High-Intensity Interval Training in the Age of Artificial Intelligence
A Data-Driven, Adaptive, and Health-Centered Paradigm for Cardio, Bodyweight Exercise, Sport for All, and High-Performance Sport
Final academic English adaptation
ABSTRACT
High-intensity interval training (HIIT) should not be understood merely as a short and difficult workout. It is a family of adjustable protocols in which periods of relatively intense activity alternate with periods of recovery. The ability to modify intensity, work duration, the work-to-recovery ratio, exercise mode, the number of intervals, and weekly volume makes HIIT an adaptable training architecture. This same parametric structure produces repeated and interpretable data on heart rate, speed, power, performance decline, perceived exertion, and recovery, thereby creating a natural interface with wearable devices, computer vision, virtual coaching, and artificial intelligence. Using a narrative review and conceptual-analysis approach, this article proposes a five-pillar framework for cardio HIIT: (1) intelligent and personalized; (2) evidence-based and measurable; (3) applicable and inclusive; (4) health-centered and safety-manageable; and (5) scalable and future-oriented. Current evidence, particularly for improvements in cardiorespiratory fitness, supports the use of HIIT in healthy, inactive, and selected clinical populations. However, HIIT is not superior to moderate-intensity continuous training or resistance training for every outcome, and its safety depends on screening, protocol design, progressive overload, technical competency, and appropriate supervision. The central conclusion is that HIIT is not the sole future of exercise; rather, it is one of the training architectures most prepared for a future in which sport, medicine, media, data, and artificial intelligence operate within a shared decision-making loop.
Keywords: high-intensity interval training; cardio; cardiorespiratory fitness; cardiovascular health; bodyweight exercise; sport for all; high-performance sport; artificial intelligence; adaptive exercise; data-driven training; wearable technology; virtual coach; sports medicine; digital health
1. Introduction
The contemporary world faces a persistent paradox: evidence supporting the benefits of physical activity is more extensive than ever, yet lack of time, cost, limited access, declining motivation, and difficulty sustaining behavior continue to prevent a large proportion of the population from exercising regularly. World Health Organization guidelines emphasize aerobic activity, muscle strengthening, and reduced sedentary behavior across age groups and clinical populations; however, translating those recommendations into sustainable behavior requires exercise formats that are both scientifically grounded and feasible in daily life [1].
Because of its time efficiency, protocol diversity, potential for home-based delivery, and adjustability to individual capacity, HIIT represents one important response to this challenge. A 2024 umbrella review encompassing 24 systematic reviews, 429 primary studies, and nearly 13,000 participants found relatively consistent evidence that HIIT improves cardiorespiratory fitness in adults compared with no exercise and, in many analyses, compared with moderate-intensity continuous training, although the methodological quality of the included reviews was not uniform [2].
The expansion of research across healthy, athletic, inactive, older, and clinical populations has made HIIT one of the most productive and interdisciplinary areas of exercise physiology [3]. Its scientific position is strengthened, however, only when promotional language is replaced by precise claims. Statements such as “the best exercise,” “completely safe,” “guaranteed fat burning,” or “appropriate for everyone” are inconsistent with protocol heterogeneity and individual variability. The argument advanced here is strong but defensible: because HIIT is adaptable, measurable, relevant to cardiovascular health, compatible with bodyweight exercise, and highly amenable to artificial intelligence, it is one of the most important training models of the digital-health era.
2. Definition and Conceptual Boundaries
HIIT is not the name of one fixed workout. It refers to a family of sessions in which periods of relatively high-intensity activity alternate with rest or lower-intensity activity. Intensity may be prescribed as a percentage of maximal heart rate, heart-rate reserve, maximal power, speed, ratings of perceived exertion, or performance-test values. Consequently, a four-by-four-minute cycling session, short running intervals, a low-impact program for older adults, and a bodyweight cardio circuit may all use an interval structure while imposing very different physiological and mechanical demands.
Conceptual boundaries matter. HIIT is not synonymous with all-out sprint interval training, high-intensity functional training, CrossFit, or any session that feels difficult. Exercise mode, relative intensity, interval duration, recovery, and the targeted physiological adaptation determine where a protocol belongs. This distinction is essential for research reporting, clinical prescription, coach education, and injury surveillance.
