São Paulo State University - UNESP Campus of Presidente Prudente Post-Graduate Program in Movement Science PEDRO HENRIQUE NARCISO PICCHI IMPACT OF HIIT ON BONE HEALTH IN ADOLESCENT GIRLS: CLINICAL TRIAL ON THE INFLUENCE OF MECHANICAL IMPACT Presidente Prudente - SP 2025 São Paulo State University - UNESP Campus of Presidente Prudente Post-Graduate Program in Movement Science PEDRO HENRIQUE NARCISO PICCHI IMPACT OF HIIT ON BONE HEALTH IN ADOLESCENT GIRLS: CLINICAL TRIAL ON THE INFLUENCE OF MECHANICAL IMPACT Master thesis presented to Postgraduate program in Movement Sciences of School of Technology and Sciences of São Paulo State University (UNESP) - Presidente Prudente as part of the requirements for obtaining the title of Doctor in Movement Sciences. Supervisor: Rômulo Araújo Fernandes, PhD Co-supervisor: Ricardo Ribeiro Agostinete, PhD Presidente Prudente - SP 2025 N222i Narciso, Pedro Henrique Impact of HIIT on bone health in adolescent girls: clinical trial on the influence of mechanical impact / Pedro Henrique Narciso. -- Presidente Prudente, 2025 106 p. Dissertação (mestrado) - Universidade Estadual Paulista (UNESP), Faculdade de Ciências e Tecnologia, Presidente Prudente Orientador: Romulo Araujo Fernandes Coorientador: Ricardo Ribeiro Agostinete 1. Osso. 2. Adolescencia. 3. Exercício Físico. 4. Citocinas. I. Título. Sistema de geração automática de fichas catalográficas da Unesp. Dados fornecidos pelo autor(a). SAO PAULO STATE UNIVERSITY Câmpus de Presidente Prudente ACCEPTANCE CERTIFICATE Thesis title: IMPACT OF HIIT ON BONE HEALTH IN ADOLESCENT GIRLS: CLINICAL TRIAL ON THE INFLUENCE OF MECHANICAL IMPACT Candidate: PEDRO HENRIQUE NARCISO PICCHI Supervisors: ROMULO ARAUJO FERNANDES, PhD / RICARDO RIBEIRO AGOSTINETE, PhD Approved as part of the requirements for obtaining the title of Master in Movement Sciences by the Examining Committee: FABIO SANTOS DE LIRA, PhD (President) Sao Paulo State University – UNESP, Brazil. PANAGIOTA KLENTROU, PhD (online participation) Brock University, Canada. DIMITRIS VLACHOPOULOS, PhD (online participation) University of Exeter, England. Presidente Prudente, 08th May, 2025 Digitally signed by Dimitris Vlachopoulos Reason: I am the author of this document Date: 09 May 2025 07:57:35 DN: CN=Dimitris Vlachopoulos, O=University of Exeter, E=D.Vlachopoulos@exeter.ac.uk, C=GR Panagiota Klentrou Digitally signed by Panagiota Klentrou Date: 2025.05.09 08:33:47 -04'00' São Paulo State University - UNESP Campus of Presidente Prudente AKNOWLEDGEMENT This dissertation was only possible thanks to the privileges guaranteed by my parents, because, through their tireless efforts, they never stopped investing in my education and training, professional and personal, in addition to providing all the unconditional support so that I could dedicate myself to the pursuit of my goals. To my brothers and Liliam, for being references for me and for always being present in the moments when I needed them. I thank my advisors, official and unofficial. First, to Prof. Rômulo Fernandes, who from the first moment trusted my work and gave me the opportunities to grow as a student, in addition to having "bought" my academic madness throughout this process. His advice and guidance were fundamental for my brief training so far. To Prof. Ricardo Agostinete, with whom I recently had the privilege of officially working as co-advisor, a role he has performed impeccably over the years, always demanding from me in the discussions, as well as giving me confidence, opportunities and support for me to get here. In addition, I still gained the privilege of his friendship. To Prof. André Werneck, who saw potential in me that even I didn't know existed (and to this day I doubt they are there), and who, from my first moment in this academic journey, dedicated himself to teaching me and helping me whenever necessary, being a fundamental part in every small academic achievement I obtained. In addition to being a professional and personal reference, he is someone I am proud to have as a friend. The three of you are and always will be personal and professional references for me. I thank my girlfriend, Kely, for the partner she has been during this period of the master's degree in which we have been together. Her emotional support, her help with daily demands, her understanding in difficult phases and her affection were essential for me to continue pursuing my dreams. It is a privilege to have her on this journey. To my friends from "Casinha", Alemão, Yuri, Popó and Santiago, thank you for having welcomed me into your group and for the sincere friendship we have cultivated in recent years. In addition to the staff, near or far, they were always ready to help me academically. You are rare people, who I will carry forever on the "left side of my chest", as Milton Nascimento would say. I hope to continue sharing the best moments with you. São Paulo State University - UNESP Campus of Presidente Prudente I thank Fábio, for having made himself available so many times to discuss ideas, give guidance and for having provided his laboratory to carry out the project, in addition, of course, to the friendship built during this period. I thank Ana Morano, who in the most difficult moment of my master's degree was there with me making things happen. His help was essential for this project to get off the ground. I thank Bárbara, for making herself available to conduct the biochemical analyses of the project. To my psychologist, Ethyenne, I thank not only for welcoming me during a critical moment in my mental health, but for making herself available to help me with the recruitment of volunteers for the project. Likewise, I thank Carol Muniz, for practically guaranteeing the sample of my study. I thank the nurses who helped on the collection weekends, with excellent professionalism and dedication. To CAPES and FAPESP, for having financed me since the beginning of my graduation, as well as my projects and my trip abroad. And, above all, I thank each of the volunteers who accepted to be part of this project, giving time, effort and, especially, blood. I am grateful to the friends I made during my research internship period at Brock University. First, to my supervisor, Nota Klentrou, for having received me so well, making her time available to guide me when I needed it and trusting my work since then. To Bareket, who always made a point of being present and caring about me during the period I was in St. Catharines. To Anthony, McKenna, Andrew, José Juan and Madi, for being family to me, and for the professional and personal moments shared, which undoubtedly made this period much lighter. To the other laboratory friends, Ilektra and Sewar, friends of EEL, John, Mohammed, Matthew and Ramnick, thank you for the partnership and teachings during the period I was at Brock. To the little pieces of Brazil that I found in St. Catharines, Júlia, Gus and Carol, thank you very much for your friendship in icy lands. Last but not least, I thank the employees of Unesp, the professors I have had throughout my life, and my friends who, despite the circumstances of life having taken us away from daily life, I carry with me the gratitude for having them in my life. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 001 - and in part by the São Paulo Research Foundation (FAPESP) (grant 2022/08875-5 and 2023/10153-1). São Paulo State University - UNESP Campus of Presidente Prudente “Life is an eternal dance and like a dance, the more rigid I became, the harder it got. […] It is this eternal dance that separates human beings from angels, from demons, from gods. And I must not forget, we must not forget that we are human beings" (Ren Gill) São Paulo State University - UNESP Campus of Presidente Prudente RESUMO O objetivo desta tese foi analisar o efeito do exercício intervalado de alta intensidade, com e sem impacto gravitacional, nos marcadores de remodelação óssea e osteocinas de adolescentes do sexo feminino não treinadas. Além disso, analisamos o efeito de cada modalidade (com e sem impacto) em citocinas e adipocinas potencialmente associadas aos marcadores de remodelação óssea e osteocinas. Por fim, avaliamos como cada modalidade de exercício influencia a percepção afetiva das adolescentes e qual delas foi mais agradável. Este foi um estudo de design cruzado, no qual 11 adolescentes do sexo feminino participaram de dois testes de exercício intervalado de alta intensidade (corrida intervalada de alta intensidade [HIIR] e ciclismo intervalado de alta intensidade [HIIC]) em ordem aleatória. A composição corporal, ingestão nutricional e histórico de atividades físicas relacionadas aos ossos foram avaliados utilizando métodos específicos. As participantes realizaram um teste incremental progressivo para determinar a carga máxima de trabalho na esteira e no cicloergômetro. Em dias diferentes, pela manhã, elas completaram os testes, que envolveram oito séries de um minuto de esforço máximo, seguidas por um minuto de recuperação passiva entre as séries. Amostras de sangue foram coletadas antes do exercício (linha de base), 5 minutos e 60 minutos após o exercício. O sangue sérico foi utilizado para analisar marcadores ósseos de formação (como P1NP e OPG) e reabsorção (como CTX-I), osteocinas (como esclerostina, OPG e RANKL), citocinas (como IL-6, IL-10 e TNF-α) e cortisol. Adipocinas (como leptina e adiponectina) e outros marcadores metabólicos (como glicose e LHD) foram medidos no soro. A percepção afetiva foi registrada pela Escala de Sentimentos (FS) antes e depois de cada teste, e o nível de prazer foi registrado pela Escala de Aproveitamento da Atividade Física (PACES) após cada teste. A análise estatística incluiu modelos mistos lineares (LMM) para avaliar o efeito fixo do tempo (pré, 5 minutos e 60 minutos após o exercício), modalidade (HIIR e HIIC) e interação entre os fatores (tempo*modalidade), considerando o efeito aleatório de cada indivíduo. Covariáveis (densidade mineral óssea, massa gorda, atividade física prévia relacionada aos ossos e valores dos marcadores sanguíneos na linha de base) foram testadas individualmente para cada modelo, considerando a correlação com o resultado. A diferença na pontuação total do PACES e nos itens individuais entre as modalidades foi testada usando um teste t independente. O Capítulo 4 mostra que a HIIR induziu uma expressão significativamente maior de P1NP em comparação à HIIC, embora CTX-I e outras osteocinas tenham sido diferentes entre os grupos. Além disso, a HIIR foi mais agradável do que a HIIC, e a percepção afetiva foi negativamente impactada pela última. O Capítulo 5 mostra que tanto a HIIR quanto a HIIC tiveram respostas semelhantes em marcadores anti-inflamatórios e adipocinas; no entanto, a HIIR induziu maior secreção de TNF-α, possivelmente devido ao impacto mecânico e ao possível maior dano tecidual. Assim, a HIIR e a HIIC induziram respostas semelhantes em adolescentes do sexo feminino não treinadas, apesar de pequenas diferenças em P1NP e TNF-α. Mais importante ainda, a HIIR foi percebida como mais agradável, com menor impacto na percepção afetiva das adolescentes. Portanto, a HIIR parece ser a melhor estratégia para melhorar os resultados de saúde relacionados aos ossos e à inflamação em adolescentes do sexo feminino. Palavras-chave: adolescência, exercício aeróbico, osteocinas, citocinas São Paulo State University - UNESP Campus of Presidente Prudente ABSTRACT The aim of this thesis was to analyze the effect of high-intensity interval exercise with and without gravitational impact on bone turnover markers and osteokines of untrained adolescent females. Additionally, we analyzed the effect of each mode (with and without impact) on cytokines and adipokines that are potentially associated with bone turnover markers and osteokines. Lastly, we assessed how each mode of exercise impacts the affective perception of the adolescents and which mode was more enjoyable. This was a cross-over design study, where 11 adolescent females participated in two HIIE trials (high-intensity interval running [HIIR] and high-intensity interval cycling [HIIC]) in random order. Body composition, nutritional intake, and bone-specific physical activity background were assessed using specific methods. They performed a progressive incremental test to determine the peak workload for the treadmill and cycle ergometer. On different days, in the morning, they completed the trials, involving eight bouts of one minute of maximal effort, followed by one minute of passive recovery between bouts. Blood samples were collected before exercise (baseline), 5-min, and 60-min post-exercise. Serum blood was used to analyze bone markers of formation (i.e., P1NP and OPG) and resorption (i.e., CTX-I), osteokines (i.e., sclerostin, OPG, and RANKL), cytokines (i.e., IL-6, IL-10, and TNF-α), and cortisol. Adipokines (i.e., leptin and adiponectin) and other metabolic markers (i.e., glucose and LHD) were measured in serum. Affective perception was reported by the Feeling Scale (FS) before and after each trial, and enjoyment was reported by the Physical Activity Enjoyment Scale (PACES) after each trial. Statistical analysis included linear mixed models (LMM) to assess the fixed effect of time (pre, 5- min, and 60-min post-exercise), mode (HIIR and HIIC), and interaction between factors (time*mode), while accounting for the random effect of each individual. Covariates (areal bone mineral density, fat mass, previous bone-specific physical activity, and blood markers values at baseline) were tested individually for each model considering the correlation with the outcome. Difference in PACES total score and their individual items between modes was tested using an independent t-test. Chapter 4 shows that HIIR induced significant greater expression of P1NP compared to HIIC, although CTX-I and other osteokines were different between groups. In addition, HIIR was more enjoyable than HIIC, and the affective perception was negatively affected by the latter. Chapter 5 shows that both HIIR and HIIC similar response in anti-inflammatory markers and adipokines; however, HIIR induced greater secretion of TNF-α, potentially because of the mechanical impact and eventual greater tissue damage. Thus, HIIR and HIIC induced similar responses in untrained female adolescents, despite small differences in P1NP and TNF-α. More importantly, HIIR was perceived as more enjoyable, with a lower impact on affect perception of the adolescents. Thus, HIIR seems to be the best strategy to improve health outcomes related to bone and inflammation in adolescent females. Keywords: adolescence, aerobic exercise, osteokines, cytokines São Paulo State University - UNESP Campus of Presidente Prudente LIST OF FIGURES Figure 1.1 - Bone mass across the lifespan with optimal and suboptimal lifestyle choices. ......... 13 Figure 1.2 - Bone (re)modeling thresholds. ................................................................................... 15 Figure 1.3 - Theoretical framework of bone functional adaptation at the diaphysis. .................... 16 Figure 1.4 - Interaction between cytokines and bone cells. .......................................................... 20 Figure 3.1 - Flowchar of the study ................................................................................................ 25 Figure 3.2 - Study protocol and timeline. ...................................................................................... 26 Figure 4.1 - Flowchart of the study. ............................................................................................... 34 Figure 4.2 - Study timeline and procedures. .................................................................................. 35 Figure 4.3 - RPE and HR (mean±SEM) change during HIIR and HIIC (n=11). .......................... 40 Figure 4.4 - Bone turnover makers and osteokines response (mean±SEM) to HIIR and HIIC. ... 41 Figure 5.1 - Flowchart of the study. ............................................................................................... 60 Figure 5.2 - Study protocol and timeline. ...................................................................................... 61 Figure 5.3 - Cytokines response to HIIR and HIIC (n=11). .......................................................... 65 Figure 5.4 - Adipokines ratio response to HIIR and HIIC. ........................................................... 66 São Paulo State University - UNESP Campus of Presidente Prudente LIST OF TABLES Table 4.1 - Samples' characteristics (n=11). .................................................................................. 39 Table 4.2 - Macro and micronutrients intake over the week preceding HIIE trials (n=11). .......... 39 Table 4.3 - Effect of HIIR and HIIC on affective perception and enjoyment (n=11). .................. 42 Table 4.4 – Cortisol response (mean±95%CI) to HIIR and HIIC (n=6). ...................................... 42 Table 5.1 - Participant's characteristics (n=11). ............................................................................. 64 São Paulo State University - UNESP Campus of Presidente Prudente GLOSSARY OF TERMS APHV – Age at Peak Height Velocity BALP – Bone Alkaline Phosphatase BMC – Bone Mineral Content BMD – Bone Mineral Density BMI – Body Mass Index BPAQ – Bone-specific Physical Activity Questionnaire BTM – Bone Turnover Markers CI – Confidence Interval CTX-I – C-terminal Telopeptide of Type I Collagen DXA – Dual-energy X-ray Absorptiometry EDTA – Ethylenediaminetetraacetic Acid ELISA – Enzyme-Linked Immunosorbent Assay ES – Effect Size FDR – False Discovery Rate FM – Fat Mass FS – Feeling Scale FX – Fracture Strain HIIC – High-Intensity Interval Cycling HIIE – High-Intensity Interval Exercise HIIR – High-Intensity Interval Running HR – Heart Rate São Paulo State University - UNESP Campus of Presidente Prudente IL-6 – Interleukin 6 IL-10 – Interleukin 10 IOF – International Osteoporosis Foundation LDH – Lactate Dehydrogenase LMM – Linear Mixed Models LST – Lean Soft Tissue MESm – Minimum Effective Strain for Modeling MESp – Minimum Effective Strain for Pathological Loading MESr – Minimum Effective Strain for Remodeling OPG – Osteoprotegerin P1NP – Procollagen Type I N-Terminal Propeptide PA – Physical Activity PACES – Physical Activity Enjoyment Scale PBM – Peak Bone Mass PHV – Peak Height Velocity RPE – Rate of Perceived Exertion RANKL – Receptor Activator of Nuclear Factor κB Ligand SD – Standard Deviation SEM – Standard Error of Mean TBLH-aBMD – Total Body Less Head areal Bone Mineral Density TNF-α – Tumor Necrosis Factor alpha São Paulo State University - UNESP Campus of Presidente Prudente TABLE OF CONTENTS 1. THEORETICAL FOUNDATION ...................................................................................... 12 2. THESIS OBJECTIVES ...................................................................................................... 23 2.1. GENERAL OBJECTIVE ................................................................................ 23 2.2. SPECIFIC OBJECTIVES ............................................................................... 23 3. METHODS ........................................................................................................................... 24 3.1. PARTICIPANTS ........................................................................................ 24 3.2. ETHICAL ASPECTS .................................................................................... 25 3.3. STUDY DESIGN ....................................................................................... 25 3.4. INTERVENTION PROTOCOL ......................................................................... 26 3.5. BASELINE MEASUREMENTS ......................................................................... 27 3.5.1. Body composition .......................................................................... 27 3.5.2. Questionnaires .............................................................................. 28 3.6. OUTCOME VARIABLES ............................................................................... 28 3.6.1. Blood analysis ............................................................................... 28 3.6.2. Bone turnover markers ................................................................... 28 3.6.3. Cytokines and adipokines .............................................................. 29 3.6.4. Metabolic markers ......................................................................... 29 3.6.5. Affective perception and enjoyment ................................................ 29 3.7. STATISTICAL ANALYSIS .............................................................................. 30 4. ACUTE EFFECT OF HIGH-INTENSITY INTERVAL CYCLING AND RUNNING ON BONE TURNOVER MARKERS AND OSTEOKINES IN FEMALE ADOLESCENTS 31 4.1. ABSTRACT ............................................................................................. 31 4.2. INTRODUCTION ....................................................................................... 32 4.2.1. Participants ................................................................................... 34 4.2.2. Ethical aspects .............................................................................. 35 4.2.3. Study Design ................................................................................. 35 São Paulo State University - UNESP Campus of Presidente Prudente 4.2.4. Intervention protocol ..................................................................... 36 4.2.5. Baseline measurements ................................................................ 36 4.2.5.1. Body composition ..................................................................... 36 4.2.5.2. Questionnaires ......................................................................... 37 4.2.6. Outcome variables ........................................................................ 37 4.2.6.1. Blood analysis .......................................................................... 37 4.2.6.2. Affective perception and enjoyment ........................................... 38 4.2.7. Statistical Analysis ......................................................................... 38 4.3. RESULTS ............................................................................................... 39 4.3.1. Bone Makers ................................................................................. 40 4.3.2. Affect and enjoyment ..................................................................... 41 4.4. DISCUSSION .......................................................................................... 43 4.5. REFERENCES ......................................................................................... 49 5. CYTOKINE AND ADIPOKINE RESPONSE FOLLOWING HIGH-INTENSITY INTERVAL RUNNING AND CYCLING IN FEMALE ADOLESCENTS. .......................... 56 5.1. ABSTRACT ............................................................................................. 56 5.2. INTRODUCTION ....................................................................................... 57 5.3. METHODS ............................................................................................. 59 5.3.1. Participants ................................................................................... 59 5.3.2. Study design .................................................................................. 60 5.3.3. Incremental test and exercise protocol ........................................... 61 5.3.4. Baseline measurements ................................................................ 62 5.3.5. Blood analysis ............................................................................... 62 5.3.6. Statistical analysis ......................................................................... 63 5.4. RESULTS ............................................................................................... 64 5.5. DISCUSSION .......................................................................................... 