All NSAIDs Linked to Increased MI Risk

Pauline Anderson

May 10, 2017

Use of nonsteroidal anti-inflammatory drugs (NSAIDs), including naproxen, considered by some as one of the safest drugs in this class, is associated with a significantly increased risk for myocardial infarction (MI), results of a new patient-level meta-analysis show.

The analysis showed the heightened MI risk occurred as early as the first week of use and the risk was greater with higher doses.

"This new research on NSAIDs reinforces what physicians know already, that patients should use the smallest possible dose for the shortest possible time," Michèle Bally, PhD, an epidemiologist in the Department of Pharmacy and Research Center, Centre hospitalier de l'Université de Montréal, Quebec, Canada, told Medscape Medical News.

Even though the research suggests the increased MI risk lessened over time, "the findings were not conclusive enough about longer duration," said Dr Bally.

But for most patients, the risk is very small, she noted. "If you average people with different baseline heart risks, the risk specifically due to an NSAID is only about 1% per year, so out of 100 people treated continuously for a year, there will be one extra heart attack."

The research was published online May 3 in BMJ.

Although previous studies have linked both traditional and cyclooxygenase-2 (COX-2) selective NSAIDs to increased MI risk, the timing of the risk, the effect of dose and treatment duration, and the comparative risks between different drugs are poorly understood, the researchers note.

The recently reported Prospective Randomized Evaluation of Celecoxib Integrated Safety Vs Ibuprofen Or Naproxen (PRECISION) study found that at moderate doses, celecoxib was noninferior to ibuprofen or naproxen with regard to cardiovascular safety.

After an electronic literature search, researchers selected studies to include in this analysis. The four administrative data sets they used were Régis de l'assurance maladie du Québec, the universal, computerized public insurance database of Quebec; a population-based, case-control study from Finland; and two nested case-control studies: one from the United Kingdom using data from the General Practice Research Database and the other using data from the Canadian province of Saskatchewan.

Identifying information on individual patients was removed, said Dr Bally.

Researchers performed an individual-patient data (IPD) meta-analysis that aimed to emulate the design of a large, pragmatic randomized trial. In general, IPD meta-analyses are more "generalizable" than a clinical trial and better reflect the "real world" when studying a drug-related adverse event, said Dr Bally.

"In a clinical trial, you're assigned to a group and you take the NSAID on a continuous basis, and you stay on that dose, whereas in practice, patients may use lower doses or varying doses, use a drug on and off, and switch between drugs," she said.

NSAIDs of interest for the current study were celecoxib, the three main traditional NSAIDs (diclofenac, ibuprofen, and naproxen), and rofecoxib. Studies were limited to those conducted before rofecoxib (Vioxx, Merck) was withdrawn from the market because of safety concerns of an increased risk for cardiovascular events.

IPD meta-analyses allow researchers to "harmonize" data, to make the exposure and comorbidities as similar as possible before pooling the data, said Dr Bally.

The pooled data in this case included 61,460 cases of MI and 385,303 controls for a total of 446,763 individuals.

By using the Bayesian statistical approach, which indicates probability, researchers compared adjusted odds ratios (ORs) of acute MI for past use, recent use, and five dose-duration categories of current NSAIDs, with nonuse of any NSAID in the year before the index date, for each study and for the pooled studies.

Overall, for current NSAID use, the increase in MI risk compared with not using these drugs was 20% to 50%. The authors noted that the significantly increased risk for MI associated with current use was found for all traditional NSAIDs, including naproxen. The risk for acute MI for celecoxib did not appear to be greater than that for traditional NSAIDs and was lower than that for rofecoxib.

Rapid Onset

There was a rapid onset of MI risk in the first week of NSAID use. In this time frame, the probability of increased risk (posterior probability of OR > 1.0) was greater than 90% for all NSAIDs: 92.4% for celecoxib, 97.3% for ibuprofen, 98.6% for diclofenac, and 98.8% for both naproxen and rofecoxib.

The almost immediate elevated risk "was somewhat surprising," commented Dr Bally. "We didn't anticipate that necessarily."