3. The Five-Pillar Cardio HIIT Framework
To integrate the twelve characteristics developed in this project, this article proposes a five-pillar Cardio HIIT Framework. The framework is conceptual and has not yet been validated as a psychometric scale or clinical protocol. It should therefore be tested in future research. Its value lies in converting a dispersed set of claims into five domains that can be evaluated systematically.
1. Intelligent and personalized: This pillar encompasses adaptability, data orientation, statistical monitoring, and alignment with artificial intelligence. Nearly every principal HIIT variable – intensity, work duration, recovery duration, number of repetitions, exercise mode, and weekly volume – can be modified. An intelligent system can compare the prescribed load with the individual’s actual response and adjust the next session according to progress, fatigue, and recovery quality.
2. Evidence-based and measurable: HIIT provides a compact setting in which core exercise-physiology principles can be observed: overload, specificity, fatigue, recovery, cardiorespiratory adaptation, and interindividual variation. Metaphorically, it may be described as an “applied synopsis of exercise physiology,” not as a formal scientific term, but because metabolic, cardiorespiratory, and performance responses can be recorded together within one structured session.
3. Applicable and inclusive: This pillar links sport for all with specialized sport preparation. HIIT can be designed for an inactive adult, a school-based program, a home workout, combat-sport conditioning, or team-sport preparation. Bodyweight formats reduce cost and equipment dependence and enable delivery at home, in parks, schools, workplaces, and gyms. Inclusion does not mean one protocol for everyone; it means the capacity for scientifically guided scaling.
4. Health-centered and safety-manageable: The connection with cardiorespiratory fitness, sports medicine, and cardiac rehabilitation forms this pillar. The word “safe” is scientifically too absolute; a more defensible formulation is “designable with manageable risk.” Evidence from cardiac-rehabilitation settings suggests that prescribed HIIT can be delivered to stable, screened patients under professional supervision with a low rate of major events [5, 6, 17, 21]. This conclusion should not be generalized to unscreened, unsupervised high-intensity exercise in all individuals.
5. Scalable and future-oriented: The simple logic of work and recovery makes basic instruction accessible, while the range of possible movements is nearly unlimited. Cardio HIIT, bodyweight HIIT, dance intervals, low-impact formats, and combat-oriented formats are suited to media, applications, and virtual coaching. This scalability supports expansion through social platforms and home exercise, but it must be paired with level-based instruction, technical guidance, safety warnings, and content-quality control.
4. Twelve Operational Characteristics
In practice, the five pillars can be translated into twelve observable and assessable characteristics. These should be treated not as slogans, but as criteria for program design, monitoring, and evaluation.
1. Adaptive: Intensity, volume, the work-to-recovery ratio, exercise selection, and training environment can be changed according to the individual and the objective.
2. Data-driven: The session is divided into clear time units and generates repeated data on physiological response and recovery.
3. Science-centered: Exercise-physiology principles such as overload, recovery, and specificity are directly represented in program design.
4. Public and specialized: Applications range from population health to high-performance sport preparation.
5. Media-compatible: Timers, movement variety, and short duration make the format suitable for video, applications, and home broadcasting.
6. Low-cost and accessible: No-equipment and low-equipment versions can be delivered in diverse settings.
7. Diverse and extensible: Running, cycling, dance, combat exercise, functional exercise, and bodyweight movements can all serve as delivery modes.
8. Connected to medicine and cardiovascular health: Improved cardiorespiratory fitness and controlled use in rehabilitation form its principal medical connection.
9. Risk-manageable: Low-impact alternatives and gradual progression may reduce risk, but absolute safety cannot be claimed.
10. Teachable: The core work-recovery logic is simple, although technical instruction and intensity prescription require professional competence.
11. Measurable and monitorable: Heart rate, time, speed, power, repetitions, perceived exertion, and performance decline can be recorded.
12. Intelligence-ready: Its digital parameters are well suited to personalization, prediction, and AI-supported decision systems.
5. HIIT as an Adaptive Architecture Rather Than a Fixed Program
The principal structural advantage of HIIT is multidimensional adjustability. A program can be progressed or regressed by changing interval duration, the number of intervals, the activity-to-recovery ratio, relative intensity, movement complexity, weekly volume, recovery mode, or the training environment. This feature allows progressive overload to be implemented more precisely than in a rigid, one-size-fits-all format.
