The Transfer Due Diligence Twin: Biological Digital Twins as Asset Valuation Instruments in Professional Sports Medicine

Nicolas Waern WINNIIO AB, Gothenburg, Sweden ORCID: 0000-0001-7970-2707

Corresponding author: ceo@winniio.io

Version: 1.0 | Date: 2026-04-14 | License: CC-BY-4.0

Cite as: Waern, N. (2026). The Transfer Due Diligence Twin: Biological Digital Twins as Asset Valuation Instruments in Professional Sports Medicine. WINNIIO AB. Zenodo. https://doi.org/10.5281/zenodo.PENDING

Prior works in this series: Waern (2026a). DOI: 10.5281/zenodo.19586835 [Reality Construct] Waern (2026b). DOI: 10.5281/zenodo.19586851 [Sovereign Body]

Abstract

Professional sports transfers involve extraordinary information asymmetry: the athlete's body IS the reality, yet every scouting report and pre-transfer medical is a shadow of it -- a static projection of a dynamic biological system. The acquiring club does not lack data; it lacks legibility. This position paper proposes the transfer due diligence twin [proposed term]: a biological digital twin that makes the athlete's physiology legible to all transaction parties. The sports performance analytics industry -- Catapult, STATSports, Zone7, Playermaker, and the FIFA Football Medicine programme -- has made significant progress in workload monitoring and statistical injury prediction. This paper's contribution is not to claim novelty in the concept of athlete risk assessment but to argue that a qualitative epistemological upgrade is required: from statistical pattern recognition to mechanistic simulation using Large Quantitative Models (LQMs) [proposed term]. Drawing on boundary object theory (Star & Griesemer, 1989), the reality construct framework (Waern, 2026a), and the sovereign body principle (Waern, 2026b), we propose a three-tier product architecture and situate it within a six-segment market. We examine the regulatory strategy of wellness positioning under EU MDR 2017/745 and the GDPR implications of processing special category health data for athletes whose bodies are the asset under transaction. The paper is a position paper grounded in a single anonymised product roadmap discussion. All novel terminology is flagged.

Keywords: biological digital twin, transfer due diligence, sports medicine, boundary objects, large quantitative models, injury risk, athlete valuation, data sovereignty, wellness platform, GDPR, mechanistic modelling, musculoskeletal simulation

1. Introduction: The Transfer Problem

1.1 Information Asymmetry in Elite Athlete Markets

European football transfer spending routinely exceeds five billion euros annually (Poli et al., 2023). Each transaction is a bet: the acquiring club wagers that future performance will justify the fee, while the selling club has an incentive to maximise the price and may not disclose the full physiological profile of the asset.

This is a textbook information asymmetry (Akerlof, 1970). The selling party possesses private information -- injury history, chronic niggles, recovery patterns, workload tolerance -- that the buying party cannot fully observe. The pre-transfer medical examination captures a single physiological snapshot over twenty-four to forty-eight hours. It cannot model how a knee that passed clinical assessment today will respond to the acquiring club's training regime, pitch surface, and match density. It cannot simulate the interaction between a repaired ACL and the compensatory movement patterns developed since surgery.

The consequences are visible in every transfer window. Players acquired for record fees suffer injuries within weeks. The unnamed case that prompted this paper -- a transfer valued around one billion SEK, followed by significant injury -- is neither unusual nor unforeseeable. It is the predictable outcome of evaluating a dynamic biological system with static diagnostic tools.

1.2 The Analogy: Due Diligence in Other Asset Classes

Due diligence -- systematic investigation prior to a transaction -- is well established in other asset markets. Corporate M&A due diligence spans months (Howson, 2003). Vehicle history reports (Carfax, HPI) aggregate longitudinal data from multiple sources. Neither involves a static snapshot. Yet in the sports transfer market -- where asset values frequently exceed those of SME acquisitions -- the standard remains a medical examination supplemented by video analysis.

