Relationships Between the Body Image and Physical Activity on Spanish Adolescents

AUTOR: Manuel Dorado Guzmán http://orcid.org/0000-0002-8579-7367 EDITA: Publicaciones y Divulgación Científica. Universidad de Málaga Esta obra está bajo una licencia de Creative Commons ReconocimientoNoComercial-SinObraDerivada 4.0 Internacional: http://creativecommons.org/licenses/by-nc-nd/4.0/legalcode Cualquier parte de esta obra se puede reproducir sin autorización pero con el reconocimiento y atribución de los autores. No se puede hacer uso comercial de la obra y no se puede alterar, transformar o hacer obras derivadas. Esta Tesis Doctoral está depositada en el Repositorio Institucional de la Universidad de Málaga (RIUMA): riuma.uma.es Didáctica de las Lenguas, las Artes y el Deporte Facultad de Ciencias de la Educación Relationship between Body Image and Physical Activity on Spanish Adolescents. Relación entre la imagen corporal y la actividad física en adolescentes españoles. TESIS DOCTORAL Manuel Dorado Guzmán Dirección de Tesis: Dr. D. Elvis Álvarez Carnero Dr. D. José Luis Chinchilla Minguet Málaga 2016 Autorización de los Directores de la Tesis para su presentación El Dr. José Luis Chinchilla Minguet y el Dr. Elvis Álvarez Carnero como Directores de la Tesis titulada: “Relationship Between the Body Image and Physical Activity on Spanish Adolescents. Relación entre la Imagen Corporal y la Actividad Física en Adolescentes Españoles”. Por Manuel Dorado Guzmán en el Departamento de Didáctica de las Lenguas, las Artes y el Deporte, autorizan su presentación a trámite dado que reúne las condiciones necesarias para su defensa. En Málaga, a 09 de Noviembre del 2015 Dr. D. José Luis Chinchilla Minguet. Dr. D. Elvis Álvarez Carnero A la que fue ( ) y a la que es la mujer de mi vida, A mi Madre y a mi Mujer. AGRADECIMIENTOS Me faltan palabras y no sabría redactar mi gratitud a mi director de Tesis Dr. D Elvis Álvarez Carnero por hacerme creer desde el principio que podía conseguirlo, por animarme en momentos de dudas y sobre todo por saberme guiar en el camino de la investigación y el aprendizaje, para mi todo un ejemplo docente e investigador. De corazón, mil gracias. A mi director Dr. D. José Luis Chinchilla Minguet por ofrecerse desde el principio a dirigir mi trabajo y ponerme en contacto con las personas que podían ayudarme, gracias. A mis compañeros del grupo de investigación GEOS, Lorena, Ramón Alvero, Jesús Barrera… por ofrecerse de forma desinteresada a compartir momentos, conocimientos y cualquier detalle necesario para facilitar todo este trabajo, en especial a Norma que en estas últimas semanas ha sido una gran compañera de tablas, figuras, capítulos etc. A todos los alumnos y alumnas que se prestaron con sus esfuerzos a compartir su tiempo en mi trabajo. A todos los profesores y profesoras que ayudaron en la organización del estudio. A mis amigos, aquellos que les quito tanto tiempo y que a pesar de todo son el mejor brazo donde agarrarme, tendré siempre una deuda con vosotros. A mi padre Manuel, y mis hermanos, David, María del C. y Elena, por ser de forma invisible un referente permanente en mis decisiones, en mis actos y en mi aprendizaje. Os quiero. A mi madre, la persona que mas orgullosa estaría en estos momentos de vivir cada paso a mi lado, de mi mano, por ti hago cada día estas pequeñas cosas. A mis dos hijas Marta y Manuela. Pequeñas de edad y grandes de corazón, nos quedan muchos momentos para disfrutar de la vida juntos. A ti, mi mujer, que desde que te conocí te has volcado en mí, me has permitido crecer a tu lado y no te has quejado de mis inquietudes, mis defectos y mis debilidades. Nunca podré agradecerte ni en esta vida ni en la otra todos tus esfuerzo para que yo y tus hijas seamos una gran familia. Gracias por dejarme ser parte de tu proyecto de vida, por creer en mi y por no dejar que mis manos vayan solas. Para ti, todo esto, toda mi vida. Table contents. Table. 1. Summarize of cross-sectional studies 59 Table 2. Cont. Summarize of cross-sectional studies 60 Table 3. Summarize of follow-up studies 61 Table 4. Summarize of Trials studies 63 Table 5. Effects practice PA or Sport for BI 64 Table. 6. Gender distribution in the final sample size 82 Table 7. Sample size by socio-demographic characteristics, academic level and out-of school sport practice and school 83 Table 8. Summary of the assessments following the order of execution 88 Table 9. Biological characteristic of the sample. 102 Table 10. Intention-to-treat analysis by differences on morphological variables, age, physical activity and body satisfaction 103 Table.11. Distribution of participants by body satisfaction categories and school 104 Table.13. Mean values of body satisfaction (BSQ scores) across academic level and gender 106 Table. 14. Body dissatisfaction descriptive results 107 Table. 15. Differences between school and OSSP NOSSP 109 Table 16. Spearman’s coefficient correlations between body satisfaction variables and morphological variables, and age 110 Table 17. Distribution of groups with after-school-sport practice (OSSP) and without (NOSSP) by gender 113 Table 18. Comparison of medians between after-school-sport practice (OSSP) and without (NOSSP) groups 114 Table 19. Proportions of after-school sport practice (OSSP) group across perceived BMI categories 118 Table 20. Logistic regression model for predicting body satisfaction 120 Table 21. Changes in morphological variables, total daily physical activity, body satisfaction and body self-perception after1-year of follow-up 149 Figure Contents. Figure 1. Flowchart of selected studies by inclusion criteria 57 Figure 2. Flowchart representing, final sample size, drop-outs and adherence to the study 81 Figure 3. Age distribution in the final sample size 82 Figure 4. Bioimpedance analysis (BIA) equipment 92 Figure 5. Anatomical model drawings 93 Panel 1. Body Dissatisfaction categories across academic level 105 Figura 6. Body Self-Perfection Distortion 108 Figure 7. Mean ranks on body dissatisfaction 111 Figure 8. Relationship between body dissatisfaction score 111 Figure 9. Mean ranks on body dissatisfaction 115 Figure 10. Body satisfaction differences between adolescents with after-school sport practice (OSSP) and without (NOSSP). 116 Figure 11. Differences of BMI between adolescents with Out school sport practice (OSSP) and without (NOSSP) 117 Figure 12. Differences between BMI_A and BMI_P across PBMI 119 Figure 13. Absolute and relative changes of body satisfaction 151 Abbreviations ACSM: American College of Sports Medicine. ANOVA: Variance Analysis. BCA: Body Composition Analysis BSQ: Body Shape Questionnaire. BI: Body Image. BIC: Body Image Concern. BIP: Body Image Perceptions. BD: Body Dissatisfaction. BDD: Body Dimorphic Disorders. BMI: Body Mass Index. BMI_A: Actual Body Mass Index. BMI_P: Perceived Body Mass Index. BS: Body Satisfaction. ED: Eating Disorders. EHC: European Health Council. EYHS: European Youth Heart Study. ERIC: Educational Resource Information. ESO: Education Secondary Obligatory. %FM: Fat Mass Percent. FFM: Fat Free Mass. FFMI: Fat Free Mass Index. FMI: Fat Mass Index. FDR: Fernando de los Rios. FRE: Figure Rating Scale. g: grams. HBSC: Health Behavior School Children. IES: Institute of Secondary Education. JDR: Juan de la Rosa. Kcal: kilocalories. MVPA: Moderate-vigorous Physical Activity. NASPE: National Association for sport and physical education. OSSP: Out of School Sport Practice. NOSSP: Non Out of School Sport Practice. PA: Physical Activity PAQ-A: Physical Activity Questionnaire for Adolescents. PE: Physical Education. SD: Standard Deviation. USA: United States of America WHO: World Health organization. NBCA: Students without Body Composition Assessment. NC: Normal Concern. MOC: Moderate Concern. MIC: Mild Concern. MAC: Marked Concern Introduction 23 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Introduction Since the end of the XX century the relationship between physical activity (PA) and health has been growing in relevance as consequence of several psychological and physiological benefits, which has been associated with a regular practice of PA (ACSM, 1991; Pastor, Y. & Pons, 2003; Taylor, Moore, & Holland, 1985; Van Amersfoort, 2004; World Health Organization, 2005). Although the importance of PA for health was firstly more investigated in adults, nowadays there is not a matter of doubt that it is one of the most important issues for a healthy development during childhood and adolescence. So, it has been suggested that an appropriated and wellprescribed PA program will allow wide range of healthy psychological and physiological benefits such as an improved cardiovascular, respiratory, motor, cognitive, affective and social development (Neil Armstrong & Welsman, 2007). Cardiovascular, energy metabolism and body composition (BC) are the most relevant physiological concerns influenced by PA during adolescence, which may be additionally influenced by psychological alterations during the puberty. As consequence PA has been proposed as health markers during childhood and adolescence (Ortega, Ruiz, Castillo, & Sjöström, 2008; World Health Organization, 2005; Held et al., 2012). Nevertheless, several theories and/or inferences associated with PA during this part of the vital cycle remain to be confirmed. The fact that children and adolescences are generally in a good health status, it has been Introduction 24 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS one of the main limitations to study health consequences associated with absence of PA. However the importance of PA behavior goes above and beyond its early benefits and will affect adulthood health. In spite of this growing scientific knowledge of PA-related benefits for youth, research evidence has reported a decrease in participation in PA during adolescence (Kimm et al., 2000; Telama & Yang, 2000; Ritterman Weintraub, Fernald, Adler, Bertozzi, & Syme, 2015) and an increased rate of sedentary in developed countries such Spain (Ortega, F.B., Ruíz, J.R., Castillo, M.J., Moreno, L.A., González-Gross, M., Warnberg, J., & Gutiérrez, 2005; Fernández García et al., 2008; Hernández de Vera, n.d.; Serra, 2008; Martínez-gómez et al., 2009; WHO, 2010; Martínez et al., 2011; López, 2008.; Labrado, 2011). As consequence, rates of overweight and obesity continue to be above 40% in United States and several countries in Europe, including Spain (WHO, 2010; Flegal, Carroll, Kit, & Ogden, 2012; Mura G, Vellante M, Nardi AE, Machado S, 2015). Moreover, it is paradigmatic that young people are concerned with weight control and body image (BI) as they become older (F. M. Berg, 1999; Cash, T. Pruzinsky, 1990; L Smolak & Levine, 2001), which are problems not only related with the excess of adiposity but also with bulimic syndrome and anorexia. Education for a healthy lifestyle may provide useful tools for the future life. In this regard, physical education (PE) classes offer good opportunities to modify and improve attitudes about election of healthy diet and PA plans among children and adolescents, which may be a helpful “lifelong Introduction 25 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS learning”. However, most part of the research related with health benefits of PA among pediatric and adolescent population has been traditionally conducted out of the school time (Blair, Cheng, & Holder, 2001) which had to do with a physiological performance perspective for the success of PE (Simons-Morton DG, 1987). Currently, this approach has been modified and PE subject covers a wider range of topics and concerns and its outcomes are not only physiological but also psychological and social, and most of the times eclectic. Metabolic and psychological disorders such as obesity and anorexia are serious preoccupations in Spain and around the world, which can be susceptible to be detected and treated early in the adolescence. So, identifying risk factors and means of treatment must be a main goal for adolescences education. Body self-perception variables have been proposed as risk factors to develop several behavior disorders, which affect BC (Ferron, Narring, Cauderay, & Michaud, 1999). BI and body dissatisfaction (BD) have been two of the most important variables studied in the field of body weight disorders. In adults, BD affects