An updated conceptual framework on physical performance and training in soccer: a narrative review

A new narrative review published in Sports Medicine – Open argues that the central problem in soccer physical monitoring is not a shortage of definitions for training load, external load, internal load, or fitness status, but the lack of an integrated interpretative structure explaining how these established constructs should be differentiated when evaluating physical performance. The same monitoring variable, the authors contend, can serve entirely different purposes depending on whether a practitioner is quantifying training exposure, evaluating an observed performance output, characterizing a physiological response, or interpreting an athlete’s capacity to produce or sustain that output. Most existing frameworks, they argue, never make that purpose explicit.

The review, led by Guglielmo Pillitteri together with colleagues including Filipe Manuel Clemente, Marco Beato, Alessio Rossi, Hugo Sarmento, Christopher Carling, Antonino Bianco, and Giuseppe Battaglia, proposes an integrated interpretative framework built around three questions: what did the athlete do, how did the athlete respond, and what is the athlete capable of doing? These correspond, respectively, to performance output, physiological response, and fitness status or performance capacity. Training load sits separately as the exposure or dose associated with training and competition. Crucially, the authors stress that these are interpretative dimensions, not new constructs meant to replace the established concepts of external and internal load.

The distinction matters because a single variable can serve multiple purposes. High-speed running distance, for example, is an external load measure when used to quantify accumulated work or exposure. The same distance becomes a performance output when evaluated against an individual, positional, or match reference value to determine whether a player has reproduced a specific competitive demand. Meanwhile, heart rate or rating of perceived exertion characterizes the internal response associated with producing that work, revealing whether the intended physiological stimulus was actually elicited. Without separating these purposes, the authors argue, practitioners risk drawing conclusions the data cannot support.

To build the framework, the team conducted a narrative review of publications retrieved from PubMed, Google Scholar, Scopus, and Web of Science from database inception to July 20, 2026, combining soccer-related terms with concepts including physical performance, training load, external load, internal load, fitness, fatigue, recovery, readiness, athlete monitoring, and conceptual framework. No restrictions on date, language, publication status, or study design were applied. Selection was purposive rather than exhaustive, prioritizing publications that defined target constructs, developed or revised conceptual frameworks, examined the validity of monitoring measures, or applied these concepts in soccer. The authors compared definitions and construct boundaries across publications to identify consistent usage and instances where the same indicator had been used to represent different constructs, refining the final framework through discussion among the authors.

The review situates itself on the shoulders of several influential predecessors. Impellizzeri and colleagues made the first attempt to implement a generic conceptualization framework for physical training, clarifying and refining the concept of training load through operational definitions of both external and internal load to avoid misapplications of the originally proposed concepts. Jeffries and colleagues integrated that framework with Banister’s classic fitness-fatigue model and the PerPot computational metamodel, replacing the terms fitness and fatigue with positive and negative training effects and noting that both can occur simultaneously. A strength session, for instance, may trigger increased muscle protein synthesis while also inducing neuromuscular fatigue that temporarily depresses readiness before supercompensation improves performance outcomes.

Gabbett and colleagues contributed the athlete monitoring cycle, a practical sequence in which practitioners track the external load performed, observe the athlete’s response, judge whether the load is being tolerated through perceptual well-being or other readiness measures, and decide whether the athlete is prepared for a further physical stimulus. Building on this, Gabbett and Oetter emphasized that different tissues recover at different rates: cartilage may recover within roughly 30 minutes after activity, while eccentric muscle actions can necessitate recovery periods of 72 hours or more due to the structural and inflammatory response associated with muscle fiber damage, and tendons subjected to high stretch-shortening cycles may require at least 48 hours before reloading. Beato and colleagues, meanwhile, proposed a readiness framework combining objective measures such as countermovement jump and heart rate with athlete-reported outcome instruments to identify patterns or alerts related to a player’s health condition.

What each of these frameworks leaves open, the new review contends, is an explicit link between the monitoring question and the appropriate construct. Athletes may produce nearly identical external outputs while exhibiting substantially different physiological responses and underlying capacities, which has direct implications for training interpretation, readiness evaluation, and individualized decision-making. The framework therefore treats external load volume, the total external work accumulated over a defined period, and external load intensity, the rate at which that work is performed, as complementary dimensions that should not be collapsed into a single arbitrary metric. A session may reproduce a different proportion of match volume than match intensity, so considering both dimensions can provide a more complete description of external exposure than relying on either alone.

The physiological-response dimension comes with important caveats. When external work is standardized or sufficiently comparable, differences in internal response may offer contextual information about how individual athletes are coping with that work. But the authors warn that such relationships should never be read as direct measures of fitness status, because internal load is influenced by fatigue, recovery, environmental conditions, psychological state, and other individual and contextual factors. Fitness status, defined in the framework as a chronic capacity developed through adaptation, should be assessed with appropriate independent measures of the underlying capacity of interest, such as maximal oxygen uptake, countermovement jump performance, or lactate threshold, while biomarkers like creatine kinase, a non-specific marker of exercise-induced muscle damage, and perceptual soreness scales reflect acute fatigue-related states instead.

Readiness, in this scheme, is not treated as a fully validated latent construct but as an emergent state arising from the interaction between fitness capacity and acute fatigue. The evidence linking wellness measures to training load is complex and context-dependent. Moalla and colleagues found a significant negative relationship between training load and perceived sleep stress, fatigue, and delayed-onset muscle soreness, and Pillitteri and colleagues observed that the higher the training load on a given day, the lower the expected recovery and availability on the following day. Yet Duignan and colleagues showed in a real-world setting that wellness measures are unlikely to be influenced by load in isolation, being shaped also by recovery, psychological, and social influences, and some studies have found that pre-training wellness had mainly trivial effects on subsequent training output.

The authors are candid about limitations. The framework is a conceptual and interpretative synthesis whose proposed relationships have not been empirically validated, and no single framework can capture the full multidimensional nature of soccer performance, in which technical, tactical, psychological, cognitive, environmental, and contextual factors all interact with physical performance. Between-player, between-position, between-match, and within-player variability in match physical performance metrics, especially for high-speed running, must also be considered when evaluating changes. Machine learning approaches may eventually support the interpretation of complex monitoring data, but their practical and predictive value requires further empirical validation. The review’s central message, however, is immediately actionable: before acting on any number, a practitioner should first identify whether it is being used to quantify exposure, evaluate an external performance target, confirm an intended physiological stimulus, or assess underlying capacity. Matching the measure to the question, the authors suggest, is the first step toward more coherent monitoring and training prescription.

Subject of Research: A conceptual framework for interpreting training load, performance output, physiological response, and fitness status in soccer monitoring

Article Title: An updated conceptual framework on physical performance and training in soccer: a narrative review

Article References: Pillitteri, G., Clemente, F. M., Beato, M., Rossi, A., Sarmento, H., Petrucci, M., Carling, C., Bianco, A., & Battaglia, G. (2026). An updated conceptual framework on physical performance and training in soccer: a narrative review. Sports Medicine – Open, 12(1), Article 158. https://doi.org/10.1186/s40798-026-01127-5

Image Credits: AI Generated

DOI: 10.1186/s40798-026-01127-5

Keywords: soccer, training load, external load, internal load, fitness status, athlete monitoring, readiness, performance output, physiological response, narrative review, sports science, high-speed running