Max Heart Rate Formula Research: 220 Minus Age vs Tanaka, Nes, and Gulati

Almost every fitness wearable on your wrist quietly assumes one number. Max heart rate. From that single beats-per-minute ceiling, the watch derives every zone, every recovery score, and every "you're in fat burn" nudge. And for half a century, the default number has come from the same shorthand. 220 minus your age.

That formula sits in textbooks, on gym treadmill placards, and inside the firmware of Garmin, Apple Watch, Whoop, and Polar. It's also been quietly retired by the exercise physiologists who study it. The reasons are unflattering. The equation was never derived from a clean study. Its standard deviation is wide enough to mislabel an easy ride as a threshold workout. And women, older adults, and well-trained athletes are systematically misestimated.

Better equations exist. Tanaka, Nes, and Gulati each ran large studies and published cleaner numbers. None of them are perfect. But knowing where 220 minus age came from, what the alternatives say, and when each one is useful is the difference between training to a number that means something and training to a number you made up. Let's walk through the research.

The Research: What Studies Show

Four pieces of work do most of the heavy lifting in the max heart rate conversation. They span a generation of exercise physiology and answer different questions: where did 220 minus age come from, what does a real prediction equation look like, what changes when you sample women, and how accurate is any of this at the individual level?

Robergs and Landwehr 2002: Where 220 Minus Age Came From

The cleanest takedown is the easiest to read. Robergs and Landwehr (2002), in the Journal of Exercise Physiology Online, went hunting for the original source of 220 minus age. They didn't find one. The formula traces back to a 1971 chapter by Fox, Naughton, and Haskell that plotted a hand-drawn regression line through data from 11 cobbled-together studies. The line was an observation, not a tested prediction. Fox himself never claimed it was clinical-grade.

Robergs and Landwehr then pooled the available research to estimate the actual error. The standard deviation of the 220 minus age estimate sits between 7 and 12 beats per minute, depending on the sample. A standard deviation of 10 means roughly a third of healthy adults have a true max heart rate more than 10 beats above or below the estimate. Some people miss by 20. For a 40-year-old, that's the difference between training at zone 2 and training at the lactate threshold while watching the same number.

Their conclusion was blunt. The formula has limited scientific validity. It has stuck around because it's easy to remember, not because it works.

Tanaka, Monahan, and Seals 2001: The Better General Equation

The most cited replacement formula came out one year before the Robergs critique. Tanaka, Monahan, and Seals (2001) in the Journal of the American College of Cardiology pooled 351 studies covering 18,712 subjects, then ran a separate lab validation on 514 healthy adults. They produced a clean linear equation. 208 minus 0.7 times age.

The Tanaka formula tracks the actual lab-measured max better than 220 minus age across the adult lifespan. For a 25-year-old, the two equations agree closely. By age 40, Tanaka predicts 180 versus 180 for 220 minus age. By age 60, Tanaka says 166 while 220 minus age says 160. By age 75, Tanaka says 155.5 versus 145. That late-life gap matters. The 220 minus age formula systematically underestimates older adults, which can lead to underprescribed training intensities for the population that needs aerobic gains most.

Tanaka also found something useful about sex. The same equation worked for men and women within the limits of their pooled sample, with no statistically significant interaction by sex. That finding is genuine, but it lived inside a sample that skewed toward men and toward healthy adults. The story gets more complicated when you sample women specifically.

Gulati 2010: A Female-Specific Number

The St. James Women Take Heart Project is one of the largest exercise-stress-testing cohorts of women ever assembled. Gulati and colleagues (2010) in Circulation followed 5,437 asymptomatic women through symptom-limited treadmill testing and tracked them for cardiovascular outcomes. The peak heart rate they hit on the test, as a function of age, gave them a clean female regression. 206 minus 0.88 times age.

The slope matters. Female max heart rate falls slightly faster with age than the general Tanaka slope, and the intercept is lower. For a 50-year-old woman, 220 minus age predicts 170. Tanaka predicts 173. Gulati predicts 162. That's an 8-to-11 beat gap from the male-anchored formulas. Use the wrong one and you'll think you can sustain "zone 2" at a heart rate that's actually closer to threshold work.

