Heart-rate and effort-based training
Evidence: moderate
Heart rate is a useful but noisy guide to intensity. Estimated maximum heart rate is unreliable, the popular zone models disagree because they are anchored to different things, and heart rate lags and drifts within a run. Anchor zones to a threshold field test rather than an age formula, cross-check against perceived effort and the talk test, and treat heart rate as a directional trend. For the very fit, a heart-rate cap stops being the thing that limits easy running.
Maximum heart rate is a weak anchor
Most heart-rate zones are built from an estimated maximum, which is the weakest link. The familiar “220 minus age” has no firm basis and misestimates badly; “208 minus 0.7 times age” predicts better, but even the best formula carries an individual spread of around 10 bpm (Tanaka et al. 2001). Measured in runners, that scatter means an age-estimate can be 10 to 20 bpm wrong for a given person (Nikolaidis et al. 2018). A runner whose true maximum is a full zone away from the formula will train in the wrong zones all the way down.
Why the zone models do not agree
A second problem is that “Zone 2” means different things in different systems, because the models are anchored to different reference points and then sliced differently.
- Percentage of maximum heart rate. The common five-zone watch model takes fractions of HRmax. It inherits all the error in the HRmax estimate.
- Percentage of heart-rate reserve (Karvonen). Intensity is set as a fraction of the span between resting and maximum heart rate, added back to resting. This tracks oxygen-uptake reserve more closely and gives a higher target heart rate than the same percentage of maximum (Karvonen et al. 1957).
- Percentage of threshold heart rate. Friel’s running zones anchor to lactate-threshold heart rate, estimated from a 30-minute time-trial, so the boundaries sit relative to the threshold rather than the maximum (Friel).
- Threshold-anchored three-zone model. The model used in most training-distribution research defines zones by the two physiological thresholds: easy below the first, moderate between them, hard above the second (Stöggl & Sperlich 2015). The seven-zone power model is finer still.
The practical upshot: a “Zone 2” set as a percentage of maximum heart rate is not the same intensity as the “Zone 2” between the two thresholds, and neither matches a reserve-based or threshold-heart-rate zone. Comparing zone numbers between a friend, a watch and a study is meaningless unless they share a model. Anchor to a lactate or threshold field test where possible (see lactate testing), and read any zone scheme as belonging to its own model.
Heart rate lags and drifts
Even with good zones, heart rate is a slow, drifting signal. It takes minutes to catch up to a change in pace, so it is poor for short intervals. Over a long run it climbs at a fixed pace as core temperature rises, fluid is lost and stroke volume falls, the phenomenon of cardiovascular drift (Coyle & González-Alonso 2001). Holding a fixed heart-rate cap on a long or hot run therefore forces a progressive slowdown. The flip side, comparing the pace-to-heart-rate relationship early and late in a steady run, is a rough field marker of aerobic durability.
Effort is the cross-check
Because heart rate is noisy, perceived effort is the anchor that travels everywhere. The talk test is a reliable field marker: the point where speaking in full sentences becomes uncomfortable falls in the threshold region, so easy running should leave you able to hold a conversation (Quinn & Coons 2011). Rating of perceived exertion correlates moderately with lactate and heart rate and is good enough to steer sessions, if imprecise (Chen et al. 2002). Heart rate, the talk test and RPE used together catch each other’s errors better than any one alone, and reduce the temptation to trust the watch number over how the run feels.
Where Zone 2 runs out
Low-intensity “Zone 2” running is a sensible base-building intensity and a good cap on easy-day load, but the popular claim that a narrow Zone 2 is uniquely optimal for mitochondria and fat-burning outruns the evidence; for lower training volumes, higher intensities are at least as effective (Storoschuk et al. 2025; San Millán & Brooks 2018).
There is a further, less obvious limit for the very fit, and it is why a heart-rate cap eventually stops being the useful constraint on easy running. As aerobic fitness rises, the pace that corresponds to an easy heart rate gets fast. Metabolically the runner handles it with ease, low lactate, plenty of reserve, but that fast pace lands a lot of impact. For highly trained runners the binding constraint on easy days shifts from the metabolic side (heart rate and lactate, which are no longer limiting) to musculoskeletal load and total recovery. Elite marathoners race with almost no cardiac reserve, so even their decoupling late in a race reflects slowing down rather than a heart-rate ceiling (Takayama & Aoyagi 2025). The implication, that easy-day pace for the very fit should be governed by feel, recovery and impact load rather than by chasing a Zone 2 heart rate, is reasoned inference from the elite-reserve data and the bone-stress load literature rather than a directly tested finding, but it is why elite easy days are run well below the metabolic ceiling and capped by durability and recovery, not by a number on the wrist.