Two people may use the same timer while receiving entirely different prescriptions. A beginner may perform brisk walking with extended recovery; an elite athlete may use power- or acceleration-based intervals; and a person with joint limitations may use cycling, rowing, or low-impact movements. “High intensity” should therefore be defined relative to individual capacity rather than by an absolute speed or exercise.
School-based research and studies across age groups indicate that interval logic can be scaled [4]. Nevertheless, the prescriptions for a young athlete, a novice older adult, and a patient with cardiovascular disease should not be identical. Screening, medical history, medication, pain, baseline fitness, and training goals should influence prescription.
6. Cardio HIIT, Cardiovascular Health, and Sports Medicine
Cardio and HIIT are closely related but are not identical. Cardio refers to exercise that targets the capacity of the heart, lungs, circulation, and skeletal muscle to transport and use oxygen. HIIT is one method of organizing cardio and may simultaneously include elements of strength, speed, and muscular endurance.
The strongest scientific case for HIIT concerns cardiorespiratory fitness. Large reviews indicate that HIIT can improve indices such as peak oxygen uptake in diverse populations [2, 5]. In cardiac rehabilitation, some reviews have reported larger gains in cardiorespiratory fitness than with moderate-intensity continuous training; however, protocol and population heterogeneity require careful wording, such as “in selected stable patients under appropriate supervision” [5, 6].
This cardiovascular connection makes HIIT relevant to sports medicine and to the concept of exercise as part of treatment. Exercise is not an automatic substitute for medication, surgery, or clinical evaluation; rather, it can function as an important non-pharmacological intervention alongside standard care.
7. Body Composition and the Fat-Loss Question
The media popularity of HIIT is partly driven by claims about fat loss. Evidence indicates that HIIT can support body-composition management when combined with appropriate nutrition, sleep, and total physical-activity volume, but “magical fat burning” is not a scientific concept.
A meta-analysis of adults with overweight or obesity found that HIIT and moderate-intensity continuous training both produced modest reductions in body fat and waist circumference, with no definitive difference across many body-composition outcomes. A major advantage of HIIT in that analysis was lower time commitment [7]. Time efficiency should therefore be distinguished from universal superiority for fat loss.
A complete fitness program commonly requires some combination of interval training, light-to-moderate cardio, resistance training, nutrition, sleep, and stress management. Resistance training contributes to strength, lean mass, and bone health in ways that HIIT alone may not fully provide.
8. Bodyweight Exercise and Accessibility
Combining HIIT with bodyweight exercise is one of the most important routes to accessibility. Squats, lunges, push-ups, planks, mountain climbers, locomotor drills, and low-impact alternatives can be organized in interval form. This approach reduces dependence on machines and gym facilities.
One study found that simple bodyweight training requiring little time and no specialized equipment could improve cardiorespiratory fitness in inactive adults [8]. Online interventions and reviews of home-based HIIT also support the feasibility of this approach, although sample size, intensity control, and generalizability remain limitations [9, 10].
No equipment does not mean no instruction. Fatigue, jumping movements, and poor execution may increase joint loading. Public programs should include a warm-up, technical instruction, low-impact alternatives, stopping criteria, and a progression guide.
9. Why HIIT Is Data-Driven and Highly Monitorable
Every form of physical activity can be measured, but the interval structure of HIIT offers a special advantage: the body’s response is observed repeatedly within a single session across comparable cycles. Heart rate during each interval, rate of heart-rate rise, heart-rate recovery, speed, distance, power, repetitions, performance decrement, perceived exertion, and session completion can all be recorded.
These data do more than confirm that a person exercised; they reveal the difference between the prescribed session and the individual’s actual response. Two people may share the same target intensity, while one demonstrates pronounced performance decline and slow recovery by the third interval. That distinction is central to personalization.
Wearable-based research has shown that combinations of activity signals and individual characteristics can be used with machine-learning models to estimate cardiorespiratory fitness in free-living conditions [11]. These findings strengthen the scientific basis for scalable monitoring, but they do not replace laboratory assessment or clinical judgment in every context.
10. Convergence with Artificial Intelligence and Virtual Coaching
Artificial intelligence can transform a fixed program into a feedback loop: assessment, prescription, execution, measurement, analysis, adjustment, and reassessment. Within this loop, training history, sleep, heart rate, perceived exertion, movement quality, and session response can inform the next decision.
