The role of body image in internalizing mental health problems in Spanish adolescents: An analysis according to sex, age, and socioeconomic status

fpsyg-10-01952 August 21, 2019 Time: 17:31 # 1 ORIGINAL RESEARCH published: 22 August 2019 doi: 10.3389/fpsyg.2019.01952 Edited by: Antonio Cepeda-Benito, The University of Vermont, United States Reviewed by: Siân McLean, Victoria University, Australia Amanda Nerini, University of Florence, Italy *Correspondence: Pilar Ramos [email protected] Specialty section: This article was submitted to Eating Behavior, a section of the journal Frontiers in Psychology Received: 19 April 2019 Accepted: 08 August 2019 Published: 22 August 2019 Citation: Ramos P, Moreno-Maldonado C, Moreno C and Rivera F (2019) The Role of Body Image in Internalizing Mental Health Problems in Spanish Adolescents: An Analysis According to Sex, Age, and Socioeconomic Status. Front. Psychol. 10:1952. doi: 10.3389/fpsyg.2019.01952 The Role of Body Image in Internalizing Mental Health Problems in Spanish Adolescents: An Analysis According to Sex, Age, and Socioeconomic Status Pilar Ramos1*, Concepción Moreno-Maldonado1, Carmen Moreno1and Francisco Rivera2 1Department of Developmental and Educational Psychology, Faculty of Psychology, University of Seville, Seville, Spain, 2Department of Experimental Psychology, Faculty of Psychology, University of Seville, Seville, Spain During adolescence there is a relatively high prevalence of weight problems and eating disorders. Furthermore, body image plays an important role in weight control and eating behaviors as well as in mental health. This study analyses the influence of body mass index, perception of being overweight, and body image satisfaction (BIS) on internalizing symptoms related to mental health in adolescents. In addition, sex, age, socioeconomic status (SES), dieting, and physical activity are taken into consideration. This research is based on the international study Health Behaviour in School-aged Children (HBSC). The sample consists of 4531 Spanish adolescents from 13 to 18 years old. Participants were selected through random multi-stage sampling stratified by conglomerates. Two instruments were employed: the HBSC questionnaire and the Youth Self-Report (Achenbach System of Empirically Based Assessment, ASEBA). Results demonstrated that BIS –the emotional component related to body image– was the main predictor of adolescent internalizing symptoms. In addition, results show double-inequalities according to the interaction effects of sex, age, and SES. Likewise, interesting results are shown regarding how dieting behaviors to lose or gain weight/volume and physical activity relate to body image perception and satisfaction, as well as with internalizing symptoms. This study highlights important body image aspects relevant to intervention and prevention of internalized mental health problems in adolescence. Keywords: body image, body mass index, dieting, physical activity, internalizing symptoms, sex, age, socioeconomic status INTRODUCTION Adolescence is a period characterized by physiological, emotional, cognitive, and social changes that lead to a greater concern for physical appearance. The physical changes that accompany the onset of puberty demand a constant restructuring of the adolescent’s perception of their body, provoking an increased preoccupation for this image that in many cases leads to a decrease in self-esteem Frontiers in Psychology | www.frontiersin.org 1August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 2 Ramos et al. Body Image and Internalizing Symptoms (Harter, 2006;Mills et al., 2011). The propensity of adolescents to be unhappy with their body image has been addressed in many studies, arguing that this problem with body image can threaten their health and wellbeing (Shagar et al., 2017). Body image refers to the multifaceted psychological experience of embodiment that encompasses one’s bodyrelated self-perceptions and self-attitudes, including thoughts, beliefs, feelings, and behaviors. Body image has often been defined as the self-perception of the physical self and the feelings and thoughts that result from that perception (Cash, 2004; Grogan, 2006). Disturbance in any of these domains is referred to as body image concerns or negative body image. Previous findings have shown a relationship between body image concerns and the development of psychopathology in adolescence, especially with internalizing symptoms (IS) – i.e., anxiety, depression, or social withdrawal (Thompson et al., 1995;McClintock, 2001;Siegel, 2002;Ter Bogt et al., 2006; Patalay and Hardman, 2019)–, which in turn is a risk factor for the development of eating disorders (Graber and Sontag, 2009;Puccio et al., 2016). For example, a previous study of more than 7000 Dutch adolescents showed an association between deviance from normal weight and IS, however not with externalizing problems (Ter Bogt et al., 2006). According to the authors, “lower body mass index [BMI] scores relate to greater problems, more specifically to anxiety/depression and social problems, higher BMI is associated with withdrawnness. However, body weight perception is a more substantial predictor of adolescent distress (. . .). Perception of being too heavy is the most substantial predictor of a wide range of problems and both the overweight and the young people with normal BMI are affected” (p. 32). Appraisals of physical development or body image are areas of self-evaluation that are particularly salient to adolescents and their risk for internalizing problems (Graber and Sontag, 2009). Children with higher BMI may feel discriminated against due to their image, which would lead them to present internalizing symptoms in the long-term. This is related to an increased awareness and internalization of social attitudes around weight throughout childhood, which is manifested in more psychological problems in adolescents with worse body image (Puhl and Latner, 2007). Specifically, body image concerns can often be attributed to social and cultural pressures (Bessenoff and Snow, 2006). According to the systematic review conducted by Sobrino-Bazaga and Rabito-Alcoón (2018), sociocultural factors had the most influence on body image dissatisfaction. In addition, having a body image close to the socially accepted ideal of beauty is conceptually associated with possessing other positive characteristics such as being a successful or healthy person (Tiggemann et al., 2000;Neziroglu et al., 2008; Mischner et al., 2013). In explaining human development, and in this case adolescent development, certain theoretical models highlighting the importance of an individual’s actions in guiding their own evolutionary development are currently gaining more importance. In this respect, the recently proposed Relational Developmental Systems (RDS; Overton, 2015) metatheory collects those theories defending that change throughout the life cycle occurs through mutually influential relationships between individuals and their contexts. Thus, this perspective states that biological and environmental entities influence human development through a process of coaction. Specifically, Overton (2014) propose this metatheory based on the process-relational paradigm according to which any part of the developmental system is inextricably embodied by all other parts of the system. “Embodiment refers to the way individuals behave, experience, and live in the world by their being active agents with particular kinds of bodies; the body is integratively understood as form (a biological referent), as lived experience (a psychological referent), and as an entity in active engagement with the world (a sociocultural referent)” (Lerner, 2018, p. 25). Therefore, attending to the importance of an individual’s actions in the relationship between body image and internalizing symptoms, it is imperative to understand the role that behaviors related to weight control and body image play when analyzing the different components of body