Original episode & show notes | Raw transcript
When athletes push their limits, they walk a fine line between optimal adaptation and overtraining. A central question in exercise science is what, precisely, happens at a cellular level when we cross that line? A 2021 study by Flockhart et al. published in Cell Metabolism made waves with a provocative headline suggesting that excessive training leads to “mitochondrial dysfunction.” This finding seemed to offer a tangible, mechanistic explanation for the fatigue and performance decline associated with overreaching.
However, as the Empirical Cycling podcast explores, this headline conceals a more nuanced and arguably more fascinating story. This document will guide you through that story, following the scientific journey presented in the podcast. We will deconstruct the Flockhart study, examine a critical follow-up paper by Granada et al., and synthesize the findings to build a sophisticated understanding of how our mitochondria—the powerhouses of our cells—actually respond to extreme training stress. This is not a story of dysfunction, but one of intelligent, non-stoichiometric adaptation.
The Flockhart study was designed to test a straightforward hypothesis: does a period of excessive exercise overload impair the function of mitochondria in human skeletal muscle?
Participants: The study used 11 recreationally active individuals (6 female, 5 male). Crucially, they excluded highly trained athletes. This is a common and logical strategy in such studies; using moderately trained participants ensures that the training stimulus is potent enough to induce measurable adaptations in everyone, avoiding a scenario where elite athletes might not respond at all.
The Training Protocol: The intervention was a four-week block of High-Intensity Interval Training (HIT), designed to progressively overload the participants.
Week 1 (Light): Two sessions of 5x4-minute intervals.
Week 2 (Moderate): Three sessions (two 5x8-min, one 5x4-min).
Week 3 (Excessive): Five HIT sessions in one week (three 5x8-min, two 5x4-min). This was the “hell week” designed to induce an overreached state.
Week 4 (Recovery): A reduced-load week.
The Measurements: To track the effects, researchers took muscle biopsies, conducted performance tests, and administered oral glucose tolerance tests (OGTT) at baseline and after each training block.
1. “Mitochondrial Dysfunction”
The study’s primary claim was based on a metric called intrinsic mitochondrial respiration (IMR).
What it is: Researchers took the muscle biopsy, mechanically ground it up, and used a centrifuge to isolate the mitochondria. They then placed these isolated mitochondria in a solution with saturating levels of fuel (substrates) and ADP (the “on” switch for energy production) and measured their maximum rate of oxygen consumption. This is known as “State 3 respiration.”
The Result: The study found that IMR increased from the baseline to the moderate training week, but then dramatically decreased by about 50% during the “hell week.” After the recovery week, it rebounded but remained below the initial baseline level.
The Nuance: This sounds bad, but as the podcast highlights, this measurement is of isolated mitochondria, normalized to their own mass. It tells us the oxygen consumption per unit of mitochondria. It does not necessarily tell us the total respiratory capacity of the entire muscle.
2. Impaired Glucose Tolerance
The oral glucose tolerance test (OGTT) involves consuming a standard glucose drink after an overnight fast and measuring blood sugar over two hours.
The Result: During the excessive training week, the “area under the curve” for blood glucose was significantly higher, indicating that the body struggled to clear sugar from the blood.
The Nuance: The impairment was temporary and most pronounced between 60 and 90 minutes into the test. By the two-hour mark, blood glucose levels had returned to normal. The podcast speculates this could be due to temporarily fatigued beta cells in the pancreas being sluggish in releasing insulin, rather than a fundamental problem with the muscles’ ability to take up glucose.
3. Decreased Performance
The Result: The study’s headline graphic suggested that performance decreased during the excessive training week.
The Nuance: A closer look at the actual data, as pointed out in the podcast, shows a different story. The power output during the HIT sessions did not decrease; it stagnated. It stayed flat compared to the moderate week, and then supercompensated to a new peak after the recovery week. Furthermore, the athletes’ VO2peak (maximal oxygen uptake) continued to increase even during the hell week. This is a critical point: if mitochondria were truly “dysfunctional,” we would expect the body’s ability to use oxygen to decline, not increase.
The podcast astutely notes that the same dataset could be used to generate very different, even contradictory, headlines:
“Excessive Training Increases Whole-Body Fat Oxidation.” (This was observed).
“Excessive Training Increases Muscle GLUT4 Expression.” (GLUT4 is the primary transporter that brings glucose into muscle cells; its expression went up).
“Excessive Training Does Not Impair Performance, but Stagnates It Temporarily.”
