CarboPlanner

Energy Expenditure in Football: From Calorimeters to GPS-Driven Models

A research-first history—how we moved from sealed chambers to on-pitch models that respect football’s stop–start reality.

By Lukas Rodrigues Bjerre

Technology

12 min read

Energy Expenditure in Football: From Calorimeters to GPS-Driven Models

A research-first history for football nutritionists and performance staff—how we moved from sealed chambers to on-pitch models that respect the sport’s stop-start reality.

Clean horizontal timeline from calorimetry to GPS/IMU and portable VO₂ systems

Calorimetry → DLW → HR telemetry → Accelerometers → HR+accel → GPS/IMU → Metabolic power → Portable metabolic carts (K5)


Why measuring energy expenditure in football has been hard

Football is intermittent, with near-constant changes of speed and direction. Steady-state assumptions that work on a treadmill break down on grass. The science has inched toward ecological validity: from laboratory indirect calorimetry (gold standard per breath) to doubly labeled water (DLW) for free-living truth (per day), then to field signals—heart rate, accelerometry, and GPS/IMU—and finally to hybrid models that try to translate movement and internal load into energy expenditure in football. The storyline below follows that progression and the evidence behind each step.


Foundations: calorimetry and DLW set the anchors

Indirect calorimetry. From Douglas bags and room calorimeters onward, oxygen consumption and carbon dioxide production gave us rigorous conversions to metabolic energy—exceptionally precise, but historically confined to lab tasks and steady locomotion.¹

DLW. The isotope method made it possible to quantify total daily energy expenditure (TDEE) in free-living athletes over 1–3 weeks. It is the gold standard for daily totals, but has no within-session resolution and is costly.²

Together, these methods define “truth” at two ends of the spectrum: per-breath in the lab and per-day in real life. Everything that follows tries to bridge the gap for the pitch.¹²


First on-pitch estimates: heart rate telemetry (1970s–1990s)

Portable HR telemetry gave practitioners a continuous internal load signal. With individual HR–VO₂ calibration, HR can proxy metabolic rate outside the lab—a premise validated against whole-body calorimetry at the group level.³

Applied to football, early field studies used HR during match play (calibrated with treadmill VO₂) to produce the first credible match-cost numbers: approximately ~1,510 kcal per 90′ in professionals, later converging near ~1,360 kcal, and broadly consistent with VO₂-based summaries around ~5 MJ (~1,200 kcal).⁴⁵⁶

What HR got right. Scalable and continuous, sensitive to internal stress.
Where it struggled. Drift (heat, arousal and hydration levels), non-steady efforts, and limited permissions for match-day wear for many years.³⁵⁶


Movement sensors and fusion: accelerometers → HR+accel (1980s–2000s)

Accelerometers turned movement into “activity counts” that correlated with energy cost when validated against calorimetry—excellent for population-scale monitoring, less so for complex sport actions where movement ≠ metabolism (e.g., resisted work).⁷

Fusion devices (e.g., Actiheart) combined HR with accelerometry using branched algorithms, improving intensity and energy estimation beyond single-sensor inputs.⁸ This was the template for modern multi-signal inference: combine external and internal signals to reduce ambiguity.


GPS/IMU: quantifying external load in context (mid-2000s–2010s)

As sampling moved from 1→5→10 Hz and inertial sensors were embedded, GPS/IMU became standard in field sports for distance, velocity, accelerations, and changes of direction.⁹ Subsequent policy changes enabled broader use in match contexts, making external-load datasets routine in professional football environments.¹⁰

Payoff. High-resolution external load under real tactical constraints.
Open question. How to convert kinematics—especially short accelerations/decelerations—into accurate energy expenditure in football terms.


A new lens on intensity: metabolic power (2010s)

Building on di Prampero’s equivalence (accelerated running on flat ≈ uphill running at constant speed), Osgnach et al. computed instantaneous metabolic power from speed and acceleration, revealing how much energetic cost hides outside speed thresholds. In elite matches, only ~26% of distance occurred above 20 W·kg⁻¹, yet those moments accounted for ~42% of total energy cost.¹¹ A systematic review later supported the methodological validity while underscoring assumption sensitivities (efficiency, braking, anaerobic contribution) in change-of-direction (COD) play.¹²

This reframed what “high intensity” means for fueling and recovery: not just running fast, but how often and how sharply you speed up, slow down, and turn.¹¹¹²


Portable metabolic carts on the pitch: closing the loop (1990s–2020s)

Miniaturised systems culminated in COSMED K5, enabling breath-by-breath VO₂ in ecologically valid football tasks (drills, SSGs, friendly match contexts).¹⁶ This provides the session-level ground truth needed to train or validate models built from GPS/IMU and HR, rather than relying on steady-state treadmill surrogates.