image in adolescence. In this regard, BMI represents the physiological aspect of body changes during adolescence, body image perception (BIP) represents the cognitive component, and body image satisfaction (BIS) corresponds to the emotional component of body image. Moreover, these domains are also related to a behavioral component –dieting or exercising– that contributes to control weight and body image (Banfield and McCabe, 2002; Ramos et al., 2012). Furthermore, some studies warn that these components – physiological, cognitive, emotional, and behavioral– interrelate differently depending on the adolescent’s sociodemographic factors [i.e., sex, age, and socioeconomic status (SES)]. Firstly, concern for body image differs according to sex; whereas boys tend to be more concerned with having a muscular body image, girls are more likely to have beauty ideals that are inseparable from thinness, in most cases below a healthy size (Neumark-Sztainer et al., 1999;Cafri and Thompson, 2004). In addition, although a higher proportion of girls tend to have a normal BMI compared to boys, they also show a greater prevalence of suffering misperceptions about their body image leading to a greater number of weight-control behaviors (Barker and Galambos, 2003;Larson et al., 2009; Ramos et al., 2013). Systematic literature reviews about sex inequalities in body dissatisfaction, such as those conducted by McCabe and Ricciardelli (2004) and Sobrino-Bazaga and Rabito- Alcoón (2018), support this perspective. However, there is less research regarding sex differences in the relationship between body image problems and IS in adolescence. Current studies such as that published by Bacopoulou et al. (2018) point out that girls showing higher anxiety on the internalizing subscale of the Youth Self-Report (YSR) have a higher risk of developing eating disorders than girls without anxiety, whereas this does not occur in their male peers. Secondly, with regards to age researchers agree that both boys and girls are less satisfied with their body image throughout the period stretching from adolescence to adulthood (Sobrino- Bazaga and Rabito-Alcoón, 2018). However, during adolescence it appears that only girls decrease their body satisfaction with age (e.g., Bully and Elosua, 2011). This same tendency Frontiers in Psychology | www.frontiersin.org 2August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 3 Ramos et al. Body Image and Internalizing Symptoms has been found in other variables such as BIP and weight control (Neumark-Sztainer et al., 1999;Barker and Galambos, 2003;McArthur et al., 2005). These results indicate that during adolescence girls’ association of ideal beauty with being thin, in most cases below a healthy size, is strengthened. However, there are no results that directly address how the relationship varies between body image and IS according to the age of the adolescents, and therefore remains an important unanswered question. Thirdly, adolescents from families with a low SES tend to present higher rates of overweightness and obesity (O’Dea and Caputi, 2001;Cheung et al., 2011;Pereira et al., 2011). In addition, pertaining to families with a low SES also contributes to the persistence of meal-related health problems during this life stage (e.g., Wadsworth and Achenbach, 2005). Nevertheless, there is little research exploring the impact of SES on IS, or how SES moderates the association between body image and IS, and results are inconsistent. For example, in the study conducted by Ter Bogt et al. (2006) both boys and girls across all SES showed to experience elevated levels of IS when perceiving their body as too thin or (more frequently) too heavy. Another study (Zeller et al., 2004) conducted with 121 obese children and adolescents and their mothers found that in mothers with distress, low SES related to higher IS in their obese children. However, this study did not find a relationship between SES and IS self-reported by the children themselves. In Spain, 17.2% of 11–18 year old adolescents show signs of being overweight or obese, with this percentage being clearly higher in boys and adolescents with a lower SES (Moreno et al., 2016). This study also shows that despite girls having a lower index of being overweight or obese, they are less satisfied with their body image and diet more frequently to lose weight, a behavior that also increases with age. Regarding internalizing symptoms, Oliva et al. (2015) underscores the higher proportion of internalizing symptoms in girls. In addition, unlike their male peers, in girls the internalizing symptoms increased with age and with a lower SES. Therefore, in the context of the current research, it is important to further study the roles of sex, age, and socioeconomic level in the relationship between body image problems and IS in adolescence. With this objective we first propose to understand how the physiological (BMI), cognitive (BIP), and emotional (BIS) components of body image interrelate when predicting the adolescents’ internalizing symptoms. In addition, we will attempt to analyze the specific role that the variables sex, age, and socioeconomic level play in the aforementioned relationship. Along these lines, we assume that the exposure to double/triple/multiple factors of inequalities is more harmful than the exposure to a single one. Therefore, following an interesting approach from social sciences of exploring multiple inequalities and its intersectionality, mostly initiated from a gender perspective (Acker, 2006;Walby, 2011), this present study examined interactions between sex, age, and SES as sources of inequalities. Secondly, this study proposes to further research those adolescents with worse scores in the specific component (physiological, cognitive, or emotional) that in the previous objective was the most important for IS. In this case, this study proposes to analyze in detail the profile of these adolescents, not only based on the physiological, cognitive, or emotional components, but also according to how the configuration of these variables relates with the behavioral component, measured through dieting and the frequency of physical activity. MATERIALS AND METHODS Data Data comes from the Health Behaviour in School-aged Children (HBSC) cross-sectional survey. The HBSC study is an international network supported by the World Health Organization that collects data in almost 50 countries in Europe and North America. The survey is conducted every 4 years with the aim of increasing knowledge about health-related behaviors, lifestyles, wellbeing, and developmental contexts of youth. Specifically, the data used in this paper is obtained from the 2018 edition of the HBSC study in Spain. Participants This study is based on a nationally representative sample composed of 4531 adolescents aged 13–18 years who participated in the 2018 edition of the HBSC study in Spain. They were selected using random multi-stage sampling stratified by conglomerates according to age, type of school (public or private), and region (Moreno et al., 2018). The sample was divided equally between boys and girls. The percentage of adolescents according to each age group was 16.7% who were 13–14, 51.3% were 15–16, and 32% were 17–18 years old. Likewise, the global sample was distributed between 63.3% that attended public schools, 31.7% that attended charter schools, and 5% that attended private schools. Regarding SES, 18.9% reported a low level, 48.8% a medium level, and 32.3% a high level. Lastly, the sample was distributed between the different regions of Spain in the following percentages: 6.7% Andalucia, 4.1% Aragon, 4.6% Principado de Asturias, 7.2% Illes Balears, 5.5% Canarias, 3.7% Cantabria, 3.3% Castilla y León, 6.9% Castilla-La Mancha, 10.2% Cataluña, 6.6% Comunitat Valenciana, 5.7% Extremadura, 3.5% Galicia, 8.5% Comunidad de Madrid, 4.4% Región de Murcia, 6.4% Comunidad Foral de Navarra, 7.2% País Vasco, 3.6% La Rioja, and 1.9% Ceuta y Melilla. Instruments The HBSC questionnaire