A NIRS-Informed Performance Model for HYROX

By Ayda Page

April 2026 11 minutes read

Executive thesis

HYROX is officially standardized as eight 1-km runs alternated with eight fixed stations: SkiErg, sled push, sled pull, burpee broad jumps, rowing, farmers carry, sandbag lunges, and wall balls. In the first sport-specific physiological study, athletes completed a simulated race in about 86.5 minutes, spent more than 60% of total time running, operated mostly at hard to very hard intensities, and showed higher lactate and perceived exertion during stations than during the runs. A recent review from the National Strength and Conditioning Association[1] literature synthesizes the event as a blend of aerobic capacity, anaerobic power, local muscular endurance, maximal strength, and technical proficiency. That combination makes HYROX poorly described by any single endurance number. VO₂max, lactate thresholds, and heart-rate zones still matter, but they do not fully explain where an athlete loses seconds, how transitions degrade pace, or which muscle groups become locally limiting under fatigue. [2]

What the race actually demands

The official race structure matters because it fixes the problem the athlete has to solve. Unlike open-skill formats, HYROX always presents the same sequence, so the coaching task is not preparing for the unknown; it is optimizing a known pattern of repeated locomotion, heavy cyclic work, carries, lunges, and high-repetition terminal fatigue. In the 2025 Frontiers study, station segments were more physiologically taxing per unit time than the runs, wall balls produced the highest end-race strain, and the heaviest sled stations were completed quickly but imposed high metabolic stress. This is not a steady aerobic race with brief interruptions. It is a repeated transition event, where the athlete must repeatedly rebuild economical running after localized muscular disruption. [3]

The same study also shows why simplistic interpretation is risky. Overall performance correlated with VO₂max, endurance training volume, and body-fat percentage, but those associations were driven primarily by the running components. Hand-grip strength, muscle-mass percentage, and resistance-training volume did not correlate with station time in that small recreational Open-division sample. Read carefully, that does not mean strength is unimportant; it means generic strength tests may not identify the real bottlenecks in HYROX, which are more likely to involve lower-body force production, local muscular endurance, transition efficiency, and technique under fatigue. The NSCA review reaches a broader conclusion than the single primary study: aerobic capacity is foundational, but anaerobic power, local muscular endurance, maximal strength, and technical skill all contribute to performance. [4]

Why the traditional endurance model is incomplete

The traditional endurance model was built to describe continuous or relatively stable workloads. Even in endurance science, there is no universally accepted gold-standard method for defining the entire exercise-intensity spectrum, and different threshold models partition effort differently. In intermittent and multidimensional sports, heart-rate measures capture only a limited slice of the training response and should be interpreted with context and with additional markers rather than as direct readouts of total fatigue or performance readiness. Reviews on athlete monitoring in complex sports explicitly argue for multivariate interpretation that combines physiological, perceptual, and contextual information. [5]

That logic applies strongly to HYROX. Heart rate can tell a coach that the systemic load is high, but not whether the athlete’s next 1-km split is slowing because of deteriorating running economy, incomplete peripheral re-oxygenation, upper-to-lower body blood-flow redistribution, local quadriceps oxygen debt from lunges, or a breakdown in technical coordination. A recent functional-fitness monitoring review makes the same argument explicitly: traditional metrics often neglect internal strain, energy-system engagement, and neuromuscular fatigue in mixed-modal exercise, and a more integrated physiological model is needed. Critical power and W′ add useful severe-domain information, but they are still predominantly systemic constructs; they do not identify where oxygen is being wasted locally inside the race. [6]

What NIRS is actually measuring

Near-infrared spectroscopy is valuable here because it adds a peripheral layer to the picture. The core physiology is straightforward: NIRS is a noninvasive tool for examining the match between oxygen delivery and oxygen utilization in working tissue. In skeletal muscle, it provides semi-quantitative information on hemoglobin-plus-myoglobin oxygenation and local hemoglobin volume, which makes it useful for monitoring oxygen availability, extraction, and re-oxygenation kinetics during exercise and recovery. Systematic-review work further indicates that, in appropriate settings, NIRS-derived muscle-oxygenation measures agree with gold-standard high-energy-phosphate measures and can act as meaningful markers of skeletal-muscle oxidative capacity. [7]