67 5.6. REFERENCES ......................................................................................... 71 6. THESIS CONCLUSION ..................................................................................................... 76 São Paulo State University - UNESP Campus of Presidente Prudente 7. THESIS REFERENCES ..................................................................................................... 77 8. DESCRIPTION OF ACADEMIC ACTIVITIES (COURSES, PUBLICATIONS, CONFERENCES) DURING MASTER’S COURSE ............................................................... 84 8.1. COURSES (N=8) ..................................................................................... 84 8.2. COMPLEMENTARY ACTIVITIES (N=5) .............................................................. 84 8.3. PARTICIPATION IN SCIENTIFIC CONFERENCES/PUBLICATIONS ON BOOKS OF ABSTRACTS (N=6) 85 8.4. PUBLISHED AND SUBMITTED PAPERS (N=9) ..................................................... 86 8.4.1. First author ................................................................................... 86 8.4.2. Coauthor ...................................................................................... 87 9. APPENDICES ...................................................................................................................... 88 9.1. ETHICAL APPROVAL ................................................................................. 88 9.2. CONSENT AND ASSENT TERMS ..................................................................... 91 9.3. QUESTIONNAIRES .................................................................................... 94 9.4. FAPESP SCHOLARSHIPS (MSC AND BEPE) ................................................ 102 9.5. ELAP SCHOLARSHIP ............................................................................. 105 9.6. PERMISSIONS TO REPRODUCES FIGURES ...................................................... 106 São Paulo State University - UNESP Campus of Presidente Prudente 12 1. THEORETICAL FOUNDATION Osteoporosis is a systemic disease and is characterized by changes in bone micro and macroarchitectur, decreased mass and stiffness and, consequently, increased bone fragility (GÓMEZ et al., 2021). This can result in fractures that are frequently seen in the forearm, hip, and spine, which have significant implications for public health (more hospitalizations, reduced quality of life, loss of functional capacity, and potentially lower life expectancy) (CLYNES et al., 2020; DALY et al., 2019; KANIS et al., 2019). Considering its significant impact on the quality of life during old age, added to the trend of aging of the world population, there is a growing concern on the impact of osteoporosis in society (SHEN et al., 2022) The incidence of osteoporosis does not only increase with age, but it impacts more significantly women than men (CLYNES et al., 2020; GUERRA et al., 2016; KANIS et al., 2019; SHEN et al., 2022; ZERBINI et al., 2015). This is because changes in reproductive hormones during menopause, such as decreased estrogen production, negatively affect bone metabolism, contributing to increased bone fragility among women (CAULEY, 2015; ZALLONE, 2006). Data from 1990 to 2019 shows that women consistently experienced higher rates of disease burden (101.4% vs 86.2%) and mortality (116.9% vs 105.3%) due to low bone mineral density (BMD) compared to men (SHEN et al., 2022). In Brazil, the number of fractures among women is expected to increase by 63% between 2015 and 2030, with an estimated 24% mortality rate within one year following hip fractures (AZIZIYEH et al., 2019; GUERRA et al., 2016), highlighting the gender disparities related to bone health. Additionally, the risk of osteoporosis is also influenced by peak bone mass (PBM), which is the greatest amount of bone accrued until early adulthood (WEAVER et al., 2016; ZERBINI et al., 2015). Although genetics play a major role in PBM, 38–54% is influenced by external factors, primarily mechanical loading through body weight and physical activity (BONJOUR et al., 2009; KRALL; DAWSON‐HUGHES, 1993). Furthermore, 95% of adult bone mass is achieved after 4 years following the peak in bone accretion (BAXTER-JONES et al., 2011). During this period of mass gain, bone is more susceptible to external influences; thus, it is a critical moment for optimizing peak bone mass (WEAVER et al., 2016). For girls, this is particularly significant because these gains during growth represent twice the amount of bone mass that will be lost São Paulo State University - UNESP Campus of Presidente Prudente 13 subsequently during the postmenopausal years (BAXTER-JONES et al., 2011). Therefore, understanding the best strategies to enhance peak of bone mass of adolescent girls is essential to reduce their risk of developing osteoporosis later in life (FIGURE 1.1). Figure 1.1 - Bone mass across the lifespan with optimal and suboptimal lifestyle choices. Note: Reproduced from (WEAVER et al., 2016) under the Creative Commons CC-BY License (Attribution- Noncommercial). Physical activity (PA), i.e., bodily movement produced by skeletal muscles that results in energy expenditure (CASPERSEN; POWELL; CHRISTENSON, 1985), is one of the most important determinants of bone health across life (KOHRT et al., 2004). The mechanical forces involved in PA stimulate bone adaptations through direct impact loading (e.g., ground reaction forces from activities such as jumping and running) and indirect loading via muscle contractions (BONNET; FERRARI, 2010). Studies have shown that adults who were physically active during adolescence present an 8-10% greater bone mineral content (BMC) compared to their less active peers (BAXTER-JONES et al., 2008). Furthermore, among girls, maintaining or increasing PA levels throughout adolescence resulted in greater gains in BMC and bone mineral density (BMD) compared to the girls whose PA levels declined during the same period (RAUTAVA et al., 2007). São Paulo State University - UNESP Campus of Presidente Prudente 14 These changes in bone structure are referred to as mechanoadaptation, a process that occurs in response to mechanical demands, as described by Wolff’s Law (WOLFF, 1893). Also known as “Mechanostat Theory”, this model posits that bone adapts to mechanical loads based on specific thresholds (FROST, 2003). When strain exceed the modeling threshold, it initiates an adaptive process that enhances the strength of the stressed area by reorganizing the bone microarchitecture or adding additional tissue. However, excessive stress may lead to microdamage and eventually fractures may occur to the bone; in contrast, when loads fall below a remodeling threshold (i.e., conditions of disuse), bone resorption accentuates, and its strength and mass consequently decrease (FROST, 2003). In this sense, bone adaptation would be a consequence of the balance between bone modeling (bone formation independent of osteoclast resorption) and remodeling (coupled action of osteoclasts and osteoblasts resorbing and forming bone, respectively, within a multicellular unit), and the latter would be the dominant mechanoadaptive response in bone (FROST, 2000) (FIGURE 1.2). São Paulo State University - UNESP Campus of Presidente Prudente 15 Figure 1.2 - Bone (re)modeling thresholds. Note: The bottom line represents typical peak bone strains, ranging from zero to fracture strain (FX), and the location of the remodeling, modeling and microdamage strain thresholds (MESr, MESm, and MESp, respectively). The horizontal axis indicates a balance between gain and loss in bone mass or strength. When strains are below this line, thus, bellow the MESr range, bone remodeling resorbs bone. When the strains enter or exceed the MESm range, modeling would increase bone strength and mass. If strains exceed the MESp, bone fracture may occur. DW = disuse; AW = the adapted window as in normally adapted adults; MOW = the mild overload window as in healthy growing mammals; POW = the pathological overload window. Source: adapted from (FROST, 2000). However, recent advances in bone mechanobiology have suggested that both bone modeling and remodeling may occur as a response to heightened mechanical loading (i.e., formation modeling and targeted remodeling) and as a response to a state of disuse (i.e., resorption modeling and disuse-mediated remodeling). In either case, bone modeling and remodeling could lead to distinct bone adaptations (morphological and tissue-level changes), which would depend on the mechanical loads applied to the bone and its inherent mechanical properties (e.g., bone stiffness) (Hughes et al., 2020). (FIGURE 1.3). This framework is centered on the role of the osteocytes, the mechanosensory cells in the bone, and their influence on the osteoblasts and osteoclasts. São Paulo State University - UNESP Campus of Presidente Prudente 16 Figure 1.3 - Theoretical framework of bone functional adaptation at the diaphysis. Note: Mechanical loading [1] on a bone of a given stiffness [2] creates a strain stimulus [3]. Bone formation modeling [A] and targeted remodeling [B] may be triggered when the stimulus is greater than costumery. Formation modeling results in new bone deposited on the endocortical and periosteal surfaces within the diaphysis, while targeted remodeling aims to removed bone from the damaged area. In a scenario of disuse or unloading, disuse- mediated remodeling [C] and resorption modeling [D] may be triggered on or near the endocortical surface of the diaphysis. In the disuse-mediated remodeling, there is a couple action of resorption and formation, while resorption modeling occurs independent of formation. Remodeling affects mainly tissue-level mechanical properties [5] and modeling alters bone morphology [7]. Consequently, all these processes impact bone stiffness [2], thereby affecting the strain response to subsequent bouts of mechanical loading. Reproduced with permission from Hughes et al., (2020) under the Creative Commons CC-BY License (Attribution-Noncommercial). The osteocytes are the most abundant cell type within the bone matrix. Through their extensive lacuna-canalicular network, they serve as key mechanosensory cells, transducing mechanical signals such as fluid shear stress, hydrostatic pressure, electric fields, and tissue strain (i.e., shear, compression, and torsion forces) into biochemical signals, i.e., osteokines, that coordinate the activity of osteoblasts and osteoclasts, thereby regulating bone formation and resorption (JAALOUK; LAMMERDING, 2009; ROBLING; CASTILLO; TURNER, 2006; YOU et al., 2008). São Paulo State University - UNESP Campus of Presidente Prudente 17 Osteocytes almost exclusively express sclerostin, a glycoprotein that acts as an antagonist in the Wnt singling pathway to inhibit osteoblast activity under unloading conditions (LIN et al., 2009; ROBLING et al., 2008). Studies have shown that mechanical loading is responsible to inhibit sclerostin in osteocytes, facilitating bone formation (ROBLING et al., 2008). Osteocytes also promotes bone resorption via secretion of Receptor Activator of Nuclear Factor κB Ligand (RANKL), a factor that binds to its receptor RANK on the membrane of the osteoclast’s precursors, facilitating osteoclastogenesis (YOU et al., 2008). Secretion of RANKL is an example of targeted- remodeling, commonly observed in fatigue-loaded bone by apoptosing osteocytes surrounding the damaged area (KENNEDY et al., 2012). However, osteocytes can also regulate bone resorption via expression of osteoprotegerin (OPG), a decoy receptor that can bind to RANKL and block osteoclastogenesis (ROBLING et al., 2008; YOU et al., 2008). In this sense, the balance between OPG and RANKL (OPG:RANKL ratio) is frequently used as a marker of bone anabolism. While osteokines may indicate whether an intervention is inducing a signling response favoring formation of osteoblasts or osteoclasts, the activity of these cells can be measured assessing bone turnover markers (BTMs). The BTMs are biochemical markers