Aspirin Risks and Benefits

A new review published in the BMJ once again opens the question of the risks vs benefits of daily aspirin as a prevention for heart attacks and strokes. The reviewers looked at nine randomized trials involving over 100,000 patients and found that aspirin is effective in reducing heart attacks and strokes, but also increases the risk of gastrointestinal bleeding and that in some patients this risk outweighs the benefit.

This is an old and enduring controversy, and one with significant public health ramifications. Aspirin is an anti-platelet agent – it inhibits platelets, the cell fragments in the blood that are the first line against bleeding, from aggregating (clumping together). Platelets aggregate in order to quickly stop bleeding from damaged veins or arteries. But they can also aggregate around cholesterol plaques in arteries, causing a large thrombus (blood clot) that can block off the artery, or that can break off and lodge in a downstream artery (an embolus) and cause a stroke or heart attack.

By inhibiting platelet aggregation daily aspirin reduces the risk of forming a thrombus or embolus, and thereby reduces the risk of heart attack or stroke. Of course, the real story is always more complex than our straightforward explanations. There is some research to suggest that the anti-inflammatory effects of aspirin may also be important to their role in reducing vascular risk. The relative contribution of anti-platelet and anti-inflammatory effects have not been fully teased out. Further, the anti-inflammatory effects of daily aspirin may have non-vascular benefits, like reducing the risk of some cancers.

By inhibiting the normal function of platelets, however, aspirin also increases the risk of bleeding. Patients taking regular aspirin will often notice that they bruise easily, or bleed for a long time from minor cuts (such as cuts from shaving). In fact these symptoms can be used as a rough guide to how much of an anti-platelet effect an individual patient is getting from their dose of aspirin. Sometimes we will cut back on the dose of aspirin, for example, if patients have excessive bruising. The real concern is not about simple bruising, however, but of significant bleeding. Aspirin can cause or exacerbate gastric ulcers, and if they bleed they will bleed more severely because of the blood-thinning effect of the aspirin. Also, if a patient does have a stroke that stroke is more likely to bleed, and that bleeding is likely to be more severe. Also, patients taking aspirin who suffer trauma are likely to have more severe bleeding from that trauma.

In other words – taking daily aspirin has both risks and benefits. Researchers have for years tried to figure out exactly what the relationship is between aspirin risks and benefit at specific doses of aspirin and in specific populations of people. What is not controversial is that daily aspirin is effective for secondary prevention of heart attacks and strokes. Secondary prevention means (in this case) people who have already had a heart attack or cerebrovascular event (stroke or transient ischemic attack – TIA). There is still some dithering about dose (81mg per day vs 325mg per day), but overall the benefits outweigh the risks.

The enduring controversy is over aspirin for primary prevention, in those who have not already had a vascular event – a heart attack or stroke. The risks of aspirin are generally the same in these two groups. Risk of bleeding complications increases with age, but is not significantly different in the primary vs secondary prevention groups.

What is different is the benefit of aspirin. As a general principle the benefit of any preventive intervention increases as the incidence of the disease or event it is meant to prevent increases. Aspirin has increased benefit for secondary prevention because the population of people who have already had an event are at higher risk of having a subsequent one than those who have never had a vascular event. Heart attacks and strokes are relatively uncommon in people who have never had such an event, and so the benefit of aspirin, while significant as a relative decrease in risk, is very minor as an absolute reduction of risk in the low risk population.

The question is – where do the lines of risk vs benefit cross? This is the question the recent study sought to illuminate.

What they found is that for primary prevention those taking regular aspirin were about 30 percent more likely to have a serious GI bleed. So-called “all-cause mortality” was not different in the aspirin vs no-aspirin groups. Aspirin reduced the risk of heart attacks and strokes (more heart attacks in men and more strokes in women) but less than the increase in GI bleed.  For every 162 people taking regular aspirin, one nonfatal heart attack was prevented but there was an increase of about two GI bleeds.

When translated to the general population we are talking about millions of heart attacks, strokes, and bleeds over several years in the US alone, so the public health consequences are huge. In other words, obsessing over small percentage differences has a big effect when translated to the overall population.