more females than males (Carlson Jones, 2004; Davison & McCabe, 2006; Miranda, Conti, de Carvalho, Bastos, & Ferreira, 2014) and BI distortion is more prevalent in girls than boys, and this is probably related with a social model for females mainly related with thinness (Bener & Tewfik, 2006), which has made in the past the most part of the studies had been conducted in females. Nevertheless, boys may experience their specific difficulties during adolescence. Theoretical framework 32 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 1999) suggested several terms such as BS, body ideal size or body size perception in order to define more precisely the BI concept. Others have even suggested that a networking of several BI´s (Cash, T. Pruzinsky, 1990). However, most of the times the term scientists use is more related with the component of the BI construct they want study or the technique they use to assess BI. As a consequence, it is difficult to summarize the BI concept just in a definition and nowadays it seems that a multidimensional theoretical construct is the most accepted way to define BI. Considering BI as a multidimensional concept, the most recent literature defines BI based in two main constructs (Burrowes, 2013): 1. Body perception. This is an individuals’ self-assessment of the physical proportions of their bodies and the extent to which this assessment is accurate. This capacity is well described in terms of normative references and in the worst scenario individuals suffer from body dimorphic disorder (BDD), a psychological disorders where by people cannot accurately perceive their body size. 2. Body satisfaction. This is the extent to which an individual is content with their body size and shape. Incorporated into this theme are terms such as body confidence, body esteem, and body dissatisfaction. In this way measuring body perception or defining BS can assess BI. Body perception assessment relies more in quantifying the actual whole body shape or size and contrasting it with the self-assessment (see assessment Theoretical framework 33 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS tools section). Regarding BS, it is more focused in specific parts of the body and qualitative and subjective aspects such as appearance. Alterations in one or both dimensions may take to BI disturbances and finally life-threating psychological disorders as anorexia or morbid obesity. BI disturbances are organized in two categories distorted patterns of thinking (BI distortion, BD, BDD or distorted body perception) and maladaptive behaviors (Reas & Grilo, 2004). This thesis is in the scope of BI disturbances and does not include the study or assessment of serious syndromes (i.e. bulimia nervous, anorexia and so on). Specifically, we are concerned with BD disturbance and body perception in healthy adolescents. 1.2. Body Satisfaction and dissatisfaction.- BS, BI perception and health status are not always in agreement. Two scenarios of BI disturbance could be plausible; firstly individuals with BD, unhealthy BC (for example excessive thinness) and distorted body perception (overestimation of actual weight status); on the other had, individuals with BS, unhealthy BC (for example obesity) and distorted body perception (underestimation of actual weight status). Both syndromes may interact and be affect by PA and diet, although they are commonly designated as eating disorders (ED) and are far to be an exclusive problem of adults. Adolescents are a risk population for these disturbances since they experience ample physical changes during the period of puberty and Theoretical framework 34 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS face dynamic perceptions of BI (Alwan, Viswanathan, Paccaud, & Bovet, 2011; Swaminathan, Selvam, Pauline, & Vaz, 2013), which are additionally influenced by media and social stereotypes with a deep impact in BS and psychological wellbeing among adolescents (Durkin & Paxton, 2002; Groesz, Levine, & Murnen, 2002). Currently, this situation growing in importance and a high rate of adolescents are worried with BI issues, which could be in the origin of ED (Camacho MJ, Fernández E, 2006). Regarding the gender, BI distortion is more prevalent in females than in males (Carlson Jones, 2004; Davison & McCabe, 2006; Miranda et al., 2014), so the former maybe more affected by social stereotypes such as thin beauty ideal and thinness (Groesz et al., 2002) and as consequence it seems that eating-related disorders at earlier ages (Bener & Tewfik, 2006). Although, BD it is not exclusive of female, and different social models may influence male adolescents, mainly those related with a mesomorfic BC and it has been suggested that males show more self-esteem, body shape and physical condition (Fernandez, J.1999) and females present higher body mass index (BMI) (values) during adolescence than males (Cole, Bellizzi, Flegal, & Dietz, 2000). However, girls have a leptosomic BC ideal pattern of and health care concerns (Striegel-Moore, 1997). All together, these previous characteristics make women tend to overestimate their body weight, specifically in waist, breast and head regions in order to adjust their actual figure to beauty expectancies, which enable more probabilities of BD (Lowery et al., 2005) (Ansari, Dibba, & Stock, 2014). Theoretical framework 35 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 1.3. Non-modifiable Risk Variables Associated with Body Image Development.- 1.3.1Gender Traditionally, BI is a more significant issue for life women than men (Koff, Rierdan, & Stubbs, 1990; V, 1999) and as consequence the former have given more importance to physical appearance and weight management (Goñi, A. y Zulaika, 2000). Although, BI care is gaining interest progressively among men and health-threating risk behaviors have been already described nowadays (Cafri et al., 2005), however it seems that women continue to be more influenced by social model of beauty based in an extreme thinness (Franzoi & Shields, 1984). Across the vital cycle, females are at higher risk for negative BI than males (Elgin & Pritchard, 2006; Feingold & Mazzella, 1998). In general, when the most port of female with BI concerns are dissatisfied because they feel overweight (Thompson, J. K., Heinberg, L. J., Altabe, M., & Tantleff-Dunn, 1999), and they would prefer a slender shape, which may be strongly influenced by the dramatic changes in body fat accumulation in hips since adolescence, which may interfere with their satisfaction later in life (AAUW report, 1992). In contrast, reasons in men are more heterogeneous, with almost 50% indicating they are weight dissatisfied and want to gain weight and muscularity (Cafri et al., 2005; Mellor et al., 2004; Theoretical framework 36 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Marrodán Serrano et al., 2008). This previous female pattern may be related with a lower self-concept among women than men (Freedman, 1984; Brooks-Gunn, 1992; Kearney-Cooke, 1999) which in turn could bring women to higher likelihood of ED. While there is not a strong mechanistic evidence to explain this women pattern, female biology, familiar environment and sociocultural context have been suggested to have influence (Gardner, Stark, Friedman, & Jackson, 2000; Leung, Thomas, & Waller, 2000; Thompson & Chad, 2000). All these differences seem to be exacerbated during adolescence, where girls show an impaired perceptive capacity and difficult in accepting their physical appearance (Corbin, 2002; Marcotte, D., Fortin, L., Potvin, P., & Papillon, 2002). This variability along the cycle vital may confirm age as a risk factor for BI disturbances to bear in mind when working with children, adolescents and young adults. 1.3.2 Age As pointed out in the previous section, age is an important cause for the changes of BI perception and related consequences, and childhood and adolescence (particularly puberty transition) has been described as sensible time periods for the development of disturbances. In children, (Mellor et al., 2004) found about 40% of girls and 25% boys of elementary school are dissatisfied with their body. BD continually increases towards adolescence both females and males, with girls continuing to report higher negative BI than boys (Paxton, Eisenberg, & Neumark-Sztainer, 2006; Muth Theoretical framework 37 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS & Cash, 1997; Zaimin Wang, Byrne, Kenardy, & Hills, 2005) which must be related with physiological and morphological changes that happen during puberty, so while girls start to accumulate more body fat, male gain muscle mass, strength and improved motor skills (González-Montero de Espinosa, M. André, A. García-Petuya, E. López-Ejeda, N. Mora, A. Marrodán, 2010). So, a conflict between pubertal growth and esthetic ideal has been suggested as the critical determinant for the highly prevalent BD among youth (Eisenberg, Neumark-Sztainer, Story, & Perry, 2005; Blaivas, Levine, & Murnen, 2002). Moreover, it seems that BS diminished from 13 to 18 years old to keep steady afterwards (Tiggemann, 2004). Nevertheless, adults continue reporting higher BI concerns than younger populations, which may indicate that BI dissatisfaction during adulthood is mainly determined in adolescence. 1.4. Adolescence as Risk Stage for Body Image Development.- Adolescents are constantly worried about their weight, seeking a beauty ideal, and the lack of acceptance of the body makes them feel excluded from society. As described in the previous paragraphs, BI perception and satisfaction may be developed earlier in life. Recent evidence suggest that BI issues rise since childhood and throughout adolescent, so 55% of girls between 7 and 12 years old wish to be thinner, Theoretical framework 38 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS while among adolescent females the prevalence is 80%, which is also associated with low self-esteem (Singh, Ashok, Binu, Parsekar, & Bhumika, 2015). In the same way, 50% of British females (12-18 years old) would like to lose weight, 60% believed they must modify dietary habits and 20% were performing severe caloric restriction (Furnham A, 1986). In fact, there is a higher percentage of adolescents insatisfaction with their BI than satisfied (Gómez Peresmitré, 1995; Pineda. G, 2000). Many of the studies that found a link between negative BI (both BS and body perception) and depression were conducted with adolescents. The relationship between depression and BI was found for both girls (Pesa, Syre, & Jones, 2000) and boys (Cohane & Pope, 2001), although girls are more likely to experience depression along with negative BI than boys (Sujoldzić & De Lucia, 2007). Adolescents who are overweight are more likely to experience low BS and depression (Chaiton, Cohen, O’Loughlin, & Rehm, 2009). Research indicates that low BS may be the link that explains this association between weight and depression. (Pesa et al., 2000) found that there was no statistical relationship between weight and depression when the influence of BS scores was controlled, implying that BI is important in understanding the relationship between weight and depression. In terms of other psychological factors that are related to BI, (Izgiç, Akyüz, Doǧan, & Kuǧu, 2004) found that individuals who had previously suffered from a social phobia were at increased risk of suffering from Theoretical framework 39 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS lower BI. (Cafri et al., 2005) found that ‘internalizing the thin idea’ (taking on board societal thin ideals as standards for yourself) were key components of negative BI. Increased levels of social comparison (an increased tendency to evaluate yourself by comparing yourself to others) is linked to an increased risk of low BS (Myers & Crowther, 2009). In a study that explored the interaction between parental and peer relationships with BS (Holsen, Jones, & Birkeland, 2012) found that good quality relationships with parents and peers were associated with higher BS whereas poor quality relationships with parents and peers were associated with lower BS. The quality of these relationships also predicted BS over time, with individuals who had good quality relationships in childhood also experiencing higher BS in adulthood. Many authors highlight the importance of the role of parents when considering any attempts to improve BI amongst young people. Mothers and fathers are able to influence BI in positive directions through words of encouragement and negative directions through criticism (Rodgers et al., 2009). This is especially the case for girls (Crespo, Kielpikowski, Jose, & Pryor, 2010; P. van den Berg, Wertheim, Thompson, & Paxton, 2002). Parents are also able to influence the BS of their children by acting as role models. In a study focusing on the impact of mothers’ attitudes towards their own bodies (van den Berg et al., 2010) found that mothers who are concerned about their weight are more likely to have daughters who are dissatisfied with their bodies and more likely to have both daughters and Theoretical framework 40 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS sons who try to control their weight. Girls who miss-perceive their actual body shape or size, and estimate that they are heavier than their actual weight, are more likely to use extreme weight control behaviors such as vomiting (Liechty, 2010). Lower BS appears to increase the chances of adolescents using unhealthy weight control behaviors (such as crash dieting) that are likely to result in weight gain and poorer overall health (Neumark-Sztainer et al., 2006). These papers all indicate that BI may have a key role in predicting risk of ED behavior. 