is a broad survey that measures adolescents’ healthy habits, wellbeing, and their developmental contexts from a multidisciplinary perspective. The questionnaire uses a variety of validated scales and items. Spanish-language versions of the scales were used, that were translated/backtranslated between English and Spanish following the guidelines and cooperation of the international coordination of the HBSC study (Roberts et al., 2009). The specific variables employed in this study –selected according to their scientific basis and study objectives– are detailed below. Frontiers in Psychology | www.frontiersin.org 3August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 4 Ramos et al. Body Image and Internalizing Symptoms Demographic variables were sex (students had to self-identify as a boy or girl) and SES evaluated using the Family Affluence Scale (FAS III, Schnohr et al., 2013). The FAS score was calculated as the sum of the responses from the six items included in the latest version of the instrument: “Does your family own a car, van or truck?” (0 = no; 1 = yes, one; 2 = yes, two or more); “Do you have your own bedroom for yourself?” (0 = no; 1 = yes); “How many computers do your family own (including laptops and tablets, not including game consoles and smartphones)?” (0 = none, 1 = one, 2 = two, 3 = more than two); “How many bathrooms (room with a bath/shower or both) are in your home?” (0 = none,1=one, 2 = two,3=more than two); “Does your family have a dishwasher at home?” (0 = no; 1 = yes); “How many times did you and your family travel out of Spain for a holiday/vacation last year?” (0 = not at all,1=once,2=twice,3=more than twice). The terciles (low, medium, and high), with cut-off points of 6 and 9 (in a range of 0–13) were used as a categorical variable. This paper analyses different body-related variables: –Body Mass Index is the ratio between self-reported weight and height (kg/m2). To use it as a categorical variable BMI was recoded according to the cut-off points proposed by Cole et al. (2000, 2007), with the options thin, normal, and overweight/obese. –Body Image Perception was assessed with an item created for the HBSC study. Specifically, students were asked “Do you think your body is?” and their response options on a 5- point Likert scale ranged from 1 much too thin, to 5 much too fat. Responses were recoded into 3 categories: thin (1 too thin, and 2 a bit thin), normal (3 I have the correct size) and overweight (4 a bit fat, and 5 too fat). –Body Image Satisfaction was assessed through the feelings and attitudes toward the body subscale of the Body Investment Scale (Orbach and Mikulincer, 1998). This subscale consists of 6 items (“I am frustrated with my physical appearance,” “I am satisfied with my appearance,” “I hate my body,” “I feel comfortable with my body,” “I feel anger toward my body,” and “I like my appearance in spite of its imperfections”). It is answered on a 5- point Likert scale from 1 totally agree, to 5 totally disagree. The Cronbach’s alpha in the analyzed sample was 0.89. In order to use it as a categorical variable, the sample was divided into terciles (low, medium, and high) with the cut-off points 3.6 and 4.5 (in a range of 1 to 5). When it is used as a continuous variable, high scores reflect high body satisfaction. Two variables were included regarding the behavioral component: (1) a measure that was created by the international HBSC protocol to explore the behavior of dieting, asking adolescents whether they were currently dieting to lose or gain weight/volume; and (2) adolescents were asked about their level of Moderate to Vigorous Physical Activity (MVPA), as indicated by the number of days in which they felt physically active during a total of at least 60 min a day over the last 7 days. The response options ranged from 0 to 7 days (Prochaska et al., 2001). In addition, their level of Vigorous Physical Activity (VFA) was assessed as the frequency with which they engaged in some type of physical activity that made them sweat or out of breath in their free time outside of school hours. The response options on a Likert scale ranged from 1 never, to 7 every day. Lastly, the variables related to internalizing mental health problems were assessed using the licensed version of the Spanish version of the YSR (Achenbach and Rescorla, 2001). The three internalizing syndrome scales employed were: anxious/depressed (13 items, with a 0 to 13 response range), withdrawn/depressed (8 items, with a 0 to 8 response range), and somatic complaints (10 items, with a 0 to 10 response range). All the items are Likert-type response format comprising three categories (0 = not true, 1 = somewhat true or sometimes true, 2 = very true or often true) and refer to symptoms and experiences from the past 6 months. The application guidelines marked by Achenbach (2013) were followed in the application of this instrument within the HBSC study in Spain. The Cronbach’s alpha in the present study was 0.79 for anxious/depressed, 0.75 for withdrawn/depressed, 0.76 for somatic complaints, and 0.87 for the global scale of IS. In order to use IS as categorical (for the second objective of the study) low (12–24), medium (24.50–32) and high (32.50–80) terciles were used. When it is used as a continuous variable, high scores reflect high internalizing symptoms. Procedure The international coordination of the HBSC study stipulates three basic conditions that must be complied with during data collection: the questionnaires must be answered by the students themselves; the anonymity of the answers must be strictly guaranteed; and the questionnaires must be administered within the school context (Inchley et al., 2016). Thus, the data was always collected in the school setting under the supervision of teachers. In all the schools, instructions were given to the teachers who would be supervising the classroom when the adolescents responded to the questionnaires. New information and communication technologies (ICT) based on a CAWI (Computer-Assisted Web Interviewing) model were used in data collection. The guided computerized procedure has the advantage of immediately receiving and incorporating the students’ responses in the database, thus reducing possible errors from the data entry process, as well as helping to maintain the anonymity of the responses. In addition, instructions for the students were included at the beginning of the questionnaire to guarantee homogeneity amongst all the participants. In all the participating education centers, permission was requested from the committee board of each school and they handled the request of permission of the adolescents and their families according to each centers’ protocols. This study was approved by the Ethical Research Committee of the regional government of Andalusia, Spain (Junta de Andalucía). Sample selection was carried out between February and June 2018. Once data were collected, they were cleaned and controlled according to the international HBSC protocols offered by the Norwegian Centre for Research Data (Currie et al., 2014). In this procedure, from the initial sample composed of 5356 adolescents, Frontiers in Psychology | www.frontiersin.org 4August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 5 Ramos et al. Body Image and Internalizing Symptoms subjects who showed a response rate below 40% of the total questionnaire (9.5% of the initial sample), who did not give basic sociodemographic information (0.9% of the initial sample), who showed an overly elevated pattern in their answers (4.3% of the initial sample), or who had not responded to all the analyzed variables (3.5% of the initial sample) were eliminated. Due to the large remaining sample (4531 adolescents), the decision made was to not impute the missing values and instead analyze only those subjects with valid values in the analyzed variables. Statistical Analysis Bivariate analyses including chi-square, t-test and ANOVA were used, requiring a significance level of 0.01. The corresponding