This review is not the final word on this question. It leaves us with the same bottom-line conclusion that we had prior to this review (it is a review, so the studies it looked at were already taken into consideration in forming guidelines). That is – for primary prevention aspirin use should be individualized. When looking at the general primary prevention population the risks seem to outweigh the benefits, but perhaps there are subpopulations that are at high risk for a stroke or heart attack in whom aspirin would be a net benefit. It is difficult to study subpopulations for this question because we need very large numbers to get statistically significant results (because the base risk is so low).

But – for those who have diabetes, high blood pressure, high cholesterol, or a strong family history of heart attack or stroke, the benefit of daily aspirin may still outweigh the risks. This means if you were placed on aspirin by your doctor and you have one or more of the above risk factors, this latest study should not make you stop taking aspirin – although you might want to revisit the question with your doctor to see if their recommendations for you have changed.

Official recommendations are to individualize the decision whether or not to use aspirin for primary prevention based upon age, risk of bleeding, and risk factors for vascular disease.

This is a question, with many sub-questions, that was an interesting controversy when I was in medical school two decades ago, has been the subject of ongoing research since then, and researchers continue to refine the data on this question finer and finer. It is an excellent example of evidence-based medicine at work. It is also an example of the real process of individualized medicine -making treatment recommendations based upon the history and risk factors of the individual patient. It further demonstrates the attention and effort that science-based medicine dedicates to preventing disease and morbidity. All of this stands in direct contrast to the propaganda of those who oppose science-based medicine.

Posted in: Public Health

Leave a Comment (18) ↓

18 thoughts on “Aspirin Risks and Benefits

  1. WilliamLawrenceUtridge says:

    I’m pretty sure this is a combination of pedantry, grammar trolling and a spell-checker gone awry, but “aspiring” appears a lot in the article. I’m assuming this is “aspirin” getting converted to the gerund of “aspire” by spellcheck? The other, less likely option is that “aspiring” is a genuine verb used to refer to people who take a preventive dose of aspirin. Which would be kinda awesome.

    Feel free to delete this comment if I’m way off base.

  2. Thanks – I will happily blame my spellchecker/autocorrect. I think I caught them all now.

  3. cervantes says:

    I’m not a real doctor (I’m a doctor of philosophy) but it seems to me that one thing that’s missing from this analysis is the relative severity of the events — heart attack vs. GI bleed. It seems to be treating them as equivalent, but is that really correct? My naive assumption would be that I’d rather have a bleed than a MI, because MI permanently damages the heart and can lead to progressive and severe disability; whereas a bleed is presumably repairable and Bob’s your uncle. Is this wrong?

  4. WilliamLawrenceUtridge says:

    Sorry Dr. Novella, I know Dr. Gorski hates spelling and grammar nit-picking and that it can be annoying. I must admit I’m a tad disappointed I didn’t get to add a new and cool word to my vocabulary. Perhaps I should coin “aspirining” – that should clear up any confusion :)

    Cervantes, I don’t know if you can do a straight death-to-death comparison as medicine also attempts to reduce risk factors overall. The ideal solution would be a dosage or formulation that reduced the risks of cardiovascular events without causing bleeding. An imperfect comparison would be Vioxx – pulled due to previously-unappreciated increased risk of death despite being a very effective pain medicine. Another would be the whole-cell pertussis vaccine – very effective vaccine but increased the risk of bowel problems, and was pulled and replaced with a less effective but safer version.

    Probably bad comparisons, but I would also be interested to hear the epidemiologist and MD’s perspective on that issue too…

  5. David Gorski says:

    Sorry Dr. Novella, I know Dr. Gorski hates spelling and grammar nit-picking and that it can be annoying.

    Well yes, but a bit less so now that our regular copy reader is out for a while, which might leave it to me to proofread Mark Crislip’s posts. As Dr. Zachary Smith would say, “Oh, the pain, the pain!” :-)

  6. Janet Camp says:

    I always wonder how risk factors are calculated when they are well-controlled? I have family history, high bp (well-controlled through lifestyle and meds), high chol/trigl (also controlled by same) and had been diagnosed with type II which precipitated the needed lifestyle changes. I lost 45 lbs (maintained for over five years now) and all the numbers vastly improved and blood sugar is normal. I still take low dose aspirin based on my risk factors, though there was no heart attack due to stent placement.