This leads to a crucial question: If VO2peak is up, fat oxidation is up, and key proteins are being expressed more, what explains the drop in isolated mitochondrial respiration? The answer likely lies in how the mitochondria are being built.
The Granada et al. study, published in Nature Communications, provides the key to unlocking the puzzle. Using more advanced techniques and an even more brutal training protocol, they investigated the composition of mitochondria during adaptation.
If the Flockhart study was a “hell week,” the Granada study was “hell weeks,” plural. Participants endured 40 HIT sessions over 20 consecutive days. This extreme protocol ensured a massive stimulus for mitochondrial growth.
The key innovation was their measurement technique. Instead of just measuring a single function like respiration, they used mass spectrometry-based proteomics. This allowed them to identify and quantify hundreds of individual proteins within the mitochondria simultaneously, giving them a complete picture of its changing composition.
The traditional view in exercise physiology has been that when mitochondria grow, they do so stoichiometrically—that is, all of their component parts are built in a constant, fixed ratio, like assembling a car with a set number of wheels, engine parts, and seats.
The Granada study definitively proved this assumption is wrong.
Their data revealed that during periods of rapid mitochondrial biogenesis (growth), the body prioritizes. The expression of proteins in the electron transport chain (ETC)—the very “engines” measured in the Flockhart study—temporarily lags behind the overall growth in mitochondrial volume and mass. In contrast, the expression of mitochondrial ribosomes (the machinery to build new proteins) and proteins involved in importing fuel (like fats and lactate) were ramped up.
This is the answer to the puzzle.
The Flockhart study saw a decrease in respiration per unit of mitochondria because that unit had changed. The mitochondria had grown larger, but were temporarily less dense with ETC proteins. It’s like building a larger factory floor (more mitochondrial volume) before all the new assembly lines (ETC proteins) have been fully installed. The total number of assembly lines in the muscle was likely still increasing (as evidenced by the rising VO2peak), but the density per factory had decreased.
This is not dysfunction. This is a logical and efficient adaptive strategy. The body prioritizes increasing the surface area for importing fuel, as this is often a limiting factor, knowing the ETC itself is rarely operating at its absolute maximum in a healthy, exercising person.
The response letter by esteemed exercise scientists John Hawley and David Bishop, along with the Granada data, helps us form a complete picture.
The term “dysfunction” implies pathology. True mitochondrial dysfunction is seen in severe genetic diseases that lead to blindness, seizures, and diabetes. What was observed in these training studies was not a disease state, but a transient, adaptive, and functional remodeling of the organelle in response to an extreme stimulus.
Both the podcast and the Hawley/Bishop response emphasize that the “dysfunction” in the Flockhart study disappears when the data is normalized to citrate synthase activity, a long-standing proxy for mitochondrial mass. This, combined with the proteomics data from Granada, strongly suggests the finding was an artifact of a changing mitochondrial composition, not a sign of impairment.
While the science is complex, the practical takeaways reinforce exercise fundamentals:
Overtraining is Real, But “Dysfunction” is the Wrong Word: The fatigue from overreaching is undeniable. However, at the cellular level, it seems to be driven by a complex remodeling process, not by mitochondria simply “breaking.”
Adaptation Takes Time: The Granada study showed that even after 72 hours of recovery, the mitochondrial protein ratios had not returned to baseline. This underscores that supercompensation is not an overnight process. Real recovery from a hard block of training takes days or even weeks.
Fundamentals Remain King: These studies do not suggest a new way to train. They reinforce the principles of progressive overload followed by adequate recovery.
Listen to Your Body (and Your Power Meter): For the athlete, the most reliable signs of overreaching are not found in a muscle biopsy. They are found in your performance data and your perceived exertion. If your power at threshold feels significantly harder than it should (elevated RPE), or your peak sprint power is consistently down, these are your body’s signals that you need more recovery.
The journey from the headline of the Flockhart paper to the deep proteomic analysis of the Granada paper is a perfect illustration of the scientific process. An initial, provocative observation of “dysfunction” gave way to a more sophisticated understanding of non-stoichiometric adaptation.
When pushed to the extreme, our bodies do not build new mitochondria according to a fixed blueprint. They intelligently prioritize expanding the capacity to import fuel, even if it means temporarily diluting the density of the internal respiratory machinery. This is a feature, not a bug. It is a testament to the elegant and efficient way our bodies respond to the stress of training, a process that is far more complex and remarkable than a simple story of “dysfunction” could ever capture.