The advantage of HIIT for algorithms is its parametric structure. A system does not always need to replace the entire workout; reducing one interval, extending recovery, substituting a jump with a low-impact movement, or changing the target intensity may be sufficient. A 2026 study proposed a machine-learning framework for individualized exercise prescription based on body mass index and physical-fitness testing and evaluated it in a population of young men [12]. The result is promising, but its age, sex, and setting limitations preclude universal generalization.
This article uses the proposed term “AI-native exercise” to describe exercise whose core variables can be digitized, monitored, analyzed, and algorithmically adjusted. HIIT is a strong candidate for this concept, not because artificial intelligence created it, but because its structure is already divided into decision-ready units.
A virtual coach is not necessarily intelligent. A prerecorded video merely displays a session; an AI-enabled virtual coach should receive user information, analyze performance quality or physiological response, and modify feedback. A smartphone camera, motion sensors, a smartwatch, and subjective user reports may all contribute to such a system.
11. The Proposed Cardio HIIT-AI Model
The proposed model contains seven stages. It is a research and design framework rather than a clinically validated protocol.
1. Profiling and screening: Age, training history, goals, disease, medication, injury, movement limitations, baseline fitness, equipment, and environment are documented.
2. Format selection: An appropriate cardio, bodyweight, low-impact, dance, combat, cycling, running, or hybrid format is selected.
3. Initial prescription: Relative intensity, work duration, recovery, number of intervals, weekly frequency, and stopping criteria are defined.
4. Execution monitoring: Heart rate, time, speed, power, repetitions, movement quality, perceived exertion, and session completion are recorded.
5. Data-quality control: Signal quality, probable sensor error, missing data, and device compatibility are evaluated.
6. Intelligent analysis: Response, fatigue, progress, recovery, possible intolerance, and technical deviation are analyzed.
7. Adjustment, referral, and learning: Load is increased or reduced, movements are modified, recovery is added, or warning signs trigger referral to a coach or clinician.
12. An Intelligent Comparison with Peer Training Methods
To preserve readability across word-processing software, the comparison is presented as a numbered analysis rather than a wide table.
1. HIIT as the reference architecture: Its primary advantages are time efficiency, improvement of cardiorespiratory fitness, and the ability to adjust multiple variables. Its limitations include relatively high perceived exertion and the need for precise intensity prescription. Its AI-integration potential is high because work and recovery cycles can be compared. Scientifically, it is a flexible architecture, not the universal winner for every outcome.
2. Moderate-intensity continuous cardio: Greater tolerability and the ability to accumulate prolonged aerobic volume are major strengths. It may require more time for some goals, but heart rate, speed, and distance are readily monitored. It should not be framed as an obsolete competitor; for many individuals it is either preferable to or complementary with HIIT.
3. Resistance training: Resistance training is essential for strength, muscle mass, bone health, and neuromuscular performance. It may require weights, machines, or specialized instruction. Artificial intelligence can analyze load, repetitions, movement velocity, and fatigue. Resistance training cannot be fully replaced by cardio HIIT.
4. Bodyweight training: It is inexpensive, accessible, and feasible in varied settings. Resistance quantification and technique control can be more difficult, while computer vision has substantial potential for repetition counting and movement-pattern recognition. Bodyweight exercise is one of the most suitable delivery formats for cardio HIIT.
5. Functional training: Functional training targets transfer to daily activity and movement skill, but standardization is more difficult. Artificial intelligence may analyze range of motion, balance, and pattern quality. Functional exercises can be combined effectively with HIIT.
6. Combat and team sports: These sports include technical, tactical, and competitive dimensions and should not be reduced to physical conditioning. Technology can analyze load, movement, acceleration, and decision-making, but HIIT remains a conditioning tool rather than a substitute for sport-specific skill. A review of Olympic combat sports supports its complementary role [13].
7. Dance and group programs: Music, enjoyment, and social participation may strengthen motivation, while equalizing intensity across participants is difficult. Virtual coaching and movement feedback may help. Dance HIIT is a media-compatible example of combining interval structure with a group experience.
13. Interdisciplinary Role in High-Performance Sport
The pattern of intense activity and recovery resembles the structure of many sports. Combat athletes experience cycles of attack, movement, engagement, and relative recovery; team-sport athletes require acceleration, change of direction, and repeated high-intensity efforts. HIIT can therefore be used as a specific conditioning tool [13].