effect size was reported in every case. Thus, phi, Crammer’s V and Tau-b were calculated to determine the effect size in the comparison of proportions, with the following cut-off points: 0– 0.09 = negligible, 0.10–0.29 = small, 0.30–0.49 = medium, 0.50 and above = high. Cohen’s dwas calculated in the comparison of two groups’ means, with the following cut-off points: 0– 0.19 = negligible, 0.20–0.49 = small, 0.50–0.79 = medium, 0.80 and above = high. Eta-squared was calculated in the comparison of more than two groups’ means, with the following cut-off points: 0–0.009 = negligible, 0.010–0.059 = small, 0.060–0.139 = medium, 0.140 and above = high. Pearson’s correlation was calculated in the relationship between two quantitative variables, with cut-off points 0–0.090 = negligible, 0.10–0.239 = small, 0.24–0.369 = medium, 0.37 and above = high (Cohen, 1988). Linear regression analysis was used to examine the relationship between sociodemographic variables (sex, age, and FAS) and the indicators of body image on the total scale of IS and each one of the subscales (anxious/depressed, withdrawn/depressed, and somatic complaints). The principal effects were analyzed first, and afterward the previous models were replicated adding the interaction effects. Pearson’s correlations were analyzed to determine the significant interaction effects, segmented according to the variables implicated in said interaction. Differences in correlations were compared using the Fisher’s Z-test, establishing the cut-off point in values above 2.33 (in absolute values), corresponding to a 99% unilateral confidence interval. To analyze the characteristics that define the two groups established for the second objective of the present study, a logistical regression model was created which incorporated the value of the odds ratio with its corresponding confidence interval. Statistical analyses were conducted using the IBM SPSS Statistics 25.0 software. RESULTS Analysis of the Bivariate Relationship Between Variables Firstly, the distribution of the sample in the three components of body image (BMI, BIP, and BIS) is shown in Table 1. These variables presented a high correlation between them. TABLE 1 | Sample description in the three variables related to body image, dieting, and physical activity. NPercentage BMI Thin 617 13.6% Normal 3197 70.6% Overweight/obese 717 15.8% BIP Thin 803 17.7% Normal 2195 48.4% Overweight 1533 33.8% BIS Low 1625 35.9% Medium 1673 36.9% High 1233 27.2% Diet to lose No 3941 87.2% Yes 581 12.8% Diet to gain No 5098 96.2% Yes 200 3.8% MVPA Null 316 7.4% Low 2739 64.5% High 1191 28.0% VPA Null 527 12.4% Low 2319 54.6% High 1405 33.1% BMI, Body Mass Index; BIP, Body Image Perception; BIS, Body Image Satisfaction; MVPA, Moderate-Vigorous Physical Activity; VPA, Vigorous Physical Activity. Specifically, a higher BMI was positively related with overweight BIP (r= 0.514, p<0.001) and negatively associated with BIS (r=−0.268, p<0.001). Psychological components (BIP and BIS) were also negatively correlated, indicating that overweight BIP is associated with a lower BIS (r=−0.410, p<0.001). These three components of body image varied according to the adolescents’ sex, age, and socioeconomic level. As shown in Table 2, BMI was higher in 15–16 year old adolescents compared to the 13–14 year olds, as well as being higher in adolescents with a low FAS score. Regarding overweight BIP, this was higher in girls and in adolescents with low FAS. Lastly, BIS was lower in girls, 17–18 year old adolescents, and those with low FAS. Differences were also found in these three components of body image in their relationship with indicators of IS, as shown in Table 3. Specifically, the adolescents with higher BMI showed higher scores in withdrawn/depressed and in the global scale of IS. However, a relationship was found between all the indicators of IS and BIS and BIP: those adolescents with overweight BIP and low BIS were more anxious/depressed, withdrawn/depressed, had more somatic complaints, and had a higher global IS score. Analysis of IS distribution according to sociodemographic variables (shown in Table 3) also revealed inequalities according to sex (girls had higher scores in all indicators analyzed), age (differences were lower than in sex, however older adolescents had higher scores in the withdrawn/depressed subscale and in the global IS score), and FAS (adolescents with lower family affluence scored higher in the three subscales, but the global score did not show relevant differences). Examining behaviors related to weight control and body image, Table 1 shows the distribution of the sample in these Frontiers in Psychology | www.frontiersin.org 5August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 6 Ramos et al. Body Image and Internalizing Symptoms TABLE 2 | Means (standard deviations), significance test and effect sizes of sex, age, family affluence, dieting, and physical activity with the three variables related with body image. Body Mass Index Body Image Perception Body Image Satisfaction Descriptive statistics Significance tests Descriptive statistics Significance tests Descriptive statistics Significance tests Sex Boys 21.38 (3.93) t(4529) = 4.08, 3.04 (0.81) t(4529) = 11.10, 4.13 (0.80) t(4529) = 15.53, Girls 20.93 (3.47) p<0.001;d= 0.12 3.31 (0.83) p<0.001;d=0.33 3.72 (0.96) p<0.001; d=0.46 Age 13–14 year 20.31 (3.55) F(4528) = 36.99, 3.16 (0.79) F(4528) = 0.30, 4.04 (0.88) F(4528) = 10.76, 15–16 year 21.08 (3.61) p<0.001; η2= 0.02 3.17 (0.82) p= 0.74; η2<0.01 3.94 (0.91) p<0.001; η2= 0.01 17–18 year 21.71 (3.87) d=0.21/0.17/0.37 3.19 (0.88) d= 0.01/0.02/0.03 3.85 (0.92) d= 0.11/0.10/0.21 FAS Low 21.92 (4.48) F(4528) = 29.57, 3.27 (0.90) F(4528) = 9.70, 3.77 (0.98) F(4528) = 18.66, Medium 21.19 (3.58) p<0.001; η2= 0.01 3.18 (0.82) p<0.001; η2= 0.01 3.94 (0.90) p<0.001; η2= 0.01 High 20.66 (3.20) d= 0.19/0.15/0.34 3.11 (0.80) d= 0.11/0.09/0.20 4.02 (0.87) d= 0.18/0.09/0.26 Diet to lose No 20.75 (3.38) t(4765) = −19.87, 3.07 (0.80) t(5233) = 23.78, 4.02 (0.85) t(5007) = 1.96, Yes 23.90 (4.65) p<0.001;d=0.88 3.90 (0.68) p<0.001;d=1.05 3.27 (1.03) p<0.001;d=0.86 Diet to gain No 21.21 (3.76) t(4765) = 4.07, 3.22 (0.83) t(5233) = 13.33, 3.90 (0.93) t(5007) = 1.96, Yes 20.07 (3.05) p<0.001;d=0.30 2.41 (0.83) p<0.001;d=0.98 4.04 (0.84) p= 0.050; d= 0.15 MVPA Null 21.81 (4.79) F(4528) = 18.20, 3.35 (1.00) F(4528) = 28.55, 3.47 (1.08) F(4528) = 157.99, Low 21.31 (3.67) p<0.001; η2= 0.01 3.22 (0.85) p<0.001; η2= 0.01 3.83 (0.90) p<0.001; η2= 0.07 High 20.65 (3.43) d= 0.13/0.18/0.30 3.03 (0.69) d= 0.15/0.24/0.42 4.28 (0.74) d=0.39/0.53/0.98 VPA Null 21.25 (4.73) F(4528) = 0.60, 3.19 (0.95) F(4528) = 9.13, 3.64 (1.04) F(4528) = 74.10, Low 21.20 (3.74) p= 0.55; η2<0.01 3.22 (0.84) p<0.001; η2<0.01 3.87 (0.90) p<0.001; η2= 0.03 High 21.08 (3.26) d= 0.01/0.03/0.05 3.04 (0.81) d= 0.03/0.22/0.18 4.14 (0.82) d=0.25/0.31/0.56 FAS, Family Affluence Scale; MVPA, Moderate-Vigorous Physical Activity; VPA, Vigorous Physical Activity; y, years-old. In contrasts of d of Cohen with variables of three values: the d in first position corresponds to the contrast of the 1st and 2nd value, the d in second position corresponds to the contrast of the 2nd and 3rd value and the d in third position corresponds to the contrast of the 1st and 3rd value. The bold values indicate small (0.20–0.49), medium (0.50–0.79), or high (0.80 and above) effect size values. variables. Table 3 shows no relationship with the behavior of gaining weight/volume but does with the weight-loss diets. Specifically, adolescents that dieted to lose weight/volume had higher scores in all the subscales and in the global scale of IS. This same relationship was found with the two indicators of physical activity, meaning that those adolescents who did less physical activity (either MVPA or VPA) showed higher scores in all the internalizing indicators. Returning to the body image variables, Table 2 also shows the bivariate relationship between the three components of body image and