The fundamental insight is this: the cave already exists. The athlete's body IS the reality -- the ground truth of the asset under transaction. Every scouting report, every match statistic, every pre-transfer medical is a shadow of that reality: a partial, static projection of a dynamic biological system. The twin does not create the athlete's biology; it makes it legible to all parties in a transfer. This paper proposes that the emerging capability of biological digital twins creates the possibility of a fundamentally different instrument: a transfer due diligence twin [proposed term] that provides the acquiring party with a dynamic, simulatable model of the athlete's physiology, enabling prospective injury risk assessment, workload tolerance modelling, and recovery trajectory prediction before the transaction is completed.

1.3 Scope and Position

This is a position paper developing a conceptual framework grounded in a single anonymised product roadmap discussion. It does not report clinical validation. Novel terminology is flagged with [proposed term]. The geographic context is primarily European, though the principles extend to any sport with high-value player transactions.

2. The Current Landscape: Statistical Approaches to Athlete Risk

A substantial industry already exists around athlete monitoring and injury prediction. This paper's contribution is not to claim novelty in athlete risk assessment but to argue that the prevailing epistemological approach has inherent limitations that mechanistic modelling can address.

2.1 The Statistical Paradigm

GPS-based tracking (Catapult/Genius Sports, STATSports, Playermaker) is ubiquitous in elite sport, generating rich data on training load and injury incidence (Rossi et al., 2018; Carey et al., 2018). The UEFA injury studies have established robust epidemiological baselines over two decades (Ekstrand et al., 2011, 2020). AI-driven platforms like Zone7 (claiming 70%+ injury prediction accuracy) and the FIFA Football Medicine programme represent the statistical frontier, though systematic review reveals heterogeneous results with limited external validation (Claudino et al., 2019).

2.2 The ACWR Debate

The acute:chronic workload ratio (ACWR), central to Gabbett's (2016) training-injury prevention paradox, illustrates the epistemological limits of statistical approaches. Impellizzeri et al. (2020) identified a mathematical coupling artefact producing spurious load-injury associations. This controversy does not invalidate workload monitoring but underscores the need for models capturing causal mechanisms linking load to tissue response.

2.3 The Epistemological Gap

All statistical approaches identify correlations across populations; none model the causal chain from load through tissue mechanics to injury for a specific individual. The due diligence twin occupies a different epistemological category: mechanistic simulation built on platforms such as OpenSim (Delp et al., 2007) and AnyBody Technology (Damsgaard et al., 2006). However, a significant gap remains between these research-grade platforms and a deployable tool operating at transfer-negotiation speed and reliability. Acknowledging this gap is essential.

3. The Product Ladder: From Aggregation to Simulation

3.1 Three Tiers of Health Intelligence

The anonymised discussion that prompted this paper articulated a product architecture with three ascending tiers, each representing a qualitative increase in the depth of physiological understanding provided to the user. We formalise this architecture here, situating each tier within the SMILE methodology (Waern, 2026a), which inverts the conventional data-driven paradigm by beginning with desired outcomes and working backward through actions, insights, information, and data.

Tier 1: Aggregated Health Record (Free). Consolidates fragmented health data -- club medical records, wearables, lab results -- into a unified longitudinal profile. This addresses what Carlile (2002) terms the syntactic knowledge boundary: creating shared vocabulary across siloed systems.

Tier 2: AI-Assisted Risk Prediction (Premium). Applies machine learning to generate predictive risk scores and connects athletes with relevant specialists. This addresses Carlile's semantic boundary: providing shared interpretive frameworks across disciplines. However, ML models trained on population data predict average outcomes for similar athletes -- they cannot model the specific dynamics of an individual's musculoskeletal system under specific loading (see Section 6).

