Empirical Cycling Community Notes

Watts Doc 46: Why Low Glycogen Training Probably Doesn't Work

Original episode & show notes | Raw transcript

Introduction: The “Train Low” Hypothesis

For years, a popular hypothesis in endurance sports has been that training with low muscle glycogen stores—often called “training low”—could be a shortcut to superior aerobic adaptations. The central idea is that exercising in a glycogen-depleted state creates a significant cellular stress, which in turn amplifies the molecular signals that drive improvements in endurance, such as the creation of new mitochondria (mitochondrial biogenesis).

This document unpacks the science behind this hypothesis, focusing on a key signaling pathway known as p38 mitogen-activated protein kinase (MAPK). We will examine the foundational research, critique the human performance studies, and ultimately arrive at an evidence-based conclusion about the efficacy of this training strategy, as detailed in the provided podcast.

The Molecular Machinery of Adaptation

To understand the debate, we must first understand the key molecular players involved in how our muscles adapt to exercise.

p38 MAPK: The Cellular Stress Sensor

PGC-1α: The Master Regulator of Mitochondrial Biogenesis

Examining the Evidence: From Mice to Men

The “train low” hypothesis hinges on the idea that low glycogen potentiates p38 MAPK activity, which in turn drives PGC-1α and enhances adaptation. Let’s follow the scientific trail.

A highly-cited 2005 paper by Akimoto and colleagues provided the foundational evidence linking p38 MAPK to PGC-1α.

  1. Correlation in Mice: Researchers had mice perform a bout of exercise and observed that the activation of p38 (measured by its phosphorylation) occurred over the same time frame as the increase in PGC-1α gene expression (mRNA). This established a correlation.

  2. Proving Causation in a Dish: To prove causation, they used cultured muscle precursor cells (myoblasts).

    • They introduced a luciferase reporter gene attached to the PGC-1α gene. Luciferase is a protein that glows (like in fireflies); if the PGC-1α gene was being transcribed, the cells would glow, providing a direct measure of its activity.

    • When they added a chemical that activates the p38 pathway, the cells glowed brightly, indicating high PGC-1α transcription.

    • Crucially, when they added both the p38 activator and a p38 inhibitor, the glow disappeared. This demonstrated that p38 MAPK activation is necessary for PGC-1α transcription.

  3. Confirmation in Living Muscle: Finally, they created transgenic mice whose muscles had a constantly active p38-stimulating protein. These mice showed a large increase in PGC-1α protein and other mitochondrial markers, confirming that this pathway works in living tissue.

Conclusion from Akimoto: The p38 MAPK pathway is a direct and potent activator of PGC-1α and mitochondrial biogenesis. This makes it a plausible mechanism for the “train low” hypothesis.

Part 2: Testing the “Train Low” Twist in Humans

With the pathway established, the next step was to see if low glycogen actually potentiated this signal in humans and led to better performance. A key study investigated this using a clever protocol.

The Fat Oxidation Red Herring: The researchers noted that the low-glycogen group exhibited about 30% higher rates of fat oxidation. This finding is often highlighted by proponents of “train low.” However, as the performance data shows, this increased fat burning did not translate to superior endurance. The body simply adapted to use the fuel that was most available. It is a change in substrate utilization, not an enhancement of performance capacity.

Synthesis and Conclusion: Why “Train Low” Fails the Test

When all the evidence is considered, the “train low” strategy is not the performance-enhancing shortcut it’s claimed to be.

  1. The Core Mechanism is Flawed: The foundational premise that low glycogen potentiates p38 MAPK signaling was not observed in human studies. While low glycogen is a stressor, it does not appear to amplify this specific pathway more than normally fueled, high-quality training.

  2. Performance is Compromised: Training in a glycogen-depleted state demonstrably reduces the quality of high-intensity work you can perform. Since the primary driver of adaptation is progressive overload—challenging the body with a demand it’s not used to—reducing power output is counterproductive.

  3. Homeostasis and Training Status: A well-trained athlete’s body is remarkably good at maintaining cellular homeostasis. For a given workload, they experience less cellular stress than an untrained person. This is why some studies show less p38 activation in trained individuals. To elicit further adaptation, they must increase the training stress (ride longer or harder), not compromise it by removing fuel.

  4. The Cost Outweighs the (Lack of) Benefit: Training with low glycogen has significant downsides:

    • Reduced training quality.

    • Increased perceived exertion.

    • Impaired recovery.

    • Negative mood and irritability.

    • Potential for compromised immune function.

Given that the evidence shows no superior performance benefit, these costs are not justified.

Final Actionable Advice

In summary, while the molecular pathways like p38 MAPK are fascinating, the attempt to manipulate them through glycogen depletion is an ineffective strategy. The most reliable path to improvement remains the one proven by decades of coaching and application: fuel yourself properly and train hard.