Should you train fasted as an endurance athlete?

What does fasted training actually mean?

In popular usage, 'fasted training' usually means training first thing in the morning without eating — a pre-breakfast run or ride. Physiologically, this roughly corresponds to training with depleted liver glycogen (which drains overnight because the brain keeps pulling glucose from circulation) but usually with reasonably full muscle glycogen (because muscle glycogen only drains meaningfully through exercise, not sleep).

The sports-science literature uses a more precise vocabulary. 'Train-low' refers to training with low carbohydrate availability in muscle, liver, or both. It includes multiple specific protocols: 'fasted' (train before breakfast), 'sleep-low' (do a hard session in the evening, don't eat carbs after, do an easy session the next morning still in a depleted state), 'twice-a-day' (do two sessions close together with limited carbs in between), and 'low-carb diet training' (chronically low-carb diet + all training done low). These are not the same thing, and they have different effects and different risks.

Most amateur athletes who say they 'train fasted' are doing the first and simplest version — a pre-breakfast easy run or ride. This is the version with the most favorable research base, for the right population, at the right intensity. It is not the same as doing sleep-low for weeks on end or living on a chronically low-carb diet, both of which carry more risk for less benefit in the controlled studies.

What does the research actually support?

The mechanistic case for train-low is real. When muscle glycogen is low, several signaling pathways that drive aerobic adaptation are amplified — in particular the AMPK / PGC-1α pathway that triggers mitochondrial biogenesis and the CaMKII pathway that enhances fat oxidation. Training in this state theoretically produces a larger adaptive signal than training with the same workload in a fed state.

Controlled studies going back to Hansen et al. (2005) and elaborated by Hawley's group consistently show that train-low protocols produce measurably greater increases in mitochondrial enzyme activity, fat oxidation rates, and certain endurance markers compared to train-high protocols with the same training load. The effect is real, it's repeatable, and it's been demonstrated in both untrained and well-trained athletes.

The harder question is whether that amplified adaptive signal actually produces better performance. The research here is more mixed. Most controlled studies have failed to show a reliable performance advantage for train-low over train-high when total training load is matched — the extra molecular signaling does not always translate into race results. A 2017 review by Burke concluded that the performance evidence 'does not consistently support the molecular evidence', and that train-low should be treated as a tool to target specific adaptations, not a general-purpose upgrade.

The current consensus in the sports nutrition community is roughly: train-low enhances some aerobic adaptations, especially fat oxidation and mitochondrial content, but the performance benefit is uncertain and context-dependent. It is a useful tool in specific hands for specific athletes. It is not a free upgrade to everyone's training, and it is not the right default.

Why does fasted training usually hurt hard sessions?

The other side of the train-low story is that it works for low-intensity aerobic sessions and actively fails for high-intensity ones. The mechanism is simple: high-intensity exercise depends almost entirely on carbohydrate fuel. When you push above the aerobic threshold into tempo, threshold, or VO2max territory, your muscles are burning glycogen at rates that cannot be sustained by fat oxidation alone. If glycogen is low, you cannot hit the watts or the pace, you cannot sustain the effort, and the training stimulus of the session degrades.

Controlled studies of fasted high-intensity training show exactly this pattern. Athletes can't hold target power in fasted intervals; perceived effort is higher; total work completed is lower; and the adaptive signal from the session is actually worse than the same session done fed. You're getting less training out of the same time, with worse data and worse outcomes. The cardinal sin of fasted training is using it on the hard days.

This is why the practical protocol, when fasted training is used at all, is to apply it only to easy Zone 2 sessions. Zone 2 can be done from fat oxidation alone for most trained athletes; it's by design aerobic enough that muscle glycogen isn't the limiting fuel. Doing an easy Zone 2 run or spin before breakfast is biochemically coherent. Doing a VO2max session before breakfast is not, and the data is unambiguous about it.

The honest rule: fasted for easy, fueled for hard. Every train-low expert will say some version of this, and it's the single most important practical takeaway from the literature.

Who should never train fasted?