found in blood or urine that reflect the rate of bone formation and resorption (SCHINI et al., 2023). They have been used to monitor acute adaptations from interventions either in clinical or research settings. The International Osteoporosis Foundations (IOF) recommends that Procollagen I N-propeptide (PINP) and Type I Collagen C-Telopeptide (CTX-I) should be used as primary markers of osteoblasts and osteoclasts activity, respectively (VASIKARAN et al., 2011). Roughly, increased serum P1NP would be a consequence of increased bone formation, while increased serum CTX-I would be the result of increased bone resorption, and the balance between them would express the bone turnover (SCHINI et al., 2023). Together with the osteokines (e.g., sclerostin, RANKL, OPG), BTMs would provide a more comprehensive understanding of the effect PA on bone modeling and remodeling. However, studies investigating the effect of PA on bone markers present high heterogeneity and are mostly focused on adult men and the elderly (DOLAN et al., 2020, 2022, 2024), limiting our understanding of how those markers respond and adapt to mechanical loading, mainly in younger population. The impact of PA - including exercise, a structured and planned form of PA - on bone markers is influenced by different factors, including age, sex and training variables, i.e., duration, São Paulo State University - UNESP Campus of Presidente Prudente 18 intensity and gravitational impact (DOLAN et al., 2022; KOHRT et al., 2004; SCHINI et al., 2023). Kish et al. analyzed the acute effect of plyometric exercise on bone turnover markers (bone alkaline phosphatase [BALP], OPG, RANKL, and CTX-I) of boys and men and found that although the intervention positively affected those markers, the pattern of increase differed between groups (KISH et al., 2015). This corroborates the “window of opportunity” concept for building stronger bones during growth (HAAPASALO et al., 1998; PEARSON; LIEBERMAN, 2004). Additionally, high-impact exercise induces an osteogenic effect in boys and girls; however, this response might occur through different signaling pathways, which, in girls, would involve the suppression of catabolic pathways, i.e., inhibition of RANKL expression and CTX-I (DEKKER et al., 2017; KLENTROU et al., 2018; KURGAN et al., 2020). On the other hand, engaging in exercise modalities that do not involve gravitational impact may not benefit the bone tissue. Research indicates that cyclists and swimmers exhibit similar or lower bone mineral density to their non-athletic peers, with lower bone density than athletes in high-impact sports (AGOSTINETE et al., 2020a, 2020b; GONZÁLEZ-AGÜERO et al., 2017; OLMEDILLAS et al., 2018). However, recents studies on the impact of cycling on bone turnover markers have shown that, when exercise is performed at high-intensity, it may exhert an anabolic effect on bone (BORZOOEIAN et al., 2023; KOUVELIOTI et al., 2019a, 2019b). The long duration of exercise (>60min), usually at moderate-to-vigorous intensity, would lead to a disturbance in calcium metabolism that seems to be responsible for increasing osteoclasts activity, thus increasing bone resorpotion (BARRY et al., 2011; WHERRY; SWANSON; KOHRT, 2022). Thus, shortening the exercise length and increasing intensity using a high-intensity interval exercise (HIIE) strategy would benefit an anabolic response on bone as this type of training is know to impose great tension on musculoskeletal system, consequently, increasing load on bone tissue (BUCHHEIT; LAURSEN, 2013). Yet, only one study have compared the effect of HIIE with and without impact on bone metabolic markers, using a cross-over design. In this study, recreationally active women were submited to two sessions of HIIE on a bycicle and on a treadmill, and the acute response of sclerostin, P1NP and CTX-I were measured through serum before and after each session. Different from what was expected, the authors did not observe any differences in markers of formation and resorpotion following both HIIE, although sclerostin have increased significantly 5-min post-exercise and decreased to baseline levels 1h post- São Paulo State University - UNESP Campus of Presidente Prudente 19 exercise (KOUVELIOTI et al., 2018a). Thus, not only the effect of high-intensity exercise with and without impact remain controversial, but it remains unclear whether those modes of exercise (cycling and running) could benefit adolescent females whose bones are still growing. Furthermore, individuals not engaged in any exercise routine may present a bone that is less adapted to mechanical loading, increasing its responsiveness to exercise (HUGHES et al., 2020). Potentially, HIIE is a great strategy for bone tissue, as it combines short periods of high intensity and high muscle demand, intervaled with periods of recovery, important to reestablish the responsiveness of bone cells to the mechanical stimuli applied (BUCHHEIT; LAURSEN, 2013; ROBLING et al., 2002). Additionaly, high-intensity exercise may impact bone metabolism via non- mechanical pathways, such as modulation of cytokines and inflammation (FURMAN et al., 2019; KURGAN et al., 2020; NEMET et al., 2002; PETERSEN; PEDERSEN, 2005; TAKAYANAGI, 2007), potentially mediated by the adipose tissue and skeletal muscle (KIRK et al., 2020). In this scenario, RANKL is one of the main mediators of the relationship between immune and bone cells (TAKAYANAGI, 2007; YOU et al., 2008). This is because RANKL, which is secreted by osteocytes and osteoblasts, also belongs to the TNF family and is secreted by activated T-cells (ANDERSON et al., 1997). IL-6 and TNF-α are pro-inflammatory cytokines (although muscle-derived IL-6 plays an anti-inflammatory role) that are negatively associated with bone by acting to stimulate the expression of RANKL, thus increasing osteoclasts’ differentiation and bone resorption (UDAGAWA et al., 1995). Conversely, the anti-inflammatory cytokine IL-10 suppresses the expression of both TNF-α and IL-6, blocking the signaling of RANKL, therefore limiting bone resorption (FURMAN et al., 2019; PETERSEN; PEDERSEN, 2005) (FIGURE 1.4). São Paulo State University - UNESP Campus of Presidente Prudente 20 Figure 1.4 - Interaction between cytokines and bone cells. Note: Reproduced with permission from Mezil et al. (2015) under the Creative Commons CC-BY License (Attribution-Noncommercial). Adipose tissue plays an important role in energy metabolism and systemic inflammation through the secretion of adipokines, the main ones being leptin and adiponectin (KERSHAW; FLIER, 2004). The first is a key regulator of energy balance, but high-levels of leptin is associated with a pro-inflammatory state via increased expression of IL-6 and TNF-α. Adiponectin, on the other hand, has an anti-inflammatory effect, as it increases insulin sensitivity and promotes IL-10 expression (KERSHAW; FLIER, 2004; REN et al., 2022). The effect of exercise on these cytokines and adipokines is dependent on intensity and type of exercise. For instance, intermittent and aerobic exercise promote a greater increase in IL-10 compared to resistance exercise (CABRAL-SANTOS et al., 2019). Kouvelioti et al. (2019) compared the effects of high-intensity interval running (HIIR) and cycling (HIIC) on IL-10 and São Paulo State University - UNESP Campus of Presidente Prudente 21 TNF-α in active young adults, finding distinct responses between the two exercises (KOUVELIOTI et al., 2019b). Regarding adipokines, the effect of exercise on them is more complex. Studies on trained adults have showed that short-term intense exercise declined leptin (JÜRIMÄE; JÜRIMÄE, 2005) and increased adiponectin concentrations (JÜRIMÄE; PURGE; JÜRIMÄE, 2005); however, this effect could be influenced by training status and amout of muscle involved in the activity (BOUASSIDA et al., 2010). Although it’s widely spread that bone adaptations depend on the gravitational impact, other non-mechanical pathways may influence bone response to exercise as well. Beyond impact, HIIR and HIIC also differ in the number of muscles recruited (greater in running) and muscle strain (greater in cycling), which could impact mediators of bone remodeling, such as cytokines (i.e., IL- 6, IL-10 and TNF-α) and adipokines (i.e., leptin and adiponectin). Exploring the differences and potential benefits of high-intensity exercise without gravitational impact may benefit adolescents with injury-related limitations or who require impact-free exercise. Additionaly, assessing the preference of the adolescents for the type of intervention (HIIR vs HIIC) may also help the professionals responsible for prescribing exercise for this population. Considering that the practice of PA among adolescents has been decreasing over the years (COOPER et al., 2015), designing the strategy that promotes motivation and adherence to exercise in this population, mainly through feelings of enjoyment and well-being, seems to be as important as identifying the dose (e.g., intensity) that promotes the greatest health benefit (DISHMAN; SALLIS; ORENSTEIN, 1985; EKKEKAKIS; PARFITT; PETRUZZELLO, 2011). A disavantadge for HIIE is that it may lead to unpleasant feelings, which would affect negatively the adherence in physical activity (BIDDLE; BATTERHAM, 2015; EKKEKAKIS, 2009; RHODES; KATES, 2015). The Dual-Mode Theory explains this phenomenon, stating that exercise affects emotions through a combination of cognitive processes (like self-efficacy and goals) and physical sensations (such as breathing, acidity levels, and core temperature) (EKKEKAKIS, 2009). However, despite the arrousal in perceived feelings, a positive feeling of pleasure, activated by the reward system, may occur as a result of increased self-efficacy and excitement after completing an intense exercise session (MALIK et al., 2018; ROSE; PARFITT, 2007). In addition to the contrasts in the literature on this topic, no study was found that compared the effect of HIIC and HIIR on the affective perception and enjoyment of adolescent girls. São Paulo State University - UNESP Campus of Presidente Prudente 22 To the best of our knowledge, no study has investigated the effect of HIIT and HIIC on bone markers, osteokines, cytokines, and adipokines in adolescent females not engaged in any exercise program using a cross-over design trial. Moreover, we believe that in addition to investigating the effect of different exercise modes on health outcomes, it is important to understand exercise mode preference and how each of them can affect mediators of adherence to physical exercise, such as affective perception and enjoyment. We hypothesize that HIIE and HIIC would impact bone through different pathways, but leading to similar bone turnover (i.e., ratio between bone formation and resorption markers) and inflammatory response. However, HIIT would be less detrimental to the affective perception of the exercise, while it would be more enjoyable. São Paulo State University - UNESP Campus of Presidente Prudente 23 2. THESIS OBJECTIVES 2.1. General objective To investigate the effect high-intensity interval exercise with and without impact on markers of bone metabolism, and potential mediators of these markers, such as pro- and anti-inflammatory cytokines and adipokines, assessing the impact of each mode of exercise on affective perception and enjoyment. 