ChatGPT road running training prescription: An expert validation analysis

Abstract

The use of online tools for training guidance, especially ChatGPT, is a common practice among road runners. However, there is no information on the quality of the information provided. This study aimed to analyse the guidance and prescriptions obtained by ChatGPT for road running, under validation by professional coaches. This is an exploratory descriptive study, carried out in two stages: 1) presentation of prompts for evaluation of ChatGPT regarding training guidelines in three different cases (beginner runner; intermediate runner; advanced runner); 2) evaluation, using a Likert scale (1:completely correct; 4:completely incorrect) of the responses generated by ChatGPT, by road running coaches. Most of the responses were considered “completely correct.” For the first simulated case, it was noted that some experts classified questions three (training volume) and four (training intensity) as “more incorrect than correct.” In comparison to the simulated cases two and three, most experts classified the generated responses as “more incorrect than correct.” The highest frequencies of the “more incorrect than correct” option were observed for questions two and five at the intermediate level, and questions one, two, and four at the advanced level. For the questions related to the third case (advanced level), it was observed that question five was considered “completely incorrect” by 6.3% of the experts. The coaches’ suggestions were directed to the race strategy, hydration and supplementation. ChatGPT provided realistic running training information based on user prompts, but the use of AI tools to prescribe specific road running activities has both potential and limitations.

Keywords

Artificial intelligence endurance hydration race strategy

Introduction

Running is one of the most practiced activities around the world.^1,2 Some studies have described that running can have protective effects on physical,^3,4 cognitive,⁵ and mental health.⁶ This activity provides an opportunity for people to become more active due to its low motor skill requirements, low equipment cost, and lack of necessity for specific sports facilities.⁷

However, for the sustainability of the practice, as well as the promotion of health and well-being, the prescription of more effective and targeted training becomes essential.⁸ Nonetheless, not all runners have access to professional trainers and medical care.⁹ In the current context, due to technological advances, various training plans are available online for free. These plans offer training models for different activity levels, distances, and intensities, along with general tips (e.g., www.runnersworld.com/uk/training-plans/). Additionally, more recently, new tools such as artificial intelligence, specifically ChatGPT, can assist in training prescription.

ChatGPT is a specific version of the GPT (Generative Pre-trained Transformer) language model. It is designed to interact with users conversationally, providing responses to a wide range of queries and questions (https://openai.com/). ChatGPT is trained on a vast amount of textual data to understand and generate text on various topics, enabling it to provide information, perform tasks, and engage in natural conversations (https://openai.com/).^10,11 This tool has already been used, validated, and discussed in various contexts related to public health,¹² exercise prescription for individuals with different conditions and fitness goals,¹³ strength resistance training prescription,¹⁴ and physical activity counseling.¹⁵ It has also been used to evaluate the development of running training plans.¹⁶

However, as ChatGPT is available for public consultation, there is a need to verify the validation, quality, and safety of the information generated on the platform in various contexts and “prompts” for running training, especially considering the evolution of training itself. In this context, it is not yet well established whether the information provided to runners and athletes is reliable, aligns with the latest scientific evidence, and adheres to sports training principles.

Despite these promising applications, the reliability, quality, and safety of ChatGPT-generated training advice for runners remain largely unverified. It is unclear whether such guidance aligns with current scientific evidence, adheres to training principles, or can safely support runners and athletes in achieving performance and health goals.¹⁷ We hypothesize that the training recommendations provided by ChatGPT for runners are not fully aligned with expert knowledge and current scientific evidence. Therefore, the aim of this study was to analyze the guidance and prescriptions provided by ChatGPT for running practice under the validation of professional trainers.

Methods

This study is characterized as descriptive and exploratory, conducted through two stages: 1) presenting prompts for the evaluation of ChatGPT; 2) evaluating the responses generated by ChatGPT by road running coaches. As this is a validation conducted by experts, the study was not submitted to an Ethics Committee anccording the Resolution National Health Council (CNS) No. 510/2016. However, the coaches/experts voluntarily consented to participate.