the weight control and body image behaviors. In general, adolescents with higher BMI more frequently dieted to lose weight/volume and did MVPA. Adolescents with an overweight BIP showed a higher frequency of dieting to lose weight/volume, and a lower tendency to practice either type of physical activity, as well as a lower frequency of dieting to gain weight/volume. Those adolescents with lower BIS dieted more to lose weight/volume and were more inclined to not practice either of the two types of physical activity (MVPA or VPA). Regarding the relationship between these weight and body image behaviors, differences were found according to sex, age, and FAS score, as shown in Table 4. Specifically, dieting to lose weight/volume was more frequent in girls. On the contrary, dieting to gain weight/volume was more frequent in boys, those higher in age, and with high FAS. On the other hand, both MVPA and VPA were more frequent in boys, of a lower age, and with high FAS. The Predictive Capacity of Body Image Variables (BMI, BIP, and BIS) on IS, Considering the Influence of Sociodemographic Variables (Sex, Age, and Family Affluence) Table 5 shows the results of the linear regression analyses conducted for each of the four measures of IS (anxious/depressed, withdrawn/depressed, and somatic complaints) and for the global scale of IS. Two types of linear regression analyses were conducted for each internalizing indicator: one simple and another global. Whereas the simple analysis only measured the direct influence of the independent variables, the global analysis controlled possible interaction effects of the sociodemographic variables in the relationship between body image and the internalizing indicators. Regarding the individual predictive capacity (without interaction) of the sociodemographic and body image variables on the four internalizing indicators, results showed BIS as the variable with the most weight. Therefore, BIS showed to be the most important predictor of IS, whether in the global score (η2 P=0.140) or in the three subscales: anxious/depressed (η2 P=0.166), withdrawn/depressed (η2 P=0.127), and somatic complaints (η2 P=0.068). The adolescents’ sex was the second most important dependent variable, both in the global scale and in the anxious/depressed and withdrawn/depressed scales. Age was the second highest predictive variable of withdrawn/depressed and the third in the global scale. Frontiers in Psychology | www.frontiersin.org 6August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 7 Ramos et al. Body Image and Internalizing Symptoms TABLE 3 | Means (standard deviations), significance test and effect sizes of the three variables related with body image, sex, age, family affluence, dieting, and physical activity with anxious/depressed, withdrawn/depressed, somatic complaints, and total of internalizing symptoms. Anxious/depressed Withdrawn/depressed Somatic complaints Internalizing symptoms (total) Descriptive s. Significance tests Descriptive s. Significance tests Descriptive s. Significance tests Descriptive s. Significance tests BMI Thin 6.34 (4.01) F(4777) = 7.59, 3.60 (2.81) F(4777) = 16.68, 3.61 (3.12) F(4777) = 5.83, 13.77 (10.56)F(4777) = 48.97, Normal 6.05 (4.07) p= 0.001; η2<0.01 3.65 (2.91) p<0.001; η2<0.01 3.49 (3.05) p= 0.003; η2<0.01 17.00 (9.60)p<0.001; η2= 0.02 Overweight/obese 6.65 (4.43) d= 0.07/0.15/0.08 4.31 (3.12) d= 0.01/0.22/0.23 3.90 (3.40) d= 0.05/0.14/0.09 18.97 (10.28)d=0.33/0.23/0.72 BIP Thin 6.25 (4.00) F(5248) = 162.23, 3.81 (2.97) F(5248) = 135.79, 3.49 (2.94) F(5248) = 114.44, 16.9(9.61)F(5248) = 136.94, Normal 5.24 (3.57) p<0.001; η2= 0.06 3.10 (2.62) p<0.001; η2= 0.05 2.97 (2.75) p<0.001; η2= 0.04 14.71 (8.73)p<0.001; η2= 0.05 Overweight 7.50 (4.55) d=0.29/0.56/0.27 4.64 (3.12) d= 0.15/0.31/0.22 4.48 (3.48) d= 0.19/0.46/0.27 19.96 (10.90)d=0.33/0.20/0.49 BIS Low 8.15 (4.63) F(5022) = 402.48, 5.01 (3.16) F(5022) = 311.59, 4.71 (3.61) F(5022) = 221.54, 21.49 (11.00)F(5022) = 375.15, Medium 5.53 (3.40) p<0.001; η2= 0.14 3.38 (2.61) p<0.001; η2= 0.11 3.17 (2.64) p<0.001; η2= 0.08 15.42 (8.35)p<0.001; η2= 0.13 High 4.48 (3.20) d=0.66/0.30/0.89 2.59 (2.41) d=0.56/0.29/0.83 2.63 (2.53) d=0.51/0.19/0.65 12.76 (7.87)d=0.64/0.29/0.87 Sex Boys 5.09 (3.68) t(5410) = −19.11, 3.44 (2.86) t(5410) = −7.37, 2.74 (2.66) t(5410) = −19.36, 14.77 (8.77)t(5410) = −14.57, Girls 7.28 (4.26) p<0.001;d=0.52 4.06 (2.99) p<0.001;d=0.20 4.41 (3.31) p<0.001;d=0.53 18.98 (10.59)p<0.001;d=0.40 Age 13–14 year 5.97 (3.90) F(5410) = 2.44, 3.23 (2.75) F(5410) = 18.10, 3.51 (3.03) F(5410) = 7.55, 14.24 (10.03)F(5410) = 47.61, 15–16 year 6.13 (4.17) p= 0.087; η2<0.01 3.72 (2.90) p<0.001; η2<0.01 3.43 (3.06) p= 0.001; η2<0.01 16.85 (9.63)p<0.001; η2= 0.02 17–18 year 6.38 (4.17) d= 0.03/0.05/0.08 4.07 (3.06) d= 0.15/0.10/0.24 3.83 (3.24) d= 0.02/0.12/0.10 18.29 (10.13)d=0.28/0.12/0.39 FAS Low 6.73 (4.33) F(4783) = 12.15, 4.31 (2.99) F(4783) = 27.04, 4.15 (3.36) F(4783) = 18.41, 18.04 (10.79)F(4783) = 7.77, Medium 6.11 (4.09) p<0.001; η2<0.01 3.76 (2.93) p<0.001; η2= 0.01 3.4 (2.91) p<0.001; η2<0.01 16.69 (9.69)p<0.001; η2<0.01 High 5.83 (3.95) d= 0.13/0.07/0.21 3.35 (2.81) d= 0.19/0.15/0.34 3.42 (3.16) d=0.20/0.05/0.21 16.35 (9.44)d= 0.12/0.08/0.14 Diet to lose No 5.99 (3.99) t(5297) = −8.38, 3.66 (2.90) t(5297) = −5.37, 3.42 (3.01) t(5297) = −8.61, 16.55 (9.64)t(5297) = −5.72, Yes 7.52 (4.77) p<0.001;d=0.37 4.36 (3.13) p<0.001;d=0.22 4.61 (3.66) p<0.001;d=0.36 19.07 (11.59)p<0.001;d=0.26 Diet to gain No 6.20 (4.12) t(5297) = 0.19, 3.74 (2.93) t(5297) = −0.81, 3.57 (3.11) t(5297) = −0.94, 16.84 (9.95)t(5297) = −1.96, Yes 5.97 (4.37) p= 0.058; d= 0.01 3.90 (3.27) p= 0.417; d= 0.06 3.73 (3.49) p= 0.348; d= 0.07 17.90 (9.99)p= 0.051; d= 0.14 MVPA Null 7.74 (4.92) F(4902) = 94.18, 5.17 (3.52) F(4902) = 154.33, 4.68 (3.73) F(4902) = 95.49, 21.96 (11.55)F(4902) = 169.97, Low 6.49 (4.08) p<0.001; η2= 0.04 4.04 (2.92) p<0.001; η2= 0.06 3.80 (3.13) p<0.001; η2= 0.04 17.89 (9.69)p<0.001; η2= 0.07 High 5.00 (3.64) d=0.35/0.35/0.71 2.67 (2.40) d=0.43/0.46/0.93 2.68 (2.6) d=0.33/0.36/0.72 13.05 (8.46)d=0.44/0.48/0.95 VPA Null 7.65 (4.83) F(4909) = 66.01, 5.05 (3.38) F(4909) = 97.22, 4.23 (3.41) F(4909) = 49.70, 20.15 (11.56)F(4909) = 58.70, Low 6.31 (3.99) p<0.001; η2= 0.03 3.80 (2.82) p<0.001; η2= 0.04 3.77 (3.12) p<0.001; η2= 0.02 17.17 (9.66)p<0.001; η2= 0.02 High 5.35 (3.79) d=0.33/0.21/0.53 3.13 (2.73) d=0.44/0.21/0.64 2.92 (2.82) d= 0.17/0.25/0.42 15.00 (9.04)d=0.32/0.29/0.50 BMI, Body Mass Index; BIP, Body Image Perception; BIS, Body Image Satisfaction; FAS, Family Affluence Scale; MVPA, Moderate-Vigorous Physical Activity; VPA, Vigorous Physical Activity; y, years-old; s, statistics. In contrasts of d of Cohen with variables of three values: the d in first position corresponds to the contrast of the 1st and 2nd value, the d in second position corresponds to the contrast of the 2nd and 3rd value and the d in third position corresponds to the contrast of the 1st and 3rd value. The bold values indicate small (0.20–0.49), medium (0.50–0.79), or high (0.80 and above) effect size values. Frontiers in Psychology | www.frontiersin.org 7August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 8 Ramos et al. Body Image and Internalizing Symptoms TABLE 4 | Percentage, significance test and effect sizes of sex, age, and family affluence with dieting and physical activity. Diet to lose Diet to gain MVPA VPA Descriptive Significance Descriptive Significance Descriptive Significance Descriptive Significance statistics tests statistics tests statistics tests statistics tests No Yes No Yes Null Low High Null Low High Sex Boys 90.0% 10.0% X2(1) = 32.73, p<0.001; phi = 0.085 94.2% 5.8% X2(1) = 45.57, p<0.001; phi =0.100 5.3% 58.1% 36.6% X2(2) = 161.71, p<0.001; V=0.195 8.1% 47.1% 44.8% X2(2) = 281.65, p<0.001; V=0.257 (5.7) (−5.7) (−6.8) (6.8) (−5.2) (−8.7) (12.3) (−8.5) (−9.6) (16.2) Girls 84.3% 15.7% 98.1% 1.9% 9.5% 70.9% 19.6% 16.7% 61.9% 21.4% (−5.7) (5.7) (6.8) (−6.8) (5.2) (8.7) (−12.3) (8.5) (9.6) (−16.2) Age 13–14 