2.2. Specific objectives I) To compare the effect of HIIR and HIIC on bone turnover markers (i.e., P1NP, CTX-I, and sclerostin) and osteokines (i.e., RANKL and OPG) of adolescent girls (Chapter 4). II) To investigate the impact of each mode on their affective perception and enjoyment (Chapter 4). III) To compare the effect of HIIR and HIIC on inflammatory markers (i.e., IL-6, IL- 10 and TNF-α) and adipokines (i.e., leptin and adiponectin) of adolescent girls (Chapter 5). São Paulo State University - UNESP Campus of Presidente Prudente 24 3. METHODS 3.1. Participants Female adolescents were recruited for the study through digital media and visits to public and private schools in the city of Presidente Prudente-SP/Brazil. Inclusion criteria were: (i) aged between 14 and 19 years old; (ii) non-smokers; (iii) having already gone through menarche; (iv) not presenting any pre-diagnosed osteometabolic disease; (v) free of musculoskeletal injury that would impair exercise, (vi) not having used glucocorticoids for more than 3 months (COSMAN et al, 2014); (vii) not engaged in any sport modality or exercise routine in a structured way (more than 3 hours of weekly training) in the previous 6 months; (viii) presented the consent and assent form signed by the adolescent and their legal guardian. Exclusion criteria were: (i) not passed 2 years from PHV, (ii) could not draw blood. A total of 21 adolescents were assessed for eligibility; however, 7 adolescents did not meet the inclusion criteria (e.g., participation in organized sports or resistance training programs [5] or declined to participate [2]), and 3 adolescents withdrew after being assigned to the study (e.g., fear of blood draws [1], inability to draw blood [2]). Consequently, 11 adolescents successfully completed the study (FIGURE 3.1). São Paulo State University - UNESP Campus of Presidente Prudente 25 Figure 3.1 - Flowchar of the study Note: HIIR, high-intensity interval running; HIIC, high-intensity interval cycling. Source: from the author. 3.2. Ethical aspects The study was conducted per the Declaration of Helsinki and received ethics approval from the Research Ethics Boards of the Sao Paulo State University (CAAE 67621222.8.0000.5402 / Approval 5.951.255). 3.3. Study design This was a cross-over randomized designed study. Participants assigned to the study performed two high-intensity exercise trial in random order: a high-intensity interval running (HIIR) trial on the treadmill and a high-intensity interval cycling (HIIC) trial on a cycle ergometer. The order of trials was block randomized using the Excel software. Participants had to visit the laboratory at São Paulo State University twice over two weeks. During their first visit, participants completed baseline assessments such as body composition and anthropometrics, answered São Paulo State University - UNESP Campus of Presidente Prudente 26 questionnaires, and performed an incremental test on either the bicycle or treadmill according to their assigned exercise mode. The second visit happened 48h to 72h later, when they performed the first high-intensity interval trial of the same exercise mode as the incremental test performed earlier in the week. The third and fourth visits followed the same procedures – with the exception of the baseline assessments – but alternating the exercise modes (cycling and running) (FIGURE 3.2). Figure 3.2 - Study protocol and timeline. Note: HIIR, high-intensity interval running; HIIC, high-intensity interval cycling. Source: from the author. 3.4. Intervention protocol Before the incremental test performed during the first and third visits, the protocol was explained in detail, security information was given, and they were allowed to familiarize themselves with running or cycling at different intensities. In the treadmill, partipants warmed up for 5-min at 5 km/h with 1% slope. Then, speed was increased to 6km/h and increments of 1km/h were done every minute until voluntary fatigue. Similarly, the incremental test on the bicycle started with a 5-min warm-up at 35 W, followed by an increase to 60 W and increments of 25 W every minute until voluntary fatigue. Heart rate (HR) was continuously monitored with an OH1 Sensor (Polar®), and rate of perceived exertion (RPE) was evaluated using the Borg Scale, recording maximum values. São Paulo State University - UNESP Campus of Presidente Prudente 27 After 48 to 72 hours of the incremental test, in the morning (between 07:30 and 10:30) the participants performed the high-intensity interval trial in the exercise mode (running or cycling) they were tested in that week. Both trial sstarted with a 4-min warm-up followed by 8 bouts of 1- min effort at the maximal workload achieved in the incremental test, separated by 1-min passive recovery intervals. The morning before each trial, participants were given a standardized breakfast containing orange juice and whole wheat bread (butter was allowed upon request). On the morning of each trial, blood samples were collected pre-exercise (1.5 hours post-prandial), 5-min and 60-min post- exercise. 3.5. Baseline measurements 3.5.1. Body composition The height (cm) of the adolescents were measured using a wall stadiometer (E120A, Tonelli, Brasil)with 0.1 cm accuracy, and body mass (kg) using a digital scale (Welmy, model W300A), with 0.1 kg accuracy. Based on the values obtained, the peak height velocity (PHV) of these adolescents, an important indicator of somatic maturation, was estimated based on the mathematical models (MOORE et al., 2015). The equation provides the time (in years) left (negative values) or past (positive values) for the PHV, then, subtracting the value of age, we can estimate the age at which PHV occurred (APHV). The bone density (BMD [g/cm2]) and content (BMC [g]) were measured by the bone densitometry technique [dual-energy X-ray absorptiometry - DXA], (General Electrics brand, model WH - Prodigy Primo). The device was calibrated before the beginning of the measurements, following the manufacturer's recommendations, and three examinations were performed: total body, lumbar spine, and femoral neck. The measurements were conducted by a single, previously trained evaluator and, during the exam, all participants were wearing light clothing, barefoot, with no metal objects attached to their bodies. The total body analysis, besides providing the variables of bone density and content, also provided the values of lean soft tissue and body fat, both in percentage and absolute values. São Paulo State University - UNESP Campus of Presidente Prudente 28 3.5.2. Questionnaires The adolescents answered the Bone-specific Physical Activity Questionnaire (BPAQ) (WEEKS; BECK, 2008). It collects information regarding historical bone-relevant physical activity in two domains: exercise loading history from birth (past-BPAQ [pBPAQ]) and from the previous 12 months (current-BPAQ [cBPAQ]). Over the two weeks of assessments and tests, a trained dietitian applied the 24-hour multiple-pass dietary recall with participants via online calls. Participants were previously instructed to follow their usual diet. The calls were made on Mondays, Wednesdays, and Saturdays, accounting for two weekdays and one weekend day, and they were asked to report all foods and beverages consumed on the day before the call. A photographic manual was used to reduce memory bias of portion sizes. The information collected was used to estimate their mean energy intake over the week and the quantities of micronutrients and macronutrients consumed in the week preceding each trial, using a food composition table (BARANOWSKI, 2012). In addition, information about the date on which menarche occurred and whether the adolescents are on contraceptive methods were collected. If they did, information about the method used was collected. 3.6. Outcome variables 3.6.1. Blood analysis Approximately 3 mL of whole blood was collected into pre-chilled Vacutainer SST tubes and EDTA from participants’ median cubital vein in the antecubital fossa while lying in a supine position. Samples were collected using a standard venipuncture technique by a trained nurse during each session at pre-exercise, and 5 min and 60 min post-exercise. Blood samples were inverted 8- 10 times before sitting for 30 min at room temperature. The samples were then centrifuged at 3000 g for 15 min at 4° before the serum and plasma were aliquoted into 1.5ml microcentrifuge tubes (Eeppendorfs®) and stored at -80°C until analysis. 3.6.2. Bone turnover markers The analyses of P1NP and CTX-I were carried out by a private laboratory that adheres to the Brazilian standards set by the Ministry of Health of the Brazilian government. P1NP was São Paulo State University - UNESP Campus of Presidente Prudente 29 analysed from serum and the CTX-I, from the plasma EDTA; both used the electrochemiluminescence technique. Additionally, OPG, RANKL and sclerostin were analysed from serum by the enzyme- linked immunosorbent assay (ELISA) with the following kits: #DSST00, P416285; #DY626, P412568; #DY805, P356096 (R&D System, Minneapolis, USA). The intra-assay coefficient of variation for the standard curve in each assay were 3,7% for sclerostin; 0.9% and 1.1% for RANKL; and 0.7% and 1.0% for OPG. 3.6.3. Cytokines and adipokines Serum IL-6, IL-10 and TNF-α were measured in triplicate with ELLA single assay kits (cat.# SPCKBPS-003028, ST01B-PS-000276, ST01B-PS-002803, respectively; ProteinSimple, San Jose, CA). The inter-assay coefficients of variation were 9.9% for IL-6, 6.6% for IL-10 and 8.7% for TNF-α. The intra-assay coefficients of variation for IL-6, IL-10 and TNF-α were 4.6%, 6.69% and 2.54%, respectively. Serum concentrations of leptin and adiponectin were acquired by the enzyme-linked immunosorbent assay (ELISA) using commercial kits (R&D System, Minneapolis, USA). 3.6.4. Metabolic markers In order to control for potential differences in the metabolic demand between the exercise trials, serum glucose and LDH from were measured using the colorimetric method with Labtest commercial kits (Lagoa Santa, Minas Gerais, Brazil). 3.6.5. Affective perception and enjoyment Before and 5 min after each trial, the adolescents answered the Feeling-Scale (FS), a 11- point single-item bipolar measure ranging from +5 to -5 (ALVES et al., 2019). The difference between pre- and post-exercise was used to estimate exercise impact on self-reported affect and arousal from exercise. Then, 10 min after completing each trial, the adolescents answered the Physical Activity Enjoyment Scale (PACES), which quantifies enjoyment of physical activity through a 17-item bipolar scale that is separated by a 7-point scale (e.g., 1 = “I enjoy it”; 7 = “I hate it”, 4 = “neutral”) (ALVES et al., 2019). São Paulo State University - UNESP Campus of Presidente Prudente 30 3.7. Statistical Analysis Normality and variability of the residuals were tested using graphic visualization (i.e., histograms and Q-Q plots). Variables that violated the normality of the residuals were log- transformed. Linear mixed models were used to assess the fixed effects of exercise mode and time (repeated measure) and the interaction between these factors on the outcomes, while accounting for the individual variability (random effect). Kenward-Roger's method was set to adjust the degrees of freedom in the model. Baseline level of the outcome was included as a covariate in the model, and total-body less head BMD and pBPAQ were also tested as covariates. For the models where adipokines were the outcomes, body fat (%) was also included as a covariate. When interaction was significant, False Discovery Rate (FDR) was used to adjust the p-values for multiple comparisons. An independent T-test was used to analyze differences in nutritional variables between weeks and differences between modes on PACES score. Additionally, we also tested the effect of each mode on each item of the PACES questionnaire. Effect size was calculated using Cohen’s-d: . Statistical significance was set at an alpha level of 0.05 and analysis were carried using “stats”, “lme4”, “emmeans” and “effectsizes” packages in R. São Paulo State University - UNESP Campus of Presidente Prudente 31 4. ACUTE EFFECT OF HIGH-INTENSITY INTERVAL CYCLING AND RUNNING ON BONE TURNOVER MARKERS AND OSTEOKINES IN FEMALE ADOLESCENTS 4.1. Abstract The purpose of this study was to compare the acute effect of high-intensity interval exercise running (HIIR) and cycling (HIIC) on bone turnover markers and osteokines in adolescent girls, assessing the impact of the exercise on affective perception and enjoyment. In a cross-over design trial, 11 adolescent females, aged 15 to 19 years, completed two high-intensity interval trials (HIIR and HIIC) in random