Taking statins was associated with a higher incidence of cataract diagnosis in a propensity score–matched cohort and in other analyses of a military health care database in Texas. The report was published online Sept. 19 in JAMA Ophthalmology.
This is yet another observational study to show a possible association between statin use and cataracts – this time a negative one – continuing the ongoing controversy in this area of research.
Investigators said that their propensity score–matched analysis was the first study of its kind, and also one of the largest, including 45,000 patients who were followed up longitudinally within the same health care system, with similar health care coverage and access to care and medication.
The study was extremely well done, Dr. Kim Allan Williams Sr., chair of cardiology at Wayne State University, Detroit, said in an interview. "This is one of the larger studies, and the thing that distinguishes this one is how well the propensity matching was done. Dr. Williams, who was not involved in the study, added that the study wasn’t a definite and final answer to the question, and didn’t negate other studies that showed a protective or no association between statins and cataracts.
Researchers conducted the analysis on all adult patients who were enrolled in the San Antonio Military Multi-Market Area as Tricare Prime or Plus, analyzing the data from October 1, 2003, to March 1, 2010. All patients were enrolled in the system throughout the study.
Dr. Kim Allan Williams Sr.
With a computer model, they selected a group of statin users and nonusers who looked very comparable to each other.
"One of the biggest caveats in observational studies is that, patients’ baseline characteristics may be different," Dr. Ishak Mansi, the study’s principal investigator, said in an interview." And when you look after 6 or 7 years, you may have higher incidence of cataract in some, not because of the statins, but because some of the patients who were given statins were sicker from the very beginning. So we have to make sure we’re comparing apples to apples."
They divided the participants according to statin use, using their pharmacy records: There were 13,626 statin users, who had received at least one 90-day supply of a statin; and 32,623 nonusers, who had never received a statin throughout the study.
To conduct a propensity score–matched cohort analysis, researchers used 44 variables that were expected to increase the likelihood of receiving a statin prescription, as well as increasing the risk for cataract. Those variables included age, sex, 17 comorbid conditions, obesity, alcohol dependence/abuse, use of 14 medication groups, and more.
As for statins, simvastatin made up 73% of the prescriptions, atorvastatin 17.4%, pravastatin 7%, rosuvastatin 1.7%, and fluvastatin or lovastatin 0.24% of the prescription. Nearly 34% of the statin users received maximal doses of statins.
For the primary analysis, researchers matched 6,972 pairs of statin users and nonusers, with no significant differences in their baseline characteristics. Among the statin users, cataracts occurred in 2,477 (35.5%), compared with 2,337 (33.5%) of the nonusers, yielding a statistically significant odds ratio of 1.09 (JAMA Ophthalmol. 2013 Sept. 19 [doi:10.1001/jamaophthalmol.2013.4575]).
"Then we said, let us assume that the computer model may not have been good in selecting patients, and let’s see if we can get the same result by slicing the data differently," said Dr. Mansi, professor of medicine at UT Southwestern, Dallas.
The secondary analysis included 6,113 statin users and 27,400 nonusers with no prespecified comorbidities. Among them, 33.7% of statin users and 9.4% of nonusers developed cataracts, a significant 20% increased risk, the researchers reported.
Further analysis showed that statin use was an independent predictor of cataract (adjusted OR, 1.43), and statin use continued to be an independent predictor of cataract, when researchers repeated the analysis using backward stepwise elimination (adjusted OR, 1.42).
They also examined the relationship between cataract and LDL cholesterol and HDL cholesterol, and found that the mean LDL cholesterol level was inversely related to risk for cataract (adjusted OR, 0.997; P = .009), but mean HDL cholesterol was not (adjusted OR, 1.002; P = .16).
They also analyzed the data according to years of statin use (2, 4, and 6 years). "Whatever approach we used, the result was consistent," said Dr. Mansi, a staff internist at VA North Texas, Dallas.
Dr. Williams, vice president of the American College of Cardiology, said that the study puts the issue "into real question again, and we find ourselves wanting a randomized controlled prospective trial, which would be very difficult to do. These are drugs that you just don’t randomize people to. If we’re stuck with observational trials, no matter how large or how well done they are, there will always be questions."