Don’t be a slave to heart rate training zones!

Heart rate monitors may not be the best way to predict VO2 max and monitor overtraining

Article at a glance:
  • Evidence is presented for the suitability or otherwise of using heart rate monitors to predict VO2max, measure exercise intensity, determine training zones and monitor overtraining;
  • The importance of lactate measurement for assessing optimum training intensities is emphasised;
  • Examples are given where use of perceived exertion and other methods may be more suitable than using HRMs.

Athletes have used heart rate training successfully for years. However, slavish adherence to heart rate training zones won’t allow an athlete to reach his or her true potential. Gary O’Donovan and Romain Denis discuss the strengths and limitations of heart rate monitoring and heart rate training.

For centuries, heart rate monitoring consisted of placing an ear or a stethoscope on the chest. The heart’s electrical activity was first recorded in the 19th century and American biophysicist Norman Holter invented a portable device for recording the heart’s activity in 1961. The Holter monitor can record the heart’s electrical activity for 24 hours or more, but it is too large for recording heart rate during exercise. Watch-like heart rate monitors were first developed in the 1980s, and since then most endurance athletes have used heart rate training(1). It is claimed that today’s heart rate monitors can perform all sorts of wonders, from predicting maximal oxygen uptake to detecting overtraining; in this article we’ll examine the validity of these claims.

Prediction of VO2max

Strengths: Some heart rate monitors can be used to predict VO2max (maximal oxygen uptake). For example, it has been shown that the Polar S810 heart rate monitor can accurately predict VO2max, which is a strong predictor of endurance performance and the best measure of health-related fitness. In different tests, most participants’ estimated VO2max scores were within 5ml/kg/min of their directly measured scores(2). The S810 is also highly reliable, yielding consistent results test after test(1). What’s more, S810-predicted VO2max was favourably associated with cholesterol and blood pressure in a study of 3,820 men and women aged 25-64 years(3).

Limitations: In the example above, the Polar S810 predicts VO2max from age, height, weight, gender, self-reported physical activity level, and resting heart rate measurements. Strict adherence to testing procedures is required because, as the manufacturer acknowledges, a number of factors influence resting heart function, including noise, temperature, time of day, diet, alcohol, smoking, exercise, and pharmacological stimulants. Be aware also that self-reported physical activity level is prone to error.

It is also noteworthy that the heart rate monitor penalises the ‘fat fit’ by expressing VO2max relative to body weight. When we directly measure maximal oxygen uptake in our lab, we express the results relative to body weight and independent of body weight (millilitres of oxygen per kilogram of body weight per minute and litres of oxygen per minute, respectively). In our lab, for example, we might determine that a 110-kg middle-aged man has a VO2max of 4 litres per minute, which is excellent compared to an untrained man of the same age. When this value is expressed relative to body weight, the same individual has a VO2max of 36ml/kg/min, which is no better than an untrained man of the same age. Unfortunately, VO2max is rarely expressed independent of weight outside of the laboratory.

Progress monitoring

Strengths: A reduction in heart rate for a given intensity is usually indicative of an improvement in fitness. In our experience, previously inactive individuals and those returning from a period of injury are often delighted to see a reduction in heart rate for a given intensity after only four weeks of aerobic training.

Limitations: A number of factors other than a change in fitness might explain why heart rates can differ from one test to another; for example: natural biological variation is such that heart rate can vary by 2-4 beats/min from one day to the next; dehydration can increase heart rate by up to 7.5%; heat and humidity can increase heart rate by around 10 beats/min; and, altitude can increase heart rate by 10-20%, even with acclimatisation(4). Trained individuals are unlikely to experience discernible reductions in submaximal heart rates and improvements in fitness are best identified from changes in the blood lactate response to exercise (see ‘determining training zones’ below).

Measuring exercise intensity

Strengths: In normal individuals, there is a linear relationship between heart rate and intensity during incremental exercise(5). Therefore, exercise intensity can readily be expressed as a percentage of predicted or directly measured maximum heart rate. The near linear relationship between heart rate and oxygen uptake is such that exercise intensity can also be expressed as a percentage of VO2max.