Stage 1 – Presentation of prompts to ChatGPT

The ChatGPT search strategy (ChatGPT Mar 23 Version) from OpenAI was used on April 27, 2023. Version 3.5 of ChatGPT, a freely accessible software via the site https://chat.openai.com/,¹⁸ was utilized. Prompts were submitted to ChatGPT simulating questions about running training plans for individuals at different practice levels, as follows:

Case 1 – beginner level (intending to run a 5 km race);

Case 2 – intermediate level (intending to run a race between 10 km and 21 km);

Case 3 – advanced level (intending to run a 42 km race).

The questions submitted to ChatGPT were developed by three of the authors of this manuscript, who are specialists in running training prescription with over 10 years of experience, owners of sports coaching businesses, and hold master's and doctoral degrees. The questions were presented according to the specifics of the aforementioned training levels, leading to the implementation of a 4-month (16 weeks) training cycle. The questions were formulated to identify ChatGPT's response regarding the planning of the cycle, volume, frequency, and intensity. Additionally, the formulation of the questions considered the hypothesis of practitioners entering prompts into ChatGPT. The prompts submitted to ChatGPT are described in Table 1.

Table 1.

Prompts submitted to ChatGPT according to the athlete's level.

Case 1	Level: beginner
I am a 35 year old man and I have never done physical activity. I'm 1.83 m tall and 80 kg. I want to run 5 km in 4 months. In light of this, I would like to know the following:
1- Do I need to take any exam to start practicing the sport?
2- How many days (frequency) per week do I need to train to reach my goal safely?
3- What is the volume of weekly training (in km) in the first 2 months of periodization?
4- What intensity of training should I do during 3 and 4 months of preparation?
5- What types of training will I do throughout the preparation (in the 4 months)?

Case 2	Level: intermediate
I'm a 35-year-old woman and I've been running for about 5 years and have done several 10 km races. I would like to do a half marathon safely in 4 months. So I would like to know the following:
1- Do I need to take any tests to make my training safer?
2- How many days (frequency) per week do I need to train to reach my goal safely?
3- What is the volume of weekly training (in km) in the first month of periodization?
4- What is the intensity of training that should be carried out during the specific period of preparation and what is the ideal number of weeks for that specific period?
5- What types of training will I do throughout the preparation (in the 4 months)? Please cite and mention how to carry out:

Case 3	Level: advanced
I'm a 35-year-old man, I've been running for about 10 years and I've done 5 marathons. In 4 months I want to do the 42 km under 3 h 30. Therefore, I would like to know the following:
1- Do I need to take any specific exam to make my training viable and make it easier to reach my goal?
2- How many days (frequency) per week do I need to train to reach my goal safely?
3- What is the volume of weekly training (in km) in the month with the highest periodization volume?
4- What intensity should I perform the interval training (sprints) during the preparation, and at what point in the cycle should I do them?
5- What is the best strategy for marathon day regarding race pace, and what precautions should I take during the run to avoid hitting the wall?

The questions were entered into the ChatGPT system by one of the researchers sequentially, and the generated responses are provided in the Supplementary Material of this manuscript.

Stage 2 – Evaluation of responses by coaches

After obtaining the responses generated by ChatGPT, they were organized and sent via email using an electronic form to running training specialists between June and September 2023. The coaches were recruited through the Running Coaches Association of Santa Catarina (ATC/SC). The form was sent to 32 coaches, owners of sports coaching businesses specialized in running, and 16 responses were received.

Coaches were asked to validate the information presented in each case (beginner, intermediate, and advanced) according to the understanding of correctness (content accuracy), clarity (being comprehensible and coherent), and conciseness (the degree to which all information is delivered).¹⁹ For this purpose, a Likert scale with 4 alternatives was provided for each response: “[1] completely correct,” “[2] more correct than incorrect,” “[3] more incorrect than correct,” and “[4] completely incorrect.” The 4-point Likert scale was chosen because it avoids a neutral midpoint and compels evaluators to classify the responses as leaning more toward correctness or incorrectness. This approach has been used in similar validation studies, as it enhances the sensitivity of expert judgment and reduces ambiguity in interpretation.²⁰ After evaluating comprehension, clarity, and conciseness, experts could provide descriptive comments they deemed relevant regarding the provided guidance. It is emphasized that the experts were not informed that the guidance was generated by ChatGPT.

Data analysis

The information was presented descriptively and through frequencies. To present the information generated by ChatGPT, a reduction of the data was carried out, emphasizing the response to the main question in an objective manner. To verify the evaluation by specialists, a frequency analysis was conducted, considering each question and its assessment independently.