A general protocol is not equivalent to a sport-specific protocol. Football conditioning should reflect running and change-of-direction demands, while combat-sport conditioning should reflect the sport’s work, recovery, striking, and movement profile. Artificial intelligence can strengthen this personalization through data on playing position, tactical role, competition calendar, and training load.
14. Media, Digital Platforms, and Home Exercise
The popularity of HIIT cannot be explained by physiology alone; its structure also aligns with the logic of media. Timers, short intervals, frequent movement changes, and home-based feasibility make it suitable for sports television, video platforms, applications, and social networks. Labels such as cardio HIIT, bodyweight HIIT, dance HIIT, and full-body HIIT target distinct audiences.
A study comparing evidence-based protocols with popular social-media workouts found that media-based sessions can elicit genuine physiological responses, but protocol design and perceived exertion differed [14]. Follower counts and visual appeal are therefore not equivalent to scientific validity.
Media can reduce access costs and bring exercise into the home, but may also disseminate exaggerated promises, inappropriate prescriptions, and poor technique. A high-quality standard should include level classification, low-impact options, explanations of relative intensity, safety warnings, technical instruction, and the freedom to stop.
15. Sport for All, Access, and Lifestyle Integration
The objective of sport for all is not to produce a difficult session; it is to make structured physical activity part of citizens’ lifestyles. To serve this purpose, a program should be as affordable, accessible, adjustable, and sustainable as possible. Bodyweight cardio HIIT addresses part of this need because it reduces time and equipment requirements.
A short session does not guarantee long-term adherence. A review of 188 studies found high attendance in supervised interventions but lower and highly variable adherence in unsupervised sessions; it also found no stable, significant adherence advantage for HIIT over continuous training [15].
Enjoyment, perceived competence, social support, variety, autonomy, realistic goal setting, and access all matter for lifestyle adoption. Artificial intelligence should not optimize intensity alone; it should also consider dropout risk, user preferences, available time, and psychological experience.
16. Limitations, Safety, and Data Governance
Scientific future-oriented analysis is credible only when limitations and safety conditions are stated explicitly.
First, HIIT is not a single protocol. Four-minute intervals, all-out sprints, bodyweight circuits, and interval dance differ in intensity, skill requirements, and joint loading. Findings from one study cannot be generalized to every format.
Second, superiority depends on the outcome. Evidence is comparatively stronger for cardiorespiratory-fitness improvement [2], whereas findings for body composition, strength, muscle mass, enjoyment, and long-term adherence are less uniform or conclusive [7, 15, 16].
Third, the phrase “suitable for everyone” should be replaced with “designable for different populations.” An interval format may be designed for an older adult, a stable patient, or an inactive person, but exercise mode, intensity, volume, and supervision must differ. Individuals with chest pain, unusual dyspnea, syncope, uncontrolled disease, or major functional limitations require professional assessment before high-intensity exercise.
Fourth, the claim that HIIT is “safe and low-injury” is not defensible as an absolute statement. Cardiac-rehabilitation data suggest that major-event rates may be low in stable, screened, supervised patients [17, 21]. High-intensity functional programs, complex exercises, and technique breakdown under fatigue present a different risk profile; injury reviews have identified the shoulder, knee, and lower back among commonly reported sites, while also noting that the quality of many studies is limited [22]. Safety should therefore be an outcome of design, instruction, and supervision, not an intrinsic property of an exercise label.
Fifth, intelligent decision quality depends on data quality. Consumer wearables are useful for monitoring trends, but heart-rate accuracy varies with device type, placement, movement, exercise intensity, user characteristics, and signal quality. Arm movement and optical artifact can produce false or missed beats, while non-standardized hardware, software, and processing methods complicate comparison [18, 19]. An algorithm should therefore incorporate signal-quality indices, uncertainty ranges, missing-data rules, and a mechanism for human verification.
Sixth, heart-rate data, movement patterns, sleep information, and health history are sensitive. Their collection and use should be based on informed consent, data minimization, secure storage and transmission, purpose transparency, deletion rights, algorithmic audit, and clearly defined responsibility. World Health Organization guidance emphasizes that ethics, human rights, accountability, and public benefit should remain central to the design and deployment of artificial intelligence for health [20].
Seventh, an AI coach should not completely replace a coach, exercise physiologist, physiotherapist, or physician. Its preferred role is decision support and access enhancement. A safe system must recognize when an automated response is insufficient and when referral to a professional is required.