Limitations: Maximum heart rate is best determined in a graded exercise test, but a maximal exercise test can be inappropriate, especially in less well-trained individuals. Maximum heart rate can be predicted from the formula 220-age, from the formula 210 – (age x 0.65), or from the formula 207 – (age x 0.7); however, all predictions are subject to error(5,6).The variation in heart rate among the normal population is such that the predicted maximum heart rate of 95% of individuals of a given age will lie within a range of 40 beats/minute(5).

It is an oversimplification to suggest that there is a linear relationship between heart rate and oxygen uptake. For example, oxygen uptake often increases relatively more than heart rate during high-intensity exercise. The relationship between oxygen uptake and heart rate can be predicted more accurately when oxygen uptake is expressed as a percentage of heart rate reserve (HRR, the difference between maximum heart rate and resting heart rate)(7). In fit people, for example HRR (beats/min) = 1.05VO2 – 4.1. Don’t concern yourself with these equations however; the next section explains that lactate threshold should be used to determine exercise intensity, not VO2 or heart rate.

Determining training zones

Strengths: Heart rate monitoring allows individuals to train at the intensity recommended to improve aerobic fitness, which is 50-90 % of maximum heart rate(8).

Limitations: Too many athletes don’t reach their true potential because they adhere to ill-conceived heart rate training zones. Untrained individuals will enjoy improvements in fitness at 50% of maximum heart rate. Trained individuals require more individualised exercise prescription.

Training zones are best determined from the blood lactate response to exercise. Figure 1 shows the blood lactate response to incremental exercise in a 46-year-old club cyclist. Notice that blood lactate does not increase above resting levels until power output reaches 210 watts. In this individual, exercise up to 210 watts is comfortable, sustainable, and ideal for long-duration or recovery training.

As exercise intensity increases, the cyclist recruits more fast-twitch muscle fibres and produces more lactic acid. His body attempts to buffer lactic acid by combining it with carbonic acid, a weaker acid that splits in the lungs into water and exhalable CO2. From 210-270 watts, the cyclist’s increased breathing is sufficient to expel CO2 and buffer blood lactate. In this individual, exercise at 210-270 watts is difficult, but it will improve his ability to tolerate and dispose of blood lactate.

At 270 watts, there is a rapid rise in blood lactate concentration. This is known as the ‘lactate threshold’. The increase in blood pH (acidity) stimulates a dramatic increase in breathing (as the body attempts to expel CO2) and exercise beyond the lactate threshold is not sustainable. Exercise beyond the lactate threshold is ideal for interval training, however. In order to improve endurance and 10-mile time trial performance in this individual, we might recommend three or four 5-minute bouts at 280 watts with 5-minute recovery bouts at 160 watts. In order to improve strength and sprinting speed, we might recommend 10 1-minute bouts at 330 watts with 3-minute recovery bouts at 160 watts.

Lactate and heart rates

The data in figure 1 were derived from a cyclist who had a power meter on his road bike. Thus, he was able to train at the correct intensity outside of the lab. Subsequent visits to the lab showed that his blood lactate curve shifted to the right and his power output at lactate threshold increased to 290 watts (see figure 2). This improvement in fitness would not have been detected if we had only used a heart rate monitor because his heart rate at each workload did not change over time.

Figure 1: Heart rate and blood lactate responses to incremental cycling

Figure 1: Heart rate and blood lactate responses to incremental cycling

If the cyclist did not have a power meter, we would have anchored much of his training around the heart rate at the lactate threshold. We would also have advised him not to expect his heart rate to recover between intervals and to ignore increases in heart rate of 5-15% during an hour of steady state exercise(4).

Figure 2: Blood lactate responses to exercise before and after 12 weeks of training

Figure 2: Blood lactate responses to exercise before and after 12 weeks of training

In runners, we often prescribe exercise intensity around the speed at lactate threshold rather than the heart rate at lactate threshold. In runners, cyclists and all athletes(4), it is important to stress that there is no predictable relationship between heart rate and lactate threshold. Lactate threshold tends to occur at around 90% of maximum heart rate in well-trained individuals, but it can occur at 50-90% of maximum heart rate(9).