Results

Regarding the necessity of medical exams, a positive response was given for all three simulated cases. In Case 2, there is also a recommendation to consult a physical education professional or physiotherapist if the practitioner has a history of injury. For Case 3, it is advised to seek an experienced coach for training planning. Regarding frequency and volume recommendations, an increase in both variables is observed according to the beginner, intermediate, and advanced levels. Specifically concerning training intensity, the recommendations vary based on the case presented, with responses tailored for beginners and advanced levels. For the intermediate level, guidelines are based on maximum heart rate. Specifically for Question 4, the recommendations include using different training methods (easy running, interval running, fartlek training, long runs, strength training, flexibility), as well as “hill running” and “tempo runs.” For the advanced level, it is recommended to use pace charts to calculate the ideal pace, maintain a steady pace during the race, control breathing, maintain posture, and avoid injuries (Table 2).

Table 2.

ChatGPT responses to the questions asked in each simulated case.

	Case 1 (Beginner level)	Case 2 (Intermediate level)	Case 3 (Advanced level)
Q1: Medical Examination	Yes	Yes	Yes
Q2: Training Frequency	3 to 4 days	3 to 5 days	4 to 6 times
Q3: Training Volume	10 to 15 km (10% progression)	15 to 20 km (first week) - 30 to 35 km (week 4)	Can vary from 60 to 100 km
Q4: Training Intensity	Include interval training, fartlek, long runs; Consult a Physical Education professional	Base period (intensity should vary from 60 to 70% of max HR); Strengthening period (intensity between 70% and 80% of max HR); Tapering period (intensity between 80% and 90% of max HR); tapering (intensity between 60 and 70% of max HR)	Interval training can be performed at different moments of the periodization. Basic phase (lighter and shorter, with more recovery); Specific phase (more intense, with longer repetitions and less recovery time)
Q5: Types of Training to be Performed	Level running, interval running, fartlek training, long runs, strength training, flexibility	Base runs, interval runs, hill runs, tempo runs, strength training	Use pace charts to calculate the ideal pace; During the race, it is important to maintain a constant pace and avoid starting too fast; control breathing, maintain posture, and avoid injuries

Table 3 presents the results of the experts’ validation of the responses generated by the prompts entered into ChatGPT. Most of the responses were considered “completely correct.” However, for the first simulated case, it was noted that some experts classified questions three (training volume) and four (training intensity) as “more incorrect than correct.” In comparison to the simulated cases two and three, most experts classified the generated responses as “more incorrect than correct.” The highest frequencies of the “more incorrect than correct” option were observed for questions two and five at the intermediate level, and questions one, two, and four at the advanced level. For the questions related to the third case (advanced level), it was observed that question five was considered “completely incorrect” by 6.3% of the experts. Question five refers to the pacing strategy to be used during the marathon, as well as the precautions that should be taken to avoid failing to complete the race within the expected time.

Table 3.

Frequency analysis of experts’ validation in the responses generated by the prompts inserted in ChatGPT.

Questions		Alternatives
Questions		Completely Correct	More Correct than Incorrect	More Incorrect than Correct	Completely Incorrect
Case 1 (Beginner Level)	Q1	68,8%	31,3%	0	0
	Q2	87,5%	12,5%	0	0
	Q3	43,8%	43,8%	12,5%	0
	Q4	56,3%	31,3%	12,5%	0
	Q5	62,5%	37,5%	0	0
Case 2 (Intermediate Level)	Q1	50,0%	37,5%	6,3%	6,3%
	Q2	43,8%	43,8%	12,5%	0
	Q3	37,5%	56,3%	6,3%	0
	Q4	43,8%	50,0%	6,3%	0
	Q5	43,8%	43,8%	12,5%	0
Case 3 (Advanced Level)	Q1	56,3%	31,3%	12,5%	0
	Q2	50,0%	37,5%	12,5%	0
	Q3	43,8%	50,0%	6,3%	0
	Q4	43,8%	43,8%	12,5%	0
	Q5	50,0%	43,8%	0	6,3%

Table 4 presents the experts’ suggestions regarding the responses to the prompts. The highest number of suggestions was observed for the beginner and intermediate levels. For the comments related to Case 1, most experts emphasized the need for guidance from a nutritionist and a physical education professional, as well as the necessity of conducting physical tests to monitor training load. Regarding training load (frequency, volume, and intensity), experts highlighted the need to work with time or hours, and the possibility of achieving the goal of running 5 km with a lower training frequency. The guidance for Case 2 is similar to what was observed in Case 1. For the response concerning training intensity, one participant noted that they would work with other variables, though they did not specify which ones. For Case 3, the highest number of suggestions was obtained for the question related to race strategy. The comments referred to the need to train pacing strategy during workouts, as well as hydration and supplementation.

