Tanaka · Journal of the American College of Cardiology 2001 · Meta-analysis and cross-sectional laboratory validation study · n=19,226 (18,712 across 351 studies in meta-analysis; 514 in validation study)

Age-predicted maximal heart rate revisited.

Cited 3786 times in the scientific literature.

Level 1 - systematic review of randomized trials

Meta-analysis of 351 studies combined with a laboratory validation study

PubMed 11153730 · doi:10.1016/s0735-1097(00)01054-8 · record verified 2026-08-30

What was done

A meta-analysis was conducted using group mean maximal heart rate (HRmax) values from 351 studies involving 492 groups and 18,712 healthy subjects to develop a generalized age-predicted HRmax equation. The resulting formula was subsequently cross-validated in a laboratory-based study measuring HRmax directly in 514 healthy subjects.

What was found

HRmax was strongly inversely related to age in the meta-analysis (r = -0.90), producing the regression equation: HRmax = 208 - 0.7 x age. The laboratory validation study yielded a nearly identical regression line: HRmax = 209 - 0.7 x age. The relationship was not modified by sex or variations in habitual physical activity level. The widely used formula (220 - age) systematically underestimates HRmax in older adults.

Why it matters

The common 220 - age formula underpredicts HRmax in older adults, potentially causing clinicians to underestimate cardiovascular stress during diagnostic exercise testing and to underprescribe target training intensities. The revised equation provides an evidence-based replacement across adulthood.

Limits

The meta-analysis relied on aggregated group mean values rather than individual participant data. The findings are restricted to healthy adults and cannot be assumed to apply to individuals with cardiovascular disease or those taking heart rate-modifying medications. Prediction intervals and standard errors of the estimate were not reported in the abstract.

Cited by