A High-Resolution Biomarker for Skeletal Muscle Mitochondrial Adaptations
Quantifying skeletal muscle adaptations with continuous non-invasive wearable devices
Liked this piece? Show your support by tapping the “heart” ❤️ in the header above. It’s a small gesture that goes a long way in helping me understand what you value and in growing this newsletter. Thanks so much!
A few weeks ago I attended an event hosted by NASA HumanWorks at the Johnston Space Center where they brought scientists, technologists, and startups working in the "Human Performance" space (no pun intended) together to collaborate on ways we can protect astronaut’s health in space. I was lucky enough to represent NNOXX as part of a collaboration with my friends at Hytro and Whoop— together, we put together a working concept model for integrating multiple biosensors, blood flow restriction, and exercise devices designed for zero-gravity to offset space-associated losses in skeletal muscle mass.
One of the discussions that frequently came up over the weekend was how mitochondrial stress acts as a central biological hub for spaceflight impact. Since then, I’ve been thinking about ways we can utilize low cost, non-invasive, monitoring technologies to analyze a muscle’s metabolic efficiency— a marker of mitochondrial adaptation1.
A potential solution I’ve been thinking about is to create a "work score"— a dimensionless metric— that quantifies the relationship between mechanical power and local muscle oxygen kinetics, as measured via near-infrared spectroscopy (mNIRS). The normalized work score (WS) follows a straightforward formula:
This ratio provides insight into two key mechanisms of muscle metabolic adaptation. First, someone can deoxygenate a muscle less at a given power output, meaning they accomplish the same external work while consuming less oxygen. Second, they can produce more power at a given level of deoxygenation, meaning they generate greater external work per unit of energy consumed. Both mechanisms suggest increased efficiency and fitness, largely driven by changes in mitochondrial and capillary density, increased skeletal muscle oxidative capacity, and improved mechano-energetic coupling.
To test this concept, I had a field sport athlete collect muscle oxygenation2 data over a 32-day training period consisting of unstructured exercise including running intervals, change of direction drills, and small-sided games3.
For each session, second-by-second data was collected with a shoe-worn power meter and muscle oximeter worn on the outer quadriceps. These data points were then used to calculate daily normalized work scores, which are depicted in the image above (each vertical column represents a day's worth of measurements, with the red points highlighting the top 2% of scores achieved each day). The dotted line in the image above traces daily mean values, while the solid trend line with its confidence interval reveals a slight but meaningful upward progression over the month-long period, indicating improved fitness and efficiency at the local tissue level.
Additionally, the work score can provide valuable real-time feedback during individual workouts. By monitoring the work score continuously, practitioners can precisely identify when fatigue begins to occur—either when someone starts to deoxygenate more at a fixed power output or produces less power at a given level of deoxygenation. This capability enables highly targeted exercise interventions and more precise exercise prescriptions.
This methodology offers particular promise for assessing changes in mitochondrial healthy and fitness in astronauts, where equipment is limited. Additionally, the normalized work score can also be applied in biotechnology applications, especially in evaluating therapeutics targeting mitochondrial function or muscle metabolism. Consider a Phase II clinical trial evaluating a novel compound designed to enhance skeletal muscle oxidative capacity. Traditional endpoints often lack sensitivity to detect localized improvements in mitochondrial function, particularly in populations with limited exercise tolerance. By implementing the work score metric, researchers could detect subclinical improvements before they manifest as global performance changes, potentially reducing required sample sizes through high-frequency sampling, establishing clearer dose-response relationships, and identifying responders versus non-responders based on cellular-level adaptations.
The longitudinal data visible in the figure above demonstrates the methodology's capacity to detect subtle trends over time, with the confidence interval narrowing as data accumulates - a critical feature for determining therapeutic efficacy in early-stage trials. This approach effectively bridges the gap between cellular metabolic assays and whole-body performance measures, providing a functional assessment of mitochondrial capacity in vivo during ecologically valid activities. For biotechnology companies developing compounds targeting cellular energetics, muscle protein synthesis, or microvascular function, this methodology offers a non-invasive, high-frequency assessment tool to evaluate therapeutic efficacy at the tissue level during functional activity.
Did you enjoy this piece? If so, you may also want to check out other articles in Decoding Biology’s Wearable Technology & Biometrics collection.
A different approach that i’ve been working on is to use PhysioNexus for environmental adaptation analysis. In practice, this means using PhysioNexus for condition-specific comparisons between earth and microgravity and comparing how causal relationships in an individual’s physiological network change over time.
For an in-depth article on interpreting variations in muscle oxygenation measurements see Dampening the Noise: Making Sense of Variability In Biometric Measurements.
This data was collected in the week leading up to, during, and after a training camp. As a result, they performed little resistance training during this period, and thus the vast majority (>90%) of their exercise was recorded and included in this dataset.