order. They performed a progressive incremental test to determine the workload in peak velocity output and peak power output, which were subsequently used to prescribe the HIIR and HIIC trials, respectively. Each trial consisted of 8 bouts of 1 min of effort at maximal workload with 1 min of passive recovery between bouts. Blood samples were collected pre-exercise (i.e., baseline), 5- and 60-min post-exercise. Serum blood was used to analyze bone markers of formation (i.e., P1NP and OPG) and resorption (i.e., CTX-I, sclerostin and RANKL), and cortisol. The adolescents answered the Feeling Scale (FS) before and after each trial, and the Physical Activity Enjoyment Scale (PACES) 10-min after each trial. P1NP significantly peaked 5-min post- HIIR (Interaction effect: β = 11.418, p = 0.027; ES = 0.28) but not following HIIC. Neither HIIR nor HIIC impacted CTX-I (Time effect: β = 0.009, p = 0.653; ES = 0.09). In contrast, sclerostin increased immediately after HIIR and HIIC, returning to baseline levels 60-min post exercise (Time effect: β = 10.889, p = 0.011; ES = 0.89). HIIR and HIIC did not change OPG (Time effect: β = 48.164, p = 0.329; ES = 0.07), RANKL (Time effect: β = 29.316, p = 0.132; ES = 0.07), or the logOPG:RANKL ratio (Time effect: β = -0.174, p = 0.207; ES = 0.11). Regarding affective and enjoyment, FS score decreased following HIIC (Δ = -4.1±0.9, p < 0.001, ES = 1.01). PACES total score was higher in HIIR compared to HIIC (75.6±5.7 vs 67.2±6.3, p = 0.004, ES = 1.48). Cortisol levels did not change following any exercise mode (Time effect: β = 0.016, p = 0.992; ES = 0). Overall, even in high-intensity exercise, bone formation seems to be condionated to the presence of impact-loading, as observed in the effect of HIIR in P1NP. Even though the response in the other markers was similar between the exercise modes, HIIC has a negative impact on affective perception and is also less pleasurable than HIIR. São Paulo State University - UNESP Campus of Presidente Prudente 32 4.2. Introduction Osteoporosis is a chronic disease defined as weakening of bones due to a decrease in bone mass and strength, consequently increasing the risk of fractures (GÓMEZ et al., 2021). Recent evidence indicates that the prevalence of osteopenia (i.e., reduced bone mass) and osteoporosis is 40% and 19.4% worldwide, respectively (XIAO et al., 2022). In addition, women, especially in postmenopausal period, are at greater risk of developing osteoporosis (WEAVER et al., 2016; XIAO et al., 2022). Although sexual hormones changes play a significant role in the rapid loss of bone mass post-menopause (CAULEY, 2015; ZALLONE, 2006), women tend to have a lower peak bone mass (PBM) in early adulthood (BAXTER-JONES et al., 2011). Among the factors capable of affecting PBM is physical activity (BAXTER-JONES et al., 2008; WEAVER et al., 2016). The effect of physical activity on bone occurs due to mechanical impact generated by ground reaction forces and muscle contractions during practice, in addition of hormones activity (BONNET; FERRARI, 2010; KOHRT et al., 2004). Despite that, it is unclear whether only gravitational impact generates bone adaptations, or the intense muscle contractions performed in high-intensity exercises are enough to stimulate anabolic bone responses. In theory, the physical exercise combining high-intensity exercise and higher ground reaction force can boost bone formation, but that is an assumption to be confirmed in paediatric groups yet. High-intensity interval exercise (HIIE) has proven to be a time-efficient and safe strategy with several health benefits for children and adolescents (CAO; QUAN; ZHUANG, 2019; EDDOLLS et al., 2017). Moreover, evidence from animal models indicates that HIIE appears to be more effective than continuous exercise, as short periods of rest between bouts are essential to restore the responsiveness of bone cells to the stimulus applied (ROBLING et al., 2002). However, research focused on bone health involving bone metabolic markers remains limited, especially studies comparing HIIE protocols with gravitational impact, e.g., high-intensity interval running (HIIR), and without gravitational impact, e.g., high-intensity interval cycling (HIIC). To date, there are two studies in the literature comparing the effect of HIIR and HIIC on bone turnover makers (BTM) and osteokine, being both studies carried out with active adult women (KOUVELIOTI et al., 2018, 2019b). São Paulo State University - UNESP Campus of Presidente Prudente 33 BTM and osteokines are considered dynamic variables of bone tissue, being considered relevant markers of anabolic or catabolic effect of an exercise intervention (DOLAN et al., 2020). The most important BTMs are procollagen type 1 N-terminal ropeptide (P1NP) and C-terminal telopeptide of type I collagen (CTX-I), which are fragments of collagen released in the blood stream by osteoblasts during formation and by osteoclasts during resorption, respectively (SZULC et al., 2017). Osteokines, on the other hand, are signalling proteins realized mainly by osteocytes, which influence osteoblasts and osteoclasts acitivation/differentiation (BURGERS; WILLIAMS, 2013). Sclerostin is a catabolic osteokine that acts inhibiting the canonical Wnt pathway, consequently decreasing bone formation (CHOI; ROBLING, 2021; GALEA; LANYON; PRICE, 2017). Another catabolic osteokine also produced by osteocytes is the Receptor Activator of Nuclear Factor Kappa-B Ligand (RANKL), that, when bounded to its receptor RANK in the membrane of the pre-osteoclasts, induces its differentiation, increasing bone resorption (YOU et al., 2008). Osteoclastogenis may also be inhibit by the over expression of osteoprotegerin (OPG), an anabolic protein that binds to RANKL and prevents it to induce osteoclasts differentiation (ROBLING et al., 2008; YOU et al., 2008). Although studies have explored these variables in paediatric populations, it is not clear what effect HIIR and HIIC has on these variables in adolescent girls not engaged in any exercise routine. Despite the positive effects of HIIE on paediatrics’ health, studies should also consider the impact the interventions have on affect (i.e., pleasure and displeasure) and enjoyment, as they might predict the adherence in physical exercise program (DISHMAN et al., 2005; DISHMAN; SALLIS; ORENSTEIN, 1985; EKKEKAKIS, 2009; RHODES; KATES, 2015). Evidence in the literature indicates that HIIE can generate unpleasant feelings, decreasing the adherence to the exercise program, due to a combination of cognitive processes and physical sensations (BIDDLE; BATTERHAM, 2015). Conversely, despite the unpleasant feelings generated by the exercise, the reward mechanism may also act, generating a positive feeling of pleasure as result of increased self-efficacy and excitement after an intense exercise session (MALIK et al., 2018; ROSE; PARFITT, 2007). Thus, the aim of this study was to investigate the impact of HIIR and HIIC on bone turnover markers and osteokines in untrained adolescent females, while assessing the effect of HIIC and HIIR on their affective perception and enjoyment. São Paulo State University - UNESP Campus of Presidente Prudente 34 Methods 4.2.1. Participants In the city of Presidente Prudente, Brazil, female adolescents were recruited from schools and via online publications to participate in this study. A total of 21 adolescents were assessed for eligibility, which inclusion criteria were: (i) be female; (ii) aged between 14 and 19 years; (iii) non- smokers; (iv) not enrolled in sports or exercise training for the previous 6 months; (v) be free of musculoskeletal injury that would impair exercise, and be able to perform running and cycling. Exclusion criteria were: (i) not passed 2 years from PHV, (ii) could not draw blood. After initial screening, 14 of them met the inclusion criteria. However, after being assigned, 3 adolescents were excluded from the study because they could not draw blood. Thus, 11 adolescents completed both trials (FIGURE 4.1). Figure 4.1 - Flowchart of the study. Note: HIIR, high-intensity interval running; HIIC, high-intensity interval cycling. Source: from the author. São Paulo State University - UNESP Campus of Presidente Prudente 35 4.2.2. Ethical aspects The study was conducted considering the Declaration of Helsinki and received ethics approval from the Research Ethics Boards of the Sao Paulo State University (CAAE 67621222.8.0000.5402 / Approval 5.951.255). 4.2.3. Study Design This was a cross-over randomized designed study. Participants assigned to the study performed two high-intensity exercise trial in random order: a high-intensity interval running (HIIR) trial on the treadmill and a high-intensity interval cycling (HIIC) trial on a cycle ergometer. The order of trials was block randomized using the Excel software. Participants had to visit the laboratory at São Paulo State University twice over two weeks. During their first visit, participants completed baseline assessments such as body composition and anthropometrics, answered questionnaires, and performed an incremental test on either the bicycle or treadmill according to their assigned exercise mode. The second visit happened 48h to 72h later, when they performed the first high-intensity interval trial of the same exercise mode as the incremental test performed earlier in the week. The third and fourth visits followed the same procedures – with the exception of the baseline assessments – but alternating the exercise modes (cycling and running) (FIGURE 4.2). Figure 4.2 - Study timeline and procedures. Note: DXA, dual-energy X-ray absorptiometry; BPAQ, bone-specific physical activity questionnaire; FS, feeling scale; PACES, physical activity enjoyment scale; HIIR, high-intensity interval running; HIIC, high-intensity interval cycling. Source: from the author. São Paulo State University - UNESP Campus of Presidente Prudente 36 4.2.4. Intervention protocol Before the incremental test performed during the first and third visits, the protocol was explained in detail, security information was given, and they were allowed to familiarize themselves with running or cycling at different intensities. In the treadmill, partipants warmed up for 4-min at 5 km/h with 1% slope. Then, speed was increased to 6km/h and increments of 1km/h were done every minute until voluntary fatigue. Similarly, the incremental test on the bicycle started with a 4-min warm-up at 35 W, followed by an increase to 60 W and increments of 25 W every minute until voluntary fatigue. After 48 to 72 hours of the incremental test, in the morning (between 07:30 and 10:30) the participants performed the high-intensity interval trial in the exercise mode (running or cycling) they were tested in that week. Both trial sstarted with a 4-min warm-up followed by 8 bouts of 1- min effort at the maximal workload achieved in the incremental test, separated by 1-min passive recovery intervals. Heart rate (HR) was continuously monitored with an OH1 Sensor (Polar®), and Rate of Perceived Exertion (RPE) was evaluated using the Borg Scale, recording maximum values.. The morning before each trial, participants were given a standardized breakfast containing orange juice and whole wheat bread (butter was allowed upon request). On the morning of each trial, blood samples were collected pre-exercise (1.5 hours post-prandial), 5-min and 60-min post- exercise. 