Crucially, football-specific validation work shows why this ground truth matters. Even high-end GPS/accelerometer trackers used in professional squads can underestimate exercise EE, with bias increasing at higher intensities when benchmarked against portable indirect calorimetry in standardized soccer drills.¹⁵ Meanwhile, consumer wearables (wrist devices) show poor EE accuracy on average—systematic review and meta-analysis evidence points to a negative bias for daily totals.¹³¹⁴

The message for practitioners: don’t trust kinematics or wrist-based calories at face value, especially in COD-heavy sessions. Preference should go to multi-signal models that are trained or checked against on-pitch VO₂.¹⁵¹⁶


Why this moment is different (and better) for fueling decisions

Physiological anchors exist at both scales. DLW pins down realistic daily energy needs across microcycles; portable VO₂ gives session truth during football-specific work.²¹⁶

Signals are ubiquitous and complementary. GPS/IMU quantify external load; HR reflects internal load; accelerometers capture high-frequency movement features.⁸⁹¹⁰¹⁷

A coherent modeling pathway. Decades of HR–VO₂ calibration, accelerometry validation, and the metabolic-power formalism define which features matter; modern multi-sensor algorithms can learn mappings that respect football’s intermittency.³⁷⁸¹¹¹²¹⁷

The literature warns against shortcuts. Device-only calories (consumer or team-sport trackers) are often biased; validation against portable VO₂ is the difference between plausible and misleading numbers.¹³¹⁴¹⁵

Nutrition relevance is immediate. Recognising that accel/decel work dominates energetic cost in key phases helps align carbohydrate periodisation and recovery with what actually happened on the pitch.¹¹¹²


References

Webb P. The measurement of energy expenditure. J Nutr. 1991;121(11):1897–1901. doi: 10.1093/jn/121.11.1897

Westerterp KR. Doubly labelled water assessment of energy expenditure: principle, practice, and promise. Eur J Appl Physiol. 2017;117(7):1277–1285. doi: 10.1007/s00421-017-3641-x

Ceesay SM, et al. The use of heart rate monitoring in the estimation of energy expenditure: a validation study using indirect whole-body calorimetry. Br J Nutr. 1989;61(2):175–186. doi: 10.1079/BJN19890107

Reilly T, Thomas V. Estimated daily energy expenditures of professional footballers. Ergonomics. 1979;22(5):541–548.

Bangsbo J. Energy demands in competitive soccer. J Sports Sci. 1994;12:S5–S12.

Shephard RJ. The energy needs of the soccer player. Clin J Sport Med. 1992;2(1):62–70.

Avons P, et al. Approaches to estimating physical activity in the community: calorimetric validation of actometers and heart rate monitoring. Eur J Clin Nutr. 1988;42:185–196.

Brage S, et al. Reliability and validity of the combined heart rate and movement sensor Actiheart. Eur J Clin Nutr. 2005;59(4):561–570. doi: 10.1038/sj.ejcn.1602118

Aughey RJ. Applications of GPS technologies to field sports. Int J Sports Physiol Perform. 2011;6(3):295–310. doi: 10.1123/ijspp.6.3.295

Hennessy L, Jeffreys I. The current use of GPS and its potential in soccer. Strength & Conditioning Journal. 2018;40(3):83–94.

Osgnach C, et al. Energy cost and metabolic power in elite soccer: a new match analysis approach. Med Sci Sports Exerc. 2010;42(1):170–178. PubMed

Brochhagen J, Hoppe MW. Metabolic power in team and racquet sports: a systematic review. Sports Med Open. 2022;8(133). doi: 10.1186/s40798-022-00510-3

Evenson KR, et al. Systematic review of the validity and reliability of consumer-wearable activity trackers. Int J Behav Nutr Phys Act. 2015;12:159. doi: 10.1186/s12966-015-0314-1

O’Driscoll R, et al. How well do activity monitors estimate energy expenditure? A systematic review and meta-analysis. Br J Sports Med. 2020;54(6):332–340. doi: 10.1136/bjsports-2018-099643

Dasa MS, et al. Accuracy of tracking devices’ ability to assess exercise energy expenditure in professional female soccer players. Int J Environ Res Public Health. 2022;19(8):4770. doi: 10.3390/ijerph19084770

COSMED. K5 wearable metabolic system—product information. 2017. cosmed.com

Costa S, et al. Quantifying the physical activity energy expenditure of commuters using a combination of GPS and combined heart rate and movement sensors. Prev Med. 2015;81:339–344. doi: 10.1016/j.ypmed.2015.09.022

Hulton AT, et al. Energy requirements and nutritional strategies for male soccer players: a review. Nutrients. 2022;14(3):657. doi: 10.3390/nu14030657

© 2025 Carboplanner. All rights reserved