Common Painkillers Increase the Risk of Heart Attacks

Introduction

A recent meta-analysis study in the British Medical Journal [1] suggests that using NSAIDs (a commonly prescribed type of painkiller) increases the risk of acute myocardial infarction (heart attack).

This study is open access and you can read (or download) it here.


What is new here?

Not much to be honest. It was already known that NSAIDs increase the risk of MI (heart attacks).

There was some debate as to whether there were differences in risk between the different classes of NSAID (COX 1 vs COX2 inhibitors) and different individual drugs.

This study suggests that all these drugs increase risk to an almost identical degree (except for Celecoxib which seems to have a lower OR).

I have summarised the main results in the table below:

Table 1 Summary of Results

I made this using the Odds Ratios from the study.

Risk increases almost immediately

The increase in risk seems to be almost immediate and peaks at one month (except for diclofenac).

The risk also appears to be dose dependent so the higher the dose of the drug being taken, the greater the increased risk of having a heart attack.

Rofecoxib not as bad for heart attacks as previously thought?

One point of note is that Rofecoxib (Vioxx) was withdrawn by the manufacturer over a decade ago.

This was following concerns over its risk of increasing heart attacks and stroke.

It is interesting that this study does not show a huge difference between Rofecoxib and the other NSAIDs.

Confounding Factors

Particular care was taken to examine and control for confounding factors — I won’t cover it here as there is a lot of information but you can read about on page 4–5 of the study.

Confounding factors are things such as age, diabetes status and so forth which could change the risk of a heart attack independently from the NSAID drugs which are being studied.

They are one of the biggest problems in medical research and if you read my blog you will often see me talking about them.

Although we can take measures to reduce their effect there is no way to completely eliminate them.

Problems

Bayesian Statistics

This study has a peculiarity that makes it more difficult to interpret.

It uses a different type of meta-analysis from that which is conventionally used.

It is called Bayesian meta-analysis and you can read more about here.

According to the Cochrane Handbook[2]:

“Bayesian statistics is an approach to statistics based on a different philosophy from that which underlies significance tests and confidence intervals [classical analysis].”

I won’t even pretend to fully understand the statistical intricacies of how this differs from the standard statistical models.

The Cochrane Handbook link I have provided above has a summary but as they state:

“Statistical expertise is strongly recommended for review authors wishing to carry out Bayesian analyses.”

I’m afraid I don’t have that kind of statistical expertise and this presents a problem in that I can’t render the same kind of analysis which I could normally.

If you are experienced with this type of analysis I would love to hear your thoughts.

Other Issues

There are some other problems here too. This study incorporated results from 4 studies for analysis — a total of 8 studies actually met the inclusion criteria however:

“four ultimately had to be excluded because of ethico-legal restrictions placed by health authorities on transfer of IPD to third parties.”

This creates a potential source of bias. A further issue is that it reduces the statistical validity of the data by reducing the potential sample size.

A smaller sample consisting of fewer studies means that the effect of any oddities or inconsistencies would be greater.

The total cohort was still quite large at 446 763 (61 460 heart attacks) so I don’t want to overplay this.

Another interesting point is mentioned by the authors towards the end of the paper:

“In particular, we could not study whether the effect of past doses of NSAIDs persisted and affected current risk nor could we determine the precise onset of any associated increased risk or the exact duration of any persistence of risk after stopping an NSAID.”
“In particular, for diclofenac, the risk of myocardial infarction with treatment for more than 30 days (blue and lavender lines in g 2 and table 3 ) is higher than with treatment for 8–30 days (brown and red lines in g 2 ), which hints at cumulative effects for this drug.”

So although it appears that the increase in risk happens quite rapidly it is not possible to quantify exactly how long it takes to occur — for example does just a single dose increase your risk?

Also how long does it take for the risk to return to normal, indeed, does it return to normal?

The findings with diclofenac suggest that the effects might persist for a longer period and more specific studies would be needed to examine these issues.


What If I am Taking these Medications?

I would not suggest taking any drastic action right away. It has already been known that this link exists so there is no reason to panic.

If you have a lot of risk factors for heart disease (e.g. previous angina or coronary artery disease, high blood pressure, diabetes, strong family history etc.) or are really concerned — discuss this issue with your doctor.

It may be possible to minimise your dosage or consider other options to reduce the potential risk.

Whilst this study would suggest reducing or stopping NSAIDs would be a good thing — we must always take into consideration the balance of benefit vs risk and the potential for unforeseen consequences.

For example, if you have chronic pain which severely limits your activity then stopping your NSAIDs may paradoxically not reduce your risk of a heart attack if it results in you becoming more sedentary and/gaining more weight.

Chronic pain itself may be a risk factor for increased cardiovascular risk (at least indirectly).

For example I have seen a number of studies that link chronic pain with depression, and depression is a significant risk factor for cardiovascular disease.

So just to re-iterate, don’t change anything without discussing it with your doctor first.


Quick note about this post

I have noticed that as my posts get longer people tend to read them less and this is understandable as most people are short of time.

I have therefore made this post more brief and used a slightly different way of breaking down the information.

I won’t always be able to do it but will try to make things more brief when I can.

It is a difficult balance to achieve though, and I hope I have got it right.

Please let me know what you think — is it too simple, not simple enough or about right?

Thank you for reading


References

  1. Bally, Michèle, Nandini Dendukuri, Benjamin Rich, Lyne Nadeau, Arja Helin-Salmivaara, Edeltraut Garbe, and James M. Brophy. 2017. “Risk of Acute Myocardial Infarction with NSAIDs in Real World Use: Bayesian Meta-Analysis of Individual Patient Data.” BMJ 357 (May): j1909.
  2. Higgins JPT, Green S., ed. 2011. “Cochrane Handbook for Systematic Reviews of Interventions: 16.8.1 Bayesian Methods.” The Cochrane Collaboration. http://handbook.cochrane.org/chapter_16/16_8_1_bayesian_methods.htm.

Want more like this? — Follow me on Twitter or connect on LinkedIn to keep up with my latest posts. The original version of this post was made on my Steemit blog here (Steemit is a new way to monetise your content find out more here).


All images are taken from my personal Thinkstock Photography account unless stated otherwise. More information can be provided on request.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.