This article is more than 5 years old. Information may no longer be current.
Statins do not reduce pneumonia risk; studies reflect 'healthy-user' bias
Click Here to Manage Email Alerts
New data suggest that previously reported protective effects of statins against pneumonia may reflect a healthy-user bias, according to researchers from the University of Iowa.
“Our results demonstrate a statistically significant protective effect for statins against pneumonia using traditional models, consistent with other findings,” the researchers wrote in Clinical Infectious Diseases. “We also show that these modest observed protective effects disappear with an instrumental variables approach. Thus, we are able to reconcile the positive and negative published studies and can conclude that the positive effects of statins [against pneumonia] are most likely due to the fact that the people who were treated with statins differ from those who were not.”
Linnea A. Polgreen, PhD, of the department of pharmacy practice and science at the University of Iowa, and colleagues, used Medicare data to build a cohort of patients hospitalized for acute MI in 2007 and 2008 who survived for at least 30 days and did not have acute myocardial infarction in the previous 12 months. They determined the patients who were treated with statins by identifying those who filled a statin prescription within 30 days after discharge. After identifying patients with a pneumonia diagnosis, they modeled the effect statin treatment had on pneumonia. One model used an instrumental variable approach to control for nonrandom treatment assignment.
There were 124,695 Medicare beneficiaries in the study: 15.3% developed pneumonia during the observation period, and 61.9% filled a statin prescription. Statin users were less likely to have a stroke, heart failure, chronic kidney disease or depression than nonstatin users, and they also were younger. Findings from an ordinary least squares model and a logit model suggest statins were associated with a 1.5% reduction in the probability of pneumonia for the full sample of patients and the 1-year survivors. In the instrumental variable approach, however, none of the statin treatment variables demonstrated a protective effect against pneumonia.
“Although we focused on pneumonia, it is possible that previously reported positive effects for a wide range of infections in observational studies are also due to a healthy user bias rather than the biologic effects of statins,” the researchers wrote. “We think that the broad range of positive outcomes provide further reasons for skepticism that the positive outcomes associated with statin use are due to pleotropic effects. Instead, it seems likely that the positive outcomes arise from statins being prescribed preferentially to healthier patients.” – by Emily Shafer
Disclosure: The researchers report no relevant financial disclosures.
Click Here to Manage Email Alerts