1.5. Media and Body Satisfaction.- One of the areas that have received a great deal of research attention is the impact of the media on BI. This area typically explores the impact of images of thin and attractive women, or men with well-defined muscles, on those who watch them. Several methodologies have been used to investigate this phenomenon, although this topic is beyond the scope of this manuscript, we brifley summarized the most important. Traditionally studies have used images of ‘idealized body shapes’ but do not source these images directly from the media. Other methods include asking participants to report how many hours a week they expend watching TV to be related with BS. Finally, Theoretical framework 41 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS structurated intervew are used to describe pressure teenagers felt from media in order to aim a perfect body shape. (Van Vonderen & Kinnally, 2012) (Benowitz-Fredericks, Garcia, Massey, Vasagar, & Borzekowski, 2012). The majority of research indicates that face to idealized BI can result in a small to moderate reduction in BS and body perception (Grabe, Ward, & Hyde, 2008). This finding has been reproduced studies using both male and female participants (Barlett, Vowels, & Saucier, 2008), although we cannot considere it as an absolute truth. So, some studies have failed to replicate the finding and have found that exposure to idealized BI has the same impact as being exposed to images of inanimate objects instead (Holmstrom, 2004). In instance, women who are only slightly bigger than the models used in the media, exposure to media images improved their BS (Holstrom, 2004). This suggested that for these women exposure to thin images may act as a motivational factor to help maintain lower body weight. The extremity range of the images utilized to assess BI is another that may influence responses from study participants, and it has been only found a relationship between idealized BI and low BS when extreme images were used (e.g. very muscular men) (Barlett et al., 2008) In addition, any pre-existing low BS in the participants appears to have an impact on the results. Individuals who already have low BS are likely to be negatively affected by images of idealized body shapes Theoretical framework 48 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS highlight the importance of BC assessment in order to understand how regional adiposity interferes with BS and PA. 3. Body Composition.- 3.1. Concepts.- BC is a study area of the biology, which refers to each component of the human body. The assessment of BC may be a cornerstone of the physical fitness assessment because it offers the ability to classify a subject with regard to health, development and performance. Additionally, several chronic adaptations that occur as a result of exercise or PA, diseases or development match with changes in BC more tan with body weight (Sardinha, 2012). Specifically BC assessment constitutes a cornerstone to assess the nutritional status in patients suffering with chronic or acute malnutrition and the clinical construct of obesity diagnosis (VALTUEÑA, S.: ARIJA, V. y SALAS, 1996). So there is not matter of doubt in considering BI disturbances a BC issue. The study of BC can be split in five levels of progressive complexity: atomic, molecular, cellular, tissue–organ and whole body. Levels are the first criteria of the hierarchical organization of human BC. The second criteria are the components or compartments of BC. These are several simple entities that can be measured in a level of Theoretical framework 49 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS organization. However, not all components can be measured easily or frequently in vivo. The Molecular level is the most commonly assessed and fat mass (FM) and fat free mass (FFM) are the most important components in this level (ZiMian Wang, Wang, & Heymsfield, 1999). FM measurement is difficult and barely performed in clinical or school settings, however anthropometry and whole-body bioelectrical impedance analysis (BIA) measurements have been developed to estimated FM and FFM in field settings (see methods section). In spite of the FM and FFM are the clinical and biological constructs more valid than body weight in the diagnosis of BI-related disturbances, body weight and BMI continue to be the most important physical constructs associated with BI. All the studies that explored the relationship between BI and body weight found that increased weight was associated with lower BS. Body weight was found to be the strongest predictor of negative BI regardless of gender or ethnicity (Xanthopoulos et al., 2011). The research suggests that overweight people are more prone to low BS compared to people of normal weight. Individuals who are obese are particularly likely to have low BS (Goldfield et al., 2011). Due to the methodology that can be used when studying the relationship between body weight and BI none of the studies are able to provide evidence that increased body weight causes lower BS, or lower BS causes increased body weight. It is possible that for some individuals an Theoretical framework 50 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS increase in body weight is associated with social pressure to lose weight, which leads to a reduction in BS (Cafri et al., 2005). For other individuals low BS may lead to ineffective or unhealthy dieting behaviors that result in weight gain (Goldfield et al., 2011). As a result, boys may become involved in efforts to improve their muscularity. Adolescents are influenced by social and cultural trends and are under the pressure to materialize the ideal body of the culture to which they belong (MJ, 2005). 4. Body Image, Physical Activity and Body Composition during Adolescence: A systematic review.- 4.1. A study Review.- Adolescence is a stage of physical and psychological changes and above all of group acceptance. Social roles are defined and they come into conflict with the current society, especially if they are not able to achieve the effects of the cultural idealism. It is the goal of the school and the physical education (PE) professionals to teach about healthy lifestyles and to make our students aware about the real importance of having a BI satisfaction and acceptance. Therefore, the importance of PE and sport practice must be included in BI and satisfaction research. Theoretical framework 51 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Adolescence period is not only cognitive, but also is related entirely to the subjective valuations of the students and it determines their social status. For adolescents, it is very important how their classmates see them and some would do anything to be part of a group and not be excluded, a situation that can cause them stress, frustration and sadness (Singh et al., 2015). BI is a term defined in literature as “the picture of our body which we form in our mind, that is to say the way in which the body appears to ourselves” (P. Schilder, 1936). It is considered as a crucial element to explain important aspects of the personality such as self-esteem or selfconcept and especially to create tools to help in diagnosing prematurely possible psychopathologies (Baile Ayensa, 2003). BS is the acceptance of each individual of their body, it is a cognitive element, as well as multifactorial, both BS and BD are shown in the reviewed literature a dependence on different variables such as age, gender and the type of the PA practiced (Jaffee, Lutter, Rex, Hawkes, & Bucaccio, 1999; McAuley & Blissmer, 2000). Previous studies revealed low levels of BS in teenagers (Wilkosz, Chen, Kenndey, & Rankin, 2011), however most of the studies that have analyzed the relationship between BI and PA did it on university students and/or young adults (Weaver & Byers, 2006; Williams & Cash, 2001; LePage & Crowther, 2010; Shriver et al., 2013). As cited by (Reel et al., 2007) in their meta-analysis, it is important and urgent to know which is the real situation of adolescents, since it is necessary to establish a common goal during this stage to avoid ED that might appear during puberty. Theoretical framework 52 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS PA, as well as sport practice inside and outside the school context, can help reduce psychic anxiety, improve personal satisfaction and accept a positive BI (Netz, Wu, Becker, & Tenenbaum, 2005; L. a Tucker & Maxwell, 1992; McAuley, Bane, Rudolph, & Lox, 1995). Regular PA leads to both physical and mental health benefits, which can make an important contribution to improve the physical and psychological quality of life (Kopcakova, Veselska, Geckova, van Dijk, & Reijneveld, 2014). So, evidence from the literature has proved an association between PA and BI (Snyder W, Cook M, Nasset E, Karhausen L, Howells G, 1975; Marsh & Peart, 1988; MEYER, 1987; COLLADO FERNÁNDEZ, 2005; Marín Regalado, 2007). Additionally, either sport practice or PA was positively related to other variables affecting BI and physical self-concept (Sonströem, R.J., Speliotis, E.D. y Fava, 1992) and general self-concept (Felker, D. W., y Kay, 1971; Deo, P., y Bhullar, 1974). Therefore the main goal of this review was to analyze the state of art of the literature analyzing the relationship between BI, BS/BD and PA/ sport practice (inside and outside the school context) during the stage of adolescence. Secondly, we explored studies analyzing the plausible relationship between improvement on BI and/or BS and enhancement BC (adiposity markers) mediated by PA. Theoretical framework 53 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS METHODS This review was conducted as a systematized review focused in mapping out and categorizing studies covered the relationship between BI or BS and PA/sport topic. We followed an organized search based on inclusion criteria and an Internet-based bibliographic search. Inclusion Criteria The criteria for including studies in this review were based on those suggested by Cochrane Collaboration (Higgins JPT & Green, 2008) so types of studies, participants, interventions and outcome measurements were defined as follow. Types of studies Cross-sectional and follow-up studies, surveys, randomized trials (control and not control) where BI and/or BS were a main outcome. Types of participants This analysis was limited to studies with adolescents (aged between 12 and 18 years) either in the school, sport or clinical context. Participants included in the studies were from all countries ethnicities around the world. Those studies where the participants were under surgery treatment, blind adolescents or mental impairment were excluded. Theoretical framework 54 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Types of Design All kinds of design were included in the review, however only those, where a measurement of PA was described, were selected as valid. Specifically, in cross-sectional and follow-up studies were needed to report validated questionnaires and monitors/sensors used to quantify PA. Additionally, ad hoc recalls where two or more categories of PA/sport participation were reported was accepted in large-cohorts of epidemiological studies. Finally, randomized control trials (RCTs), randomized trials (RTs) or clinical trials (CTs, for example, interventions with adolescences in hospitals without any control group) were included when a PA or exercise intervention was included in the treatment at least in one group. Types of outcomes Primary outcomes Many of the papers in this review use a BI ‘score’. This score can reflect body perception or BS. When BI was not accurately measured or estimated that studies were removed from this review since including them will have an adverse impact on the accuracy of the findings. There are a number of different techniques for measuring this BI score, including: 1. Self-report questionnaire. These scales include the Body Shape Questionnaire, Body Esteem Scale and the Body Shape Theoretical framework 55 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Satisfaction Scale, Physical Self-perception Profile among others. 