For coaches, the practical implication is not that NIRS replaces gas exchange, lactate, or heart rate. It is that NIRS complements them by showing where the local oxygen problem sits. Earlier reviews also report that muscle-oxygenation behavior is linked to exercise thresholds and to the delayed steady-state response above threshold, which is exactly the type of behavior expected in repeated run-station transitions. But NIRS is not a magic number: absolute readings are influenced by adipose-tissue thickness, device architecture, skin perfusion, and the mixed contribution of hemoglobin and myoglobin. The most defensible use in practice is usually within-athlete interpretation of kinetics, drift, asymmetry, minima, and re-oxygenation speed, not broad cross-athlete comparison of a single absolute SmO₂ value. [8]

The HYROX performance pillars

The pillar model below is an evidence-informed coaching framework, not yet a validated HYROX-specific consensus taxonomy. It is built by integrating official race demands, the first HYROX-specific physiology paper, the NSCA review, running-economy literature, intermittent-exercise recovery data, and NIRS methodology. [9]

Speed — oxygen cost of pace. In HYROX, speed is not purely maximal velocity. It is the ability to sit at a given race pace with the lowest possible oxygen cost and least technique decay before and after stations. Running economy is strongly associated with endurance performance and can outperform VO₂max as a predictor in trained runners, while strength training can improve running economy even when VO₂max changes little. A NIRS-informed speed metric, therefore, is not just “how fast can you run,” but “how much local deoxygenation and drift does a target pace require before and after a station.” That is closer to race reality. [10]

Endurance — repeated deoxygenation tolerance. HYROX endurance is not simply threshold durability in a continuous sense. It is the ability to tolerate repeated desaturation, restore oxygenation quickly, and reproduce output again eight times. The broader literature shows that aerobic fitness improves recovery from high-intensity intermittent exercise, and repeated-sprint research links faster muscle-oxygen-uptake recovery or post-sprint re-oxygenation with better repeatability of performance. In HYROX terms, the decisive question is not only whether the athlete can hold pace in one long effort, but whether they can repeatedly re-enter efficient running after each oxygen-costly station. [11]

Strength — force under oxygen limitation. HYROX strength is the capacity to sustain force when oxygen availability is constrained and metabolites are accumulating. The general strength literature is clear that muscular strength supports sprinting, jumping, and many sport skills, while the NSCA HYROX review specifically identifies maximal strength and local muscular endurance as relevant determinants. Yet the first HYROX-specific study also warns coaches against lazy proxy testing: hand-grip strength did not explain performance. The implication is that HYROX strength should be assessed and trained in task-relevant patterns and durations, especially lower-body and trunk force production under fatigue, not only through generic gym markers. In NIRS terms, the useful question becomes: how much force output can the athlete preserve as local oxygenation falls, and how quickly can that muscle region recover for the next run? [12]

Power — rate of deoxygenation. Power in HYROX is not just peak wattage or jump height in isolation. It is the ability to generate force rapidly at the start of sled work, accelerations, burpee broad jumps, wall-ball cycles, and transitions without causing such a steep metabolic cost that the next segment collapses. Rate of force development is a recognized determinant of explosive athletic actions and is also sensitive to fatigue. A useful NIRS lens here is the kinetics of oxygen extraction: how sharply oxygenation drops when the athlete produces high output, and whether that drop is accompanied by worthwhile mechanical gain. In high-level practice, power is best interpreted as rapid force expression with a tolerable oxygen penalty, not as explosiveness divorced from the rest of the race. [13]

Flexibility — oxygen cost of movement. Flexibility is the least defensible HYROX pillar if it is defined as “more range is always better.” The literature does not support that. In running, excessive compliance does not automatically improve economy, and older work even linked reduced flexibility or greater stiffness with better economy in some contexts. More recent mobility training reviews suggest performance benefits are possible, but they are not universal or linear. For HYROX, flexibility is better framed as sufficient mobility to achieve positions efficiently and safely: ankle dorsiflexion for lunges and running mechanics, thoracic and shoulder range for rowing and wall balls, and trunk-hip motion quality for force transfer. In that sense, the relevant variable is the oxygen cost of mechanically compromised movement, not maximal joint laxity. [14]