Table 4.

Experts’ suggestions based on ChatGPT responses.

Questions	Case 1 – Beginner Level
Q1	- It would include the need to consult a nutritionist and a professional to guide strength training/adaptation in weightlifting. - The hydration information is ambiguous. Being hydrated is different from “hydrating yourself.” - It is unrealistic for a company to require a medical exam to start an activity. This scenario is unrealistic today. - When aiming to achieve goals, a medical exam is necessary to prevent problems during the process. - Guidance on performing a stress test (treadmill, track) is missing to identify the subject's initial conditioning level. From this parameter, training loads can be determined, and a more suitable plan can be made.
Q2	- It is possible to achieve the goal with fewer weekly training sessions, but consistency and a programmed progression are important.
Q3	- I would find it more interesting to work with minutes/hours of training. - Mileage is an imprecise measure. Proper training uses time and intensity as variables, not kilometers. - In the future, the training volume will depend on the target race volume.
Q4	- The recommendation to include intensity is too generic and may not be necessary in some cases. - Increasing volume (frequency and duration) and intensity is related to the training density, which includes the rest component. To make decisions about increasing duration, frequency, or intensity, some training control parameters (HR, perceived exertion scale, and metabolic response) should be used. - Intensity work is appropriately employed in traditional periodization after the runner has reached the intended volume, meaning they can already run 5 km satisfactorily.
Q5	- All these activities should be included from the start of the program. - Include mixed workouts (interval and continuous in the same session).

Questions	Case 2 – Intermediate level
Q1	- The issue of exams is always a priority for me, regardless of the athlete's level.
Q2	- As mentioned earlier, I prefer working with minutes/hours of weekly training and assessing progress in mileage. - I am against cookie-cutter training plans. What was proposed is correct, but not for most people.
Q3	- Work based on weekly hours of training.
Q4	- It depends on the type of periodization. An inverse periodization can be used. - HR% does not always apply as the sole basis for training intensity, and phase adjustments may differ. I would add other variables as well.
Q5	- What types of training will I carry out throughout the preparation (in the 4 months)? Please cite and mention how to carry out:

Questions	Case 3 – Advanced Level
Q1	- Always include exams in the routine.
Q2	- Periodization by the volume of activity time.
Q3	- Volume based on elapsed time.
Q4	- It is interesting to train intensity based on the percentage calculated in tests.
Q5	- Very specific, and the strategy will depend on what happened during training. - The strategy could involve a negative split, but it will depend on the athlete's level and how the long runs are being conducted (pace strategy). - Hydration and supplementation should also be individualized, previously planned, and tested. - I believe that in the race pace section, it lacked mentioning the importance of the coach in planning the strategy; for the advanced level, it is possible to choose some race tactics that can be applied on race day.

Discussion

The use of artificial intelligence tools for prescribing activities of various natures has been observed over the past few years.^14,21 Considering that consultation tools for specific road running training guidance are common practice among road runners, especially beginners, the aim of this study was to verify and examine the guidance and prescriptions provided by ChatGPT for road running training under the validation of professional coaches. Our hypothesis was that the training recommendations provided by ChatGPT for runners are not fully aligned with expert knowledge and current scientific evidence, which could generate inconsistencies or potential risks if applied without professional supervision.

However, the main findings, based on the coaches’ evaluation of ChatGPT's responses, indicated that: 1) across the three training levels (beginner, intermediate, and advanced), coaches observed a consistent recommendation from the tool to seek a qualified professional, along with a tendency to increase training intensity and volume and use more varied training methods; 2) according to the experts’ assessment, the indications related to training volume and intensity, as well as the prescriptions for intermediate and advanced runners, showed a higher frequency of “incorrect” or inconsistent responses; and 3) most adjustment suggestions were directed toward the beginner case, while evaluations for advanced runners emphasized aspects such as race strategy, hydration, and supplementation.