Tier 3: Biological Digital Twin Simulation (Top Tier). The transfer due diligence twin itself: a mechanistic simulation modelling what could happen to this body under specified conditions, with quantified uncertainty. The athlete's musculoskeletal system is modelled as differential equations parameterised from individual imaging and biomechanical data. Estimates remain subject to parameter uncertainty, constitutive equation limitations, and boundary condition sensitivity (Valente et al., 2014). This tier addresses Carlile's pragmatic boundary: misaligned buyer-seller incentives require a verified, auditable simulation as trust infrastructure.

3.2 The SMILE Alignment

Each tier maps to the SMILE methodology (Waern, 2026a): Tier 1 corresponds to Situate and Map (understanding current state), Tier 2 to Illuminate (surfacing non-obvious patterns), and Tier 3 to Lead and Evolve (simulation-informed decisions, iterative model refinement). The product ladder is an epistemological progression from description to prediction to explanation.

4. The Due Diligence Twin: Concept and Comparisons

4.1 Definition

The transfer due diligence twin [proposed term] is a biological digital twin of an elite athlete, constructed from longitudinal health data, biomechanical assessment, imaging, and wearable streams, presented to a prospective acquiring party as a component of the transfer process. Its function is to reduce information asymmetry by providing a dynamic, simulatable model of the asset's physiological condition and trajectory. The concept was articulated as "a stamp -- something mandatory in connection with buying and selling players... a form of due diligence, a pass at the individual level." Like a passport -- a standardised document enabling a transaction by attesting to the bearer's status -- the twin attests to the physiological status and projected trajectory of the athlete.

4.2 Comparison with Financial Due Diligence

In corporate acquisitions, due diligence is not optional -- the acquirer must conduct reasonable investigation (Howson, 2003). The transfer due diligence twin proposes the same standard for biological assets. If a technology exists that models injury risk with greater fidelity than a forty-eight-hour medical, the failure to use it becomes a governance failure. The analogy is structural: both address information asymmetry, both require historical and current data, and both quantify otherwise opaque risk.

4.3 Comparison with Vehicle History Reports

The vehicle history report model is relevant because it demonstrates that longitudinal data aggregation can become a market norm without regulatory mandate. No law requires a Carfax report, yet its absence signals concealment. The analogy is imperfect: a Carfax report aggregates historical events, whereas a biological twin generates forward-looking simulations whose predictive accuracy remains unvalidated at scale. Nevertheless, the same dynamic could emerge in athlete transfers. If a standardised athlete health passport [proposed term] exists, clubs that present it signal confidence; clubs that withhold it signal concealment.

5. The Twin as Boundary Object in Sports Medicine

5.1 Coincident Boundaries

The transfer due diligence twin is, in Star and Griesemer's (1989) taxonomy, a coincident boundary object -- one whose boundaries are coextensive with the object of study but read differently by different communities:

1. The team physician, who reads the twin as a clinical model -- injury mechanisms, tissue loading, rehabilitation trajectories.
2. The head coach, who reads it as a tactical resource -- workload tolerance, optimal rest patterns, position-specific demands.
3. The scout/sporting director, who reads it as a valuation instrument -- risk-adjusted transfer fee, contract length implications, resale value trajectory.
4. The insurer, who reads it as an actuarial model -- premium calculation, exclusion identification, liability quantification.
5. The athlete, who reads it as a sovereignty instrument -- a verifiable record of their own body that they own, control, and can present to any prospective employer.

Each community projects its own vocabulary and decision context onto the same computational object. The twin enables cooperation without consensus (Star & Griesemer, 1989, p. 393).

Star (2010) cautioned that boundary objects lose interpretive flexibility when standardised by powerful actors. In the transfer context, the buying club's data science team may impose its preferred interpretation -- injury risk thresholds, acceptable training loads -- that override the selling club's or the athlete's own assessment, transforming the twin from a negotiation tool into an instrument of buyer power.

5.2 Absorptive Capacity and the Twin

Cohen and Levinthal (1990) defined absorptive capacity as an organisation's ability to assimilate and apply new external information. The acquiring club's absorptive capacity determines whether it can extract value from a twin. A club with sophisticated sports science infrastructure can integrate twin-derived insights into load management; one without receives it as an opaque artefact. This creates a secondary market: the platform provides the interpretive capacity (expert matchmaking, scenario simulation) that enables clubs with lower absorptive capacity to use the twin effectively.