Finally, the strength of the language should match the strength of the evidence. Critical analyses of the HIIT literature emphasize that permanent superiority, complete safety, lower injury risk, greater enjoyment, or higher long-term adherence have not been demonstrated across all populations and protocols. Phrases such as “may,” “in selected populations,” “with a specified protocol,” and “under appropriate supervision” are scientifically more accurate than universal claims [15, 16].
17. Research and Organizational Agenda
To develop this concept into a framework suitable for scientific congresses, medical institutions, and international sport organizations, the following priorities are proposed.
1. Validation of the five-pillar framework: Translate the five pillars into operational indicators and test their validity, reliability, and predictive value in real-world programs.
2. Standardized protocol reporting: Report intensity, work duration, recovery, volume, exercise mode, participant level, and adverse events in sufficient detail.
3. Testing in diverse populations: Conduct multicenter studies involving women, older adults, adolescents, people with different diseases, and low-resource settings.
4. Human coach versus intelligent system comparisons: Identify the decisions for which artificial intelligence adds value and those for which human supervision remains essential.
5. External algorithm validation: Test models on devices, languages, cultures, and populations that differ from the training data.
6. Data safety and fairness: Evaluate sensor error, population bias, privacy, security, and legal responsibility.
7. Behavioral and social outcomes: Measure long-term adherence, enjoyment, access, cost, social participation, and effects on lifestyle.
8. Organizational application: Develop educational standards, coaching certifications, media guidelines, and ethical frameworks for federations and health organizations.
18. Conclusion
HIIT is one of the most important training models of the twenty-first century, not because it replaces every other method, but because it can integrate with them. It can be cardio-based, delivered through bodyweight exercise, enriched with strength and functional elements, combined with dance or combat exercise, and adjusted for the home, gym, sport for all, high-performance sport, or rehabilitation.
The five-pillar framework developed in this article locates the strategic value of cardio HIIT in five simultaneous qualities: intelligence and personalization; scientific grounding and measurement; inclusion and accessibility; health orientation and manageable safety; and media and technological scalability.
Its parametric structure converts heart rate, work time, recovery, speed, power, repetitions, and performance decline into analyzable data. This creates a foundation for convergence with wearable devices, computer vision, virtual coaching, and artificial intelligence. Yet genuine intelligence does not mean increasing intensity continuously. An intelligent system should determine when a session should become harder, when it should become easier, when the training mode should change, and when rest or professional referral is necessary.
The most precise conclusion is therefore this: because of its time efficiency, adaptability, capacity to generate data, low-cost delivery options, connection with cardiovascular health, and readiness for integration with intelligent technologies, cardio HIIT is one of the training architectures most prepared for the future of sport, medicine, and digital health. It is not the sole future of exercise, but it may become one of the most important shared languages between exercise physiology and artificial intelligence.
References
Bibliographic note: Reference titles and publication details have been preserved in English to maintain citation accuracy.
1. World Health Organization. WHO Guidelines on Physical Activity and Sedentary Behaviour. Geneva: WHO; 2020. ISBN: 978-92-4-001512-8.
2. Poon ETC, Li HY, Gibala MJ, Wong SHS, Ho RST. High-intensity interval training and cardiorespiratory fitness in adults: An umbrella review of systematic reviews and meta-analyses. Scandinavian Journal of Medicine & Science in Sports. 2024;34(5):e14652. doi:10.1111/sms.14652.
3. Atakan MM, Li Y, Koşar ŞN, Turnagöl HH, Yan X. Evidence-Based Effects of High-Intensity Interval Training on Exercise Capacity and Health: A Review with Historical Perspective. International Journal of Environmental Research and Public Health. 2021;18(13):7201. doi:10.3390/ijerph18137201.
4. Duncombe SL, et al. School-based high-intensity interval training programs in children and adolescents: A systematic review and meta-analysis. PLOS ONE. 2022;17(5):e0266427. doi:10.1371/journal.pone.0266427.
5. Yue T, Wang Y, Liu H, Kong Z, Qi F. Effects of High-Intensity Interval vs. Moderate-Intensity Continuous Training on Cardiac Rehabilitation in Patients With Cardiovascular Disease. Frontiers in Cardiovascular Medicine. 2022;9:845225. doi:10.3389/fcvm.2022.845225.
