TrainingPeaks Now Shows You Fat vs. Carb Burn — Here's What It Still Gets Wrong

The New Feature Everyone's Talking About

TrainingPeaks recently launched Fueling Insights, a feature built with Dr. Inigo San Millan that shows how much fat vs. carbohydrate you burn during a ride. It uses power meter data and a proprietary algorithm calibrated against 250+ lab-based respiratory exchange rate (RER) tests.

It's a genuinely interesting feature. For the first time, endurance athletes can see an estimate of their substrate utilization without booking a lab test. Your post-ride analysis now includes a chart showing when you were burning mostly fat, when you crossed over into heavy carbohydrate reliance, and how your metabolic efficiency stacks up against reference profiles.

If you're a data-driven athlete, this is exciting. But if you're hoping it will tell you what to eat during your next race, you're going to be disappointed.

What Fueling Insights Actually Shows You

Fueling Insights categorises your ride into one of four metabolic profiles based on your power output, duration, and the algorithm's RER estimates. It then shows you a timeline of fat and carbohydrate contribution to your total energy expenditure.

In practical terms, you'll see something like: "At 200 watts, you burned approximately 60% carbohydrate and 40% fat. At 250 watts, that shifted to 80% carbohydrate and 20% fat."

That's useful information. It tells you roughly where your crossover point sits — the intensity at which your body shifts from primarily burning fat to primarily burning carbohydrate. Athletes with better metabolic efficiency can sustain higher intensities while still burning a meaningful proportion of fat, which spares glycogen and delays the bonk.

The Three Things It Still Can't Do

1. It Won't Build Your Fueling Plan

Knowing that you burn 80% carbohydrate at race pace is a data point, not a plan. It doesn't tell you:

• How many grams of carbohydrate per hour you should consume
• When to take your first gel
• How to split intake between drink mix and solid fuel
• Whether your gut can actually absorb the amount you theoretically need

There's a meaningful gap between "you're burning 90g of carbs per hour at threshold" and "take half a gel every 20 minutes starting at the 25-minute mark, alternating with 200ml of drink mix." The first is interesting. The second is what you need pinned to your handlebars.

2. It Ignores Hydration Entirely

Substrate utilization and hydration are treated as separate problems in TrainingPeaks. But they're not separate during a race. Dehydration of just 2% body mass reduces your ability to absorb carbohydrate through the gut, shifts your heart rate upward (which changes your metabolic profile), and accelerates glycogen depletion.

Your fueling plan and your hydration plan are the same plan. A tool that shows one without the other gives you half the picture.

3. It Only Works for Cycling

Fueling Insights requires power meter data. That means it works for cycling and indoor cycling — and nothing else. If you're a runner, triathlete on the run leg, or ultrarunner, you can't use it. TrainingPeaks has said additional sport applications are in development, but as of August 2026, runners are locked out.

This matters because running and cycling have fundamentally different metabolic demands. Running produces more mechanical stress, more core temperature elevation per unit of work, and different gut motility patterns — all of which affect fueling needs. A substrate model calibrated on cycling data doesn't transfer directly to running.

What Your Data Actually Tells You (That These Tools Miss)

The substrate utilization chart from TrainingPeaks is one input into a much larger equation. Here's what else matters for your race-day fueling plan — and what your wearable data can already tell you, if something translates it:

Your sweat rate varies by condition and fitness. A runner who sweats 1.2 litres per hour in 15°C weather might sweat 1.8 litres per hour in 28°C. That shift changes sodium needs, fluid volume, and gut absorption capacity. Your Garmin or Apple Watch estimates sweat loss. But neither one adjusts your fueling plan based on it.

Your heart rate tells you more than your substrate ratio. Heart rate drift during a long run signals dehydration, which signals reduced gut absorption, which means the 60g of carbs per hour your plan calls for might be too much for your stomach to handle in the final third of a race. A smart fueling plan adjusts dynamically. A substrate chart from last Tuesday's ride doesn't.

Your training load history predicts glycogen availability. If you ran 90km last week and your training load is elevated, your glycogen stores at the start of your next long run may be lower than usual. That changes how aggressively you need to fuel early. Your watch tracks this. Your fueling plan should respond to it.

Gut tolerance is trainable — but it has a ceiling. Research supports training the gut to absorb up to 90–120g of carbs per hour using glucose-fructose transporters. But your personal ceiling depends on training history, body weight, exercise intensity, and how consistently you've practised race nutrition. No platform currently tracks gut training progress or adjusts recommendations based on it.

The Missing Layer

Here's the pattern across every major platform in 2026:

• Garmin Connect+ now tracks food. It logs your calories, macros, and connects them to your training load. But it's daily nutrition tracking — MyFitnessPal inside your Garmin. It doesn't build a race-day fueling plan.

• TrainingPeaks now shows substrate utilization. It tells you your metabolic profile and how much fat vs. carb you burn at each power zone. But it's a post-ride chart, not a pre-race plan. And it's cycling only.

• Nix biosensors now has an AI-powered intelligence layer for enterprise clients. It gives real-time sweat sodium and fluid loss data during training. But it requires a $100+ hardware patch, it's positioned for professional teams, and it doesn't integrate with your fueling plan.

Every one of these platforms adds more data. None of them connect the data to a decision.

The gap isn't information. Endurance athletes in 2026 have more physiological data than any generation before them. The gap is translation: taking sweat rate, heart rate, training load, and metabolic profile and converting all of it into "drink this much, eat this, at this time, in this format."

That's the layer Sweatr builds. It pulls from your Apple Watch or Garmin, syncs your Strava activity data, and produces a personalised hydration and fueling plan that adjusts based on your actual training — not a lab average or a one-size-fits-all calculator.

When Substrate Data Becomes Useful

None of this means substrate utilization is useless. It's genuinely valuable information — when it's one input into a complete fueling model, not the whole model.

Knowing your crossover point helps you understand why your fueling needs change at different paces. An easy long run at 65% of max heart rate burns significantly more fat than a marathon-pace tempo run at 82%. That means your carbohydrate needs per hour are lower on easy days and higher on race day. If your plan doesn't account for that, you're either under-fueling on hard days or wasting gels on easy ones.

The problem is when athletes see a substrate chart and think the fueling question is answered. It's not. It's barely started.

What to Do With This

If you use TrainingPeaks and have a power meter, look at Fueling Insights. It's a useful data point. Notice where your crossover sits relative to your race pace. If you're burning almost entirely carbohydrate at your target intensity, you know your fueling plan needs to be aggressive — 60–90g per hour, minimum.

But don't stop there.

Test your actual fueling in training. Use your long runs and long rides to practise the specific products, quantities, and timing you'll use on race day. Weigh yourself before and after to estimate sweat loss. Note what your stomach can handle at race intensity vs. easy pace. Track the conditions — temperature, humidity — because they change everything.

Then find a tool that takes all of that data and builds you a plan. Not a chart. Not a dashboard. A plan that tells you what to consume, when, and how much — personalised to your body, your event, and the conditions on race day.

Your watch already collects the data. You just need something that knows what to do with it.