with a 13% increase in cataract incidence or cataract surgery – 1.13 (95% confidence interval (CI), 1.01–1.25), with significant heterogeneity (I2=90.5%). The pooled RRs for the case-control studies and RCTs showed that the use statins was not associated with cataract incidence or cataract surgery -1.10 (95% CI, 0.99–1.23) and 0.89 (95% CI, 0.72–1.10), respectively. The I2 showed significant heterogeneity among the case-control studies (95.5%) and low heterogeneity among the RCTs (30.5%). The sensitivity analysis showed that sequential omission of individual studies did not alter the overall effect.

In the subgroup analysis of cohort studies, there were significant associations in the subgroups of high methodological quality, outcome assessment, cataract, no older than 60 and less than 5 years follow-up duration. In the subgroup analysis of case–control studies, significant associations were observed in the subgroups of atorvastatin, lovastatin, high methodological quality, cataract surgery, CVD included model, smoking missing model, consultation rate missing model, and hypertension included model.  Authors report observing no associations in the fluvastatin, rosuvastatin, pravastatin, simvastatin, low methodological quality, studies performed in North America and Europe.

Methodology:

Study eligibility criteria for inclusion in the review consisted of: (1) the study was a case–control, cohort study, or randomized controlled trial (RCT); (2) non–statin users were included in the comparison group; (3) cataracts and/or cataract surgery was an outcome; (4) the association between statin use and the risk of cataracts/cataract surgery was investigated; (5) risk estimates of morbidity and 95% CIs were reported or the information required to calculate them was available. Basic science studies, reviews, editorials/letters, case reports, and studies without comparison groups were excluded.

Authors conducted a search on Cochrane Library, PubMed and EMBASE databases from January 1980 to January 2016 for English language publications. The search was performed using the following terms: “statins OR HMG-CoA reductase inhibitors OR Simvastatin OR Lovastatin OR Fluvastatin OR Pravastatin OR Rosuvastatin OR Atorvastatin” AND “cataract.” Authors also manually searched for relevant articles from the reference lists of the retrieved articles. When the available information was incomplete, authors attempted to contact the study investigators for additional information.

Study selection, data extraction and critical appraisal of included studies were conducted independently by two reviewers. Authors extracted data on country of origin of the population studied, patient characteristics, statin use, information source for exposure ascertainment, risk estimates and corresponding 95% CIs, and covariates adjusted for in the multivariable analysis. Authors assessed the methodological quality of the included studies based on the Newcastle-Ottawa Scale (NOS) for observational studies, which was developed to assess the quality of nonrandomized studies in meta-analysis. Using this scale, observational studies were scored across 3 categories as follows: selection (4 questions) and comparability (2 questions) of the study group and ascertainment of the outcome of interest (3 questions), with all questions having a score of 1 except for the comparability of study groups, for which separate points were awarded for controlling for age and/or sex (maximum, 2 points). A score of ≥7 points was suggestive of a high-quality study. Randomized controlled trials were assessed using the Cochrane risk of bias assessment.

Authors conducted a meta-analysis using STATA software. Heterogeneity was assessed using the Cochrane Q X2 test and the I2 statistic. Adjusted effect estimates (odds ratios, relative risks (RRs), and hazard risks) between the outcome and use of statins were extracted. In the presence of heterogeneity, authors used a random-effects model because its assumptions account for the presence of variability among studies. The association between statin use and cataract or cataract surgery risk was estimated using the RRs and corresponding 95% CIs. Because the outcomes were relatively uncommon and the odds ratios in the case–control studies were close to 1, odds ratios were considered approximations of RR.

Applicability/external validity:

Authors note that the RCTs had good internal validity, but external validity was limited indicating that the conclusion may not be applicable to the whole population but only to a population similar to those include in the RCTs. Authors also note that the observational studies involved population with different health conditions, with baseline confounders which may affect the results.  Therefore, authors state in the review that this may have led to a slightly higher effect size then it should be.

Geographic focus:

Studies included in the review were conducted in Europe, North America and USA. Authors report no association between statin use and cataract incidence for studies conducted in North America and Europe.

Quality assessment:

Authors used rigorous methods to conduct this review. Authors conducted a rigorous search to ensure that all relevant studies were included in the review. In addition, authors used appropriate methods to scree studies for inclusion and extract and appraise included studies avoiding biases.