4.2.5. Baseline measurements 4.2.5.1. Body composition The height (cm) of the adolescents were measured using a wall stadiometer (E120A, Tonelli, Brasil)with 0.1 cm accuracy, and body mass (kg) using a digital scale (Welmy, model W300A), with 0.1 kg accuracy. Based on the values obtained, the peak height velocity (PHV) of these adolescents, an important indicator of somatic maturation, was estimated based on the mathematical models (MOORE et al., 2015). The equation provides the time (in years) left (negative values) or past (positive values) for the PHV, then, subtracting the value of age, we can estimate the age at which PHV occurred (APHV). The bone density (BMD [g/cm2]) and content (BMC [g]) were measured by the bone densitometry technique [dual-energy X-ray absorptiometry - DXA], (General Electrics brand, São Paulo State University - UNESP Campus of Presidente Prudente 37 model WH - Prodigy Primo). The device was calibrated before the beginning of the measurements, following the manufacturer's recommendations, and three examinations were performed: total body, lumbar spine, and femoral neck. The measurements were conducted by a single, previously trained evaluator and, during the exam, all participants were wearing light clothing, barefoot, with no metal objects attached to their bodies. The total body analysis, besides providing the variables of bone density and content, also provided the values of lean soft tissue and body fat, both in percentage and absolute values. 4.2.5.2. Questionnaires The adolescents answered the Bone-specific Physical Activity Questionnaire (BPAQ) (WEEKS; BECK, 2008). It collects information regarding historical bone-relevant physical activity in two domains: exercise loading history from birth (past-BPAQ [pBPAQ]) and from the previous 12 months (current-BPAQ [cBPAQ]). Over the two weeks of assessments and tests, a trained dietitian applied the 24-hour multiple-pass dietary recall with participants via online calls. Participants were previously instructed to follow their usual diet. The calls were made on Mondays, Wednesdays, and Saturdays, accounting for two weekdays and one weekend day, and they were asked to report all foods and beverages consumed on the day before the call. A photographic manual was used to reduce memory bias of portion sizes. The information collected was used to estimate their mean energy intake over the week and the quantities of micronutrients and macronutrients consumed in the week preceding each trial, using a food composition table (BARANOWSKI, 2012). In addition, information about the date on which menarche occurred and whether the adolescents are on contraceptive methods were collected. If they did, information about the method used was collected. 4.2.6. Outcome variables 4.2.6.1. Blood analysis Approximately 3 mL of whole blood was collected into pre-chilled Vacutainer SST tubes and EDTA from participants’ median cubital vein in the antecubital fossa while lying in a supine position. Samples were collected using a standard venipuncture technique by a trained nurse during each session at pre-exercise, and 5 min and 60 min post-exercise. Blood samples were inverted 8- São Paulo State University - UNESP Campus of Presidente Prudente 38 10 times before sitting for 30 min at room temperature. The samples were then centrifuged at 3000 g for 15 min at 4° before the serum and plasma were aliquoted into 1.5ml microcentrifuge tubes (Eeppendorfs®) and stored at -80°C until analysis. The analyses of P1NP and CTX-I were carried out by a private laboratory that adheres to the Brazilian standards set by the Ministry of Health of the Brazilian government. P1NP was analyzed from serum and the CTX-I, from the plasma EDTA; both used the electrochemiluminescence technique. Additionally, RANKL, OPG and sclerostin were analysed from serum by the enzyme-linked immunosorbent assay (ELISA) with the following kits: #DSST00, P416285; #DY626, P412568; #DY805, P356096 (R&D System, Minneapolis, USA). The intra-assay coefficient of variation for the standard curve in each assay were for sclerostin: 3,7%; RANKL 1.1 and 0.9%; OPG: 0.7 and 1.0%. 4.2.6.2. Affective perception and enjoyment Before and 5 min after each trial, the adolescents answered the Feeling-Scale (FS), a 11- point single-item bipolar measure ranging from +5 to -5 (ALVES et al., 2019). The difference between pre- and post-exercise was used to estimate exercise impact on self-reported affect and arousal from exercise. Then, 10 min after completing each trial, the adolescents answered the Physical Activity Enjoyment Scale (PACES), which quantifies enjoyment of physical activity through a 17-item bipolar scale that is separated by a 7-point scale (e.g., 1 = “I enjoy it”; 7 = “I hate it”, 4 = “neutral”) (ALVES et al., 2019). 4.2.7. Statistical Analysis Normality and variability of the residuals were tested using graphic visualization (i.e., histograms and Q-Q plots). Variables that violated the normality of the residuals were log- transformed. Linear mixed models were used to assess the fixed effects of exercise mode and time (repeated measure) and the interaction between these factors on the outcomes, while accounting for the individual variability (random effect). Kenward-Roger's method was set to adjust the degrees of freedom in the model. Additionally, we tested the impact of TBLH-aBDM and pBAPQ in each model, but they were not significant nor improved the goodness-of-fit of the models; thus, only baseline level of the outcome was included as a covariate in the model. False Discovery Rate (FDR) was used to adjust the p-values for multiple comparisons when a significant effect was São Paulo State University - UNESP Campus of Presidente Prudente 39 found. An independent T-test was used to analyze differences in nutritional variables between weeks and differences between modes on PACES score. Additionally, we also tested the effect of each mode on individual items from PACES. Effect size was calculated using Cohen’s-d: <0.50 small effect-size, 0.50-0.79 moderate effect-size, 0.80-1.29 large effect-size, and ≥1.30 very large effect-size (COHEN, 2013). Statistical significance was set at an alpha level of 0.05 and analysis were carried using “stats”, “lme4”, “emmeans” and “effectsizes” packages in R. 4.3. Results The Table 4.1 presents the characteristics of the sample. The adolescent females presented a mean age of 17.4 (0.9) years, 37.1% of FM and BMI of 24 (5.1). Considering the nutrient intake during the week of each test, there was no difference in relation to the intake of macronutrients, i.e., protein (p = 0.634), carbohydrate (p = 0.885) and lipids (p = 0.958), and micronutrients, i.e., calcium (p = 0.652), phosphorus (p = 0.408) and vitamin D (p = 0.334) (TABLE 4.2). Table 4.1 - Samples' characteristics (n=11). Mean (SD) 95%CI Age (years) 17.4 (0.9) 16.7 to 18.0 Body mass (kg) 65.1 (14.6) 55.3 to 75.0 Height (cm) 164.8 (8.0) 159.4 to 170.2 BMI 24.0 (5.1) 20.6 to 27.4 FM (%) 37.1 (6.1) 33.0 to 41.2 LST (kg) 35.2 (5.7) 31.4 to 39.1 Maturity Offset (years) 4.4 (0.8) 3.8 to 4.9 TBLH-aBMD (z-score) 0.8 (1.1) 0.1 to 1.5 Note: SD, standard deviation; BMI, body mass index; FM, fat mass; LST, lean soft tissue, TBLH-aBMD, total body less head areal bone mineral density. Table 4.2 - Macro and micronutrients intake over the week preceding HIIE trials (n=11). Nutrients (g)/body mass (kg) Past-HIIR Mean (SD) Past-HIIC Mean (SD) p-value Protein (g/kg) 1.64 (0.59) 1.51 (0.61) 0.634 Carbohydrate (g/kg) 4.16 (1.71) 4.04 (2.22) 0.885 Lipids (g/kg) 1.05 (0.49) 1.07 (0.55) 0.958 Calcium (mg/d) 446.5 (217.0) 484.0 (163.5) 0.652 Phosphorus (mg/d) 1161.8 (360.2) 1039.2 (319.6) 0.408 Vitamin D (μg/d) 2.6 (2.5) 1.77 (1.0) 0.334 Note: HIIR, high-intensity interval running; HIIC, high-intensity interval cycling; SD, standard-deviation. São Paulo State University - UNESP Campus of Presidente Prudente 40 The HR at the end of every bout was consistently greater than 90% of participant’s maximum HR (HRmax), as determined by the incremental test. Although HR at the end of every effort stage was precisely similar between modes, the perceived exertion during these bouts differed between HIIC and HIIR. Interestingly, RPE increased faster during HIIC, becoming significantly greater than HIIR after the fourth (p=0.029) and fifth (p=0.031) bouts. However, from the sixth bout onwards, RPE became similar again between modes (FIGURE 4.3). Figure 4.3 - RPE and HR (mean±SEM) change during HIIR and HIIC (n=11). Note: HIIR, high-intensity interval running; HIIC, high-intensity interval cycling; RPE, rate of perceived exertion; HR, heart rate. Source: from the author. 4.3.1. Bone Makers A significant increase in serum P1NP was observed 5-min post HIIR (interaction effect: β = 11.418; p = 0.027; d = 0.3 [small effect]), but not following HIIC, with concentration returning to baseline levels 60-min post exercise. Neither HIIR nor HIIC changed serum concentration of CTX-I (time effect: β = 0.009; p = 0.653; d = 0.1 [trivial effect]). In contrast, both modes led to an increase in sclerostin 5-min post exercise (time effect: β = 10.886, p = 0.011; d = 0.9 [large effect]). Regarding osteokines, HIIR and HIIC did not affect OPG (time effect: β = 48.164; p = 0.329; d = 0.07 [trivial effect]), RANKL (time effect: β = 29.316; p = 0.132; d = 0.1 [trivial effect]) or OPG:RANKL ratio (time effect: β = -0.174; p = 0.207; d = -0.1 [trivial effect]) (FIGURE 4.4). São Paulo State University - UNESP Campus of Presidente Prudente 41 Figure 4.4 - Bone turnover makers and osteokines response (mean±SEM) to HIIR and HIIC. Note: HIIR, high-intensity interval running; HIIC, high-intensity interval cycling; ES, effect size, #difference between baseline and 5-min post-exercise in HIIR; *difference between baseline and 5-min post-exercise; baseline measurments were inserted as a covariate. Source: from the author. 4.3.2. Affect and enjoyment As shown in Table 4.3, HIIC negatively impacted the affective perception of the adolescents following exercise, while the score did not significantly change following HIIR (interaction effect: β = -2.818; p = 0.035; d = -1.0 [large effect]). Moreover, the enjoyment for the exercise was greater after HIIR compared to HIIC (75.6 vs 67.2; ∆ = 8.5 [95%CI 2.8 to 14.1], d = 1.5 [very large effect]). The individual analysis of the items from the PACES showed that HIIR was more exhilarating (p = 0.014, d = 1.2 [large effect]), invigorating (p = 0.040, ES = 0.9 [large effect]), energizing (p = 0.037, d = 1.0 [large effect]) and made the adolescents feel physically better (p = 0.007, d = 1.3 [very large effect]) compared to HIIC (FIGURE 4.5). Although the interaction effect was not signi�icant, cortisol has slightly increased post-HIIC but not post- HIIR (∆= 4.0, p = 0.080; d = 1.0 [large effect]) (TABLE 4.4). Universidade Estadual Paulista “Júlio de Mesquita Filho” Campus Presidente Prudente 42 Table 4.3 - Effect of HIIR and HIIC on affective perception and enjoyment (n=11). FS Fixed effects PACES Pre Mean (95%CI) Post Mean (95%CI) Time effect β±SE (p-value) Mode effect β±SE (p-value) Time-by-mode effect β±SE (p-value) Mean (95% CI) Δ (95%CI) HIIR 3.3 (2.0 to 4.6) 2.0* (0.7 to 3.3) -2.7±0.9 (p<0.001) -1.15±0.7 (p=0.104) -2.8±1.3 (p=0.035) 75.6* (71.8 to 79.5) 8.5 (3.1 to 13.8) HIIC 3.5 (2.2 to 4.8) -0.6a (-1.9 to 0.8) 67.2 (62.9 to 71.4) Note: HIIR, high-intensity interval running; HIIC, high-intensity interval cycling; FS, Feeling Scale; astatistical difference between pre- and post-exercise; *Statistical difference from HIIC. Table 4.4 – Cortisol response (mean±95%CI) to HIIR and HIIC (n=6). FS Fixed effects Cortisol (µg/dL) Pre Mean (95%CI) Post Mean (95%CI) Time effect β±SE (p-value) Mode effect β±SE (p-value) Time-by-mode effect β±SE (p-value) HIIR 7.5 (4.8 to 10.1) 9.5 (6.9 to 12.2) -2.03±1.14 (p=0.092) 2.03±1.15 (p=0.092) 4.02±2.18 (p=0.080) HIIC 9.7 (7.0 to 12.3) 11.7 (9.1 to 14.4) Note: HIIR, high-intensity interval running; HIIC, high-intensity interval cycling. astatistical difference between pre- and post- exercise; *Statistical difference from HIIC. Universidade Estadual Paulista “Júlio de Mesquita Filho” Campus Presidente Prudente 43 4.4. Discussion This study aimed to investigate whether HIIE with and without impact-loading (HIIR and HIIC, respectively) leads to similar bone marker responses in female adolescent. Our results showed that the bone marker of formation, P1NP, significantly increased immediately after HIIR but not following HIIC; however, the effect size of this change was of small magnitude. Additionally, both modes of exercise were ineffective to alter serum concentrations of important osteokines, i.e., OPG and RANKL, involved