This is the section that popular fasted-training content usually skips, and it matters enormously. Fasted training is not a neutral choice for every athlete. For meaningful populations, it is contraindicated by the evidence and by basic physiology.

• Athletes at risk of relative energy deficiency in sport (RED-S). RED-S is a syndrome of chronic under-fueling that disrupts menstrual cycles in female athletes, lowers bone density, weakens the immune system, degrades recovery, and in severe cases causes long-term hormonal damage. Fasted training adds energy-availability stress on top of whatever daily deficit an athlete may already be running. For an athlete already at risk, this tips the balance in the wrong direction.
• Female athletes who are already prone to under-fueling, which is a large population. Research by Mary Jane De Souza, Nancy Williams, and the IOC RED-S consensus group has consistently shown that female endurance athletes are often in low-energy-availability states without realizing it, and that adding fasted training sessions compounds the problem. Stacy Sims's clinical work goes further and argues that female athletes should almost always pre-fuel before training because of hormonal differences in fuel substrate use. The evidence for that specific claim is thinner than her books suggest, but the underlying concern about female under-fueling is well-supported and worth taking seriously.
• Athletes with disordered-eating history. Fasted training can easily slide into a pattern of chronic under-fueling or restrictive behavior, and for athletes with a history of eating disorders, it's a pattern worth avoiding entirely even if the sports-science case for it is positive in the abstract.
• Athletes in heavy build or race-specific training. When weekly TSS is high and the goal is performance adaptation at race pace, the fed state is the right state. Fasted training during a heavy build phase degrades session quality and recovery and offers a theoretical adaptation benefit that doesn't outweigh the real training cost.
• Athletes returning from illness, injury, or significant stress. Adding an additional stressor (fasted training) on top of an already-depleted state is a bad idea. Restore baseline first, then consider selective train-low work if you want the adaptation.

The default position for all of these populations is: don't train fasted. The default position for well-fueled, healthy, adequately-recovered endurance athletes in a base or early build phase is: one or two fasted easy sessions per week is a reasonable tool if you want the fat-oxidation adaptation. The populations are not interchangeable.

What is the sleep-low protocol and does it work?

Sleep-low is a more aggressive form of train-low that specifically targets the interaction between evening and morning sessions. The protocol is: do a hard session in the evening that depletes muscle glycogen, avoid carbohydrates after that session, go to sleep in a depleted state, and do an easy session the next morning still in a low-glycogen state before breakfast. The goal is to spend a meaningful number of hours in a glycogen-depleted condition to maximize the train-low adaptive signal.

Marquet et al. published the most-cited sleep-low studies in 2016, showing that 1 to 3 weeks of sleep-low protocols produced small performance improvements in well-trained triathletes and cyclists compared to matched train-high controls. The effect sizes were modest but statistically significant, and subsequent research has been mixed — some studies have replicated the effect, others have not.

The practical problem with sleep-low is that the night of low-glycogen sleep is genuinely unpleasant. Athletes report poor sleep quality, low mood, strong hunger, degraded next-day mental function, and in some cases worsened sleep architecture. The adaptive gain is theoretical and modest; the cost is immediate and real. Most athletes who experiment with sleep-low abandon it within a few weeks because it's miserable, and most coaches who work with high-performing amateurs do not recommend it for anything other than a short, targeted block with explicit trade-off awareness.

If you want to experiment with sleep-low, the honest framing is: do it for a specific block of 2 to 3 weeks, only in base or early build phase, only with fully fueled recovery days around it, only in athletes with no RED-S risk factors, and only if you're willing to accept degraded sleep quality. It is a tool for specific athletes in specific phases, not a general upgrade.

How should you actually use fasted training if it fits?

For the population where it's appropriate — well-fueled, healthy, adequately-recovered endurance athletes in base or early build phase who want to maximize aerobic adaptations — the practical protocol is simple and modest.

