The 40 push-up study: what it found and what it didn’t

Researchers went back through the records of Indiana career firefighters, all men (Yang 2019). 1,104 had push-up results. Their mean age was 39.6 and their mean BMI 28.7. Over 8,601 person-years of follow-up there were 37 cardiovascular events.

The push-up test was not a free max-reps set. Men pushed in time with a metronome at 80 beats per minute, and the test stopped at 80 reps, after three missed beats, or at exhaustion.

The comparison group did 0–10 push-ups. Against them, the incidence rate ratio for a cardiovascular event, roughly how many times the comparison group’s rate each group had, was:

• 11–20 push-ups: 0.36 (95% CI 0.14–0.92)
• 21–30: 0.16 (0.06–0.42)
• 31–40: 0.25 (0.10–0.62)
• 41 or more: 0.04 (0.01–0.36)

Read plainly, the men who did 11 or more push-ups had a fraction of the event rate of the men who did 10 or fewer. The pattern is not a smooth staircase: the 31–40 group did slightly worse than the 21–30 group. The intervals are wide. For 11–20 push-ups the range runs from 0.14 to 0.92, close to no difference at its upper end, and for 41 or more it runs from 0.01 to 0.36. The striking number is the 0.04 in the top group, which would mean a rate 96% lower.

The 0.04 rests on one event. There were only 37 events in the whole cohort, so each group holds a handful, and in the top group one case more would have changed the ratio a great deal.

When the researchers adjusted for age and BMI, the hazard ratio for 41 or more push-ups against 0–10 was 0.14, with a confidence interval of 0.02–1.22. Because that interval includes 1, which means no difference, the result is not statistically significant. Once age and BMI were accounted for, the data could not rule out that push-ups made no difference at all.

The authors say it directly: their results do not support push-ups as an independent predictor of cardiovascular risk. The men were mostly white and occupationally active, and there was no correction for testing several comparisons at once.

At 80 beats per minute, every rep is paced, and the test also ends after three missed beats. The count you reach depends on keeping the rhythm as well as on strength, so a metronome count and a free max-reps count are not interchangeable. The number 40 belongs to that protocol, in that population of men.

No women were studied, so nothing in this paper applies to knee push-ups or to women’s results.

The push-up categories that ACSM uses come from the Canadian Society for Exercise Physiology (CSEP-PATH 2019). Men do standard push-ups and women knee push-ups, with no time limit, until form fails. By age group they read, from excellent to needs improvement:

• Men 40–49: 25 or more, 17–24, 13–16, 10–12, 9 or fewer
• Men 50–59: 21 or more, 13–20, 10–12, 7–9, 6 or fewer
• Women 40–49 (knees): 24 or more, 15–23, 11–14, 5–10, 4 or fewer
• Women 50–59 (knees): 21 or more, 11–20, 7–10, 2–6, 1 or fewer
• Men 60–69: 18 or more, 11–17, 8–10, 5–7, 4 or fewer
• Women 60–69 (knees): 17 or more, 12–16, 5–11, 2–4, 1 or fewer

This is a better reference for an ordinary push-up test, but it is thin too. We read it in a licensed reprint, not the manual itself. It gives categories, not percentiles, stops at 69, and the sample and year behind it were not found. It is a fitness ranking, not a health outcome.

The push-up test in the app is a strict max-reps set that ends at the first rep that breaks form, with women in the knee position as in the CSEP table. It is not the firefighter test, and the app never uses 40 as a line or quotes it as a target. The study is still worth knowing, because the number travels far beyond what the paper supports.

To give a percentile, the app reads each CSEP category as a fifth of the population. That reading is our assumption, not the source’s, so push-ups carry the "estimated" label. In midlife men the app’s medians sit about two reps above the table’s "good" band, and the bands above 69 are extrapolated. The calculator shows the same estimate.

A push-up count measures upper-body pressing endurance. It is a useful thing to track, and it moves with training, but in the evidence we have it is a fitness measure, not a health marker.

What the app can honestly say is where your count sits against a published category table and against the line for your age-85 goal. It cannot say what your push-up count means for your heart. This is not medical advice.