6. Hannan AL, Hing W, Simas V, et al. High-intensity interval training versus moderate-intensity continuous training within cardiac rehabilitation: a systematic review and meta-analysis. Open Access Journal of Sports Medicine. 2018;9:1-17. doi:10.2147/OAJSM.S150596.
7. Wewege M, van den Berg R, Ward RE, Keech A. The effects of high-intensity interval training vs. moderate-intensity continuous training on body composition in overweight and obese adults. Obesity Reviews. 2017;18(6):635-646. doi:10.1111/obr.12532.
8. Archila LR, Bostad W, Joyner MJ, Gibala MJ. Simple Bodyweight Training Improves Cardiorespiratory Fitness with Minimal Time Commitment. International Journal of Exercise Science. 2021;14(3):93-100. doi:10.70252/WEQD2681.
9. Zhang Y, Zhang B, Gan L, et al. Effects of Online Bodyweight High-Intensity Interval Training Intervention and Health Education on Mental Health and Cognition. International Journal of Environmental Research and Public Health. 2021;18(1):302. doi:10.3390/ijerph18010302.
10. Tsuji K, et al. Home-based high-intensity interval training improves cardiorespiratory fitness: a systematic review and meta-analysis. BMC Sports Science, Medicine and Rehabilitation. 2023. doi:10.1186/s13102-023-00777-2.
11. Spathis D, Perez-Pozuelo I, Gonzales TI, et al. Longitudinal cardio-respiratory fitness prediction through wearables in free-living environments. npj Digital Medicine. 2022;5:176. doi:10.1038/s41746-022-00719-1.
12. Mo M, Li B, Yang Y, et al. A machine learning framework for personalized exercise prescription based on BMI and physical fitness assessment. Scientific Reports. 2026;16:13336. doi:10.1038/s41598-026-42405-2.
13. Franchini E, Cormack S, Takito MY. Effects of High-Intensity Interval Training on Olympic Combat Sports Athletes’ Performance and Physiological Adaptation. Journal of Strength and Conditioning Research. 2019;33(1):242-252. doi:10.1519/JSC.0000000000002957.
14. Hesketh KL, et al. Evidence-based vs. social media based high-intensity interval training protocols: Physiological and perceptual responses. PLOS ONE. 2021;16(9):e0257685. doi:10.1371/journal.pone.0257685.
15. Santos A, et al. Rates of compliance and adherence to high-intensity interval training: a systematic review and meta-analyses. International Journal of Behavioral Nutrition and Physical Activity. 2023;20:134. doi:10.1186/s12966-023-01535-w.
16. Ekkekakis P, et al. Extraordinary claims in the literature on high-intensity interval training: Critical analyses from an interdisciplinary lens. Psychology of Sport and Exercise. 2023. Article S1469029223000237.
17. Wewege MA, Ahn D, Yu J, Liou K, Keech A. High-Intensity Interval Training for Patients With Cardiovascular Disease-Is It Safe? A Systematic Review. Journal of the American Heart Association. 2018;7(21):e009305. doi:10.1161/JAHA.118.009305.
18. Bent B, Goldstein BA, Kibbe WA, Dunn JP. Investigating sources of inaccuracy in wearable optical heart rate sensors. npj Digital Medicine. 2020;3:18. doi:10.1038/s41746-020-0226-6.
19. Nelson BW, Low CA, Jacobson N, Areán P, Torous J, Allen NB. Guidelines for wrist-worn consumer wearable assessment of heart rate in biobehavioral research. npj Digital Medicine. 2020;3:90. doi:10.1038/s41746-020-0297-4.
20. World Health Organization. Ethics and Governance of Artificial Intelligence for Health: WHO Guidance. Geneva: WHO; 2021. ISBN: 978-92-4-002920-0.
21. McGregor G, et al. High-intensity interval training in cardiac rehabilitation: a multi-centre randomized controlled trial. European Journal of Preventive Cardiology. 2023;30(9):745-755. doi:10.1093/eurjpc/zwad039.
22. Barranco-Ruiz Y, Villa-González E, Martínez-Amat A, Da Silva-Grigoletto ME. Prevalence of Injuries in Exercise Programs Based on CrossFit, Cross Training and High-Intensity Functional Training Methodologies: A Systematic Review. Journal of Human Kinetics. 2020;73:251-265. doi:10.2478/hukin-2020-0006.
23. American College of Sports Medicine. ACSM Worldwide Fitness Trends: 2026. Official professional resource. 2026.