His first big clue came when people started hemorrhaging after chewing gum.
Lawrence Craven did tonsil and adenoid surgery in his office. And it usually went well. But in the mid-1940s, "an alarming number of hemorrhages were evidenced in disturbing frequency," he said.
He figured it was the aspirin chewing gum people were using for pain relief. Maybe it interfered with blood clotting. Craven tested the idea out on himself. He took 12 aspirin a day till he got a "profuse nosebleed." Then he did it a couple more times to be sure.
And if the drug could interfere with blood clotting, Craven figured maybe it could prevent blood clots in arteries that caused heart attacks. So he started prescribing an aspirin a day to thousands of men at high risk of heart attack and recording what happened: very few heart attacks.
It was the first appearance of the "aspirin a day" practice in the medical literature. But no one followed up on Craven's call for controlled testing of the hypothesis. Then in the 1960s, John O'Brien, a hematologist in Portsmouth, England, arrived at the same conclusion. That convinced Peter Elwood from the UK Medical Research Council to start a randomized controlled trial. The first men enrolled in 1970.
Ten years after that, the then-new science of meta-analysis came into its own. Richard Peto from Oxford analyzed multiple trials that together showed aspirin could prevent about a quarter of heart attacks. It led to a showdown at the FDA in 1983.
One FDA board member reportedly said the evidence in favor of aspirin was "flawed" and "all the mathematical fiddling in the world couldn't hide that fact." While Peto loudly whispered, "Rubbish. The man's a fool. He's an idiot."
All in all, that first meeting didn't really go all that well. It would be October 1985 before the FDA would announce that aspirin was suitable for prevention in limited circumstances.
There's a lot more evidence now. But we're still debating it. And it still gets very heated. In May this year, the FDA rejected an application to label aspirin as suitable for prevention of heart attack in adults across the board.
Low-dose aspirin may go by the cute and cuddly adjective "baby," but it's a powerful drug. Even "baby" aspirin daily can cause life-threatening bleeding in the gastrointestinal tract and hemorrhagic stroke (bleeding in the brain).
Since March 2009, the U.S. Preventive Services Task Force (USPSTF) has recommended low-dose aspirin for men between the ages of 45 and 79 whose risk of a heart attack outweighs their risk of gastrointestinal hemorrhage - and for women from 55 to 79 whose risk of an ischemic stroke outweighs their risk of gastrointestinal hemorrhage. (Ischemic strokes are caused by blockages, not hemorrhage.) And they recommended against aspirin for prevention of colorectal cancer in 2007.
Aspirin's risks remain - whatever your risk of a heart attack or stroke. So the scales on benefits versus harms can tip quite easily. It's not clear how many people use it who aren't at risk - or how many people who could really benefit don't use it. It's probably used daily by at least 1 in 5 adults in the United States - and by most people over 45 who have signs of heart disease.
Into this already complex scenario comes the issue of cancer prevention. Somehow - we don't really know why - regular aspirin has the potential to prevent cancer. Maybe.
A couple of weeks ago, a Jack Cuzick and colleagues got a lot of media attention for their conclusion that cancer prevention tips the scales in favor of aspirin. However, many aren't convinced. If you're interested in more detail on this, I've been debating the issue with one of that review's authors on PubMed Commons.
According to the Cuzick analyses, the benefits for the average person over the age of 50 would be roughly equivalent to their estimation of the benefit of statins. But because women have a lower risk of dying from some of the conditions affected by aspirin, the net benefit for us would be much lower. The mortality rate over a 20-year period (from 10 years of aspirin-taking) would go down by less than 0.5% for a woman and be about 2% lower for a man.
That would be a lot of people. But if you read one of the enthusiastic media reports, you may have thought the benefit would be more dramatic. That's because relative risk was widely reported out of context, and as though it was for all cancers, not just some - such as by Reuters: "Cancer cases and deaths could be cut by 35 to 40 percent."
"Baby" aspirin isn't going to be the new "apple a day" for everybody. But it certainly is preventing a lot of heart attacks and strokes - and maybe some cancers. That the apple has something of a poisonous side, too, though, means we can expect the discord about what to do with it to continue.
Stay tuned: the USPSTF will be debating a highly-anticipated update on evidence about daily aspirin in the coming months.
If you’re interested in more about the science of analyzing bodies of research, here are some of the related Absolutely Maybe posts: 5 Key Things to Know About Meta-Analysis and Biomedical Research: Believe It Or Not?
* The thoughts Hilda Bastian expresses here at Absolutely Maybe are personal, and do not necessarily reflect the views of the National Institutes of Health or the U.S. Department of Health and Human Services.