People on cholesterol-lowering statins may have a somewhat smaller risk of dying from pneumonia - but the link may be weaker than some studies have suggested, a new analysis finds.

A number of studies have found that statin users seem to have lower odds of developing or dying from pneumonia. But no one knows if the cholesterol drugs themselves should get the credit.

For the new analysis, researchers pulled together 13 past studies on the question.

When they combined all the results, they found that overall, statin users were one-third less likely than non-users to die of pneumonia.

But the connection got progressively weaker as other factors - like pneumonia severity and vaccinations - were taken into account.

"This shows us that there seems to be some effect at the 10,000-foot level, but when you get down on the ground and look deeper, you see a weakening in the supposed benefit," said lead researcher Dr. Vineet Chopra, of the University of Michigan Health System in Ann Arbor.

The findings, reported in the American Journal of Medicine, highlight the limits of "observational" studies.

All but one of the studies Chopra's team found were observational - meaning the researchers compared statin users with non-users, using sources like medical records.

Those studies can't prove cause-and-effect because people on statins may simply be healthier than non-users or have better access to preventive healthcare, like vaccinations.

"We're trying to understand whether this is a property of statins, or something about statin users," Chopra said in an interview.

For now, he said, "don't stop your statin if you're on one. But don't expect it to protect you from pneumonia."

WORTH ANOTHER STUDY?

There is cause for further study, though, according to Chopra. "The only way to move forward is to do a randomized controlled clinical trial," he said.

That would mean randomly assigning patients hospitalized for pneumonia to either receive a statin or have standard treatment only.

Only one of the studies in this analysis was a clinical trial: It involved 67 pneumonia patients who had developed sepsis, a life-threatening condition that usually results from an overwhelming immune response to an infection.

Statin treatment did nothing to improve their survival odds.

As for the dozen other studies, there was a lower risk of pneumonia death when all of the results were combined.

But when Chopra's team took a closer look, the potential benefit faded. In studies that accounted for pneumonia severity, for example, there was no significant difference in statin users' and non-users' risk of dying.

The same was true in studies that used "propensity scores" - which tried to estimate people's likelihood of receiving a statin based on their personal health.

A study released earlier this year found that people on statins did have a slightly lower risk of developing pneumonia. And those findings came from a clinical trial -- though the trial was not set up to specifically study the effects of statins on pneumonia risk.

But it's still not clear whether statins, themselves, have any effect on pneumonia risk, or the odds of dying from the disease.

And why would they? No one is sure.

Pneumonia is an inflammation of the lungs that usually arises after a bacterial infection or a viral infection, like the flu. Statins have mild effects on inflammation, immune system activity and cell death, and researchers have speculated that those effects could lessen the blow of pneumonia.

There is a need for new pneumonia treatments, according to Chopra.

In the U.S., the lung disease sends more than 1 million people to the hospital and kills nearly 51,000 each year - with the elderly and people with chronic health conditions, like heart or lung disease, at greatest risk.

And worldwide, pneumonia is the number-one killer of children younger than 5, according to the World Health Organization.

Chopra said he thinks it would be worthwhile to do a clinical trial testing statins as a pneumonia treatment. Even if there is just a small reduction in death risk, he said, that could save a lot of lives on the global scale.

The study was published in American Journal of Medicine.