2. Figure drawings. Typically when using this method a participant is presented with a series of drawings of body shape and asked to identify their ‘ideal’ body shape or the body shape that they feel best reflects their actual body shape. Figure Rate Scale, Gardner Silluete, Collins’ Child Figure Drawings, Contour Drawing Rating Scale, among others. Secondary Outcomes, - Anthropometric variables: Body mass index (BMI, weight/height2), circumferences and skinfolds. - Body composition components: FM, FFM and FM percent (%FM) assessed by anthropometry, BIA or other techniques based in a molecular-approach. Search Methods An Internet-based search was carried out using the following sources: PubMed, Educational Resource Information Center (ERIC), Web of Knowledge and Dialnet. A combined set of two blocks of keywords was used. The first block included: BI, body perceptions and BD/BS. The second block: PA, sport and exercise. All studies in the final selection included at least one word from each block in the title or abstract. Additionally, studies included participants younger than 12 years and older than 18 Theoretical framework 56 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS years were discarded. Additionally, publications included in meta-analysis and reviews, which were not found in the previous search, were reviewed for eligibility. Finally, the list of references met inclusion criteria, undertook a qualitative analysis for further evaluation since the abstract information could not be enough to describe the study characteristics as valid. So, so full articles were red in order to know if they offer enough information to be included in the final analysis. Briefly, many of those articles have not been included in the final list due to not having enough statistic information, not report physical activity data and duplicate of the same study. Finally, the results from all papers met the inclusion criteria were organized in three categories: cross-sectional analysis, follow-up studies and trials (RTs, RCTs and CTs). Also, classifying each of the previous categories was performed an evidence-based analysis, so a summary of studies found a statistically significant positive or negative effect of PA on BI perception or BS were categorized. In the first stage of the search 2313 articles were found, and after applying the inclusion criteria a total of 35 cross-sectional analysis, 14 follow-up and 2 trials were selected (Figure 1). Theoretical framework 57 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS BI was the main variable in n=39, BD/BS: n=9. A total number of 47.846 participants have taken part in all the reviewed studies. Also, the majority of the investigations were conducted in females, 91,5% in crosssectional; 6,4 % in follow-ups and 2,1% in trials. Regarding the methods used to assessed BI the most common tool was BSQ (n= 2 studies) and other questionnaires (n=31 studies), and FRS was used n=3. PA, energy expenditure and dimensions related with PA or sport participation were mainly estimated using PAQ-C/A (n=4 studies) others research used movement sensors or quantifications of exercise training (n=2 trials). Figure 1. Flowchart of selected studies by inclusion criteria. Theoretical framework 64 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS In the table 5 is summarized overall studies organized by their positive, negative or neutral effect on BI variables. The most part of the studies analyzed proved a positive effect of PA on BI, however almost all are cross-sectional surveys as described previously. Table 5. Table summary of studies reporting positive, neutral or negative relationship between practice PA/sport practice and BI, Table IV. Relationship between PA/BI studies with positive or neutral/negative association. Sport & Physical Activity = Body Image Positive Sport & Physical Activity = Body Image Neutral or Negative Autors Year Autors Year Alley TR. 1991 Shaw. 1991 Rodin & Larson. 1992 McDonald & Thompson. 1992 Tucker & Mortell. 1993 Williamson, et al 1995 Jarfee & lutter. 1995 Petrie. 1996 Croasen & Raymore. 1997 Powers & Johnson. 1996 Parks & Read. 1997 Hubbard et al. 1998 Sonstroem. 1997 Tiggeman & Williamson. 2000 Parks P, et al. 1997 Camacho MJ. 2001 Waaler. 1998 Hausenblas & Symons. 2001 Ebbeck & Gibbons. 1998 Dunca M.J, et al. 2004 Rodriguez et al. 1999 Smolak et al. 2000 Phelps, et al. 2000 Williams & Cash 2001 Martinez et al. 2001 Abellan A. 2003 Katzmarzyk P, et al. 2003 Camacho MJ. 2005 Tornero et al. 2008 Campbell & Hausenblas. 2009 Urrutia S, et al. 2010 Slater M, et al. 2011 Finne C. et al. 2011 Kantanista A. Et al. 2015    Theoretical framework 65 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 4.2 Discussion.- This review was concerned with the relationship between PA/ exercise training and BI perception and BS during adolescence. The main finding of our searching was that the most part of the studies, were from cross-sectional studies. Also, it appeared to be PA and exercise training had a positive effect on BI perception. This result is in accordance with previous evidence reported in young adults, therefore three meta-analyses have been published with a wide range of effect size (Campbell & Hausenblas, 2009; Reel et al., 2007; H. a. Hausenblas & Fallon, 2006). (H. a. Hausenblas & Fallon, 2006) found small sized effects based on 121 studies; on the other hand (Reel et al., 2007) described a moderate effect based on 35 studies. Finally, (Campbell & Hausenblas, 2009) revealed that some exercises resulted in an improved BI perception, they included 57 publications in this meta-analysis. We can consider this relationship as strong evidence in young adults since all three meta-analyses where the exercise effect, as consequence a causality relationship can be strongly suggested. Nevertheless, we cannot conclude the same in adolescents since only 2 publications were conducted in trials. Researches analyzing the effect of age are scarce, we found large effects on older collectives compared to younger collectives; these facts agree with the review made by (Vaquero-Cristóbal, Alacid, Muyor, & LópezMiñarro, 2013), in which they note that researches that have compared children and adolescents with adults show that the issue increases over Theoretical framework 66 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS the years, since the perception of one’s own BI becomes more distorted, however relevant studies come to the conclusion that in the adolescence a 80% of the girls are unhappy about their bodies over a 55% of young female university students. This is especially significant among women, since they are more vulnerable to be influenced by socially established body shapes. (Strasburger, Jordan, & Donnerstein, 2010). Body Satisfaction. BI dissatisfaction may be the reason for choosing PA and exercise as a strategy for obtaining the optimal image, especially in adolescents boys and girls (Mellor et al., 2004; Smith et al., 1998) and inversely, BI dissatisfaction as also related with less engagement in PA. Regardless the association between BI dissatisfaction and PA has not yet been properly quantified, as previous research was carried out only among specific gender or age subgroup. Body Composition. Few studies have used real measurements of BC as a way to control their analysis, despite the fact that it has been proven that the use of real measurements provides more accuracy to the analysis of results. (Fingeret, Gleaves, & Pearson, 2004) (Williamson, Gleaves, Watkins, & Schlundt, 1993). This review does not include those sports practiced by adolescents nor their intensity level, as there are works that separate team sports as a clear positive complement on the student’s BS (Jaffee et al., 1999). On the contrary, (Shaw & Kemeny, 1989) found that the participation in individual aerobic activities, for example physical fitness or aerobics, is associated with a negative BI. Theoretical framework 67 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Gender: Most of the review studies take the female gender as a reference for the analysis of the BI. In our selection of review studies we considered as criterion that each study included males, females or mixed. Since we can obtain important data about the reality of the perception of the BI of each gender, and so be able to analyze the relation between them. It has been written about the previous knowledge that the female gender shows a larger distortion about their BI, hence higher levels of BD (Sallis, Prochaska, & Tayl o r, 20 0 0; A rroyo e t a l. , 2 00 8 ; R i va ro l a, 2 0 03 ; Dav i s, 1 99 1 ). Boys present better levels of perception of their BI and they are displayed as more active toward the practice of sports (Davis, 1991; Lantz, Hardy, & Ainsworth, 1997) and this can be a primary reason as to why their BI is positive and it does not show high anxiety levels for their body, which reduces other possible consequences of an alteration of their BI. (ED, depression, school dropout rate, etc). Physical activity: It is a potential element of our review, knowing the current state of sport practice of adolescent students outside school hours, if they are active or passive regarding their spare time and if said activity/practice can be considered fairly active or very/highly active. The suggested reviews regarding the analysis of BI and PA find two opposite tendencies. In the one group we find the positive inclination of PA and/ or sport practice on the personal self-concept. PA causes a drop in the body fat percentage and a more balanced BMI, which provokes a more appropriate adjustment on the current aesthetic canon. (El Ghoch, Soave, Theoretical framework 68 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Calugi, & Dalle Grave, 2013; Bartlewaki et al., 1996). In the other line we saw those studies that focus on the negative effect on the perception of the BI. Some factors associated with that effect are the need to practice some PA in order to lose weight and to fit into a culturally established aesthetic model, so consequently there is a premise of an unwanted BI, not accepted and unhealthy. In practices such as high performance sports and competitive sports there is an increase of the social pressure over the adolescents as it is associated to a result and connected to a few body requirements for specific sports, with the goal of achieving better results individually as well as in group (H. A. Hausenblas & Fallon, 2002; Camacho MJ, Fernández E, 2006). Regarding the type of sport or PA some studies have found that both aerobic and anaerobic exercises are effective to reduce the worries about BI (H. A. Hausenblas & Fallon, 2002; Caruso & Gill, 1992). There is in fact a big consensus by which the participation in team sports is associated with a healthy BI in opposition of the participation in individual sports. (Parks & Read, 1997; Rodríguez et al., 1999). This review section has explored the state of art of the connection between BI perception and physical appearance/BS, BC and PA. As conclusion, we can summarize that the current excessive concern about BI as a consequence of diverse factors (for example, social pressure and media) is causing a large amount of alterations over the perception, which Theoretical framework 69 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS consequently, in many cases, leads to the use of diets and alterations like eating disorders, as a way of trying to adjust the BI as much as possible to those ideals established by society. There are also other factors that have an influence over the BI and its perception, such as the practice of PA, although the results of the connection of both factors can be contradictory. Due to the big increase these alterations are undergoing in the present society, it is necessary to give an in-depth examination to this topic, and create tools to detect disturbances and go into detail and develop a program for the prevention and intervention for teenagers and young women, as these groups are the most affected by those factors, although alterations of the BI affect people of all ages. Variables, aims and hipothesis 71 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS CHAPTER II. VARIABLES, AIMS AND HIPOTHESIS. Variables, aims and hipothesis 73 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 5. VARIABLES, AIMS AND HYPOTHESIS.