• One to two Zone 2 sessions per week done before breakfast. The sessions should be 45 to 90 minutes, strictly aerobic, and conversational throughout. If effort starts to creep above Zone 2, eat something immediately or cut the session short.
• Never more than 90 minutes fasted. Beyond roughly 90 minutes, liver glycogen is depleted enough that you'll start to see performance degradation, CNS effects, and increased cortisol. The adaptation curve flattens past this point and the downside grows.
• Never fasted before hard sessions. Threshold intervals, VO2max work, race-pace efforts, tempo runs, and any form of high-intensity training should be done fed. The rule is strict. Breaking it wastes the session and produces a worse adaptive signal.
• Always refuel immediately after the fasted session. Within 30 to 60 minutes of the fasted session, eat a real meal with carbohydrates (1 to 1.2 g/kg) and protein (20 to 40 g). The fasted window is not about extending the under-fueled state — it's about training in that state for the session and then restoring it.
• Never fasted the day before or the day of a key quality session or race. The adaptation benefit is nowhere near the race-performance cost. Fuel normally, race normally.

This is a surprisingly narrow application of fasted training for how much popular content talks about it. That's because the honest version of the research supports a surprisingly narrow application. The aggressive daily-fasted-training regime some content recommends is not in the evidence.

What about fat-adapted and low-carb endurance athletes?

A related but different topic is chronic low-carb / ketogenic training, popularized by the 'fat-adapted athlete' framing from researchers like Jeff Volek and Stephen Phinney. The argument is that a chronically low-carb diet shifts the body's fuel preference toward fat and allows longer efforts at higher intensities without depending on glycogen, which would be an enormous endurance advantage if true.

The sports-science evidence is cautiously negative on this. The FASTER study (Volek et al. 2016) showed that well-trained, chronically low-carb ultra-runners could reach very high rates of fat oxidation during exercise — two to three times higher than carb-fed runners. That's impressive and the biochemistry is real. But when researchers followed up and measured actual performance — Burke et al. 2017, race walkers at the Australian Institute of Sport — the low-carb group performed worse than the high-carb group at racing intensities, because high-intensity efforts still depend on carbohydrate fuel, and chronically low-carb athletes struggle to access that fuel when they need it.

The more nuanced reading is that extreme long, slow, low-intensity events (very long ultra-distance trail runs, for example) may benefit from enhanced fat oxidation, while anything involving high-intensity efforts — marathon, 70.3, Ironman cycling, any race with surges — is probably hurt by chronic low-carb adaptation. This is why essentially no elite marathon or Ironman program uses a ketogenic protocol, even though the fat-adapted messaging has been loud in certain endurance communities for a decade.

The honest synthesis is that chronic low-carb training is a niche intervention with a narrow use case and clear downsides for most endurance performance. Train-low as a targeted adaptation tool is different from low-carb as a lifestyle, and the two should not be confused.

What are the most common fasted training mistakes?

Five mistakes catch most amateurs who try fasted training.

• Doing it every day. The adaptive signal does not accumulate linearly — more fasted training does not mean more adaptation, and after a certain point it means more under-fueling, more fatigue, and worse recovery. One or two sessions per week is the practical ceiling for most athletes.
• Doing it on hard days. The single worst mistake. Fasted threshold or VO2max intervals produce a worse training effect than the same session done fed. The athlete feels like a hero for 'training hard on empty' but is actually producing less adaptation per session.
• Doing it while already under-eating. Athletes who are already running a calorie deficit and adding fasted sessions are walking into RED-S territory. If you're already losing weight, already feeling tired, or already missing menstrual cycles, fasted training is not the intervention you need.
• Doing it for long runs over 90 minutes. The first 60 to 90 minutes of a Zone 2 session is where the fasted-state adaptation signal is strongest. Beyond that, you're mostly just depleted, and the training quality drops without additional adaptation gain.
• Forcing the practice because a book or podcast said so. Fasted training is one of the clearest 'right tool for the right athlete' situations in endurance nutrition. If your body is telling you fasted training makes you feel awful, recover worse, or train worse, the data supports trusting that signal. Not every protocol works for every athlete, and fasted training is more individual than the popular advocates suggest.