year 87.3% 12.7% X2(2) = 0.82, p= 0.664; V= 0.013 97.7% 2.3% X2(2) = 19.77, p<0.001; V= 0.066 4.3% 61.3% 34.4% X2(4) = 62.72, p<0.001; Tau -b =−0.106 8.4% 55.9% 35.7% X2(4) = 31.02, p<0.001; Tau -b =−0.070 (0.1) (−0.1)(2.5) (−2.5) (−3.5) (−2.0)(4.1) (−3.6) (0.8) (1.6) 15–16 year 87.5% 12.5% 96.8% 3.2% 6.3% 64.3% 29.4% 12.0% 53.1% 34.9% (0.7) (−0.7) (2.1) (−2.1)(−2.9) (−0.3) (2.0) (−0.8) (−2.0)(2.6) 17–18 year 86.5% 13.5% 94.4% 5.6% 10.9% 66.5% 22.6% 15.2% 56.2% 28.7% (−0.9) (0.9)(−4.3) (4.3) (5.9) (1.9)(−5.4) (3.7) (1.4)(−4.1) FAS Low 86.7% 13.3% X2(2) = 0.22, p= 0.894; V= 0.007 96.6% 3.4% X2(2) = 6.81, p= 0.033; V= 0.041 10.0% 66.4% 23.7% X2(4) = 28.58, p<0.001; Tau -b = 0.075 17.0% 56.3% 26.7% X2(4) = 47.29, p<0.001; Tau -b = 0.093 (−0.5) (0.5) (0.7) (−0.7)(3.1) (1.2)(−3.1) (4.4) (1.0)(−4.1) Medium 87.4% 12.6% 96.7% 3.3% 7.3% 65.7% 27.0% 11.9% 56.6% 31.5% (0.2) (−0.2) (1.9) (−1.9) (−0.1) (1.5) (−1.5) (−0.8) (2.3) (−1.8) High 87.3% 12.7% 95.0% 5.0% 5.9% 61.8% 32.4% 10.2% 51.1% 38.7% (0.1) (−0.1)(−2.6) (2.6) (−2.5) (−2.6) (4.2) (−2.8) (−3.3) (5.4) The value of the corrected standardized residual is included in parentheses. The bold values indicate small (0.10–0.29), medium (0.30–0.49), or high (0.50 and above) effect size valuesphi and V- and corrected standardized residuals -SRc- higher to 2.33 (one-sided confidence interval of 99%). Frontiers in Psychology | www.frontiersin.org 8August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 9 Ramos et al. Body Image and Internalizing Symptoms TABLE 5 | Linear regression analysis of the simple model, only with main effects, and the global model, with interaction effects, on anxious/depressed, withdrawn/ depressed, somatic complaints, and total of internalizing symptoms. Anxious/depressed Withdrawn/depressed Somatic complaints Internalizing symptoms (total) Simple model (without interaction effects) R Squared 0.245 0.172 0.154 0.228 Sex η2 P= 0.037; p<0.001 η2 P<0.001; p= 0.173 η2 P= 0.045; p<0.001 η2 P= 0.024; p<0.001 Age η2 P<0.001; p= 0.850 η2 P= 0.005; p<0.001 η2 P<0.001; p= 0.309 η2 P= 0.013; p<0.001 FAS η2 P= 0.003; p= 0.001 η2 P= 0.009; p<0.001 η2 P= 0.003; p= 0.001 η2 P<0.001; p=0.375 BMI η2 P<0.001; p= 0.182 η2 P<0.001; p= 0.311 η2 P<0.001; p= 0.878 η2 P= 0.012; p<0.001 BIP η2 P= 0.001; p= 0.017 η2 P= 0.001; p= 0.025 η2 P<0.001; p= 0.961 η2 P= 0.007; p<0.001 BIS η2 P= 0.166; p<0.001 η2 P= 0.127; p<0.001 η2 P= 0.068; p<0.001 η2 P= 0.140; p<0.001 Global model (with interaction effects) R Squared 0.253 0.174 0.163 0.235 Main effects Sex η2 P= 0.005; p<0.001 η2 P<0.001; p= 0.573 η2 P<0.001; p= 0.246 η2 P= 0.006; p<0.001 Age η2 P= 0.001; p= 0.016 η2 P= 0.005; p<0.001 η2 P<0.001; p= 0.483 η2 P= 0.003; p<0.001 FAS η2 P= 0.001; p= 0.103 η2 P<0.001; p= 0.635 η2 P= 0.002; p= 0.002 η2 P<0.001; p=0.929 BMI η2 P= 0.004; p<0.001 η2 P<0.001; p= 0.389 η2 P= 0.001; p= 0.016 η2 P= 0.002; p=0.001 BIP η2 P<0.001; p= 0.852 η2 P<0.001; p= 0.511 η2 P<0.001; p= 0.374 η2 P= 0.002; p=0.007 BIS η2 P= 0.002; p= 0.004 η2 P= 0.012; p<0.001 η2 P= 0.002; p= 0.001 η2 P= 0.010; p<0.001 Interaction 1st order effects Sex x BMI n.s. n.s. η2 P= 0.002; p=0.002 n.s. Sex x BIP η2 P= 0.002; p= 0.004 η2 P= 0.002; p= 0.003 η2 P= 0.002; p= 0.003 η2 P= 0.003; p<0.001 Sex x BIS n.s. η2 P= 0.003; p= 0.001 n.s. n.s. Age x BMI η2 P= 0.004; p<0.001 n.s. n.s. η2 P= 0.002; p=0.007 Age x BIP n.s. n.s. n.s. n.s. Age x BIS n.s. n.s. n.s. n.s. FAS x BMI η2 P= 0.002; p= 0.005 n.s. n.s. n.s. FAS x BIP n.s. n.s. n.s. n.s. FAS x BIS n.s. n.s. n.s. n.s. Interaction 2nd order effects (Sex x Age) x BMI n.s. n.s. η2 P= 0.003; p= 0.001 n.s. (Sex x Age) x BIP n.s. n.s. n.s. n.s. (Sex x Age) x BIS n.s. n.s. n.s. n.s. (Sex x FAS) x BMI n.s. n.s. n.s. n.s. (Sex x FAS) x BIP η2 P= 0.003; p= 0.001 η2 P= 0.003; p= 0.005 n.s. η2 P= 0.005; p<0.001 (Sex x FAS) x BIS η2 P= 0.004; p<0.001 η2 P= 0.003; p= 0.001 η2 P= 0.003; p=0.001 η2 P= 0.008; p<0.001 (Age x FAS) x BMI n.s. n.s. n.s. n.s. (Age x FAS) x BIP n.s. n.s. n.s. n.s. (Age x FAS) x BIS n.s. n.s. n.s. n.s. FAS, Family Affluence Scale; BMI, Body Mass Index; BIP, Body Image Perception; BIS, Body Image Satisfaction; n.s., not significant. Following this initial linear regression analysis, interaction of the sociodemographic variables was included both independently –as shown in Table 6– as well as simultaneously combining two of these demographic variables –as shown in Table 7. This exhaustive control of demographic variables increased the percentages of explained variance of the four regression analyses. Thus, the global scale showed a 23.5% predictive capacity, 25.3% for anxious/depressed, 17.4% for withdrawn/depressed, and 16.3% for somatic complaints. In addition, congruent with the aforementioned, the main effects of these global linear regression models once again demonstrated the importance of BIS. Specifically, BIS is the only dependent variable amongst the first three positions in the prediction of all the dependent variables, that is to say, both in the global scale and the three subscales. Closely analyzing the first-order interaction effects (see Table 6), BMI was found to predict somatic complaints only in girls (r= 0.138), but not in their male peers (r= 0.041). However, delving deeper and considering the second-order interaction effects, upon combining the influence of sex and age, BMI showed a double inequality with respect to its influence on somatic complaints. Specifically, the positive relationship between BMI and somatic complaints occurred both in boys and girls at 13–14 years old, without differences between them, with a Pearson’s correlation of 0.165 and 0.161, respectively. However, from 15 years old onward a relationship between Frontiers in Psychology | www.frontiersin.org 9August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 16 Ramos et al. Body Image and Internalizing Symptoms In the first part of this study, dedicated to the bivariate results, an association was also found between a higher frequency of physical activity and higher BIS, and with a lower presence of IS. Along these lines Hausenblas and Fallon (2006) and Holmqvist and Frisen (2019) found that regular physical activity is a protective factor against the development of body image problems. In fact, the results of our study found that even a low frequency of physical activity is beneficial in comparison to its complete absence. Likewise, in the logistical regression shown in the last part of the Results, it was found that MVPA, even on a low level, is a protective factor amongst adolescents with low BIS, who more frequently show IS. These results are especially relevant for intervention given the health promotion recommendations to increase healthy food consumption and the frequency of physical activity for longterm body weight control and to maintain a healthy weight (Centers for Disease Control and Prevention, 1997;European Commission for Health, and Consumer Protection, 2003). However, according to the displacement hypothesis (Andersen et al., 1998), it is increasingly difficult for adolescents to dedicate time to physical activity given the increase in sedentary activities in this developmental stage, such as in time dedicated to school work or recreation related to small screens like watching videos or playing videogames. Therefore, these sedentary activities would compete for the adolescents’ available time for physical activity. However, the fact that our study has demonstrated that a low frequency of physical activity (compared to none) had a higher predictive capacity on internalizing symptoms of the adolescents with low BIS, gives hope that this behavior could really be carried out during this developmental stage on a beneficial level. In this sense, health programs trying to promote physical activity in adolescents –as part of a healthy lifestyle– should take into consideration this question related to the effectiveness of physical activity also on a low level. Limitations and Strengths of This Study This study has some limitations that should be taken into