- 5.1. Aims.- Considering the lack of Spanish and international studies related with PA, BC and BI in school contexts. It was our purpose to: - Evaluate the BI perception and BD in Spanish male and female adolescents in a school context. - Describe the prevalence and grade of BI distortion and dissatisfaction among Spanish male and female adolescents. - Describe the prevalence of BI distortion and dissatisfaction across three grades of weight status (thinness, normal weight and overweight and obese). - Explore the associations between total daily PA, BI and BD. - Compare differences on BI and BD between adolescents who were involved OSSP and NOSSP. - Estimate determinants or predictors of BD (PA, BC, age and gender). - Evaluate if there will be significant changes in BI and BD after the first year in the compulsory secondary education. Study design 80 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS male and female adolescents between 12 and 18 years old were accepted to participated in the study always they me the next inclusion criteria: - Hand signed written informed consent in. - Not suffer any acute or chronic disease by the time of the study. - Not be under medications that could interfere with BC and perceptive capacity. - Students with partial or complete amputation where permitted to participate but their were not included in the final analysis. - Agree to complete all assessments. Study design 81 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS An schematic representation of the recruitment process and the final sample size is shown in the figure 2. Figure 2. Flowchart representing, final sample size, drop-outs and adherence to the study. FDR, Fernando de los Ríos School; JDR, Juan de la Rosa School. NIC, non informed consent Study design 82 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS A gender and age distribution of the final sample size can be observed in table 6 and figure 3 respectively. Figure 3. Age distribution in the final sample size. Table. 6. Gender distribution in the final sample size. Sex n= % Male 405 49 Female 424 51 Study design 83 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Sample size proportions by gender, ages and academic level and OSSP by school are shown in table 7. Table 7. Sample size by socio-demographic characteristics, academic level and out-of school sport practice and school. FDR, Fernando de los Ríos School; JDR, Juan de la Rosa School; OSSP, Out School Sport Practice; NOSSP, Non Out School Practice. Variables FDR n= 659 JDR n= 168 TOTAL Socio-demographic variables Gender Male 310 98 408 Female 342 78 420 Age ( years) Male Female Male Female 12 24 29 1 5 59 13 67 67 34 30 198 14 92 110 40 22 264 15 88 98 8 3 197 16 38 39 0 11 88 17 2 3 6 1 12 18 2 0 4 3 9 Course Male Female Male Female 1º ESO 33 41 15 19 108 2º ESO 41 31 20 13 105 3º ESO 87 107 42 25 261 4º ESO 149 163 22 21 355 Physical Activity variables OSSP 351 101 452 NOSSP 308 68 376  Methods and protocols 85 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS CHAPTER IV. METHODS AND PROTOCOLS Methods and protocols 87 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 8. METHODS.- Once we obtained students and parents signed written informed consent a schedule of assessments were send to the administrative team of the school. All assessments of BC, PA and body self-perception, were performed at the begging of academic year. A free teaching day was provided to carry out the assessments, and when some of them could not be performed that day, it was permitted to complete it along the same week. An assessment day was as follow, students arrived to the sports complex facility 8:30 a.m. and a general overview about the organization was explained by the leader of research team, in summary the evaluations follow the next order: BC, PA and BI (table 8) Students were in fasting conditions and they were permitted to eat after BC assessment was performed. Each student filled all questionnaires personally and trained researcher was always on-site to provide help and confirm all items were properly answered. This study was part of a larger study (GEOS study stated in 2010, funded by the Spanish government of Economy and Competitiveness, code: DEP2011-30565), which was approved by an institutional board of the Departments of Didactics of Music, Plastic and Corporal Expression, and Sports and Physical Education. Also, all procedures were in agreement with “The Helsinki Declaration of Ethical Principles for Medical Research Methods and protocols 88 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Involving Human Subjects” (World Medical Association 2013) and “Standards and Operational Guidance for Ethics Review of Health-Related Research with Human Participants.” (WHO, 2010). Table 8. Summary of the assessments following the order of execution. Methods and protocols 89 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 9. TOOLS AND PROTOCOLS.- 9.1. Body Composition.- BC was assessed using anthropometric method. All procedures followed the next protocol: All students were in fasting conditions, although water was permitted until one-hour before doing the BIA measurements (except for body weight). Participants refrained from taking tea, coffee, chocolate or any other kind of stimulants; also they did not perform any intense exercises or efforts during the previous 24 hours before the tests. Anthropometry. The height and weight were measured to the nearest 0.1 cm 0.1 kg respectively with a stadiometer (Tanita® Leicester) and scale (Tanita®, model UM-060). An inextensible tape (Rosscraft, Canada) was used to obtain waist (WC) with a precision of 0.1 cm. All measurements were carried out according to the standardized procedures described in the literature (Lohman, T. G., Roche A. F., Martorell, 1988) and in accordance with guidelines of The International Society for the Advancement of Kinanthropometry (Stewart A, Marfell-Jones, Olds, & Al., 2011). Based on the previous measurements BC components were calculated as follow. BMI was calculated using the formula (weight (kg) / squared height (m2)). Bioimpedance Analysis. Individual components of BC (%FM, FM and FFM) were obtained based on a classical two-component model (Carnero & Sardinha, 2015), (equation 1). Weight was obtained from anthropometric measurement and FFM was estimated from BIA measurement. The BIA Methods and protocols 96 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS total score (PA-S) was used to describe weekly PA. Additionally, the item numbers 4 and 5 were utilized to classify adolescents involved in OSSP or not NOSSP. PAQ-A has a good concurrent validity as proved from accelerometer studies (Bervoets et al., 2014; Kathleen F. Janz, Lutuchy, Wenthe, & Levy, 2008); in our study PAQ-A score was positively associated with light, moderate, vigorous, MVPA and steps measured by accelerometry (rho=0.33, rho=0.21, rho=0.39, rho=0.36 rho=0.41, respectively; P<0.001). Regarding construct validity PAQ-A has a good correlation with indicators of fatty tissue, bone marrow, and variances in heart beat as well as certain psychological factors such as participation in competitive sports, corporal satisfaction, and anxiety (Crocker, Bailey, Faulkner, Kowalski, & McGrath, 1997; Brunet & Sabiston, 2009). These later studies may confer external validity (Bell, 2002) to this questionnaire. Additionally, we confirmed the external validity of PAQ-C in a subset of the sample of our project (n=99 boys, n=79 girls), so a positive correlation between VO2Max and the PAQ-A was confirmed (rho = 0.28, P<0.01). Methods and protocols 97 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 10. Statistical Analysis.- Continuous variables were expressed as mean and standard deviation (SD) when normally distributed. Always normality was not confirmed continuous variables were treated as median, ranks and intervals of confidence. Categorical variables were reported as frequencies by school and separately for females and males. The Kolmogorov-Smirnov test was used to confirm the normality of the distribution. Chi-square tests were computed in order to compare rates of students between gender categories, school, education level and OSSP categories to confirm proportionality across all possible combinations of categories. Pearson’s coefficient correlation was used to analyze associations between BI variables and independent variables. Spearman’s correlation coefficient was utilized when non normal distribution was detected. These previous correlations were conducted for the total sample and for NOSSP and OSSP groups adjusted for age and in the same way for BS and BD. These treatments were performed to confirm our hypothesis of relationship between PA, BC and BI. The Mann-Whitney test was used to compare BD (BSQ score) and BD image (difference between measured BMI and perceived BMI) between the two groups of OSSP. Also, BC variables and PAQ-A score were compared between these groups. This treatment may confirm if those students with PA and sport practice out of the school have better BI perception and BC profile. Methods and protocols 98 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS One-way ANOVA was carried to compare differences in PAQ-A and distortion score across different grades of perception (figures drawings). This analysis may help to understand if those students with an extreme body self-perception have more or less distortion and additionally if they were more or less active than those in the central categories. An ANCOVA analysis was applied then to confirm if this differences were dependent of PAQ-A score and FMI (both were covariates). After this, four separate multifactorial logistic regression analysis examined the odds ratios for the levels of the dependent variable (BS and BD) and independent variables gender, PA adjusted for age and BMI for the first model. Afterwards a second model substituting BMI by FFMI and FMI were also carried out. Finally, a related-sample Wilcoxon signed rank test was utilized to compare differences of main variables (BI and PA) at the first and last time points of the follow-up period. All statistical procedures were executed with statistical package SPSS 20.0 and the an type I error or alpha value P<0.05 was accepted to confirm significance for all analysis. A post hoc power analysis allow as to confirm that if we expected to detect correlation coefficient as low as 0.25, which assuming an alpha value of 0.001 and 80% of statistical power can be found with a sample size of 95 individuals. We could recruit 859 participants, which allowed us enough statistical power to confirm or refuse our hypothesis. Results 99 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS CHAPTER V. RESULTS Results 101 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 11. RESULTS.- 11.1 Sample characteristics.