Table. Acute MI Risk Associated With NSAID Use


Odds Ratio (95% Credible Interval)



1.24 (0.91 - 1.82)


1.48 (1.00 - 2.26)


1.50 (1.06 - 2.04)


1.53 (1.07 - 2.33)


1.58 (1.07 - 2.17)


Use of higher doses were also associated with increased MI risk. Use for 8 to 30 days at a high dose was particularly harmful for ibuprofen (>1200 mg/day), naproxen (>750 mg/day), and rofecoxib (>25 mg/day).

But a longer duration of treatment generally did not seem to be associated with greater probability of increased risk for MI. "For duration of treatment, we observed that the risk doesn't go away, but it does not increase further," said Dr Bally.

This, said the authors, may be because those more susceptible to acute MI were selected out of the cohort at earlier time points.

"But let's interpret that with a big note of caution because we did not measure reinfarctions," stressed Dr Bally.

The elevated MI risk linked to NSAIDs is thought to be related to the selective inhibition of the COX-2 enzyme, which may increase blood pressure.

As with all studies sourced from databases, this analysis used measures of drug dispensing or drug prescribing but not actual drug intake. Dr Bally noted that ibuprofen estimates may be "more fragile" because it's available over the counter and in this analysis its use "as needed" was more prevalent than for other NSAIDs.

However, Dr Bally also pointed out that in Quebec, people are reimbursed for NSAIDs, so they have an incentive to get a prescription for this drug rather than buying it over the counter.

In addition to prescribing the smallest amount of NSAIDs and for the shortest possible time, physicians should counsel patients to learn their cardiac risks, said Dr Bally.

"They should get a good idea of what their baseline risk of a heart attack is and decide if they want to use these drugs or not."

However, the overall risk is small for most patients, she noted. It's also important to be aware of the risks of taking opioids as an alternative to relieve pain, said Dr Bally.

Patients may want to consider nondrug approaches to address their pain. Such approaches could include topical analgesics, exercise, or physiotherapy.

"Extreme Reluctance"

Dr Bally pointed out that four good-quality studies that were eligible for inclusion in this analysis could not be included because access to the relevant administrative databases was denied, and there was "extreme reluctance" to allow access to databases that were included.

"In total, this research took 10 years from the original idea; it was great for me because it turned out to be my PhD thesis, but if it were not for the PhD, this study would never have been done."

Dr Bally is passionate about researchers having better access to administrative databases to study drugs, just as they now have easier access to databases in the field of genetics.

"We need to be able to follow what happens to people, in a deidentified or a data-protected framework," said Dr Bally. "If we had access to patient data, we could look at patient trajectories over various types of encounters — from being in the community, to being hospitalized, to being discharged."

For a comment, Medscape Medical News approached Charlotte Warren-Gash, MRCP, PhD, associate professor, London School of Hygiene & Tropical Medicine, United Kingdom. Dr Warren-Gash's work has focused on public health and acute MI, and she has published commentary on research linking NSAIDs to MI risk.

She noted that in contrast to some other studies, this new research found that naproxen — traditionally believed to be one of the safest NSAIDs — showed similar effects to other nonselective NSAIDs.

"The study adds to the growing body of evidence that both COX-2 inhibitors and nonselective NSAIDs are associated with a small increase in cardiovascular risk, especially at higher doses."

The strengths of this study include its large sample size and the use of "real-world" observational data, said Dr Warren-Gash.

But despite these strengths, the results should be interpreted with caution, she said.

"Although the authors controlled for a range of potential confounders, there may have been residual confounding by factors that are not recorded well or at all in routine health databases, such as the reasons for prescribing particular drugs."

That the MI risk did not continue to increase after 30 days of use for most of the NSAIDs "seems biologically implausible," said Dr Warren-Gash.

"The authors suggest that an alternative modelling approach might be needed to examine the effect of cumulative NSAID exposure over time."   

Dr Bally reports grants from McGill University Health Centre Research Institute during the conduct of the study and personal fees from Institut national d'excellence en santé et services sociaux, outside the submitted work. Dr Warre-Gash has disclosed no relevant financial relationships.

BMJ. Published online May 3, 2017. Abstract

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.