Popular Heartburn Medication Linked to Increased Stroke

November 29, 2016

Proton pump inhibitors (PPIs) — used widely to reduce stomach acid and treat heartburn — have been linked to an increased risk for ischemic stroke in a new study.

The Danish nationwide observational study, presented at the recent American Heart Association (AHA) 2016 Scientific Sessions, showed a dose-related increased risk for ischemic stroke with all four PPIs investigated, but no increased risk with histamine-2 (H2) blockers, which are used for similar indications.

The researchers, led by Thomas Sehested, MD, Danish Heart Foundation, Copenhagen, Denmark, note that their findings add to a growing body of evidence linking PPIs with cardiovascular disease.

He reported that preclinical studies have shown PPIs reduce the production of nitric oxide leading to endothelial dysfunction, and several observational clinical studies have linked their use to cardiovascular disease. "But to our knowledge, this is the first study to look at the effect of PPI drugs on ischemic stroke."

"Our results should be regarded as preliminary as they come from an observational study which has not yet been published," Dr Sehested commented to Medscape Medical News, "but I think we can say this study adds to the evidence questioning the cardiovascular safety of PPIs."

He says the evidence is not yet strong enough to support stopping use of PPI drugs if they are needed.

"However, we would recommend that people should not take these drugs unless there is a clear indication for them,' he said. "Many people are taking them unnecessarily or they are continuing to take them long-term when they don't need to. I would urge doctors to review their patients on PPIs and look at why they are taking these drugs and consider whether they really need them or if they could take a lower dose."

He pointed out that PPIs are available over the counter in many countries and many patients take them off label, which "is a concern."

Real Effect or Marker of Risk?

However, commenting on the study for Medscape Medical News, David A. Johnson, MD, professor of medicine/chief of gastroenterology, Eastern VA Medical School, Norfolk, Virginia, suggested that PPIs are most likely a marker of increased risk rather than a direct cause of cardiovascular or neurovascular adverse outcomes.

He pointed out that all the evidence suggesting cardiovascular harm with PPIs has come from observational studies, but a clinically significant effect was not observed in a randomized/propensity score–matched population (Open Heart. 2015;2:e000248).

"This suggests that the perceived risk associated with PPIs may be dependent on selection bias and different patient baseline characteristics," he said. 

Dr Johnson also noted that data suggesting PPIs may induce endothelial dysfunction by reducing nitric oxide synthesis come from animal studies, but in an open-label, crossover pilot study conducted among healthy persons and patients with coronary disease, PPI use did not significantly influence vascular endothelial function (Vasc Med. 2015;20:309-316).

"When appropriate, clinicians should consider use of PPIs in patients for reduction in upper GI [gastrointestinal] bleed risk without concern about increasing adverse cardiovascular or ischemic outcomes," he concluded.  

Also commenting for Medscape Medical News, Rhanderson Cardoso, MD, University of Miami-Jackson Memorial Hospital, who has been involved in some of the research on the PPI cardiovascular risk controversy, said, "The methods of this cohort study appear to be exceptional. The authors secured objective and reliable ascertainment of exposure, confounding factors were identified, and statistical methods were undertaken to make the groups more comparable."

He added, however, that "This is still an observational study, and due to the possibility of residual confounders, this association by no means indicates a causative effect. Further studies are indicated to study this association. An important point of this study and other observational studies of PPI use and adverse cardiovascular events is that patients should not be taking PPIs without a clear indication."

In his presentation, Dr Sehested noted that prescription of PPIs has increased rapidly during the past decade and the prevalence of PPI use is high in the general population, although many patients are taking these drugs with no obvious indication.

For the current study, the researchers used nationwide Danish registries to identify all individuals older than age 30 years who had elective gastroscopy between 1997 and 2012. Patients with prior cardiovascular disease at baseline were excluded. This database was linked to the Danish registry for medicinal products to identify patients taking various medications, including PPIs and H2 blockers.

A total of 244,679 individuals were included in the study (mean age, 57 years). Approximately 44% had filed a prescription for a PPI. During follow-up, there were 9489 (3.9%) events of first-time stroke.

Results showed that the crude stroke incidence rates per 10,000 person-years were 88.9 for PPI use vs 55.7 for no PPI use.

After adjustment for age, sex, atrial fibrillation, hypertension, diabetes, heart failure, peptic ulcer, cancer, chronic kidney disease, and use of nonsteroidal anti-inflammatory drugs, current use of a PPI was associated with a 20% increased risk for stroke, with an incidence rate ratio (IRR) of 1.19 (95% confidence interval [CI], 1.14 - 1.24; P < .0001).

H2 receptor antagonists showed no association with stroke risk, with an IRR of 1.05 (95% CI, 0.88 - 1.23; P = .60).

On further investigation, a clear dose-response relationship between PPI use and risk for stroke was seen for all 4 PPIs. At the highest dose for these 4 PPIs, stroke risk increased from 33% for lansoprazole to 79% for pantoprazole.

"Although the study is observational, so no causal effect can be proven, we think the dose response we've seen strengthens our findings," Dr Sehested commented. "We didn't see any increased risk at low doses. The risk started to rise at the medium doses, but most patients taking these drugs were taking medium doses or higher."

The lack of any effect on stroke risk with H2 blockers also strengthens the findings with regard to PPIs, Dr Sehested explained. "The null finding with H2 blockers reduces the likelihood of confounding by indication. If it was something to do with the indication that was increasing the stroke risk then we would probably have seen it with the H2 blockers too."

The current study did not include patients who already had cardiovascular disease. "In theory this group would be at higher risk, but we can't say anything on this from our data," Dr Sehested said.

He added that future analyses would look at absolute risks. "So far we have only reported an increase in relative risk of stroke. The population studied had a low risk of stoke so we need to work out the absolute increase in risk. This will be the subject of further analyses."

This study was funded by the Danish Heart Foundation. Dr Sehested has disclosed no relevant financial relationships. Dr Johnson has served as an advisor or consultant for Centocor/Janssen, CRH Medical, Pfizer, and Covidien.

American Heart Association (AHA) 2016 Scientific Sessions. Abstract 18462. Presented November 15, 2016.

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.