- The biological characteristics of the final participants are shown in table 9. Although 829 adolescents accepted to participate in the study we could only have assessment of BC in 293 of them; and 125 completed the two assessments of the follow up study. However, we could confirm that our data were not biased by this fact since our ITT analysis proved that there were not significant differences in the main variables between the group with BC analysis and without it (table 10). Nonetheless, correlation analysis between PAQ-A and BSQ scores showed that only the group with non-BC analysis had significant correlation between both variables (BC group rho = -0.92, p =0.118 vs. NBC group rho = -0.169, p<0.001); as consequence this categorical variable will be included in the regression models, which will be carried out later in the analysis. Additionally, there were not significant differences in age, weight, BSQ, BMI_P or PA scores at first time point between responders and non-responders of the follow-up study. Also, none of the independent variables were normally distributed in the whole sample as confirmed in the Kolomogorov-Smirnof test, so all-statistical treatments were non-parametric, except for those where a normal distribution is not a mandatory requirement. Results 102 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Table 9. Biological characteristic of the sample at baseline. Variables n Mean SD Min Max Median Range Age (years) 829 14.3 ± 1.2 12.0 18.5 14.0 6.5 Height (m) 827 1.63 ± 0.09 1.37 1.87 1.63 0.50 Weight (kg) 828 57.6 ± 12.3 32.0 122.9 55.7 90.9 %FM (%) 293 26.8 ± 8.0 3.3 50.2 26.4 46.9 FM (kg) 293 16.1 ± 7.7 2.1 50.1 14.2 48.0 FFM (kg) 293 42.1 ± 8.0 26.0 79.9 40.3 54.0 BMI (kg/m 2 ) 827 21.48 ± 3.62 15.04 40.27 20.73 25.24 FMI (kg/m2) 292 6.05 ± 2.72 0.69 18.86 5.48 18.17 FFMI (kg/m2) 292 15.78 ± 1.99 11.01 23.10 15.57 12.09 PAQ-A 829 2.4 ± 0.7 1.0 4.0 2.2 3.0 BMI_P (kg/m 2 ) 829 22.0 ± 2.4 17.0 29.0 21.0 12.0 BSQ 826 60.7 ± 23.1 6.0 176.0 56.0 170.0 SD. Standard Deviation; Min. minimum; Max. maximum; Sig. statistical significant; %FM. fat mass; FFM. fat free mass; BMI. body mass index; FFMI, fat free mass index; FMI, fat mass index; PAQ-A, physical activity questionnaire for Adolescents; BMI_P. body mass index perceive; BSQ, body shape questionnaire; BCA, students with body composition analysis; NBCA, students without body composition assessment Results 103 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Table 10. Intention-to-treat analysis by differences on morphological variables, age, physical activity and body satisfaction between students who complete (BCA) and not complete (NBCA) the body composition analysis. Results 104 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 11.2. Body Dissatisfaction and Satisfaction Prevalence.- The prevalence of moderate and marked concerns (dissatisfaction) was 34 students (4.1%). These categories of BS were equally distributed across schools (Chi-square, p>0.05; table 11). Regarding the analysis by academic level of education, a proportional distribution of BS categories across the educational level was observed only for boys (Chi-square = 9.402; p>0.05; Panel 1 figure 1). Nevertheless, higher prevalence for mild and marked concerns categories were detected (Chi-square = 16.984; p = 0.049; Panel 1 figure 2) in the 3rd (3ºESO= 31.8% and 44.4% for MOC and MAC respectively) 4th (4ºESO = 63.5% and 55.6% for MOC and MAC respectively) grades than in the others (1.6% in MOC for 1ºESO and 0% for 1º ESO and 2º ESO in the other categories). Also, MIC was higher in 4th grade (55.4%) than in 1st (11.5%), 2nd (15.4%) and 3rd (17.3%). Table.11. Distribution of participants by body satisfaction categories and school. Variables C. Fernando de los Rios n= 659 C. Juan de la Rosa n= 168 TOTAL No concern 475 139 614 Mild Concern 158 22 180 Moderate Concern 24 4 28 Marked Concern 3 3 6  Results 105 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Panel 1. Body Dissatisfaction categories across academic level (compulsory secondary education) in females (figure 1) and males (figure 2). NC. Normal concern; MOC. Moderate concerns; MIC. Mild concerns; MAC; marked concerns. $, indicates significant different prevalence of MIC with the other groups. *, indicates significant different prevalence of MOC with 1ºESO and 2ºESO. #, indicates significant different prevalence of MOC with 1ºESO and 2ºESO. Results 112 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS All together this results may indicate lower BD (lower BSQ scores) higher underestimation (positive Dif_BMI_A-BMI_P value) or higher BD (higher BSQ scores) higher overestimation (negative Dif_BMI_A-BMI_P value). In the scatterplots of figure we can observe this trend, which was similar either boys and girls (figure 8). Figure 8. Relationship between body dissatisfaction score (assessed by body shape questionnaire BSQ)) and body self-perception distortion (assessed as the difference between actual body mass index (BMI_A) and perceived BMI). Red dots represent boys (n=403) and blue dots girls (n=421). Slopes were not significantly different between boys and girls (b = -1.128 vs. b= -2.146, p>0.05) 112    All together this results may indicate lower BD (lower BSQ scores) higher underestimation (positive Dif_BMI_A-BMI_P value) or higher BD (higher BSQ scores) higher overestimation (negative Dif_BMI_A-BMI_P value). In the scatterplots of figure we can observe this trend, which was similar either boys and girls (figure 7). Figure 8. Relationship between body dissatisfaction score (assessed by body shape questionnaire BSQ)) and body self-perception distortion (assessed as the difference between actual body mass index (BMI_A) and perceived BMI). Red dots represent boys (n=403) and blue dots girls (n=421). Slopes were not significantly different between boys and girls (β = -1.128 vs. β= -2.146, p>0.05)                Results 113 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 11.6. Differences between After-School Sport Practice Groups.- Regarding OSSP groups there was a significantly lower proportion of girls in the OSSP and lower boys in NOSSP (Chi-square = 97.339, p<0.001, table 17). Although none significant differences were found for medians of age (OSSP =14.0 range 6.0 vs. NOSSP = 14.0 range 6.5; mean ranks OSSP = 425.7 vs. NOSSP = 406.12; p = 0.234). Table 17. Distribution of groups with after-school-sport practice (OSSP) and without (NOSSP) by gender. Results 114 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Mean ranks of BC variables were significantly different between OSSP, except for weight and FFMI (see Mann-Whitney U test significance at table 18). Body dissatisfaction was also significantly different and OSSP group report a lower score in the BSQ (table 18 and figure 9) and as expected higher score in the PAQ-A (table 18). Table 18. Comparison of medians between after-school-sport practice (OSSP) and without (NOSSP) groups. OSSP, Out school sports practice; NOSSP, Non-out school sports practice; Sig. statistical significant; %FM. fat mass; FFM. Fat free mass; BMI. Body mass index; FFMI, fat free mass index; FMI, fat mass index; PAQ-A, physical activity questionnaire for Adolescents; BSQ, body shape questionnaire. Results 115 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 11  Figure 9. Mean ranks on body dissatisfaction (as assessed by body shape questionnaire (BSQ)) between adolescents with (green bars) and without (blue bars) after-school sport practice. Difference was significant between groups (MannWhitney U test = 68574, p<0.001). Figure 9. Mean ranks on body dissatisfaction (as assessed by body shape questionnaire (BSQ)) between adolescents with (green bars) and without (blue bars) after-school sport practice. Difference was significant between groups (Mann-Whitney U test = 68574, p<0.001). Results 116 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Regarding gender, girls report higher values of dissatisfaction than boys, although the effect of OSSP was significant for both genders (figure 10). Also, boys in NOSSP classified themselves with similar BD as girls in OSSP group. On average, either boys or girls had values below 75 score. Figure 10. Body satisfaction differences between adolescents with afterschool sport practice (OSSP) and without (NOSSP). Error bars represent median and 95% of confidence interval. *, p<0.05 for Man-Whitney´s U test for comparison between boys and girls. #, p<0.05 for Man Whitney´s pair simple U for comparison between OSSP and NOSSP for total simple. Independent variables: Sex .BSQ, Body Shape Questionnaire. Results 117 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS In the figure 11 we can observe that there are no significant BMI differences between OSSP and NOSSP groups, and this was similar either male or females, who did not differ in BMI inside OSSP or NOSSP groups (figure 11). Figure 11. Differences of BMI between adolescents with Out school sport practice (OSSP) and without (NOSSP). Error bars represent median of international Z-scores using WHO LMS references to calculate Z-scores by sex and age and 95% of confidence interval. In respect to BI self-perception, similar proportions of after-school sport practice were found for lower (17) and upper (27 and 29) extreme categories (table 19, p<0.001 for Kendall’s tau ordinal proportions). It was interesting to note that there were more adolescents in the OSSP group in the BMI groups of those adolescents that perceived themselves as healthy (groups 19, 21 and 23; table 19) and the opposite happened in the highest groups of perceived BMI (25, 27 an 29; table19). Results 118 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Table 19. Proportions of after-school sport practice (OSSP) group across perceived BMI categories. Results 119 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS ESCRIBAELTÍTULODELDOCUMENTO119  1.7. Body Dissatisfaction and Satisfaction Determinants The differences between PBMI and actual BMI ranks were significantly reduced across PBMI level as observed after carting out Wilcoxon´s independent samples test ( p<0.05, figure 9). Figure 12. Differences between PBMI-BMIr across PBMI categories. Error bars represent median and 95% of confidence interval. *, p<0.05 for differences with 2nd, 3rd, 4th and 5th categories, Mann Withney´s pair sample test. #, p<0.05 for differences between 1st and 7th, Mann Whitney’s pair sample test. 11.7. Body Dissatisfaction and Satisfaction Determinants.- The differences between BMI_A and BMI_P ranks were significantly reduced across PBMI level as observed after carting out Wilcoxon´s independent samples test ( p<0.05, figure 9). Figure 12. Differences between PBMI-BMIr across PBMI categories. Error bars represent median and 95% of confidence interval. *, p<0.05 for differences with 2nd, 3rd, 4th and 5th categories, Mann Withney´s pair sample test. #, p<0.05 for differences between 1st and 7th, Mann Whitney’s pair sample test. Results 120 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Logistic regression analysis confirm that PA, gender, age and BC assessment group were significant predictors of BS, so boys, participants with OSSP and those students with BC assessment had significant less probability to report BS. Table 20. Logistic regression model for predicting body satisfaction. Discussion 121 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS CHAPTER VI: DISCUSSION Discussion 128 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS With other Spanish study reporting 46.2% and 37.4% of boys and girls respectively had right BMI self-perception, and 25.3% of boys and 12.3% of girls were thinner than their body self perceptions; although 50.2% of the girls and 28.6.1% of boys were heavier (González-Montero de Espinosa, M. André, A. García-Petuya, E. López-Ejeda, N. Mora, A. Marrodán, 2010). These data are similar to those obtained in our study and may be due to the sample from which this made on Spanish adolescents students. The grade of distortion (difference between perceived BMI and actual BMI) was negatively associated with BMI, which means that higher BMI higher underestimation, and the opposite might be true for extremely low BMI (excessive thinness). Our coefficients of correlation were similar with other studies (Fingeret et al., 2004; A. D. Stewart, Benson, Michanikou, Tsiota, & Narli, 2003) although the variable they analyzed was current body size (CBS) and as consequence their