5.3 Transfer Economics and the Benefits Case

The economic case is grounded in transfer failure costs. Transfermarkt data indicates Premier League clubs write off hundreds of millions annually on underperforming acquisitions. A GBP 50 million player spending half the season injured costs not only the fee but GBP 100-200K weekly wages, the squad slot opportunity cost, and cascading effects on league position. If the twin prevents one catastrophic failure per season, ROI exceeds simulation cost by orders of magnitude. Benefits accrue across medical, coaching, scouting, and finance functions -- a shared infrastructure investment (Weill & Broadbent, 1998).

6. Large Quantitative Models for Athletic Performance

6.1 The Epistemological Distinction

Statistical models learn patterns from historical data and answer: "what is the predicted injury rate for athletes like this one?" Mechanistic models encode causal structure and answer: "given this athlete's specific tissue properties and loading history, what could happen to their ACL under this programme, and with what probability?" The uncertainty in mechanistic models is structured and physically interpretable, unlike the opaque confidence intervals of statistical models.

We adopt the term Large Quantitative Models (LQMs) [proposed term] from Waern (2026a, 2026b) for mechanistic simulation engines modelling biological systems from first principles. "Large" refers not to parameter count but to the breadth of physiological subsystems -- musculoskeletal, cardiovascular, metabolic, endocrine -- coupled through conservation laws.

6.2 Injury Risk as a Simulation Problem

The dominant frameworks for understanding sports injury -- Meeuwisse's (1994) multifactorial model and its extensions (Meeuwisse et al., 2007; Bahr & Krosshaug, 2005) -- identify injury as the product of intrinsic risk factors (anatomy, prior injury, age, sex), extrinsic risk factors (equipment, playing surface, weather), and an inciting event. Bahr (2014) argued for abandoning vague injury definitions in favour of precise, standardised recording -- a prerequisite for any twin-based system that ingests injury history data. The classification of modifiable versus non-modifiable risk factors, central to prevention-oriented frameworks (Meeuwisse et al., 2007), is precisely where mechanistic models offer an advantage: they can simulate the consequences of modifying a specific factor (e.g., training load, surface type) for a specific individual.

As discussed in Section 2, Gabbett's (2016) training-injury prevention paradox and the subsequent ACWR debate (Impellizzeri et al., 2020) underscore the need for individualised load modelling: the "right" training load depends on the specific athlete's tissue adaptation history, current physiological state, and competitive demands.

A statistical model can estimate workload ratios but cannot model tissue-level consequences for a specific individual. An LQM, parameterised from the athlete's imaging (MRI, ultrasound), motion capture, and physiological measurements (lactate threshold, VO2max, HRV), simulates the consequences of a proposed training programme on specific tissues.

Consider a midfielder with an ACWR of 1.3 over the previous four weeks. A statistical model (Zone7-class) flags this as elevated risk based on population norms. A mechanistic model parameterised to this athlete's musculotendon properties, recovery profile, and training history might classify the same load as within their individual tolerance envelope -- or, conversely, flag a risk the statistical model misses because the athlete's tendon stiffness has been declining over three months despite stable external loads. The epistemological difference is between "athletes like you get injured at this ratio" and "your specific tissue is approaching its failure threshold."

6.3 Recovery Trajectories and Metabolic Load

Recovery involves tissue repair, metabolic recovery, neurological recovery, and psychological recovery -- subsystems interacting through feedback loops on different timescales. We introduce the metabolic load envelope [proposed term]: the individual-specific dynamic boundary within which an athlete can sustain competitive output without exceeding recovery capacity, dependent on training history, sleep, nutrition, and environment. The metabolic load envelope concept draws on established bioenergetic modelling (Mader, 2003) which provides the physiological basis for individual threshold estimation.