When used in accordance with the author’s instructions, Borg’s(10) rating of perceived exertion (RPE) scale can also be used to determine exercise intensity outside of the lab because an RPE of 13-14 often occurs at the lactate threshold, regardless of gender, mode of exercise, and training status(11-13). We recognise that many individuals will find RPE too subjective, and will be reluctant to train without a heart rate monitor. However, the running and cycling workouts in table 1 and table 2 are designed to show that it is possible to train effectively without a heart rate monitor (or a power meter).

Table 1: Using RPE to train effectively without a heart rate monitor

Day Time Intensity (RPE) Distance*
Mon 60 min Light (10-11) 10-11km
Tue 45 min Hard (14-16) 6 x 400m in 1:37 (1:20 recovery between reps)†
Wed Rest
Thurs 40 min Light (10-11) 7km
Fri 60 min Hard (14-16) 4 x 1000m in 4:15 (2:45 recovery between reps)†
Sat Rest
Sun 45 min Moderate (12-13) 15 min at 13.33km/h (4:40 per km, about 3.4km)‡

Table shows the first week of a four-week training programme designed for an individual aiming to run 10k in 45 minutes (time is min:sec). Do one of the following at the start of week two and week three: slightly increase the number of reps; slightly decrease the recovery time; or, slightly increase the speed. The fourth week (racing week) should be an easy one, including some intense but very short sessions. *Using a cycle and cycle computer, measure a safe path or a safe loop that you are used to running, and place a mark at 400m and at every kilometre. †Moderate and hard sessions should be accompanied by a 20-minute warm-up and a 10-minute cool-down. ‡The moderate session can be performed all year round in order to improve your ability to sustain race pace. The duration of the moderate session should be increased as fitness improves, for example: 20 or 25 minutes at 13.33km/h in this context.

The running workouts can readily be adapted for faster runners (and pace calculators are available online; for example: A runner wishing to run 10k in 39 minutes should decrease the duration of the 400m intervals to around 1:22 (with 1:10 rest intervals), decrease the duration of the 1,000m intervals to around 3:45 (with 2:45 rest intervals), and run at 15.4km/h for 15 minutes (equivalent to 3:54 per km) during the weekly moderate session. The all-out, self-paced nature of the cycling workouts is such that they will automatically adjust to changes in fitness.

Table 2: Cycling workouts without heart rate monitoring

Session Objective Format
Boardman Specials(16) Improve 1-hour time trial performance Twenty to sixty 10-second efforts at 1-hour average power with 20-second recovery intervals
All-out sprints(17) Improve acceleration and ability to tolerate blood lactate efforts Eight to twelve 30-second all-out with 4:40 recovery intervals
Eric Snider’s weekly cadence workout(18) Cycle like Lance Armstrong to avoid fatiguing large gears! Spin as fast as possible for 1 min in easy gear such as 39 x 19. Shift back to your usual pace gear for two min. Repeat 6-10 times.
CTS Time trial Improve endurance and 10-mile time trial performance Three or four 5-minute sprints at around 110 rpm with 5-minute recovery intervals at 90 rpm

Each session is best performed on a stationary cycle or turbo trainer, and should be accompanied by 1) a 10-minute warm-up with a 1-minute sprint at 03:00 and a 1-minute sprint at 05:00, and 2) a 6-10 minute cool-down. The CTS Time Trial is adapted from the excellent Carmichael Training Systems Train Right Video Series, which is available at (each DVD includes a heart rate test that can be ignored – just go by perceived effort during the intervals).

Fat burning

Strengths: None.

Limitations: Many so-called fitness professionals would have you believe that a ‘fat-burning zone’ exists at 60-70% of age-predicted maximum heart rate. This is nonsense. It is true that the rate of fat metabolism is greater during moderate-intensity exercise, but fuel utilisation cannot usually be measured outside of the lab. What’s more, when fuel use has been measured in the lab, it has been found that the optimum intensity for fat burning is different in each individual, varying from 54-92% of maximum heart rate(14).