In relation to the evaluation criteria used by the coaches content accuracy, clarity, and conciseness the findings indicated distinct patterns across the three training levels. The lowest agreement was observed for content accuracy, particularly regarding training volume, intensity, and periodization, suggesting that ChatGPT's recommendations did not always align with scientific evidence or professional practice.^{14,16,20,22,23} Nonetheless, the responses were generally considered clear and coherent, as the language and structure used by the tool were easily understandable, even when inaccuracies were present.^16,24 Regarding conciseness, coaches reported that the information was adequately summarized and straightforward, although sometimes overly simplified, especially in the beginner and intermediate cases. These observations indicate that while ChatGPT communicates training guidance in an accessible and concise manner, the accuracy of the content remains its main limitation when compared to expert evaluation.^13,17,24

In all three cases analyzed, ChatGPT itself recommended consulting a qualified professional, emphasizing the importance of professional supervision in training prescription, as well as a logic of increasing weekly volume in line with competitive level improvements. Although guidance from qualified professionals is considered a protective strategy for reducing the risk of injury development and providing qualified training, in the Brazilian context, most road runners report not having a coach.²⁵ These differences may reflect in training volume, participation in competitions, performance in target races, and maintenance of activity due to injury development or demotivation.²⁶

Regarding training volume and intensity, previous studies highlight the association between previous race time, weekly volume, and use of interval training as predictors of performance at different distances.^22,27 Despite this association being consistent in the scientific literature, the indications for volume and strategies used to monitor training intensity showed a higher frequency of “incorrect” responses from professionals. Based on the suggestions provided, it appears that ChatGPT's guidance lacks specificity, as well as a better understanding of training periodization and better strategies for load control.

In this regard, the specificity of the race and target goal is an important aspect to consider, as significant differences in training load are observed between runners of different distances and performance levels.^23,28,29 For example, marathon and ultramarathon runners tend to train more frequently, for longer periods, and cover greater weekly distances compared to runners of shorter distances (half marathon and 10 km),²³ and also exhibit differences throughout different phases of periodization.²⁹

In addition to suggestions regarding training control parameters, aspects related to hydration and race strategy were presented. Regarding nutrition, it is noted that the consumption of sports drinks and dietary supplements is a common practice among road runners^30,31; however, individualization strategies for hydration and nutrition should be emphasized. Regarding race strategies, a previous study highlighted that a positive strategy (decreasing speed throughout the race) influenced performance in half-marathons and marathons for runners of different performance levels.³² These strategies should be individualized and organized according to availability, and therefore, the guidance of a qualified professional becomes a fundamental element.

Most of the suggestions directed at the beginner case can be explained by the scientific gap regarding guidance for this audience. Although there is interest in the literature regarding participation trends,^33,34 training,^35,36 and performance,²² most of the available evidence is centered on high-level athletes. The lack of studies focused on the initiation process of training compromises understanding and proper guidance, as well as contributes to injury incidence in this population.³⁷

This study has some limitations. Although the participating coaches were recruited from a coaches’ association, no information was provided about their experience duration or specialties related to training prescription (5 km, half marathon, or marathon). Another important aspect is that the quality of the generated response is related to the amount of information used to formulate the question, such that users may receive varying levels of response. In this study, we used standard questions but did not explore variations in responses based on question formulation, which could generate different patterns.¹⁶ Finally, the use of the “beginner,” “intermediate,” and “advanced” runner levels was based on practice time and target race, which may present differences in real practice contexts. Despite of the study's limitations, such as the variability in coaches’ experiences and the influence of the formulated questions on the obtained responses, highlight the need for more in-depth future research. Thus, while artificial intelligence can complement the training guidance process, the intervention of qualified professionals is essential to ensure safe and effective training tailored to each runner's individual needs.

Conclusion

ChatGPT provided realistic information about running training based on user prompts, but the use of artificial intelligence tools for prescribing specific road running activities presents both potential and limitations. While recommendations to consult a physical education professional are appropriate for preventing injuries, the guidance on training volumes and intensities, especially for intermediate and advanced runners, showed a lack of precision and alignment with specific goals. Suggestions for beginner runners were more abundant. In contrast, the lack of targeted recommendations for advanced runners and the variability in responses indicate that these tools still need to evolve to provide more detailed and personalized advice.