correlations were positive and not negative. The correlation between BMI and CBS is often considerer one of the most important measures of concurrent validity for body selfperception (A. Stunkard, 2000). We could confirm this fact with our BC data, which confirm that more adiposity (FM and FMI) or muscularity (FFM and FFMI) lower BD at least in adolescents. Regarding gender differences, it has been commonly reported that there is more probability women perceive themselves as overweight (Halmi, 1983; Cash & Green, 1986) more worried about being or becoming fat and express greater concern with dieting and body weight (Kelly JT et Discussion 129 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS al 1983). In a figure drawing test, college-age women rated their current figures as heavier than the desired or ideal figure. In contrast, men’s ratings of their current figures corresponded to the ideal (Drewnowski & Yee, 1987). However, we could not confirm this difference between girls and boys. As pointed out in the previous paragraph, we found a high prevalence of alliterated BIP. In Spanish college students 55% of women and 63% of men had a distorted judgment of their body (Míguez Bernárdez, Montaña Miguélez, González Carnero, & González Rodríguez, 2011). Similarly, research across undergraduate college students in Austria, France and USA observed a wide discrepancy between men´s actual body satisfaction and their ideal BI (Pope H.G. et al., 2000). A plausible reason of this high rate of BI distortion among women could be the widely accepted ideal of a slim and fragile body shape makes it difficult for young girls to accept the puberty-related alterations of their bodies (Holsen et al., 2012; Linda Smolak, 2004). So, while 58.3% of women with a normal BMI category rate themselves above their ideal body mass, only 23.5% of men do it. This failure to resemble with an ideal model may cause a high degree of BD, which along with other factors are source for more serious disorders such as anorexia, bulimia and vigorexia (NeumarkSztainer et al., 2006). We did not obtain any valid measurement of ideal body shape so we could not confirm any relationship between current and ideal BI in order to confirm a physical dissatisfaction (difference between Discussion 130 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS current, perceived and ideal BI). Nonetheless, it was important to note that only 30.8% perceived themselves correctly, which is a lower prevalence than that found to BS, which could mean a wrong self-perception does not imply always a body self-perception distortion. Although our correlation coefficient analysis confirm an inverse relationship between BD and selfperception distortion, therefore there would be a possible association indicating those adolescents, whom overestimated their BMI reported higher rates of dissatisfaction; also, a BMI underestimation was associated with lower rate of BD. In fact, we barely observed participants with a BSQ score above 111 points (one case in boys), which is the cut off for moderate or marked BI concerns, among those with positive values of BMI_A-BMI_P (those with higher current BMI than perceived BMI). Additionally, there were a higher percentage of students with OSSP among those that rated themselves in the two lowest categories of perceived BMI (BMI drawings), and the opposite was true for the highest BMI category (table 18). This could be related with the grade of body self-perception distortion since the latter was higher in the extreme categories of perceived BMI (BMI drawings, figure 12) and inversely correlated with PAQ-A score (PA). All together, our data may suggest our adolescents are representative of Spanish body self-perception average, which may not be affected by sex, and is negatively associated with BD (BSQ score) and PAQ-A score; indicating volume of total daily PA could be play a role for diminishing overestimation of body self-perception. Discussion 131 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Physical Activity on Body image variables. As expected from other Spanish and European dataset (Commission, 2010) low regular involvement in PA (55.3%), although there were differences by gender and a higher percentage of males were involved in OSSP (male 71.8% vs. 39.9% female) and the majority of girls (60.1%) were not involved in any OSSP. Those data were in accordance with other study (Jankauskiene & Kardelis, 2005), that showed that only 11.6% of the girls practiced regularly PA, 29.6% were irregular and 58.8% were not engage in any form of leisure PA. In this study the main reason to engage in PA was to improve the BI (45.2%), health improvement (33.6%) and 21.2% reported other motives. Gender influences significantly on the results. Most studies show a difference in PA of about 20 % among boys and girls. Moreover, in some age groups, especially at the end of the school years, the percentage of girls who do not perform any practice of sport and PA rises to almost half of the population (about 47 %), while in the case of children this percentage decreases to approximately 25% (Hernández Álvarez, 2010; Babey, Hastert, Yu, & Brown, 2008). Although some studies have suggested that the amount of PA declines with age (van Mechelen, Twisk, Post, Snel, & Kemper, 2000) we could no find any significant correlation or difference between age or academic level and total daily PA (PAQ-A score), as consequence the relationships between PA and BI variables might not be affected by multicollinearity. Finally, our differences must not be influenced by school environment, Discussion 132 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS since we did not find different proportions of participation by school. The Relationship between BI and PA have been well documented in previous studies (Finne, Bucksch, Lampert, & Kolip, 2011; Fountoulakis, C., & Grogan, 2004). Adolescents with more positive BI may engage in PA because they do not perceive barriers to exhibit their bodies in public settings; this permitted us to speculate that PA may play a role in BIP and BS/BD. Evidence from previous studies appears to suggest that some experience in PA is associated with better BI (Alley, 1991; L. A. Tucker & Mortell, 1993; Parks & Read, 1997; Camacho MJ, Fernández E, 2006; Campbell & Hausenblas, 2009; Urrutia, Azpillaga, de Cos, & Munoz, 2010; Slater & Tiggemann, 2011; Finne et al., 2011). In our analysis by groups of after-school sport and PA practice, significant differences in mean ranks of BC variables were found, except for weight and FFMI. BD was also significantly different between PA practice groups, so OSSP group reported a lower score in the BSQ and perceived BMI. In our knowledge there are not previous studies considering OSSP and NOSSP groups as an independent variable of BI, although there have been published analysis that have considered the sport practice as indicator BS or BD (Jaffee et al., 1999), and it seems that playing on school or non-school sport teams was associated with better BS (Kopcakova et al., 2014) and Body self perceptions (Veselska, Madarasova Geckova, Reijneveld, & van Dijk, 2011). Therefore, it was hypothesized that participation in OSSP and total daily Discussion 133 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS PA has a strong relevance for an enhanced body self-perception and thus a lower likelihood of BD, we could confirm this hypothesis by the significant negative correlation between PAQ-A score and BSQ score, and also from the significant differences on BSQ score ranks between NOSSP and OSSP, which is in accordance with previous studies (Martínez-Gómez D, 2007; Martínez-Gómez et al., 2009). Gender seems to play an important role in the connection between BI and PA (Gillison, Standage, & Skevington, 2011), for example adolescent girls, who are engaged in PA practice perceive the latter as a sort of duty, and their motivation for PA may be related to their physical attractiveness, health benefits and feeling good about themselves, which may be connected with the way they perceive themselves (Gillison et al., 2011). While boys self-perception did not play such an important role as in girls, because their motivation for PA differs from girls. Boys are more engaged in team sports with the aim of being part of peer relationships, which is not connected directly with the way they perceive themselves (Veselska et al., 2011). In our adolescents, BI self-perception had similar proportions of after-school sport practice for lower (17) and upper (27 and 29) extreme categories. It was also interesting to note that there were more adolescents in the OSSP group in the BMI groups of those adolescents that perceived themselves as healthy (groups 19, 21 and 23). However, correlation analysis did not confirm a significant association between BMI_A-BMI_P difference and total daily PA (PAQ-A score), and Discussion 134 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS so we cannot confirm the hypothesis more volume of PA better body selfperception. Our result must be in accordance with findings from previous study, where the individual difference in actual body size and perceived body size or body size discrepancies have potential negative consequences for exercise motivation and this relationship differs among males and females, In contrast, other studies support the notion that participation in sports and activities that have a strong focus on appearance is related to higher levels of BIC (Slater & Tiggemann, 2011; Ingledew, Hardy, & Cooper, 1995; Ingledew & Sullivan, 2002). A probable explanation for the previous disagreement could be related with body weight status and adiposity. So, some authors suggested relationship between PA and body self-perception may be valid only for obese and overweight people, who are more dissatisfied with their body sizes than normal weight people (Schwartz & Brownell, 2004) and reported higher BI scores (Fowler, 1989). Our results supported partially this evidence since significant negative correlations were found between BMI_A-BMI_P and all indexes of adiposity (FM, BMI and FMI). Several factors may have contributed to the high BSQ scores of the obese and overweight subjects. Firstly, since they are not extremely overweight, they may have been experienced to less negative social feedbacks. On the other hand, since the overweight adolescents does not have a large amount of weight to lose. Third, views on ideal weight status as defined by the norms of current society have changed in the last few years, with Discussion 135 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS the trend towards more tolerance and acceptance of the big body shapes. The association between BID and PA has not yet been properly explored in quantitative research, and there is little evidence on the association between BI and PA in boys still. Inferences from previous studies (Kantanista, Osiński, Borowiec, Tomczak, & Król-Zielińska, 2015; Monteiro et al., 2011) concluded that higher levels of PA in adolescents had a protective effect on BD, which is independent of BMI or gender. We could confirm this evidence in our correlational analysis by an inverse relationship between PAQ-A and BSQ scores. Additionally, the logistic regression analysis, confirm PA participation is main determinant of BS independently of gender and age. Nonetheless, past research revealed that physical abilities must be the most important dimension of BS for boys (McCabe, A. E., Roberts, B. T., & Morris, 1991) and BMI is positively associated with BD (Newman, Thompson, & Roberts, 2006; Wardle & Cooke, 2005), as confirmed too in this sample of adolescents. In this line, it has been found participants between 9-23 years old, whom performed more PA and showed better self-concept of them (Moreno, Cervelló, & Moreno, 2008). Sport and leisure time activities are not part of the lifestyle of the majority of the Spanish females and the value of PA or sport for BI perception is not well understood. It has been speculated that bodyrelated perceptions may have lower importance than ability and fitness in determining female adolescents’ PA behaviors (JM Moreno, ACM Baena, Discussion 136 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 2012). As consequence, information on the main difficulties