    Is there a difference between the level of risk for managed and unmanaged groups? I realize this is not an advice column, so I will ask my doctor, but the question came up in the comments to the NY Times piece on the same study, so I’m not the only one wondering about this.

    How much does controlling risk factors affect risk factors? Are people with good control studied in these kinds of trials?

  7. CW says:

    A doctor told my father to start taking low-dose aspirin as a way to prevent a second heart attack. The doctor said that low-dose aspirin negates the effect of GI bleeding. And then I related this news story to my father, and he’s no longer taking low-dose aspirin.

    Well, at least there’s always the one beer/glass of red wine treatment. :P

  8. Harriet Hall says:


    Your father may have heard that it negates the effect, but that might not be exactly what the doctor said or meant to say.
    If the doctor really told your father it “negates the effect,” he was either misinformed or was lying. Maybe your father should look for another doctor.

    I read the same news story as well as the studies behind it, and I chose to keep taking low-dose aspirin. I am well aware of the risks, but my cardiologist and I have both concluded that the benefits outweigh the risks in my particular case.

  9. Anarres says:

    Very interesting post,

    I had a MI 6 years ago, I take 100mg daily since then.

    Thanks Steven.

  10. BillyJoe says:


    “A doctor told my father to start taking low-dose aspirin as a way to prevent a second heart attack….And then I related this news story to my father, and he’s no longer taking low-dose aspirin.”

    The article is about aspirin in primary prevention. Your father has had a heart attack, so the conclusions do not apply to him. The usefullness of aspirin in secondary prevention has been confirmed. He should continue to use aspirin.

  11. nybgrus says:


    GI bleeds can be deadly as well and even when they are not, can be much more than just a nuisance to be fixed easily. I agree that they can’t be exactly equated with MI, but considering the excellent therapies we have for MI and the potentially very serious consequences of GIB I’d say the gulf is smaller than you’d imagine.

    And as BJ said, the discussion is about primary prevention, where the baseline risk of MI is low compared to the elevated risk of GIB and the fact that many people who suffer this complication would have other comorbities that it may well exacerbate. For secondary prevention it is pretty clear that the benefits in MI reduction outweigh the increased risk of GIB.

  12. nybgrus says:

    oh, sorry, I should add that GIBs can be acutely deadly as well, if the ulcer erodes through the muscularis and get to one of the gastric arteries.

  13. Quill says:

    One other good thing about daily aspirin therapy for those whose risk profile warrants it: it is incredibly inexpensive. Every drug store has a house-brand that is something like a year’s supply of 81-mg tablets for about five to ten dollars. (Costco probably has a twin pack of 500 tablet bottles for about the same price.) This could also be why more people are on daily aspirin therapy than should be — who can pass up such cheap “insurance?”

  14. DugganSC says:

    On a side note of aspirin risks, is the use of aspirin by those under sixteen still contraindicated for fear of Reye’s Syndrome. I read an article some years back claiming that the studies which showed the link have been shown to be flawed and that the only evidence that there was any real link was the decrease in incidence of Reye’s Syndrome after people were getting warned, a decrease which was uniform even in countries where the government was no warning people to not give aspirin to children. Could you comment on that?

  15. Badly Shaved Monkey says:

    I think it does bear making the point explicitly to all the SCAM-supporting numpties who play the “Big Pharma shill” card every time a sceptic criticises their preferred voodoo, that all of this work is a perfect example of good medical science being done for the public good on a low-value off-patent drug. I think the likelihood of there being a Big Aspirin cabal is really quite small.

    Unless of course the men in black helicopters fund this kind of pro bono work precisely to conceal their other conspiracies.

    [I do worry I have spent too much time down the SCAM rabbit-hole]

  16. Mark Crislip says:

    Hey! I just noticed. Paul says I have gotten a lot better…

  17. Chris says:

    It must be the ice storms, Dr. Crislip. I have just experienced them for the first time.

    You folks in Oregon can keep them, they are nasty and annoying. I am willing to give up pretty looking trees for not risking my neck by just walking outside and having electricity disappear because ice covered branches drop off and break power lines.

Comments are closed.