This extends Impellizzeri et al.'s (2019) external/internal load framework by adding simulated load [proposed term] -- the LQM-predicted physiological response to a hypothetical future external load. Simulated load makes the twin prospective: it models not "how did this athlete respond to last season?" but "how could this athlete respond to next season's schedule, with what probability of exceeding their envelope?"

7. Regulatory Navigation: Wellness, Not Medicine

7.1 The Medical Device Boundary

The EU MDR 2017/745 broadly defines medical devices to include software intended for diagnosis, prediction, or prognosis. The regulatory strategy examined here -- positioning the platform as a wellness instrument -- frames its intended use as personal health tracking and fitness optimisation rather than clinical decision support. This mirrors the fitness wearable industry's approach: Apple Watch's heart rate monitoring is wellness; its irregular rhythm notification is a regulated medical device (FDA DEN180044). The boundary is defined by intended use, not technical capability.

7.2 Risks of the Wellness Positioning

This positioning carries risks. If clubs use twin outputs for clinical decisions, the de facto use may cross the medical device boundary regardless of stated intent (MDCG 2019-11). The transfer context creates pressure toward clinical-grade claims -- clubs paying tens of millions want certainty, not wellness insights. And the wellness/medical boundary is under increasing regulatory scrutiny, trending toward broader classification (Muhlberger et al., 2022).

7.3 GDPR and Athlete Data Sovereignty

Athlete health data is special category data under GDPR Article 9. In the transfer context, multiple controllers coexist: the selling club (employer), the athlete (data subject), the platform (processor), and the acquiring club (prospective employer). The sovereign body principle (Waern, 2026b) argues that the athlete must be the primary data controller -- the twin is their twin, and its disclosure requires explicit, informed, granular consent. This architectural commitment also addresses the power asymmetry in sports labour economics (Feess & Muehlheusser, 2003): if the twin is club-owned, it becomes a surveillance instrument; if athlete-owned, an empowerment instrument.

8. Market Architecture: Six Segments

8.1 The Addressable Landscape

The transfer due diligence twin is the flagship application of a broader platform serving six segments: (1) elite sports (clubs, federations, agencies), (2) corporate health (workforce wellness, Goetzel et al., 2014), (3) insurance (individualised underwriting, Frees et al., 2014), (4) pharmacy chains, (5) patient organisations, and (6) high-net-worth individuals.

8.2 The Equity Dimension

Women are systematically underrepresented in health data (Criado Perez, 2019), and female athletes face specific injury patterns -- higher ACL injury incidence (Arendt & Dick, 1995), menstrual cycle influences on soft tissue risk (Martin et al., 2018) -- that statistical models trained predominantly on male data handle poorly. The LQM approach partially addresses this gap because mechanistic models are parameterised from the individual's own data rather than population distributions. However, the constitutive equations and validation datasets underlying musculoskeletal models remain male-dominated (Costello et al., 2014), a limitation the field must address.

9. Limitations

This paper's limitations are substantial. No biological digital twin has been clinically validated for injury prediction in elite athletes; the capability exists in research (Erdemir et al., 2007; Lloyd & Besier, 2003; Pizzolato et al., 2015) but the translation to deployable product remains a major engineering challenge. The wellness regulatory positioning is a commercial tactic, not a settled legal position. The paper does not address subject-specific parameterisation challenges, computational cost of real-time simulation, or model sensitivity to parameter uncertainty (Valente et al., 2014) in detail. The market architecture is conceptual, not empirically validated. The single-meeting provenance limits generalisability across sports and regulatory environments.

