A higher risk of death is associated with long-term use of popular stomach acid reducers known as proton pump inhibitors, according to a new study.

These drugs are sold under brand names such as Prilosec, Nexium, Protonix, Zegerid, Aciphex and Dexilant, along with generic versions such as omeprazole, lansoprazole and pantoprazole. Originally available only by prescription, they are increasingly offered over the counter. The study looked at prescription use only.

Each year, an estimated 15 million Americans are prescribed proton pump inhibitors, commonly called PPIs. That figure doesn't include over-the-counter sales.

Previous studies have indicated that long-term use of PPIs is associated with elevated risks for heart disease, fractures of various bones and other medical problems. The new analysis goes further by linking these medications to higher death rates.

Prolonged use of PPIs was associated with a 25 percent greater risk of death, compared with people taking H2 blockers, another class of acid reducers. H2 blockers are sold under brand names including Pepcid, Tagamet and Zantac, as well as generic names such as famotidine, cimetidine and ranitidine.

For the study, researchers examined millions of military veterans' prescription records that spanned an average of nearly six years. The findings were published Monday in the journal BMJ Open; the study can be found at j.mp/acidppi. The senior author was Dr. Ziyad Al-Aly of the U.S. Department of Veterans Affairs' health system in St. Louis.

Acid reducers treat painful ailments including GERD (gastroesophageal reflux disease), heartburn and peptic ulcers. Physicians should prescribe PPIs when there's a legitimate reason, but only long enough to provide benefits that outweigh the risks, the study's authors said. The increased mortality associated with PPIs was proportionately connected to their duration of use.

Some makers of proton pump inhibitors said this week that their products are effective and safe when used for the recommended period, which is typically about two weeks per full course of the over-the-counter versions. They didn't specifically address the issue of patients often taking prescription-level PPIs for much longer periods and at higher doses.

For the main analysis, VA records of 349,312 patients were analyzed. Of those people, 275,977 were prescribed PPIs and 73,335 were prescribed H2 blockers. Between these two groups, the patients who took PPIs showed the 25 percent higher death rate.

Two secondary comparisons were also made. One found a 15 percent increased death rate for PPI users compared with patients who didn't use PPIs but may have taken an acid suppressor other than H2 blockers. Records from 3,288,092 patients were examined for that comparison.

In the other comparison, the death rate was 23 percent higher among PPI users compared with people who didn't use any acid suppressor. A total of 2,887,070 patient records were examined.

The study's comparisons were observational, so the conclusions aren't as definitive as a randomized, placebo-controlled clinical trial. Another limitation is that the VA patients were "mostly older, white male U.S. veterans," the report's authors said, so the results may not apply directly to a larger population. Also, the study didn't get information on the causes of death for the targeted patient population.

However, the newly published findings are consistent with previous studies that showed long-term use of PPIs, but not H2 blockers, is associated with higher rates of disease. These include kidney disease, dementia and infection by the antibiotic-resistant superbug C. difficile.

Proton pump inhibitors work by a different mechanism than H2 blockers, explaining the differing responses. PPIs block stomach cells called parietal cells from releasing positively charged hydrogen atoms, or protons, into the stomach. This inhibits production of stomach acid, chemically known as hydrochloric acid.

H2 blockers stop the action of histamine, which stimulates parietal cells to produce hydrochloric acid. This indirect method is less efficient than that of PPIs, but is sufficient for many patients.

A study published in the journal JAMA Neurology last year found a 44 percent increased risk of dementia among people using PPIs.

Proton pump inhibitors can cause nutritional deficiencies, because acid is needed to release essential nutrients such as vitamin B12. Since stomach acid kills pathogenic bacteria, use of acid reducers also has been linked to increased rates of food poisoning.

And a 2015 study led by Stanford University researchers found that use of PPIs, but not H2 blockers, is associated with higher risk of heart attacks.