to practice PA during leisure-time physical activities may provide clues to implement institutional actions aiming at the increase of this practice, preferably with the use of spaces at out-of school of time. BD was strongly related with PA variables, conversely BIP was not associated with PA in the whole sample, although OSSP was more prevalent among students, whom qualified themselves in healthy BMI groups, this does not imply any positive relationship. The correlation analysis by gender confirms a significant inverse relationship between BIP and PAQ-A score for boys (rho= - 0.155) but not for girls. This lack of association between PA and BI perception in girls could be explain by feelings of alienation arising from being unable to conform to a desired physique and conflict between sport participation and female ideology (N Armstrong & McManus, 1994). Also, early pubertal maturation may lead to low PA for variety of reason including both intrapersonal factor (e.g. body self-esteem, depression and perceived skillfulness) and interpersonal factor, which could be a reason to explain that early-maturing girls report a poorer BI perception than later maturing ones (Graber JA, Brooks-Gunn J, 1999). This finding could also be supported by having less discriminated function of body attractiveness for PA of girl (Hagger, Basil, & Stambulova, 1998). Discussion 137 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Body Mass Index, Body Image and Physical Activity A paradigm for a PA-related improvement in BI variables (BIP and as consequence BS) could be suggested by the influence of PA in BMI or BC variables. So, a first simplistic hypothetical mechanism would be more PA healthier BMI or BC, as consequence improved self-perception and BI satisfaction. On the other hand, exercise and PA has been proved to influence the activity of proprioceptive sense, which may help to perceive accurately body shape and body weight (Proske & Gandevia, 2012). We could not obtain any neurophysiological measurements to confirm these mechanisms, but a relationship between BI variables and BMI/BC adjusted by PA score and BI and BMI differences between OSSP groups may be a first step to prove this hypothesis. Additionally, if this mechanism is just related with BMI or BC analysis will add up additional information, it is also part of this discussion (next section of the discussion). The association between BMI, BI and PA has been widely reported along the vital cycle (Mullan & Markland, 1997; Wilson, Tripp, & Boland, 2005) as higher BMI higher BSQ score. In the vast majority of studies, where a positive effect between exercise and BI was reported, PA was the best predictor of BI Satisfaction for male adolescents (See in Table 20). Our BMI values were positively correlated with BS for both males and females, and negatively with BMI self-perception. These finding were consistent with the previous research that indicated a negative relationship between BMI and BI such as BMI perception (McCabe & Ricciardelli, 2003; Morano, Discussion 144 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS population (Elgin & Pritchard, 2006; Feingold & Mazzella, 1998). In our study, gender was a main determinant of BD, so adjusted odd ratios analysis indicated boys had 2.16 times higher probability of being classified as body satisfied than women. Age was another significant determinant of BS, with a positive odds ratio, which indicate that every year adolescents had 2.16 more probabilities to qualified themselves as body satisfied, this may be in agreement with our data of BD prevalence, where females has higher proportion of BD in the 4th level of education than the lower levels, but males did not showed this tendency. The group with BC analysis had a significant ß coefficient and was included in the BS prediction model, however we cannot give any rational explanation for this fact. Finally, PA was the most important determinant of BS, which was independent of BMI, age and gender. The mechanistic model that explains these results remains to be described, and may be an interesting starting point for future research. These age differences were also confirmed after controlling for BMI, thus suggesting that other factor were involved in these changes during the transition from childhood to adolescence. In this stage of life, not only physical transformation of the body, but also emotional changes, and social expectations and responsibilities can modify self-perceptions, attitudes and behavior in adolescents (Waylen, A., &Wolke, 2004). Inverse Hypothesis In this study we suggest PA as independent variable and BI variables Discussion 145 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS as dependent, however this is not always accepted in the literature and an opposite relationship has been suggested. BI could also identified as a barrier to exercise, negative BI has been described as a primary reason to exercise participation (Brudzynski & Ebben, 2010; Slater & Tiggemann, 2011). For example, (Biddle & Fuchs, 2009) reported that BI and appearance were important for PA participation of female adolescent. Additionally, female adolescents may need of social support for PA participations, in accordance with (Dowda, Dishman, Pfeiffer, & Pate, 2007) social support from significant others during the adolescence period may be an influencing factor for PA of females. However, we did not analyze this hypothesis, which requires a different study design and instruments, but open interesting future research line. Our results are especially important since it has been shown that the health status, perceived BI and self-esteem are significantly and positively associated (Urrutia et al., 2010) that PA and sports are means improve the health of the subject and prevent obesity (Katzmarzyk, Janssen, & Ardern, 2003) and that these practices have a positive effect on the physical aspect and the pleasure associated (Alley, 1991). Longitudinal study 147 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS CHAPTER VII: LONGITUDINAL STUDY. Longitudinal study 149 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 13. Longitudinal Changes in Body Satisfaction and Body Self-perception.- As expected age, height and weight changed after a year of followup. Conversely, BMI and total daily PA as estimated from PAQ-A score did not modify (table 21). Regarding BI variables, BS was significantly diminished as assessed by BSQ and students rated significantly lower at first year than at second year of assessment their body self-perception (table 21). Table 21. Changes in morphological variables, total daily physical activity, body satisfaction and body self-perception after 1-year of follow-up. BMI, Body mass index, PAQ-A, Physical Activity Questionnaire for Adolescents; BSQ, Body Shape Questionnaire; BMI_P, Body mass index perceive. Variables 2011 2012 Wilcoxon Signed Ranks T. Median Range Median Range - Ranks + Ranks Sig. Age (years) 13.0 6.0 13.5 6.0 0 123 *** Height (m) 1.6 0.5 1.6 0.5 5 120 *** Weight (kg) 49.2 50.9 55.0 62.5 14 108 *** BMI (kg/m 2 )19.8 17.4 20.2 16.7 72 53 - PAQ-A 2.8 2.9 2.8 3.2 60 60 - BSQ 48.0 75.0 62.0 92.0 35 63 *** BMI_P (kg/m 2 )21.0 14.0 21.0 10.0 53 37 * Longitudinal study 150 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS After calculating absolute and relative (%) changes for BS (BSQ score) and PA (PAQ-A score) and split the results by gender, we could confirm that girls increased BD (higher score higher dissatisfaction) changed significantly either as percentage or absolute values. Moreover, these increments were significantly higher in girls than boys, whom only increase significantly the relative change (figure 13). Regarding changes in PA, either boys or girls did not modify significantly their absolute total daily PA (difference PAQ-A 2011-2012 = 0.155 ±0.841 and -0.172 ±0.728, girls and boys; p>0.05 for both), but girls increase their relative PAQ-A score (10.3 ±38.3%; p<0.05). Nevertheless, change differences between boys and girls were statistically significant both absolute (0.327 ±0.140; p<0.05) and relative (13.7 ±6.2%; p<0.05). None significant correlations were found between changes in PAQ-A score and any BI variable. Longitudinal study 151 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS 19   Figure 13. Absolute and relative changes (%) of body satisfaction (assessed by body shape questionnaire (BSQ)) after 1-year followup in adolescents. Error bars represent means and standard deviation. *** and *, indicate p<0.001 and p<0.05 respectively for one sample T-test between the mean change and “0”. ###, indicates p<0.001 for paired sample T-test between boys and girls. Figure 13. Absolute and relative changes (%) of body satisfaction (assessed by body shape questionnaire (BSQ)) after 1-year follow-up in adolescents. Error bars represent means and standard deviation. *** and *, indicate p<0.001 and p<0.05 respectively for one sample T-test between the mean change and “0”. ###, indicates p<0.001 for paired sample T-test between boys and girls. Longitudinal study 152 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS Finally, we tried to explore associations between changes in BSQ and total daily physical activity but we could find none association among these variables. 14. DISCUSSION LONGITUDINAL STUDY.- Regarding de longitudinal analysis, the main findings were we observed adolescent reported significantly higher BSQ scores and lower BMI_P categories at second year than at the first. Moreover, these changes different between genders and only boys had a significant increase in absolute values; both scores were different when expressed as a percent of change, but the effect size in boys was higher than girls. Since our participants, were in the pubertal transition we can speculate that some hormonal and sociocultural variables maybe responsible for this increase in BD (BSQ score), however we had not any assessment of these variables, so we cannot confirm this hypothesis. In the cross sectional study, we describe that PA was a significant determinant of BD, which is in accordance with a meta-analysis of 57 studies (Campbell & Hausenblas, 2009), where PA intervention prove to influence changes in BD, however the changes in PA-Score was not associated with changes in BSQ, which may confirm that the quantity of PA assessed by PAQ-A did not explain changes in BD. This results do not may invalidate the utility of PA to manage BD or be in disagreement with the meta-analysis, since our adolescents did not change significantly PA. Longitudinal study 153 RELATIONSHIP BETWEEN BODY IMAGE AND PHYSICAL ACTIVITY ON SPANISH ADOLESCENTS BMI_P was reduced significantly, which means that they perceive themselves thinner although their BMI had not changed after the one-year of follow-up. The mechanism and reasons related with this misperception are not completely well described but wishes of thinner body shape may be plausible. This results is partially supported by previous studies, which suggests social issues as the origin of eating-related disorders and other psychological disturbances well-documented during adolescence, which are usually more prevalent among girls (Siegel, Yancey, Aneshensel, & Schuler, 1999; ter Bogt et al., 2006). This difference is kept even in athletes, so in (Parks & Read, 1997), 40% of females athletes from different sports overestimated 2 drawing levels with respect to their actual BMI, while that happened in 3% of males only. Considering our results we could confirm that the alteration in BMI_P and BD may be closely related. It was still paradoxical that BSQ modification was higher in boys than girls and PA was increased significantly only in girls, and the later had greater positive changes than boys. These results are in disagreement with other studies, where a decrease in the PA practice during adolescence (Hallal, Victora, Azevedo, & Wells, 2006; Gómez-López et al., 2010) was described mainly in girls (Kopcakova et al., 2014; Butt, Weinberg, Breckon, & Claytor, 2011). This could help to explain the lower levels of BSQ score in girls, and we hypothesize that could be related with the aim of preserve the weight and BI during the pubertal transition during adolescence. However, it has been described that the desire of an improved BI must