Can be, but I think the bigger companies must have some learning/ science behind them.
I can only say I had expected weight changes using Garmin calorie āburnā over a number of devices. When Iāve compared āactiveā calories in connect compared to similar spins with power, theyāve been pretty close.
I do have a reasonable estimate of my max heart rate though, which I think is the bit thatās missing for a lot of people, who just default to 220-age.
Could be device too though. When I had a Wahoo Bolt, you could set it to Work in Kj = Calories. I havenāt found that option since I changed back to Garmin, and it includes non-active calories as well, so can be a few 100 higher than the āworkā. Total calories is what defaults to strava.
Anything calorie related without a power meter is pointless guesswork. Strava reads approximately double if I donāt connect my PM.
On the head unit
Power/kj
It depends if you mean total calories or active calories
I have it set on my head unit, but iirc with the bolt you could have it kj = calories on the fit file. So when the ride syncād, KJ = Calories was the default into Strava/ MyFitnessPal
So, a lot of people in this thread equate kJās to calories, and while that works decently well, itās not actually accurate. In this article CyclingNews āDiet of professional cyclistsā, the author Alan McCubbin briefs over the fact that while most people assume a gross mechanical efficiency of 23.9% in cyclists, most scientific data puts it closer to 20.7%. After doing a bit more PubMed reading than I was planning to, most of what I can find puts cycling efficiency around that 20.7% number, and those that fall outside of that actually tend to find a lower efficiency.
While 3 percentage points may seem like a small difference, as riding time increases, it becomes quite significant. For example, my endurance ride today was 3 hours at an average of 212 watts - 2,290 kJās - or assuming a gross mechanical efficiency of 23.9% - 2,290 calories. However, if the number of calories is calculated using the more scientifically accepted efficiency of 20.7%, the calorie burn would be 2,644 calories. Thatās a difference of about 350 calories, or 15%, on an endurance ride. The higher the average power for a ride, the larger the caloric discrepancy will be. Although, it should be stated that most studies seem to find a slight increase in efficiency as power increases. Still - itās not even near 23.9%.
Conclusively, assuming a GME of 23.9% may work for athletes riding at lower power or shorter durations. Riding for 6 hours per week at an average power of 180W will yield an assumed calorie burn of 3,890 calories for the week, as supposed to 4,470 calories with a GME of 20.7%. This discrepancy of roughly 600 calories is unlikely to create a harmful caloric deficit or lead to any negative health consequences. However, for more powerful riders that spend much more time in the saddle, the 15% discrepancy in caloric burn is of much greater consequence. A rider with an FTP of, say, 330W, riding 20 hours a week at an average of 245W will perform 17,640 kJās of work in that time. If work is equated to calories, and the athlete calculates their food intake thereafter, they risk creating a caloric deficit of about 2,650 calories over the week. That is quite large, and with time, it certainly risks putting the athlete in a relative energy deficit with all that entails. This is, of course, only the case if the athlete tracks food intake and adjusts based on activity, but in the world of endurance sport, that is far from uncommon.
Maybe this is a bit too much of a deep dive for the thread, but I hope it can provide some useful information, at least for those of us that carry very high training loads and/or FTPās.
Some references:
[Cycling Efficiency in Trained Male and Female Competitive Cyclists - PMC]
[https://www.frontiersin.org/articles/10.3389/fphys.2018.00713/full]
When I had a met cart test done I was at 17% efficiency at low intensity and peaked out at just over 20%. At the intensity that I ride most of the time it was 19%. A huge difference from the efficiency that TR assumes. One question though: if you are operating at a lower efficiency is the difference made of carbs, fat or a combination of the 2.?
From what I found in studies comparing elite athletes to amateurs, this is actually the most pronounced difference between the two groups. Contrary to popular, and logically reasonable belief, professionals donāt seem to display a greater gross mechanical efficiency than those less trained, but they are able to oxidise fat more efficiently. Typically, the less trained an athlete is, the āearlierā they will start to metabolise glucose. Therefore, I suppose the ādifferenceā in efficiency at lower and higher intensities is highly dependent on athlete fitness. It is likely safe to assume that it is a mix of the two, but without individual testing, pinning it to a percentage would be quite difficult.