10. Future Research and Conclusion

10.1 Research Agenda

The transfer due diligence twin opens several research directions: (1) prospective cohort studies comparing injury incidence in twin-informed versus conventional transfers; (2) standardised parameterisation workflows for constructing athlete-specific musculoskeletal models from routinely available clinical data; (3) regulatory pathway analysis across EU, UK, and US jurisdictions; (4) validation of mechanistic models across sex, ethnicity, and body composition to address equity gaps; (5) data governance architectures implementing athlete sovereignty in multi-stakeholder transfer contexts; (6) retrospective economic impact analysis correlating pre-transfer assessment quality with transfer outcomes; and (7) empirical study of how the twin functions as a boundary object across professional communities.

10.2 Conclusion

The professional sports transfer market routinely converts biological uncertainty into financial risk through insufficient assessment. The pre-transfer medical -- a snapshot of a dynamic system -- is equivalent to valuing a company by its bank balance on a single day. The twin offers a fundamentally different instrument: a longitudinal, mechanistic model that converts biological uncertainty into quantifiable risk.

The concept is not technically ready for deployment at scale. The gap between research-grade musculoskeletal simulation platforms (OpenSim, AnyBody) and a deployable transfer due diligence tool remains substantial. But the direction is clear. As personal physiological data density increases -- continuous glucose monitors, accelerometers, proteomics, metabolomics -- the raw material for biological digital twins becomes increasingly available.

This paper argues, following the sovereign body principle (Waern, 2026b), that the twin must belong to the athlete. The athlete's body is the reality -- the ground truth that every scouting report approximates. The twin does not create the athlete's biology; it makes it legible. Any architecture that places the twin under club or platform ownership recreates the power asymmetry it should be designed to resolve.

The shift from snapshot assessment to longitudinal simulation, from statistical shadow to mechanistic legibility, will be governed not by technological capability alone but by the governance architecture within which the technology is embedded. The transfer due diligence twin is as much a governance innovation as it is a technological one.

Conflict of Interest and Funding

The author declares no conflict of interest. This work received no external funding. The author is CEO of WINNIIO AB, which develops the Life Atlas platform referenced in this paper.

References

Akerlof, G. A. (1970). The market for "lemons": Quality uncertainty and the market mechanism. The Quarterly Journal of Economics, 84(3), 488--500. https://doi.org/10.2307/1879431

Arendt, E., & Dick, R. (1995). Knee injury patterns among men and women in collegiate basketball and soccer: NCAA data and review of literature. The American Journal of Sports Medicine, 23(6), 694--701. https://doi.org/10.1177/036354659502300611

Bahr, R. (2014). Demise of the D word: The case for dichotomising sports injury definitions. British Journal of Sports Medicine, 48(7), 491--492. https://doi.org/10.1136/bjsports-2014-e4023rep

Bahr, R., & Krosshaug, T. (2005). Understanding injury mechanisms: A key component of preventing injuries in sport. British Journal of Sports Medicine, 39(6), 324--329. https://doi.org/10.1136/bjsm.2005.018341

Carey, D. L., Blanch, P., Ong, K.-L., Crossley, K. M., Crow, J., & Morris, M. E. (2018). Training loads and injury risk in Australian football -- differing acute:chronic workload ratio calculations influence injury risk. British Journal of Sports Medicine, 52(16), 1046--1052. https://doi.org/10.1136/bjsports-2016-096309

Carlile, P. R. (2002). A pragmatic view of knowledge and boundaries: Boundary objects in new product development. Organization Science, 13(4), 442--455. https://doi.org/10.1287/orsc.13.4.442.2953

Claudino, J. G., Capanema, D. O., de Souza, T. V., Serrao, J. C., Machado Pereira, A. C., & Nassis, G. P. (2019). Current approaches to the use of artificial intelligence for injury risk assessment and performance prediction in team sports: A systematic review. Sports Medicine -- Open, 5(1), 28. https://doi.org/10.1186/s40798-019-0202-3

Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128--152. https://doi.org/10.2307/2393553

Damsgaard, M., Rasmussen, J., Christensen, S. T., Surma, E., & de Zee, M. (2006). Analysis of musculoskeletal systems in the AnyBody Modeling System. Simulation Modelling Practice and Theory, 14(8), 1100--1111. https://doi.org/10.1016/j.simpat.2006.09.001