Supplemental Material

sj-doc-1-spo-10.1177_17479541251406721 - Supplemental material for ChatGPT road running training prescription: An expert validation analysis

Supplemental material, sj-doc-1-spo-10.1177_17479541251406721 for ChatGPT road running training prescription: An expert validation analysis by Daniel Rogério Petreça, Fabiano Braun de Moraes, Cyntia Maria de Holanda Martins, Mabliny Thuany and Marcos André Moura dos Santos in International Journal of Sports Science & Coaching

Footnotes

Abbreviations

ORCID iDs

Mabliny Thuany

Marcos André Moura dos Santos

Ethics approval and consent to participate

Regarding ethical procedures, invitations to specialists were sent by email, and upon acceptance, they were provided with instructions on what to analyze, how to fill out the evaluation forms, and an Informed Consent Form, ensuring data protection, confidentiality, and privacy. This study adheres to ethical principles in accordance with Article 1 of Resolution CNS 510/16, dated April 7, 2016, as described in the item IV of CONEP-SECNS-MS CIRCULAR OFFICE No. 17/2022. It is also highlighted that the data collected were limited to responses about running training and did not involve personal or sensitive information. Additionally, the interaction with specialists was conducted transparently, ensuring they were informed about the purpose of the validation and the use of their responses. The focus of the study was solely on the technical validation of the instrument, with no direct impact on the specialists, aligning with good ethical practices.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by D. R. P.; F. B. M. and M. T. The first draft of the manuscript was written by D. R. P. and all authors commented on previous versions of the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Availability of data and materials

The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.

Supplemental material

Supplemental material for this article is available online.

References

Hulteen

Smith

Morgan

, et al. Global participation in sport and leisure-time physical activities: a systematic review and meta-analysis. Prev Med 2017; 95: 14–25.

Clemence

. Global views on sports: 58% globally would like to practice more. 2024.

Wen

Wai

JPM

Tsai

, et al. Minimum amount of physical activity for reduced mortality and extended life expectancy: a prospective cohort study. Lancet 2011; 378: 1244–1253.

Tao

. Benefits of running on cardiac protection and the culture of exercise health awareness. Rev Bras Med Esporte 2022; 29: e2022_0168.

Zhang

S-S

Zhu

Peng

, et al. Long-term running exercise improves cognitive function and promotes microglial glucose metabolism and morphological plasticity in the hippocampus of APP/PS1 mice. J Neuroinflammation 2022; 19: 34.

Markotić

Pokrajčić

Babić

, et al. The positive effects of running on mental health. Psychiatr Danub 2020; 32: 233–235.

Magno

Ten Caten

Reppold Filho

, et al. Factors related to sports participation in Brazil: an analysis based on the 2015 national household survey. Int J Environ Res Public Health 2020; 17: 6011.

Lucini

Pagani

. Exercise prescription to foster health and well-being: a behavioral approach to transform barriers into opportunities. Int J Environ Res Public Health 2021; 18: 968.

Burkule

. Marathon running for amateurs: benefits and risks. J Clin Prev Cardiol 2016; 5: 113–124.

10.

Radford

Narasimhan

Salimans

, et al. Improving language understanding by generative pre-training. OALibJ 2020; 8: 7.

11.

Yenduri

Srivastava

Maddikunta

PKR

, et al. Generative pre-trained transformer: A comprehensive review on enabling technologies, potential applications, emerging challenges, and future directions. ArXiv preprint arXiv:230510435 2023.

12.

Biswas

. Role of chat gpt in public health. Ann Biomed Eng 2023; 51: 868–869.

13.

Dergaa

Saad

El Omri

, et al. Using artificial intelligence for exercise prescription in personalised health promotion: a critical evaluation of OpenAI’s GPT-4 model. Biol Sport 2023; 41: 221–241.

14.

Washif

Pagaduan

James

, et al. Artificial intelligence in sport: exploring the potential of using ChatGPT in resistance training prescription. Biol Sport 2024; 41: 209–220.

15.