Gulati's data also fed into mortality predictions. Women whose peak heart rate fell well below the formula prediction (a state called chronotropic incompetence) had higher all-cause mortality risk over follow-up. So the new equation wasn't just cosmetic. It changed which women got flagged for further cardiac workup. That's a real clinical consequence of using the right baseline.

Nes 2013: The HUNT Fitness Update

The fourth widely cited equation came from Norway. Nes and colleagues (2013) in the Scandinavian Journal of Medicine & Science in Sports measured max heart rate in 3,320 healthy adults from the HUNT Fitness Study (a subset of the much larger Nord-Trøndelag Health Study). Their equation was 211 minus 0.64 times age. It sits between Tanaka and the older Fox line, and it was specifically built on a recreationally active sample.

The HUNT formula tends to predict slightly higher max heart rate than Tanaka for fit adults. Nes also reported that body composition, sex, and self-reported fitness explained only small additional variance beyond age, meaning even rich-data models couldn't accurately predict the individual. Their take was the same as everyone else's. Age accounts for the dominant share of population-level variance, but plus-or-minus 10 beats per individual is a floor that no formula has cleared.

Shookster 2020: How Accurate Is Any of This?

The most honest summary of where the field landed came from a head-to-head accuracy test. Shookster and colleagues (2020) in the International Journal of Exercise Science measured true max heart rate in 99 healthy adults on a treadmill protocol that ended at volitional fatigue, then compared the measured number against eight published equations including Fox, Tanaka, Gellish, Gulati, Astrand, Nes, Arena, and Fairbarn. The mean bias of the Fox 220 minus age equation was smallest in this sample, but the individual error stayed near 10 beats per minute regardless of which formula was used.

That nuance is worth holding. At a population level, the choice between 220 minus age and Tanaka is mostly about which side of the true line you want to sit on. At an individual level, all of them are wrong by a chunk. The right reaction is not to chase a more precise equation. It is to stop expecting that precision from age alone.

Why This Matters for Your Fitness

The reason any of this matters is that almost every heart rate zone you see on a wearable is a percentage of your max. Zone 2 typically sits at 60 to 70 percent of max. Threshold work hovers around 88 to 92 percent. A VO2 max interval lands above 95 percent. So a 10-beat error at the top of the curve becomes a 6 to 7 beat error in every zone underneath. That's enough to convert a zone 2 session into a moderate slog, or to convert a real interval workout into one you "didn't push hard enough."

It also matters for the people most likely to read a fitness app number and act on it. Older adults, women, and previously sedentary new exercisers tend to be the populations the 220 minus age formula misses worst. A 65-year-old woman who walks briskly with her heart rate at 145 may be at the upper edge of comfortable aerobic work, not in a "dangerous" zone the watch flags as too high. That kind of misread can sour an entire fitness habit.

And there's a longevity angle. Aerobic capacity is one of the strongest predictors of all-cause mortality, a relationship we cover in detail in our piece on VO2 max and longevity. The fastest way to raise that capacity is to actually spend time at the right training intensities. Anchoring those intensities to a formula that miscalls your max by 10 beats can quietly hollow out months of training.

How to Use Heart Rate Zones in Practice

The practical answer comes in three layers, in order of accuracy. Use the formula as a placeholder. Use a wearable's auto-learned max as an upgrade. Use a real field test when the number actually matters.

The Formula Layer (Good Enough to Start)

If you have no other data, use Tanaka for a general estimate. 208 minus 0.7 times age. Women can use Gulati (206 minus 0.88 times age) instead. A recreationally active adult who runs or rides several days a week can try Nes (211 minus 0.64 times age) and see which prediction matches the heart rate they actually hit at hard efforts. Once a number is sitting in your watch, every zone derives from it.

The Wearable Layer (Better Within Weeks)

Most modern watches will quietly update the max heart rate they assume for you. Garmin and Apple Watch both track your highest recorded heart rate during workouts and adjust the number upward as your real ceiling appears. Whoop, Polar, and Coros do the same. So the easiest move is to do a few hard sessions, let the watch learn your actual peaks, and check what max it has settled on. That number is usually within a few beats of a lab-tested max.

