How reliable is this study for the relationship between heart rate and calories burned?

I am talking specifically about a paper called Prediction of energy expenditure from heart rate monitoring during submaximal exercise.

It states that within the narrow range of 90-150 bpm, there is a linear correlation between heart rate and calories burned as follows:

``````EE = -59.3954 + gender x (-36.3781 + 0.271 x age + 0.394 x weight + 0.404
V[O.sub.2max] + 0.634x heart rate) + (1 - gender) x (0.274 x age + 0.103x weight + 0.380x V[O.sub.2max] + 0.450 x heart rate)
``````

However, I have my doubts about this. In particular, how come this formula doesn't factor in the metabolic equivalent of task (MET)? Surely the exercise intensity plays an important role in how calories are burned, or is that just purely reflected through heart rate and V02max?

What I find strange is that it takes very little workout intensity to get my heart rate very high. Just today I ran on the treadmill at a mere 4.7 mph (7.5 kph) for an hour at a sustained heart rate between 145 and 155. Going by the HR formula, I burned around 1033 kcal, whereas running at this speed yields an MET of about 8.3, which yields 772 kcal (my MET is 93). That's not even close.

As another example, a game of basketball has me easily averaging a 160-170 HR (even with all the whistles, timeouts, stop & go, etc) and my HR will sustain 20-30 BPM above normal for at least another 30 mins or so after the game.

I don't think I'm necessarily out of shape. I managed to achieve a VO2max above 50 a week ago via the cooper test. I'm a big guy at 6"6, 245 lbs, roughly 20% body fat. Is it strange that my HR increases so fast from running? Does it make me an outlier with regards to this formula? Should I rely on MET values more than this HR formula for estimating caloric burn during exercise?

• Well the study only includes running and cycling, so any results are only valid to those forms of exercise, the more similar to them, the more accurate your calculations will be, but if you apply the formula to, for example, weight lifting, you'll be completely off. Also remember to subtract your bmr from the total number. Commented Jun 30, 2015 at 8:57
• If you fit into the population described in the study (your fat and mass are at the top end) then that regression equation claims to be fairly acurate, from the stated: residual (r) values (r = 0.836, P < 0.0001). The paper even gives the found deviations in Kj / min Commented Jul 1, 2015 at 0:24

Short while ago i did dinghy sailing for a couple of hours. The wind was strong but i was sailing with little effort under reduced sail. My heart rate was very high not because of physical exertion but out of alertness. My watch predicted that i burnt 1.800 calories in two and a half hours which is waay off. The formula does not include factors driven by excitement, danger, fear etc.

The analysis suffers from causality.

A measured amount of exercise to produce a heart rate does not necessarily mean the reverse.

I see that false statistic used for hot Yoga burns more calories as the heart beats faster. In heat the blood vessels in the surface of the skin dilate and blood flow increases to help cool the body. This is not blood that is delivering calories to a muscle. All heart beats are not the same. As user16435 stated other factors include excitement, danger, fear.

If it is heart rate from physical exertion then it is a good measure.

• "If it is heart rate from physical exertion then it is a good measure." What about heart rate increases after exercise? If your post-workout HR is 20 bpm above your normal heartrate and this is sustained for an hour, does this amount to additional calories burned from an increased metabolic rate? If so, does it still operate according to the paper's formula? Commented Jul 25, 2015 at 16:34
• @Amalgam54 I don't know. The blood is recovering lactic acid and I suspect it is also delivering some fuel. Commented Jul 25, 2015 at 16:38
• @Frisbee - Lactate is a fuel.
– JohnP
Commented Jul 27, 2015 at 19:51
• @JohnP OK. I meant the waste product from muscle that must be carried off. Do you know what that is? Commented Jul 27, 2015 at 19:58
• @Frisbee - It's still lactate, ethanol, CO2 and H2O, but it can be converted back into pyruvate or glucose for fuel use. The lactate threshold is where exercise is intense enough that it starts to accumulate in the blood rather than being carried off. This is different for everyone, and can be influenced/increased by specific training.
– JohnP
Commented Jul 27, 2015 at 20:55

To expand on my earlier comment:

If you fit into the population described in the study (your fat and mass are at the top end) then that regression equation claims to be fairly accurate, from the stated: residual (r) values (r = 0.836, P < 0.0001). The paper even gives the found deviations in Kj / min – arober11 Jul 1 at 0:24

But with all statistical models you could be the outlier, in the 16.4% the model does not fit, and there are a number of factors that have been found to affect the HR to VO2 relationship, have a read of: THE USE OF GPS TO PREDICT ENERGY EXPENDITURE FOR OUTDOOR WALKING, James Michael McKenzie, 2007, p:12-13 specifically:

Like accelerometry, HR monitoring is relatively non-invasive and simple for the subject, requiring only the use of a HR monitor chest strap and wristwatch receiver unit. Several studies validated the effectiveness of HR monitoring for estimating EE in free living activities (Brage et al., 2003; Hiilloskorpi, Pasanen, Fogelholm, Laukkanen, & Manttari, 2003; Livingstone, Robson & Totton, 2000; McCrory, Mole, Nommsen-Rivers, & Dewey, 1997; Schutz et al., 2001; Treuth, Adolph & Butte, 1998). As with accelerometers, HR monitors have the advantage of collecting data for long periods of time (several days, based on memory capacity). Heart rate monitoring is not without limitations though. Brage et al. (2003) noted that there could be substantial variation in HR between subjects that must be controlled with a calibration of the HR-VO2 curve for each subject for a given activity. Other variables may also influence HR, such as stress, ambient temperature, relative humidity, dehydration, and illness (Schutz & Deurenberg, 1996; Spurr et al, 1988). Heart rate can estimates EE very well from moderate to high intensity exercise (HR of 110 BPM to 85% HRMAX) due to the linear relationship between HR and VO2 in this range. Heart rate does not estimate EE well at low intensity physical activity because of the small changes in HR relative to VO2 from rest to low intensity activity (Ainslie et al., 2003; Hiilloskorpi et al., 2003; Livingstone et al., 2000; Schutz et al., 2001).