consideration when interpreting its results. Firstly, the crosssectional design means that the results must be interpreted on an associative level, and it is not possible to draw conclusions about the directionality of the relationships found. Secondly, some methodological limitations should be mentioned. Several variables were assessed by frequency, not quantity. For example, VPA was measured with a seven-point Likert scales but recoded into a three-level categorical variable to facilitated data analysis. This reduced information diversity. Thirdly, it would be appropriate to include different contextual factors in the analysis. Further research should explore how family, peers, and school environment influence adolescent body image and the relationship with internalizing symptoms. Fourthly, selfreported BMI was used as an anthropometric measure, which may lead one to question its validity. However, previous studies have shown an elevated coincidence between BMI information provided by adolescents and the real measures taken on their weight and size. For example, Strauss (1999) reported more than a 94% coincidence between the information provided by 12– 16 year old youth regarding their BMI and the scores obtained through precise and objective measures. Nonetheless, all the variables in this study were self-reported, which could imply a mono-method bias. Therefore, it could be convenient for future research to replicate these analysis based on measures obtained from other informants, not only in aspects such as BMI, but also in other questions such as, for example, the SES of the family. Despite the aforementioned limitations, this study also has important strengths. This research offers information about a large and representative sample. Specifically, the sample selection was conducted under strict methodological guidelines, obtaining a representative sample of school-aged adolescents by age, type of school (public or private), and region of Spain. This size and representativeness support our recommendations for practical implications derived from the thoroughly discussed results. In addition, the instruments used add to the strength of this research. On one hand, as has been explained in the Method, evaluation measures from classic and internationally acclaimed adolescent research, like the HBSC, have been used. On the other, this study is innovative in its addition of the YSR, a very exhaustive instrument currently recognized in psychopathological evaluation. This evaluation instrument uses a dimensional approach when assuming that the adolescents’ behaviors, as part of a continuum, could vary in their degree as well as be condensed in groups of symptoms or problems (Achenbach and Rescorla, 2001). Thus, by using this instrument this study has avoided adopting a purely taxonomic approach of mental health disorders, which has shown a higher identification of false positives (Helzer et al., 2008). Finally, this research not only considered representative variables of all body image dimensions and used a validated instrument to evaluate internalizing symptoms, but in addition their associations were analyzed to observe the effects of sex, age, and socioeconomic inequalities on their relationship allowing double social inequalities to be detected. CONCLUSION Based on a sample and quality evaluation instruments, this paper has identified the roles of sex, age, and FAS in the relationship between body image problems and IS in a representative sample of adolescents. On one hand, of the three dimensions in the scale of IS (anxious/depressed, withdrawn/depressed, and somatic complaints), it is anxious/depressed that is best explained by the combination of BMI, overweight BIP, BIS, sex, age, and FAS. On the other hand, between the different components of body image, BIS showed a higher predictive capacity both on the three subscales and on the global scale of IS. In addition, the large sample size used in this study allowed our analyses to detect possible double inequality effects. For example, this study found that whereas boys are more vulnerable in the high relationship between body image problems and IS when they have low FAS, in the case of girls they are most vulnerable in high FAS. These results require prevention programs and interventions to adapt to the specific Frontiers in Psychology | www.frontiersin.org 16 August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 17 Ramos et al. Body Image and Internalizing Symptoms combination of sociodemographic characteristics of the people with whom they work. On the other hand, this study has found interesting results regarding the role that body image behaviors have, such as dieting to lose or gain weight/volume, and physical activity, when analyzing in detail those adolescents with worse scores in BIS. In this sense, a strong relationship is found between low BIS and dieting to lose weight/volume that occurs independently from whether the adolescents show IS or not. Keeping in mind the low nutritional quality of diets that adolescents usually do (as has been discussed above), it is indispensable to prevent them from adopting these weight control behaviors as well as promote eating habits that achieve weight control in a healthier way. Likewise, this study finds that even a low frequency of physical activity, compared to none, is beneficial and protects adolescents against IS. This finding also supports interventions focused on this area, which may be able to set less demanding and more achievable goal since adolescents generally have to juggle many activities simultaneously. DATA AVAILABILITY The datasets for this study will not be made publicly available because the data belong to the HBSC study network. The Spanish Ministry of Health, Social Services and Equality has a link to request the dataset in http://www.mscbs.gob.es/, within the following links: Sanidad – Profesionales – Salud pública – Prevención y Promoción – Promoción de la salud – Salud de los Jóvenes. Requests to access the datasets should be directed to CM: [email protected]. ETHICS STATEMENT The protocol was approved by the “Comité de Ética de la Investigación en Andalucía (Sistema Sanitario Público Andaluz).” Informed consent was obtained in accordance with the Declaration of Helsinki. AUTHOR CONTRIBUTIONS All authors conceived of the study, participated in its design and helped to draft the manuscript. Likewise, all authors made suggestions and critical reviews to the initial draft and contributed to its improvement until reaching the final manuscript, which was read and approved by all authors. FUNDING This work was supported in Spain by the Spanish Ministry of Health, Social Services and Equality (PRJ201703153). ACKNOWLEDGMENTS We thank the participating adolescents and schools, the research assistants from the Spanish HBSC team who took part in this study and Ian Scionti for providing language help. REFERENCES AbouZahr, C. (2014). Progress and challenges in women’s health: an analysis of levels and patterns of mortality and morbidity. Contraception 90, 3–13. doi: 10.1016/j.contraception.2014.03.007 Achenbach, T. M. (2013). DSM Guide for the ASEBA. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families. Achenbach, T. M., and Rescorla, L. A. (2001). Manual for the ASEBA School-Age Forms & Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families. Acker, J. (2006). Class Questions: Feminist Answers. Lanham, MD: Rowman & Littlefield Publishers Inc. Andersen, R. E., Crespo, C. J., Bartlett, S. J., Cheskin, L. J., and Pratt, M. (1998). Relationship of physical activity and television watching with body weight and level of fatness among children: results from the third national health and nutrition examination survey. J. Am. Med. Assoc. 279, 938–942. doi: 10.1001/ jama.279.12.938 Andrews, P. W., and Thompson, J. A. (2009). The bright side of being blue: depression as an adaptation for analyzing complex problems. Psychol. Rev. 116, 620–654. doi: 10.1037/a0016242 Bacopoulou, F., Foskolos, E., Stefanaki, C., Tsitsami, E., and Vousoura, E. (2018). Disordered eating attitudes and emotional/behavioral adjustment in Greek adolescents. Eat. Weight Disord. 