The Field Test Layer (Most Accurate)

When you actually need to know your max, test for it. A common protocol on a treadmill or bike is a 15-minute easy warmup, then a 4 to 5 minute interval at the hardest sustainable pace, two minutes easy recovery, then a final 30-second all-out sprint. The peak heart rate you see during the sprint is a reasonable proxy for true max for most healthy adults. Repeat once a week or so until the number stops climbing.

A field test has limits. It assumes you are healthy enough to push to volitional fatigue without cardiac risk. It also assumes your motivation will let you find a real peak, which is harder than it sounds. If either is in doubt, ride the formula and let the watch learn your max over time. Both are more honest than memorizing a 50-year-old shorthand.

Common Misconceptions

Misconception 1: "There's one true max heart rate formula, and the rest are wrong."

There isn't. Each equation was built on a different sample and answers a slightly different question. Tanaka pooled mostly healthy adults across many studies. Gulati specifically sampled women. Nes drew from a Norwegian fitness-screened cohort. Shookster's head-to-head test showed that the "best" equation depends on whose treadmill you're standing on. Population-level winners exist. Individual-level winners do not.

Misconception 2: "My watch's heart rate zones are exact."

They're as exact as the max heart rate they assume. If the watch is sitting on a default of 220 minus age, every zone is off by whatever the formula's individual error is for your body. Most watches let you set a custom max, which is worth doing once you have either a field-test number or a wearable-learned ceiling. Otherwise the zones are an educated guess dressed up in color-coded bars.

Misconception 3: "If I hit a heart rate above my predicted max, something is wrong."

Usually it means the prediction was low. Healthy adults regularly hit numbers 10 to 15 beats above their age-based estimate. The "max" in max heart rate is a structural ceiling set by your physiology, not a danger threshold. If you're symptomatic (chest pain, dizziness, palpitations), stop and see a doctor. If you're just unexpectedly fit, take it as data and update your max upward.

What the Research Suggests Going Forward

The honest picture is calm. Age explains a meaningful share of the variation in max heart rate across the population. It does not explain the variation in your body specifically. Better equations exist, and the right one depends on whether you're using it for a general estimate (Tanaka), a female-specific number (Gulati), or a recreationally trained adult (Nes). All of them sit on top of an irreducible individual-error floor of about 10 beats per minute.

So the practical play is layered. Start with Tanaka. Let your watch learn your real peaks. If the number actually matters for how you train, do a field test. And whatever number you land on, remember it isn't a religious truth. Update it as your fitness changes. Update it again when your watch firmware does. The biology is durable. The formula is a tool.

The deeper takeaway is the one that bothers people. There's no single magic equation. Wearable companies sometimes imply otherwise, and the 220 minus age formula has had a long PR run for a number that was never meant to be clinical. The research community has moved on. The training community is catching up.

References

1. Tanaka H, Monahan KD, Seals DR. "Age-predicted maximal heart rate revisited." Journal of the American College of Cardiology 37.1 (2001): 153-156. doi:10.1016/S0735-1097(00)01054-8
2. Nes BM, Janszky I, Wisløff U, Støylen A, Karlsen T. "Age-predicted maximal heart rate in healthy subjects: The HUNT Fitness Study." Scandinavian Journal of Medicine & Science in Sports 23.6 (2013): 697-704. doi:10.1111/j.1600-0838.2012.01445.x
3. Gulati M, Shaw LJ, Thisted RA, Black HR, Bairey Merz CN, Arnsdorf MF. "Heart Rate Response to Exercise Stress Testing in Asymptomatic Women: The St. James Women Take Heart Project." Circulation 122.2 (2010): 130-137. doi:10.1161/CIRCULATIONAHA.110.939249
4. Robergs RA, Landwehr R. "The surprising history of the HRmax=220-age equation." Journal of Exercise Physiology Online 5.2 (2002): 1-10. Open access PDF
5. Shookster D, Lindsey B, Cortes N, Martin JR. "Accuracy of Commonly Used Age-Predicted Maximal Heart Rate Equations." International Journal of Exercise Science 13.7 (2020): 1242-1250. PMC7523886