Sandreschi

Claumann

Guimarães

JAC

, et al. A prática de atividade física e o uso do ChatGPT como uma ferramenta de aconselhamento. Rev Pesqui Qual 2024; 12: 01–20.

16.

Düking

Sperlich

Voigt

, et al. ChatGPT generated training plans for runners are not rated optimal by coaching experts, but increase in quality with additional input information. J Sports Sci Med 2024; 23: 56–72.

17.

Düking

Sperlich

Voigt

, et al. ChatGPT generated training plans for runners are not rated optimal by coaching experts, but increase in quality with additional input information. J Sports Sci Med 2024; 23: 56.

18.

Open Ai C. ChatGPT. Open Ai. https://platform.openai.com/playground, 2023.

19.

Sallam

Salim

Ala’a

, et al. ChatGPT output regarding compulsory vaccination and COVID-19 vaccine conspiracy: a descriptive study at the outset of a paradigm shift in online search for information. Cureus 2023; 15: e35029.

20.

Joshi

Kale

Chandel

, et al. Likert Scale: explored and explained. Br J Appl Sci Technol 2015; 7: 396.

21.

Dergaa

Saad

El Omri

, et al. Using artificial intelligence for exercise prescription in personalised health promotion: a critical evaluation of OpenAI’s GPT-4 model. BiolSport 2024; 41: 221–241.

22.

Vickers

Vertosick

. An empirical study of race times in recreational endurance runners. BMC Sports Sci Med Rehabil 2016; 8: 1–9.

23.

Knechtle

Tanous

Thuany

, et al. Differences in training characteristics of recreational endurance runners by race distance–results from the NURMI study (step 2). Front Psychol 2024; 14: 1269374.

24.

Zhang

Zhou

Qiao

, et al. Ethical issues in AI-generated texts: a systematic review and analysis. Int J Hum Comput Interact 2025: 1–28.

25.

Thuany

Bandeira

PFR

Vieira

, et al. The subject-environment interplay between runners from different Brazilian macro-regions. Front Psychol 2023; 14: 1134797.

26.

Stjepanovic

Knechtle

Weiss

, et al. Changes in pacing variation with increasing race duration in ultra-triathlon races. Sci Rep 2023; 13: 3692.

27.

Knechtle

Tanous

Thuany

, et al. Differences in race history by distance of recreational endurance runners from the NURMI study (step 2). Sci Rep 2023; 13: 18083.

28.

Fokkema

van Damme

Fornerod

, et al. Training for a (half-) marathon: training volume and longest endurance run related to performance and running injuries. Scand J Med Sci Sports 2020; 30: 1692–1704.

29.

Haugen

Sandbakk

Seiler

, et al. The training characteristics of world-class distance runners: an integration of scientific literature and results-proven practice. Sports Med Open 2022; 8: 46.

30.

O'Neal

Wingo

Richardson

, et al. Half-marathon and full-marathon runners’ hydration practices and perceptions. J Athletic Train 2011; 46: 581–591.

31.

Katharina

Mohamad

Derrick

, et al. Supplement intake in half-marathon,(ultra-) marathon and 10-km runners–results from the NURMI study (step 2). J Int Soc Sports Nutr 2021; 18: 64.

32.

Ristanović

Cuk

Villiger

, et al. The pacing differences in performance levels of marathon and half-marathon runners. Front Psychol 2023; 14: 1273451.

33.

Vitti

Nikolaidis

Villiger

, et al. The “New York city marathon”: participation and performance trends of 1.2M runners during half-century. Res Sports Med 2020; 28: 121–137.

34.

Lathan

. A history of jogging and running—the boom of the 1970s. Proc Bayl Univ Med Cent 2023; 36: 775–777.

35.

Sandbakk

Pyne

McGawley

, et al. The evolution of world-class endurance training: the scientist’s view on current and future trends. Int J Sports Physiol Perform 2023; 1: 1–5.

36.

Knopp

Appelhans

Schönfelder

, et al. Quantitative analysis of 92 12-week sub-elite marathon training plans. Sports Med Open 2024; 10: 50.

37.

Baltich

Emery

Whittaker

, et al. Running injuries in novice runners enrolled in different training interventions: a pilot randomized controlled trial. Scand J Med Sci Sports 2017; 27: 1372–1383.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.18 MB