Delp, S. L., Anderson, F. C., Arnold, A. S., Loan, P., Habib, A., John, C. T., ... & Thelen, D. G. (2007). OpenSim: Open-source software to create and analyze dynamic simulations of movement. IEEE Transactions on Biomedical Engineering, 54(11), 1940--1950. https://doi.org/10.1109/TBME.2007.901024

Costello, J. T., Bieuzen, F., & Bleakley, C. M. (2014). Where are all the female participants in sports and exercise medicine research? European Journal of Sport Science, 14(8), 847--851. https://doi.org/10.1080/17461391.2014.911354

Criado Perez, C. (2019). Invisible women: Exposing data bias in a world designed for men. Chatto & Windus.

Ekstrand, J., Hagglund, M., & Walden, M. (2011). Epidemiology of muscle injuries in professional football (soccer). The American Journal of Sports Medicine, 39(6), 1226--1232. https://doi.org/10.1177/0363546510395879

Ekstrand, J., Spreco, A., Bengtsson, H., & Bahr, R. (2020). Injury rates decreased in men's professional football: An 18-year prospective cohort study of almost 12 000 injuries sustained during 1.8 million hours of play. British Journal of Sports Medicine, 55(19), 1084--1091. https://doi.org/10.1136/bjsports-2020-103159

Erdemir, A., McLean, S., Herzog, W., & van den Bogert, A. J. (2007). Model-based estimation of muscle forces exerted during movements. Clinical Biomechanics, 22(2), 131--154. https://doi.org/10.1016/j.clinbiomech.2006.09.005

European Parliament and Council. (2017). Regulation (EU) 2017/745 on medical devices. Official Journal of the European Union, L 117, 1--175.

European Parliament and Council. (2016). Regulation (EU) 2016/679 on the protection of natural persons with regard to the processing of personal data (General Data Protection Regulation). Official Journal of the European Union, L 119, 1--88.

Feess, E., & Muehlheusser, G. (2003). Transfer fee regulations in European football. European Economic Review, 47(4), 645--668. https://doi.org/10.1016/S0014-2921(02)00308-2

Frees, E. W., Derrig, R. A., & Meyers, G. (2014). Predictive modeling applications in actuarial science. Cambridge University Press. https://doi.org/10.1017/CBO9781139342674

Gabbett, T. J. (2016). The training-injury prevention paradox: Should athletes be training smarter and harder? British Journal of Sports Medicine, 50(5), 273--280. https://doi.org/10.1136/bjsports-2015-095788

Goetzel, R. Z., Henke, R. M., Tabrizi, M., Pelletier, K. R., Loeppke, R., Ballard, D. W., ... & Metz, R. D. (2014). Do workplace health promotion (wellness) programs work? Journal of Occupational and Environmental Medicine, 56(9), 927--934. https://doi.org/10.1097/JOM.0000000000000276

Howson, P. (2003). Due diligence: The critical stage in mergers and acquisitions. Gower Publishing.

Impellizzeri, F. M., Tenan, M. S., Kempton, T., Novak, A., & Coutts, A. J. (2020). Acute:chronic workload ratio: Conceptual issues and fundamental pitfalls. International Journal of Sports Physiology and Performance, 15(6), 907--913. https://doi.org/10.1123/ijspp.2019-0738

Impellizzeri, F. M., Marcora, S. M., & Coutts, A. J. (2019). Internal and external training load: 15 years on. International Journal of Sports Physiology and Performance, 14(2), 270--273. https://doi.org/10.1123/ijspp.2018-0935

Lloyd, D. G., & Besier, T. F. (2003). An EMG-driven musculoskeletal model to estimate muscle forces and knee joint moments in vivo. Journal of Biomechanics, 36(6), 765--776. https://doi.org/10.1016/S0021-9290(03)00010-1

