Your Cardiologist Has a Treadmill. A Firehouse Has a Floor.
Cardiac risk assessment looks like machinery: the treadmill stress test, the blood draw, the calcium scan. A submaximal treadmill test needs a clinic, trained staff, and a billing code, so most doctors do something cheaper instead. They ask whether you exercise and write down the answer, though self-reports routinely overstate real fitness, and while the American Heart Association wants fitness treated as a vital sign, objective tools are expensive and slow, so almost nobody complies. In 2019, a Harvard team tested whether the cheapest fitness test ever invented, the push-up, could do the job.
1,104 Firefighters, 10 Years, One Metronome
Harvard researcher Justin Yang pulled a decade of medical surveillance records from 10 Indiana fire departments: 1,562 firefighters examined at baseline, 1,104 with push-up data. The group averaged 39.6 years with a mean BMI of 28.7, all men and all occupationally active. The test itself was standardized with unusual care: each firefighter did push-ups to a metronome set at 80 beats per minute, continuing until 80 reps, three missed beats, or exhaustion or symptoms. The men were split into five push-up groups and followed through 2010 for cardiovascular events: coronary artery disease diagnoses, heart failure, and sudden cardiac death, thirty-seven events across 8,601 person-years.
The Gradient Was the Story
The crude numbers are startling: compared with the 0-to-10 group, incidence rate ratios were 0.36 for 11 to 20 push-ups (64% lower), 0.16 for 21 to 30 (84% lower), 0.25 for 31 to 40 (75% lower), and 0.04 for 41 or more (96% lower), and all but one of the 37 events struck men who managed 40 or fewer. In absolute terms, 8 of the 75 men in the lowest group had an event within ten years (10.7%), versus 1 of 155 in the highest group (0.6%), and the Kaplan-Meier curves put cumulative incidence at 15% for the bottom group against 5% or less for everyone else. Each man also took a submaximal treadmill test estimating VO2max, yet in the adjusted models the aerobic fitness categories showed no statistically significant association with events at all. The two-minute floor test beat the machine.
The Arithmetic Nobody Did
Start with absolute risk: the 96% relative reduction describes a drop from 10.7% to 0.6% over ten years, about ten percentage points concentrated in 75 high-risk men. Relative risks sound like destiny; absolute risks sound like information, and the second is more useful. Second, fragility: the top-group estimate rests on one event in 155 men. One more event would roughly double the rate ratio, turning 96% lower into roughly 91% lower. The 95% confidence interval spans 0.01 to 0.36, so the data are compatible with anything from a 64% to a 99% reduction, which makes the headline the most dramatic reading the data allow rather than the most likely one. Third, the statin benchmark: the Cholesterol Treatment Trialists’ meta-analysis of 27 statin trials in 175,000 people found that each 1 mmol/L reduction in LDL cholesterol cuts major vascular events by 21%, making the crude push-up figure roughly four and a half times a statin effect, which means that if push-ups genuinely cut heart risk by 96%, they would be the most powerful cardiac intervention ever discovered, beating statins and blood-pressure drugs combined. Nobody believes push-ups are magic. So the number’s implausibility as a treatment effect is what makes it valuable as a warning signal: it is measuring overall fitness, not push-ups. Fourth, the adjusted fine print: after accounting for age and BMI, only the 21-to-30 group stayed significant (hazard ratio 0.25, 95% CI 0.08 to 0.76). The 41-plus comparison was suggestive but not significant: hazard ratio 0.11 to 0.15, p = 0.06 to 0.08. The gradient was not even monotonic, since the 21-to-30 group (rate ratio 0.16) outperformed the 31-to-40 group (0.25). The honest summary is not that 40 push-ups buys 96% protection; it is that above roughly ten push-ups, risk falls steeply, and the treadmill added nothing.
The Strongest Counterargument
The fiercest objection writes itself: the push-up test is a proxy, not a lever. The 41-plus men averaged 35 years old against 48 in the bottom group, with BMI 26.8 versus 33.1, systolic blood pressure 125 versus 137, triglycerides 116 versus 168 mg/dL, fasting glucose 85 versus 99, and 6.7% current smokers against 24%. A push-up count compresses age, weight, blood pressure, lipids, glucose, and smoking, everything a cardiologist already measures, into a floor exercise, and the authors concede the point: established risk-factor differences may explain much of the gap, so the results do not support push-up capacity as an independent predictor of risk. A marathon runner could fail this test with a superb heart, because upper-body endurance is one slice of fitness, and it favors the young and lean. Worse, the number can be gamed: eight weeks of dedicated push-up training will raise the score while the underlying risk barely moves, which is Goodhart’s law in a gym shirt. Once a measure becomes a target, it stops being a good measure.
What We Didn’t Prove
The cohort was 1,104 male Indiana firefighters, about 87% White and all occupationally active, so the results may not generalize to women, older adults, or sedentary people, and with only thirty-seven events on which to hang a gradient plus no adjustment for multiple comparisons, the authors caution that some findings could be due to chance. Nearly three in ten eligible firefighters never took the push-up test, which cost statistical power. No one has directly replicated this study: it was the first of its kind, and years later it remains the only one. The finding fits the broader literature linking muscular strength to lower mortality, but the push-up claim stands on a single cohort, and it is an association. Nothing here shows that doing more push-ups causes a healthier heart.
The Bottom Line
A careful 10-year cohort study found that firefighters who could do 40 or more push-ups had dramatically fewer cardiovascular events than those who could do fewer than 10. The 96% figure is real arithmetic built on thin data: one event in the top group, wide confidence intervals, and an adjusted analysis where only the middle group clears significance. Treat the push-up as a vital sign worth checking, not a drug worth prescribing.
What You Can Do
First, take the test: warm up, set a metronome app to 80 beats per minute, and do strict push-ups until you miss three beats or stop. Two minutes, zero dollars, one honest number. Second, build capacity the boring way: add a few sets through the day, use knee or incline push-ups while you build strength, and aim for one or two more reps each week. Tendons adapt slower than enthusiasm. Third, if you land under ten, treat it as a prompt rather than a prophecy: book the postponed checkup and get blood pressure, lipids, and glucose measured. The test earns its keep as an early warning, not a diagnosis. Fourth, keep your cardiology appointments: push-ups measure muscular endurance, not coronary arteries. Fifth, train the rest of the machine: cyclists and runners can bomb this test with superb hearts. The push-up is one vital sign, not the whole exam.