Iām not massively well read on FatMax training, but I suppose that may lead you to some answer. Since FatMax is the intensity as which you will oxidise the maximum amount of fat before switching to primarily carbohydrates, any difference in GME below FatMax should (?) presumably be āmade upā of fat, and any difference above FatMax should be āmade upā of carbohydrate. This is a bit of speculation on my end, but it would make some logical sense by examining what FatMax training is trying to achieve. Iāve seen FatMax pinned somewhere between 75-82% of FTP depending on the athlete.
This is beside the point, but it may still interest some that annoyingly, it also seems that GME decreases with time. Thus, as the duration of a session increases, we become less efficient. This study Duration and GME showed that this can be mitigated quite well with carbohydrate supplementation, so +1 for proper all the talk of fueling I guess!
Thank you. Iād often wondered how close this convenient conversion factor is to reality.
Does anyone know if any of the major platforms attempt to make the more accurate calculation when converting kJ to kcal?
Thats really interesting. What sort of training did he do in that 12 months?
I recall reading somewhere that efficiency is closely linked to muscle fibre type. The more fast twitch you have, the lower your efficiency.
Its all very fascinating but where is the real world usage? Most people donāt know their BMR, its a guestimate. Most food is labelled within a range, the weight is marginably variable, the content is marginably variable, the calculation for cals, the way it digests etc is marginably variable. So taking the extremes of all of these variables, as well as your own efficiency variable, give us a huge window. Its all ballpark at best, you then have to refine it using your own trial and error.
As an āaverageā human being, Iād much rather TR slightly under estimated the calorie burn anyway as most people could do with eating a little less.
Most equate kJ to kC even though it is a sloppy estimate. I brought this up years ago to the TR team that it is rare to have the approximately 25%efficiency for that to be true. Most people will burn more than that, but without testing data you wonāt really know what it is.
The real world usage is quite limited for most TR athletes or recreational level cyclists. However, as I mentioned, it is much more useful, or at least worth while, for higher level athletes. For someone with an FTP of 350, training 20 hours a week at an average of 260 watts, the difference in kJ and calories may be upwards of 3,000 calories. Tracking calorie intake is actually not that difficult, and even though there is margin for error, most would struggle to be off by 3,000 calories in a week without obvious negligence.
I believe the discrepancy between calories consumed based on assumed and actual calorie burn may induce symptoms of RED-s or overtraining if allowed to continue over long periods of time. Since hunger tends to be suppressed by large training volumes, simply eating to hunger may not work for all. Thus, calorie tracking may be implemented, but if the calorie expenditure is off by a large margin week after week, problems may arise.
Perhaps there could be a market for selling us doubly labelled water kits to track true calorific expenditure
I would guess that itās only in recent years that fuelling the work has really come to more prominence, so undershooting probably suited (particularly in a weight loss context).
On a quick look, even the Garmin Calorie āburnā which includes non-active calories is less than 15% more than active calories/ kj work. Strava adjusts the TR kjās down further from the actual work done!
At the moment, an underestimate helps me as Iām trying to adjust a little before race season. When I have been in longer term maintenance, Iāve adjusted intake to remain steady.
Also, actual power meters arenāt 100% accurate either!
That is true, power meters tend to be off by .5-1.5%.
Personally, I carry quite a high training load (around 18-20 hours per week atm), and I found the insights about GME to be very useful to my training. While I tend to have a good appetite and fuel well on the bike, I often find myself struggling to recover between workouts as the season progresses. By acknowledging the additional calories burned, Iāve purposefully made myself snack more and eat even more around training, even though Iām not necessarily that hungry and will get by fine until the next meal. By doing so, I have found that I recover much quicker from taxing sessions and feel better throughout the day. I would estimate that I eat an additional 300-700 calories more than I would typically. This correlates quite well with the 15% increase in caloric expenditure over kJās. It should also be noted that I have not gained any weight by implementing this over the past 4 weeks, I still maintain around 140 lbs at 5ā10.
Hmm, you donāt need a calorie number to know whether you are eating enough. Thereās nothing new in ensuring you eat enough for the lifestyle you lead. The simplest measure has always been; is your weight stable or increasing or decreasing.