Mader, A. (2003). Glycolysis and oxidative phosphorylation as a function of cytoplasmic phosphorylation state and power output of the muscle cell. European Journal of Applied Physiology, 88(4--5), 317--338. https://doi.org/10.1007/s00421-002-0676-3

Martin, D., Sale, C., Cooper, S. B., & Elliott-Sale, K. J. (2018). Period prevalence and perceived side effects of hormonal contraceptive use and the menstrual cycle in elite athletes. International Journal of Sports Physiology and Performance, 13(7), 926--932. https://doi.org/10.1123/ijspp.2017-0330

MDCG. (2019). MDCG 2019-11: Guidance on qualification and classification of software in Regulation (EU) 2017/745 -- MDR and Regulation (EU) 2017/746 -- IVDR. Medical Device Coordination Group.

Meeuwisse, W. H. (1994). Assessing causation in sport injury: A multifactorial model. Clinical Journal of Sport Medicine, 4(3), 166--170.

Meeuwisse, W. H., Tyreman, H., Hagel, B., & Emery, C. (2007). A dynamic model of etiology in sport injury: The recursive nature of risk and causation. Clinical Journal of Sport Medicine, 17(3), 215--219. https://doi.org/10.1097/JSM.0b013e3180592a48

Muhlberger, M., Drosatos, G., & Wac, K. (2022). The blurring boundary between wellness and medical apps: Legal and ethical considerations. Digital Health, 8. https://doi.org/10.1177/20552076221123439

Pizzolato, C., Lloyd, D. G., Sartori, M., Ceseracciu, E., Besier, T. F., Fregly, B. J., & Reggiani, M. (2015). CEINMS: A toolbox to investigate the influence of different neural control solutions on the prediction of muscle excitation and joint moments during dynamic motor tasks. Journal of Biomechanics, 48(14), 3929--3936. https://doi.org/10.1016/j.jbiomech.2015.09.021

Poli, R., Ravenel, L., & Besson, R. (2023). Annual review of the global transfer market. CIES Football Observatory Monthly Report, Issue 89. https://football-observatory.com

Rossi, A., Pappalardo, L., Cintia, P., Iaia, F. M., Fernandez, J., & Medina, D. (2018). Effective injury forecasting in soccer with GPS training data and machine learning. PloS One, 13(7), e0201264. https://doi.org/10.1371/journal.pone.0201264

Star, S. L. (2010). This is not a boundary object: Reflections on the origin of a concept. Science, Technology, & Human Values, 35(5), 601--617. https://doi.org/10.1177/0162243910377624

Star, S. L., & Griesemer, J. R. (1989). Institutional ecology, "translations" and boundary objects: Amateurs and professionals in Berkeley's Museum of Vertebrate Zoology, 1907--39. Social Studies of Science, 19(3), 387--420. https://doi.org/10.1177/030631289019003001

Valente, G., Pitto, L., Testi, D., Seth, A., Delp, S. L., Stagni, R., ... & Taddei, F. (2014). Are subject-specific musculoskeletal models robust to the uncertainties in parameter identification? PloS One, 9(11), e112625. https://doi.org/10.1371/journal.pone.0112625

Waern, N. (2026a). The reality construct: Digital twins as boundary-spanning artefacts for knowledge absorption, organisational evolution, and the co-authorship of intended futures. WINNIIO AB. Zenodo. https://doi.org/10.5281/zenodo.19586835

Waern, N. (2026b). The sovereign body: Personal digital twins as boundary-spanning objects across the WHO quality of life dimensions. WINNIIO AB. Zenodo. https://doi.org/10.5281/zenodo.19586851

Weill, P., & Broadbent, M. (1998). Leveraging the new infrastructure: How market leaders capitalize on information technology. Harvard Business School Press.

This paper is published under the Creative Commons Attribution 4.0 International License (CC-BY-4.0). You are free to share, copy, redistribute, adapt, transform, and build upon the material for any purpose, including commercially, provided appropriate credit is given.