in the signaling pathways of bone turnover cells, such as osteoblast and osteoclast. On the other hand, HIIR and HIIC increased the serum concentration of sclerostin immediately after exercise, with a large effect size for the change. Despite the overall effect on bone markers being similar between HIIR and HIIC, their impacts on affective responses and enjoyment differed, with HIIR eliciting more positive responses characterized by reduced affective arousal and greater enjoyment. Our results are partially in line with Kouvelioti et al (2018), whose study also investigated the influence of HIIE involving impact on bone markers but with recreationally active women. Although the authors found a rising trend in P1NP after HIIR (12% increase 5-min post-exercise, and 19% increase 1h post-exercise), this increase was not sufficient to reach significant time-by- mode interecation effect (KOUVELIOTI et al., 2018). On the other hand, the significant effect of the time-by-mode interaction found in our study on this marker showed a small effect size, but a similar change (18%). The explanation for these differences might be related to the fact that we did not account for plasma volume changes, which could inflate the differences found in the markers; however, the adolescents were encouraged to keep hydrating throughout the whole trial, potentially attenuating the influnce of this factor on the results. Besides, it is also possible that the younger and untrained sample from our study presented greater responsiveness to impact (HAAPASALO et al., 1998; PEARSON; LIEBERMAN, 2004), resulting in greater secretion of P1NP following exercise, such as we hypothesized for the study. Studies investigating age-related differences in bone markers showed that younger individuals – either boys and girls – present higher concentrations of bone markers, such as sclerostin, amino-terminal cross-linking telopeptide (NTX), and bone-specific alkaline phosphatase (bone-ALP), and the response of these markers to plyometric exercise is greater in the younger sample (DEKKER et al., 2017; KISH et al., 2015). Universidade Estadual Paulista “Júlio de Mesquita Filho” Campus Presidente Prudente 44 Nevertheless, Dolan et al (2020) have shown in a systematic review and meta-analysis that bone formation markers, such as P1NP, indead show small transient increases within 15-min post- exercise, although this would be related to exercise-induced damage of connective tissue or haemodynamic shifts (DOLAN et al., 2022). Contrasting effects of HIIE without impact on bone resorption markers are found in literature. While our study showed that CTX-I serum concentration did not change following both HIIR and HIIC, Theocharidis et al (2020) observed an increase in CTX-I concentrations following high-intensity interval swimming among adolescents (THEOCHARIDIS et al., 2020), while plyometric exercise would downgrade CTX-I following plyometric exercise on adolescent females (KURGAN et al., 2020). Conversely, CTX-I concentrations increased significantly following HIIE, regardless of impact, in young men (KOUVELIOTI et al., 2019a) but not in young women (KOUVELIOTI et al., 2018). CTX-I is a resorption marker that is more likely to increase in response to longer exercise (e.g., endurance exercise) or in exercise trials that cause disruption in calcium homeostasis (GUILLEMANT et al., 2004; HAAKONSSEN et al., 2015; SHEA et al., 2014; SHERK et al., 2017). Exercise-induced decrease in serum ionized calcium leads to increased secretion of parathyroid hormone (PTH), which is responsible for stimulating bone resorption to counteract the drop in calcium in the bloodstream. (WHERRY; SWANSON; KOHRT, 2022). Thus, exercise-induced increase in CTX-I may not necessarily predict a mechanical adaptation of the bone (e.g., an initial catabolic catabolic stimuli necessary to activate remodeling cycle), but be a consequence of hormonal and metabolic disturbances caused by the exercise, mainly related to the calcium homeostasis (DOLAN et al., 2022). In this sense, the lack of time and mode effect on CTX-I found in this study could be a positive sign that, different from previous findings on the effect of cycling on bone, HIIC may not induce a catabolic effect marked by the unchanged CTX- I concentrations in adolescent females. The activity of osteoblast and osteoclast is highly influenced by the osteocytes, the mechanosensors within the bone and major responsible for bone homeostasis (HUGHES et al., 2020). They orchestrate the bone response to mechanical loading by secreting factors and proteins, such as sclerostin, RANKL, and OPG, that will act in pathways related to osteoblastic and osteoclastic differentiation (CHOI; ROBLING, 2021; GOODMAN; HORNBERGER; ROBLING, 2015; YOU et al., 2008). In our study, HIIR and HIIC increased sclerostin similarly 5 min post- Universidade Estadual Paulista “Júlio de Mesquita Filho” Campus Presidente Prudente 45 exercise, with values returning to baseline 60 min post-exercise. Increased sclerostin following exercise has also been found in other studies involving women and different modes of exercise (GUZMAN et al., 2022; KOUVELIOTI et al., 2019b; KURGAN et al., 2020; PICKERING et al., 2017). This glycoprotein is known to increase the resorption activity of osteoclasts while inhibiting osteoblasts’ differentiation via inhibition of canonical Wnt signaling, thus increasing bone catabolism. Studies on animals have shown that mechanical loading downregulates sclerostin while it plays a crucial role in bone loss under unloading conditions (LIN et al., 2009; ROBLING et al., 2008). In this sense, one would expect that the increase in sclerostin after exercise observed in humans could be a signaling of osteocytes to initiate the targeted-remodeling process due to tissue microdamage due to high exercise intensity or repetitive impact (HUGHES et al., 2020); however, bone remodeling induced by sclerostin would also involve downregulation of OPG and increased RANKL, which were not found significant over the 60-min follow-up of the study. Interestingly, previous studies have shown that increased sclerostin following exercise might be associated with increased pro-inflammatory cytokine tumor necrosis factor alpha (TNF-α) (KOUVELIOTI et al., 2019b; OHORI et al., 2019), which was found augmented following both modes in our trial (data not shown). Regarding OPG and RANKL, neither HIIR nor HIIC had an impact on these markers following exercise. The RANKL stimulates osteoclastogenis by connecting to its receptor RANK on the surface of osteoclast precursos, while OPG, a decoy receptor, blocks this action by binding to RANKL, inhibiting the conection with its receptor (YOU et al., 2008). Thus, the balance between them is frequently used as a indicative of bone anabolism/catabolism. To the best of our knowledge, only one study investigated the effect of HIIR and HIIC on these cytokines in young female adults, and they found that both modes increased the concentration of RANKL, while OPG remained unchanged (BORZOOEIAN et al., 2023). However, a plyometric session performed by adolescent girls was effective to decrease RANKL concentrations post-exercise, while OPG was not affected (DEKKER et al., 2017; KLENTROU et al., 2018). Conversely, when studies were carrioud out in males, plyometric exercise had a positive impact on OPG in boys and men (KISH et al., 2015), while HIIC increased both OPG and RANKL in young male (MEZIL et al., 2015). Considering the similar effect that HIIC and HIIR had on bone turnover markers and ostekines, with just a small effect on P1NP in favour of HIIR, the overall efficacy of the intervention Universidade Estadual Paulista “Júlio de Mesquita Filho” Campus Presidente Prudente 46 as a source for clinical application should be considered in light of the effect it has on affect and enjoyment of the adolescent, since it may predict future physical activity behavior (DISHMAN et al., 2005; RHODES; KATES, 2015). For this reason, we assessed the impact that each intervention had on their affective perception (e.g., pleasant and unpleasant feelings) as well as how much they enjoyed each trial. Affective arousal was nearly three times greater following HIIC compared to HIIR, which did not significantly change between pre and post exercise. Additionally, post-exercise enjoyment was significatly different between modes, with a large effect size in favor of HIIR. This findings can be explained by the Dual-Mode Theory (DMT), which proposes that affective response to exercise is a product of the interplay between cognition (e.g., self-efficacy) and interoceptive cues about physiological symptoms (e.g., HR, acidosis, etc.) (EKKEKAKIS, 2009). Although HR was similar between modes, RPE almost prediceted (p = 0.09; data not shown) the differences in the FS found between HIIR and HIIC. This could have occurred because of sensations involved in each trial, such as “burning sensation” in the quadriceps, as widly reported by the participants, which they might not be used to feel based on their traning experience background. During cycling, the work rate is distributed among fewer muscle fibers compared to running; consequently, the metabolic stress on each muscle fiber may be higher, as each fiber requires a greater amount of energy (ACHTEN; VENABLES; JEUKENDRUP, 2003). In this sense, Malik et al (2018) have also reported a negative correlation between RPE following HIIR. According to Rose and Parfitt (ROSE; PARFITT, 2007) an individual may experience a positive affective response when they perceive they can complete an exercise session while being comfortably challenged. Consequently, although HIIR was also highly-demanding from the adolescents, the lower negative perceptions might have influence the enjoyment for the exercise, as shown by the PACES’s results. The results should be interpreted in light of the limitations of the study. First, the sample size of the study is particularly low and might have underpowered our analysis, specially regarding the osteokines. Although each adolescent is their on control in a cross-over design, it would be useful to have a non-exercising group to confirm that the only differences found in the study were related to the exercise and not diurnal and post-prandial fluctuations. The latter could potentially influence the bone markers, mainly CTX-I, which is usually indicated to be measured in a fasted state (QVIST et al., 2002). However, Scott et al. (2012) showed that fasting did not substantilly affect the bone metabolic response to exercise in men, despite a minor reduction of the duration of Universidade Estadual Paulista “Júlio de Mesquita Filho” Campus Presidente Prudente 47 the increase in CTX-I found in their study. Nonetheless, future studies should investigate whether this is replicable to women and pediatric population. Furthermore, each adolescent performed their trials in the same momement of the day, with all of them completing their trials between 08:00 and 11:00 in the morning, after having a standardized breakfast. In conclusion, our study shows that HIIR had a positive but small effect on the formation marker P1NP in adolescent females, while HIIC did not affect this outcome. Despite that, both modes did not change the circulating level of the anabolic osteokine OPG nor the catabolic RANKL, while a transient increase in sclerostin was found significant 5 min post-HIIR and HIIC. Overall, HIIR seems to be the most appropriate intervention for adolescent females regarding its potential benefit for the bone, but more importantly, because it was shown to be more enjoyable and less unpleasant for them. Universidade Estadual Paulista “Júlio de Mesquita Filho” Campus Presidente Prudente 48 Universidade Estadual Paulista “Júlio de Mesquita Filho” Campus Presidente Prudente 49 4.5. Acknowledgments This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC) grant to Panagiota Klentrou (2023-03572) and two grants/scholarships from the Sao Paulo Research Foundation to Pedro Narciso (2022/