Some fitness professionals have purchased expensive gas analysers and are measuring fuel use outside of the lab. However, this still misses the point. Weight loss is induced by negative energy balance; you’ll burn more calories during high-intensity exercise than during moderate-intensity ‘fat-burning’ exercise. For example, an 80-kg individual will burn around 11 calories per minute when jogging at 5mph (12 min/mile), around 16 Calories per minute when running at 7mph (8.5 min/mile), and around 22 Calories per minute when running at 10mph (6 min/mile). Advocates of the fat-burning zone also fail to realise that high-intensity training is more effective in improving one’s ability to burn fat. This ‘carbohydrate-sparing’ effect does not accompany moderate-intensity training.

Preventing overreaching and overtraining

Strengths: Overreaching is characterised by signs and symptoms that last from a few days to two weeks, including fatigue, muscle soreness, insomnia and underperformance(4). Overreaching is often utilised in a training cycle because ‘supercompensation’ may occur after an appropriate period of recovery(15). It is thought that fatigue, performance decline, mood disturbance and other symptoms are more severe in the overtrained state than in the overreached state(4). An athlete may take months or years to recover from overtraining. Some top-of-the-range heart rate monitors can detect a 5-10 beat/min increase in resting heart rate and a decrease in resting heart rate variability (see figure 3), which might be early signs of overreaching and overtraining(4).

Limitations: There is increasing evidence that heart rate variability is lower at rest in the overtrained state, but there is no diagnostic tool for overtraining; the condition is ‘diagnosed’ by excluding all other explanations for the decline in performance and mood(4). Perform any overtraining test at least 24 hours after training because heart rate variability can be influenced by a prior exercise bout.

Figure 3: Heart rate variability

Figure 3: Heart rate variability

Heart rate variability describes the subtle variations in the intervals between consecutive heartbeats. Even when heart rate is stable, the time between two consecutive beats can vary considerably. At rest, heart rate variability is larger in aerobically trained individuals than in untrained individuals – probably as a result of healthy nervous activity in the brain’s cardio-inhibitory centre. There is increasing evidence that heart rate variability is lower at rest in the overtrained state. Low heart rate variability is also a risk factor for cardiovascular disease.


Heart rate monitors can be used to estimate VO2 max, to identify changes in fitness, and, possibly, to detect the early signs of overtraining. Heart rate training zones are meaningless unless they are identified from the blood lactate response to incremental exercise. Heart rate training is particularly inappropriate during interval training. Most well trained individuals don’t need a heart rate monitor to know what constitutes an all-out one-minute interval, an all-out 5-minute interval, or an easy recovery interval!

Dr Gary O’Donovan is a lecturer in sport and exercise medicine at the University of Exeter and an exercise physiologist accredited with the British Association of Sport  and Exercise Sciences (BASES)

Romain Denis is a research student in sport sciences at Brunel University and he is pursuing accreditation with BASES


  1. Sports Med 2003; 33(7):517-538
  2. Med Sci Sports Exerc 2000; 32(S308):abstract 1535
  3. Eur J Cardiovasc Prev Rehabil 2005; 12(2): 126-131
  4. Sports Med 2004; 34(14):967-981
  5. Exercise Testing and Interpretation, A Practical Approach. Cambridge: Cambridge University Press; 2001.
  6. Med Sci Sports Exerc 2007; 39(5):822-829
  7. Med Sci Sports Exerc 1997; 29(3):410-414
  8. Med Sci Sports Exerc 1998; 30(6):975-991
  9. Exercise and Sport Science Review. New York: Franklin Institute Press; 1982:49-83
  10. Borg’s Perceived Exertion and Pain Scales. Champaign, IL: Human Kinetics; 1998
  11. Med Sci Sports Exerc 1997; 29(10): 1332-1337
  12. Med Sci Sports Exerc 1987; 19(4):354-362
  13. Med Sci Sports Exerc 1991; 23(1):88-92
  14. Med Sci Sports Exerc 2002; 34(1):92-97
  15. The Physiology of Training. London: Elsevier; 2006:1-22
  16. Velo News Magazine 1996; 68-70
  17. Med Sci Sports Exerc 1999; 31(5):736-741
Share this

Follow us