The Pill in the Medicine Cabinet
Open the medicine cabinet of almost any American over 70 and you will find a small white tablet taken every morning for no symptom at all, because for decades the daily aspirin was the cheapest insurance policy in medicine: pennies a day to keep the first heart attack away; in 2016 the U.S. Preventive Services Task Force formally advised it for adults in their fifties at elevated cardiovascular risk.
The habit stuck as the evidence shifted: a 2019 analysis of the National Health Interview Survey found 29 million Americans 40 and older without cardiovascular disease taking daily aspirin to prevent a first heart attack, 6.6 million without any doctor's recommendation, and nearly half of adults over 70 without heart disease, roughly 10 million people, taking one every day.
Then three large trials in 2018 tested the habit directly, and the largest delivered the harshest verdict.
The Trial That Finally Asked
ASPREE, the Aspirin in Reducing Events in the Elderly trial, asked the question nobody had properly asked: whether a daily aspirin extends healthy life in older adults who have never had cardiovascular disease.
Led by Monash University with the Berman Center in Minnesota, it enrolled 19,114 adults in Australia and the United States between 2010 and 2014, all at least 70, or 65 and older for Black and Hispanic Americans, all free of cardiovascular disease, dementia, and disability at entry, with half randomized to 100 milligrams of enteric-coated aspirin daily and half to placebo.
The endpoint was unusual and deliberate: disability-free survival, a composite of death, dementia, and persistent physical disability capturing the full trade-off. If aspirin prevented strokes and heart attacks, the composite would fall; if bleeding dominated, it would rise.
After a median of 4.7 years, the safety board stopped the trial early, judging there was a negligible chance that any benefit would emerge with more time.
Three Numbers That Ended an Era
The primary result was a flat null, with the composite occurring at 21.5 events per 1,000 person-years on aspirin and 21.2 on placebo, a hazard ratio of 1.01 with a p-value of 0.79, meaning nearly five years of daily aspirin had bought the average participant nothing measurable.
The second number was the harm: major hemorrhage struck 3.8 percent of the aspirin group versus 2.8 percent on placebo, a 38 percent higher rate, which translates to roughly one extra major bleed for every hundred older adults taking the pill for five years.
The third number stunned the investigators: all-cause mortality ran 14 percent higher with aspirin, 12.7 versus 11.1 deaths per 1,000 person-years, with cancer driving the gap at 3.1 percent cancer-related death on aspirin against 2.3 percent on placebo, which the authors flagged, writing that "in the context of previous studies, this result was unexpected and should be interpreted with caution."
Two companion trials that year agreed about bleeding: ARRIVE followed 12,546 adults at moderate cardiovascular risk, ASCEND 15,480 diabetic patients, and both found aspirin for primary prevention raised bleeding without convincing net benefit.
The Arithmetic Nobody Ran
Trial percentages feel abstract until multiplied by 10 million: roughly how many Americans over 70 without heart disease take daily aspirin for prevention, per the National Health Interview Survey analysis.
ASPREE's absolute bleeding increase was 1.0 percentage point over a median 4.7 years, 3.8 percent minus 2.8 percent, which spreads to roughly 0.21 percent per year.
The arithmetic is simple and stark: 10,000,000 people times 0.0021 per year equals about 21,000 extra major hemorrhages every year, the annual price of a habit the trial says buys nothing, computed from the 10-million estimate from O'Brien and colleagues and the 1.0-point gap from ASPREE, assuming survey respondents bleed roughly like trial volunteers.
The Strongest Case Against the Finding
The case against starts with the authors themselves: the 14 percent mortality rise was a secondary endpoint, and secondary endpoints in stopped-early trials are where chance findings breed, consistent with neither ARRIVE nor ASCEND finding higher mortality.
In January 2026, the ASPREE-XT follow-up reported that the cancer mortality elevation did not persist after the trial ended, and real drug effects do not usually vanish like that.
A second critique targets the endpoint itself: in 2025, statisticians led by Janet Wittes called disability-free survival an "uninformative" composite, arguing that lumping death together with dementia and disability can bury a genuine cardiovascular benefit under unrelated noise; aspirin showed a non-significant trend toward fewer vascular events that a cleaner endpoint might have caught.
The investigators rebutted in detail, but the exchange shows the null is partly a property of the endpoint choice rather than the drug alone.
A third objection is the population: volunteers healthy enough for a five-year trial, 91 percent White, median age 74, not a cross-section of American 70-year-olds. A fourth is scale: a 1.0-point absolute bleeding rise over five years is real but small, and for someone at high cardiovascular risk the calculus can still favor aspirin.
What We Didn't Prove
ASPREE did not prove that aspirin causes cancer deaths, because the signal was a secondary endpoint flagged as unexpected by its own authors, absent in two contemporaneous trials, and gone in the extended follow-up, with no proposed mechanism and 4.7 years of observation, short for a cancer story.
It did not answer whether long-term users should stop: the trial randomized starters, not stoppers, saying nothing direct about discontinuation, and nothing about adults under 70, for whom the 2022 USPSTF kept an individualized decision for ages 40 to 59.
It does not touch secondary prevention at all, where aspirin's benefit after a heart attack or stroke rests on decades of evidence that this trial leaves intact.
The bleeding arithmetic above is a projection rather than a measurement, because real-world users are sicker and more heavily medicated than trial volunteers, which could push the true number in either direction.
The Bottom Line
The daily aspirin felt like pure upside: pennies a day, no prescription, decades of reassurance. The largest trial in healthy older adults found the upside was zero and the downside measurable, with more major bleeding, no extra disability-free years, and a mortality signal that may be noise. The 2022 USPSTF recommends against starting aspirin for primary prevention at 60 and older. Ten million older Americans still take the pill every morning; most should ask their doctor whether the habit still makes sense.
What You Can Do
If you take daily aspirin without a history of heart attack or stroke, raise it at your next appointment and ask your doctor to weigh your cardiovascular risk against your bleeding risk, but do not stop or start on the basis of an article.
If you take aspirin after a heart attack, stroke, or stent, this is not about you, so keep taking it and never stop abruptly without medical guidance, because sudden withdrawal after a cardiovascular event carries its own well-documented risk.
Learn the warning signs of serious bleeding: unusual bruising, black or tarry stools, blood in urine, a nosebleed that will not stop. If a clinician suggested the habit years ago and never revisited it, the revisit is overdue. The guideline changed in 2022; the medicine cabinet should reflect the medicine, not the memory.