Coordination — oxygen waste versus oxygen efficiency. Coordination in HYROX means preserving economical movement solutions under fatigue: run mechanics after sleds, implement handling during carries, breathing-rhythm management during wall balls, and transition quality in the Roxzone. Reviews on locomotion and running biomechanics show that energetic cost both reflects and shapes movement coordination, while core-stability work emphasizes efficient force transmission and load distribution. In practice, coordination is the pillar that decides whether oxygen use is productive or wasteful. Two athletes can show similar systemic heart-rate strain, but the more coordinated athlete spends less oxygen per unit of useful work because their technique, posture, timing, and rhythm are cleaner under fatigue. That is precisely the kind of information a local oxygenation signal can help expose when paired with split data and video. [15]

A practical model for testing and programming

For high-level coaching, I would not replace the traditional model; I would layer it. The first layer remains systemic: VO₂max or vVO₂max, threshold or critical-speed/critical-power work, body composition, and race-split analysis. The second layer is sport-specific execution: run splits, station splits, transition times, and pace decay after each station. The third layer is peripheral physiology: bilateral muscle-oxygenation monitoring in the most decisive sites for the athlete, usually vastus lateralis and calf for run-station coupling, with deltoid or upper-back sites added when upper-body bottlenecks are suspected. The rationale for this layered approach is supported by the HYROX review literature, multivariate heart-rate-monitoring work, and the broader functional-fitness argument for integrated monitoring rather than single-metric control. [16]

Operationally, the most useful field protocol is probably a repeated run-station battery rather than an isolated treadmill test. For example: repeated 1-km efforts at projected race pace, each followed by a representative station or station fragment, while collecting pace, HR, RPE, lactate where feasible, and NIRS kinetics. The key outputs are not just minima but patterns: pace-specific SmO₂ drift, station-specific desaturation depth, re-oxygenation half-time into the next run, left-right asymmetry, and the coupling between rising oxygen cost and falling mechanical output. That battery can then be tied to targeted neuromuscular testing such as isometric squat or pull, countermovement jump, sled split times, and video analysis of fatigue-resistant mechanics. Such a system is far more aligned with HYROX than relying on VO₂max, threshold, and heart-rate zones alone. [17]

Programming should follow the same logic. The current evidence supports keeping a substantial aerobic-development component because running dominates total race time and overall performance tracks strongly with aerobic traits. At the same time, the strongest sport-specific recommendations point toward mixed-modal interval work, targeted circuit training, and repeated transition practice so the athlete learns to restore economical running after metabolically disruptive stations. Strength work should stay in the plan, but it should be targeted at the force qualities and contraction durations that actually show up in the race, rather than accumulated for their own sake. In other words: build the engine, raise the floor of task-specific force production, and repeatedly train the restoration of movement economy under local oxygen stress. [18]

Open questions and limitations

The main limitation is the evidence base itself. Direct HYROX physiology research is still young: the first primary study used only 11 recreational Open-division athletes, and even the NSCA review stresses that sport-specific evidence remains limited. That means any elite HYROX performance model should be presented as provisional. NIRS also brings methodological constraints, especially device differences, adipose-tissue effects, and signal-interpretation challenges. So the strongest conclusion at present is not that HYROX has already proven six fully validated physiological pillars, but that the sport clearly demands a multidimensional, transition-sensitive, and peripherally informed model, and that NIRS is one of the most promising tools for connecting local muscular oxygen behavior to the practical coaching questions that traditional endurance metrics leave unanswered. [19]

[1] [10] Factors affecting running economy in trained distance ...

https://pubmed.ncbi.nlm.nih.gov/15233599/?utm_source=chatgpt.com

[2] [3] maintain.hyrox.com

https://maintain.hyrox.com/rulebooks/HYROX_RulebookSingles_EN.pdf

[4] [9] [18] [19] https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2025.1519240/full

https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2025.1519240/full

[5] https://link.springer.com/article/10.1007/s00421-025-05748-8

https://link.springer.com/article/10.1007/s00421-025-05748-8

[6] [17] https://pmc.ncbi.nlm.nih.gov/articles/PMC12656209/

https://pmc.ncbi.nlm.nih.gov/articles/PMC12656209/

[7] https://cir.nii.ac.jp/crid/1363107370417890048

https://cir.nii.ac.jp/crid/1363107370417890048