02.15.2016
German claims data show association; stronger in men than women

by Kristina Fiore
Associate Editor, MedPage Today

Proton pump inhibitors (PPIs) were associated with an increased risk of dementia, and the relationship was stronger in men than women, according to a study of German medical claims data.

In an assessment of data from the largest German statutory health insurer, taking PPIs was associated with a significantly increased risk of incident dementia compared with not taking them (hazard ratio 1.44, 95% CI 1.36 to 1.52 (P<0.001), according to Britta Haenisch, PhD, of the German Center for Neurodegenerative Diseases in Bonn, and colleagues.

That association was slightly more pronounced in men than women (HR 1.52 versus HR 1.42), though both were statistically significant, they reported online in JAMA Neurology. The hazard ratios reflected adjustments for age, sex, polypharmacy, and comorbidities as contained in medical records.

Although the findings suggest that avoiding PPIs might prevent the development of dementia, Haenisch and colleagues noted that randomized prospective clinical trials are needed before making such a recommendation.

There’s been some evidence that PPIs, used to treat gastroesophageal reflux disease and peptic ulcers, may be linked to cognitive decline — and there is biological plausibility for the relationship, as some research has shown that PPIs can cross the blood-brain barrier and are active in the central nervous system.

In particular, animal evidence has shown that PPIs can increase both the production and degradation of amyloid in the brain, and that they can bind to tau protein. There’s also human evidence that the drugs can reduce circulating levels of B12 and other nutrients, which could promote development of dementia.

Since the use of PPIs has increased tremendously, especially among the elderly, the researchers said, they wanted to investigate whether these medications could have an impact on cognition.

They conducted a prospective cohort study using observational data from 2004 to 2011 from Germany’s biggest health insurer, Allgemeine Ortskrankenkassen, covering diagnoses and prescriptions. They focused on prescriptions for PPIs including omeprazole, pantoprazole, lansoprazole, esomeprazole, and rabeprazole.

Of the 73,679 participants age 75 and up who were free of dementia at baseline, 29,510 developed dementia over the study period, and 2,950 were receiving a regular PPI medication (mean age 84, 78% female).

In addition to finding a significantly increased risk of incident dementia among those taking a PPI — and seeing that this relationship was slightly more pronounced in men — Haenisch and colleagues found that there was a similar risk of dementia across the three most commonly used PPIs (omeprazole, pantoprazole, and esomeprazole), with a slightly more pronounced risk for those on esomeprazole (HRs 1.51, 1.58, and 2.12, respectively).

The findings are in line with recent pharmacoepidemiological analyses of primary data and with mouse models in which use of PPIs increased levels of amyloid-beta in the brains of mice, the researchers said. They were also similar to the dementia risk seen in aprevious study involving the longitudinal AgeCoDe cohort (HR 1.38, 95% CI 1.04 to 1.83).

Avoiding PPIs may prevent dementia, they concluded, with the caveat that randomized prospective trials are first needed to confirm these findings.

It’s possible, however, that the association is not causal, wrote Lewis Kuller, MD, DrPH, inan accompanying editorial. Older patients often take many drugs, which may be a function of the extent of their disease and comorbidities, and there may be other similarities such as education which are linked with both conditions and may confound the results, he noted.

Still, Kuller said the researchers “have provided an important and interesting challenge to evaluate the possible association of the use of PPIs and the risk of dementia.”

“This is a very important issue given the very high prevalence of pharmacological drugs’ long-term use in elderly populations that have a very high risk of dementia,” Kuller wrote.

But he noted that “further determinants of whether PPIs are causal for dementia requires validation in large cohorts and probably well-designed case-control studies with good measures of PPI use (especially long-term use), covariates, and especially methods to measure incidence of dementia.”

Haenisch and colleagues noted that their claims-based analysis could not exclude all potential confounders, and their data lacked specificity on specific dementia types and causes.

• Reviewed by Henry A. Solomon, MD, FACP, FACC Clinical Associate Professor, Weill Cornell Medical College and Dorothy Caputo, MA, BSN, RN, Nurse Planner

LAST UPDATED 02.15.2016