TrainingPeaks’ New ‘Fueling Insights’ Promises to Help You Estimate Your Calorie Burn (but I Have Some Caveats)
Knowing how to eat properly is critical for endurance athletes—ask me how I use Gu to run marathons . However, it took a lot of trial and error (and more than a few stomach issues) before I figured out how and when to fuel on long runs. I’d love to have some guidance as I train, or at least some baseline data to draw on for insight.
Popular fitness app TrainingPeaks has launched an intriguing — and perhaps underrated — feature that could help with just that: Fueling Insights. Developed in collaboration with exercise physiologist Dr. Iñigo San Millán, the tool promises to “transform your power meter data into accurate estimates of carbohydrate and fat burn” during cycling workouts. It’s an innovative, science-backed solution that’s likely to be misinterpreted by the very athletes it’s intended for.
Promise and Science
During exercise, muscles require a lot of fuel. The calories they burn come from two main sources: fat and carbohydrates. Fat is the primary source of energy at rest and during low-intensity exercise, but the more intensely you exercise, the more carbohydrates your body needs in addition to fat. The body does not have long-term carbohydrate stores like fat, so athletes pay attention to carbohydrate nutrition before competitions and make sure to consume carbohydrates (in the form of gels, drinks, and sometimes even solid food) during long rides or runs.
The amount of carbohydrates you need during these workouts, of course, depends on how much you burn, so doing some precise calculations will help you determine how to eat. Elite cyclists do exercise tests in the lab to assess their energy expenditure, but most runners and recreational cyclists do nothing of the sort.
The concept behind TrainingPeaks is truly impressive: instead of relying on general calorie estimates based on weight and heart rate, the Fueling Insights system uses data from more than 250 lab-calibrated metabolic tests conducted at San Millan using respiratory exchange rate (RER) measuring equipment. The system classifies athletes into four metabolic profiles based on fitness level and gender, then uses power data to estimate carbohydrate and fat oxidation in real time.
The methodology is sound in principle. San Millan’s research found patterns: recreational athletes burn fuel differently than elite cyclists, and these differences are predictable enough to form the basis of reliable models. The higher the power output, the more the body switches to burning carbohydrates instead of fat—an accepted physiological principle that the algorithm attempts to quantify in grams per hour.
When Lifehacker senior health editor Beth Squareski took a VO2max test, her lab data backed it up, showing when she was burning carbs and when she was burning fat, and how much of each. Her breathing rate indicated that during the few minutes she walked to warm up, her body was burning mostly fat. When the test transitioned to jogging, she was burning about 2 to 3 grams of carbs per minute, with 85 percent or more of the fuel coming from carbs and 15 percent or less. In the last few minutes of the test, when the workload was particularly intense, her carb burn was around 3 to 4 grams per minute.
TrainingPeaks shares how San Millan’s work with Tour de France riders helped develop the modern high-carbohydrate strategies that have become standard in professional cycling. His lab results challenged conventional wisdom and increased the recommended intake from 35–55 grams per hour to 80–125 grams per hour.
How useful is this data?
Here, enthusiasm must be balanced with a healthy dose of skepticism. TrainingPeaks essentially extrapolates lab data collected during controlled ramp tests to predict energy expenditure during real-world workouts at varying intensities, environmental conditions, and individual metabolic states. That’s a big step to take with confidence.
The company acknowledges some key limitations, though perhaps not enough. The model assumes a “fed state,” which may be less accurate when training in a fasted state. It’s currently only designed for cycling, as running power meters don’t have the same accuracy or adaptability as cycling power meters. And, crucially, the carbohydrate burn numbers it gives reflect the amount of muscle being oxidized , not the amount needed to be consumed while riding.
This last point can’t be overstated: When Fueling Insights tells you that you burned 600 grams of carbs on a four-hour ride, it doesn’t mean you need to consume 600 grams of carbs during that ride. TrainingPeaks recommends replacing about 50% of the carbs burned, given your existing glycogen stores and digestive limitations. That doesn’t make the headline half as catchy, but hopefully cyclists will understand the nuance.
Fat-burn numbers also shouldn’t be taken as recommendations for how much fat to eat. These energy estimates are based on your body’s internal processes—what calories it takes in and where they come from during exercise—not on the macronutrient balance of your overall diet.
How Data Can Shape Your Habits
TrainingPeaks acknowledges that its numbers can be easily misinterpreted, and has created a helpful FAQ to address this. However, the explanations seem inadequate for the complexity they introduce: the average cyclist, seeing “800g carbohydrate burned,” is likely to either panic about an energy deficit or attempt to consume an insane amount of carbohydrates, which could lead to serious gastrointestinal distress.
I will add that the decision to split users into four metabolic profiles adds another layer of potential confusion. Athletes are expected to choose their category, but the distinction between “competition” and “trained” is not always clear-cut, and getting it wrong can significantly skew your recommendations.
While Fueling Insights is currently limited to cycling, I hope it will help runners, too. The principle is the same: higher power output generally means more carbohydrate oxidation. However, running power measurement remains less standardized and precise than cycling power, so there are questions about accuracy. Plus, if I know runners, I know their fueling strategies are individual and specific.
Bottom line
The science behind Fueling Insights is impressive. For coaches working with elite athletes, the tool can provide valuable insight into training stress and nutrition strategies over time. However, athletes and coaches should view these numbers as estimates rather than exact recommendations. The tool is most effective when viewed as part of a broader nutrition strategy, rather than as a replacement for individual experimentation and common sense.
For now, perhaps the most promising application may be comparative analysis—tracking how energy needs change depending on the type of training, identifying particularly carbohydrate-intensive sessions, or monitoring trends over a training block. It’s possible that some relative data may be even more valuable than absolute numbers.
As with any new fitness technology, the smart approach is cautious optimism. Consider new ideas, but don’t let them override years of proven nutrition strategies and personal experience. And remember: no algorithm, no matter how complex, can replace the fundamental trial and error of figuring out what works for you.
Fueling Insights is currently available to TrainingPeaks Premium users for cycling training using power meter data.