23, 621–628. doi: 10.1007/s40519-017-0466-x Banfield, S. S., and McCabe, M. P. (2002). An evaluation of the construct of body image. Adolescence 37, 373–393. Barker, E., and Galambos, N. (2003). Body dissatisfaction of adolescent girls and boys: risk and resource factors. J. Early Adolescence 23, 141–165. doi: 10.1177/ 0272431603023002002 Bessenoff, G., and Snow, D. (2006). Absorbing society’s influence: body image selfdiscrepancy and internalized shame. Sex Roles 54, 727–731. doi: 10.1007/ s11199-006-9038-7 Borraccino, A., Lemma, P., Iannotti, R., Zambon, A., Dalmasso, P., Lazzeri, G., et al. (2009). Socioeconomic effects on meeting physical activity guidelines: comparisons among 32 countries. Med. Sci. Sport Exerc. 41, 749–756. doi: 10.1249/MSS.0b013e3181917722 Brooks-Gunn, J. (1991). “How stressful is the transition to adolescence for girls?,” in Adolescent Stress: Causes and Consequences, eds M. E. Colten, and S. Gore, (New York, NY: Aldine de Gruyter), 131–149. Bully, P., and Elosua, P. (2011). Changes in body dissatisfaction relative to gender and age: the modulating character of BMI. Span. J. Psychol. 14, 313–322. doi: 10.5209/rev_SJOP.2011.v14.n1.28 Cafri, G., and Thompson, J. K. (2004). Measuring male body image: a review of the current methodology. Psychol. Men Masc. 5, 18–29. doi: 10.1037/1524-9220.5 Cash, T. F. (2004). Body image: past, present, and future. Body Image 1, 1–5. doi: 10.1016/S1740-1445(03)00011-1 Centers for Disease Control and Prevention, (1997). Guidelines for school health programs to promote lifelong healthy eating. J. Sch. Health 67, 9–26. doi: 10.1111/j.1746-1561.1997.tb06289.x Chaplin, T. M., and Aldao, A. (2013). Gender differences in emotion expression in children: a meta-analytic review. Psychol. Bull. 139, 735–765. doi: 10.1037/ a0030737 Cheung, Y. T., Lee, A. M., Ho, S. Y., Li, E. T., Lam, T. H., Fan, S. Y., et al. (2011). Who wants a slimmer body? The relationship between body weight status, education level and body shape dissatisfaction among young adults in Hong Kong. BMC Public Health 11:835. doi: 10.1186/1471-2458- 11-835 Frontiers in Psychology | www.frontiersin.org 17 August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 18 Ramos et al. Body Image and Internalizing Symptoms Chu, T. L., Martin, S. B., Petrie, T. A., and Greenleaf, C. (2019). Relations between sociocultural pressures and weight control behavior among early adolescent boys. Psychol. Sch. 56, 163–175. doi: 10.1002/pits.22212 Coakley, J., and White, A. (1992). Making decisions: gender and sport participation among British adolescents. Sociol. Sport J. 9, 20–35. doi: 10.1123/ssj.9.1.20 Cohen, J. (1988). Statistical Power Analysis for the Behavioral Science, 2nd Edn. Hillsdale, NJ: Lawrence Erlbaum Associates. Cok, F. (1990). Body image satisfaction in Turkish adolescents. Adolescence 25, 409–413. Cole, T. J., Bellizzi, M. C., Flegal, K. M., and Dietz, W. H. (2000). Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 320, 1240–1243. doi: 10.1136/bmj.320.7244.1240 Cole, T. J., Flegal, K. M., Nicholls, D., and Jackson, A. A. (2007). Body mass index cut offs to define thinness in children and adolescents: international survey. BMJ 335:194. doi: 10.1136/bmj.39238.399444.55 Currie, C., Inchley, J., Molcho, M., Lenzi, M., Veselska, Z., and Wild, F. (2014). Health Behaviour in School-Aged Children (HBSC) Study Protocol: Background, Methodology and Mandatory items for the 2013/14 Survey. St. Andrews: CAHRU. Eisenberg, M. E., Neumark-Sztainer, D., Story, M., and Perry, C. (2005). The role of social norms and friends’ influences on unhealthy weightcontrol behaviors among adolescent girls. Soc. Sci. Med. 60, 1165–1173. doi: 10.1016/j.socscimed. 2004.06.055 European Commission for Health, and Consumer Protection, (2003). Status Report on the European Commission’s Work in the Field of Nutrition in Europe. Luxembourg: Office for Official Publications of the European Communities. Feng, X., and Wilson, A. (2016). Neighborhood socioeconomic disadvantage modifies the relationship between weight status and weight-related satisfaction. Int. J. Obes. 40, 1922–1926. doi: 10.1038/ijo.2016.120 Fismen, A.-S., Samdal, O., and Torsheim, T. (2012). Family affluence and cultural capital as indicators of social inequalities in adolescent’s eating behaviours: a population-based survey. BMC Public Health 12:1036. doi: 10.1186/1471-2458- 12-1036 Graber, J. A., and Brooks-Gunn, J. (1996). Transitions and turning points: navigating the passage from childhood through adolescence. Dev. Psychol. 32, 768–776. doi: 10.1037/0012-1649.32.4.768 Graber, J. A., and Sontag, L. M. (2009). “Internalizing problems during adolescence,” in Handbook of Adolescent Psychology, eds R. M. Lerner, and L. Steinberg, (New York, NY: John Wiley and Sons Inc.), 642–682. Grogan, S. (2006). Body image and health contemporary perspectives. J. Health Psychol. 11, 523–530. doi: 10.1177/1359105306065013 Harter, S. (2006). “The self,” in Handbook of Child Psychology: Social, Emotional, and Personality Development, Vol. 3, eds W. Damon, and R. M. Lerner, (Hoboken, NJ: John Wiley & Sons), 505–570. Hausenblas, H. A., and Fallon, E. A. (2006). Exercise and body image: a metaanalysis. Psychol. Health 21, 33–47. doi: 10.1080/14768320500105270 Hayward, C., and Sanborn, K. (2002). Puberty and the emergence of gender differences in psychopathology. J. Adolescent Health 30, 49–58. doi: 10.1016/ S1054-139X(02)00336-1 Helzer, J. E., Kraemer, H. C., Krueger, R. F., Wittchen, H. U., Sirovatka, P. J., and Regier, D. A. (2008). Dimensional Approaches in Diagnostic Classification: Refining the Research Agenda for DSM-V. Washington, DC: American Psychiatric Association. Holmqvist, K., and Frisen, A. (2019). From negative to positive body image: men’s and women’s journeys from early adolescence to emerging adulthood. Body Image 28, 53–65. doi: 10.1016/j.bodyim.2018.12.002 Hyde, J. S., Mezulis, A. H., and Abramson, L. Y. (2008). The ABCs of depression: integrating affective, biological, and cognitive models to explain the emergence of the gender difference in depression. Psychol. Rev. 115, 291–313. doi: 10.1037/ 0033-295X.115.2.291 Inchley, J., and Currie, C. (2004). Summary of Key Findings From the 2003/04 Pupil Survey. Findings from the Physical Activity in Scottish Schoolchildren (PASS) Survey. Available at: https://risweb.standrews.ac.uk/portal/en/researchoutput/report-of-findings-from-the-pass- 200304-pupil-survey(a4228636-8b81-49a8-83b2-8aca4bc2958d).html Inchley, J., Currie, D., Young, T., Samdal, O., Torsheim, T., Augustson, L., et al. (2016). Growing up Unequal: Gender and Socioeconomic Differences in Young People’s Health and Well-being: Health Behaviour in School-Aged Children (HBSC) Study: International Report from the 2013/2014 Survey. (Health Policy for Children and Adolescents, No. 7). Copenhagen: WHO Regional Office for Europe. Ingram, R. E., and Luxton, D. D. (2005). “Vulnerability-stress models,” in Development of Psychopathology: A Vulnerability-Stress Perspective, eds B. L. Hankin, and J. R. Z. Abela, (Thousand Oaks, CA: Sage Publications Inc.), 32–46. doi: 10.4135/9781452231655.n2 Janssen, I., Katzmarzyk, P. T., Boyce, W. F., Vereecken, C., Mulvihill, C., Roberts, C., et al. (2005). Comparison of overweight and obesity prevalence in schoolaged youth from 34 countries and their relationships with physical and dietary patterns. Obes. Rev. 6, 123–132. doi: 10.1111/j.1467-789x.2005.00176.x Kanaan, M. N., and Afifi, R. A. (2010). Gender differences in determinants of weight-control behaviours among adolescents in Beirut. Public Health Nutr. 13, 71–81. doi: 10.1017/S136898000900500X Larsen, P. S., Strandberg-Larsen, K., Micali, N., and Nybo Andersen, A.-M. (2015). Parental and child characteristics related to early-onset disordered eating: a systematic review. Harv. Rev. Psychiatry 23, 395–412. doi: 10.1097/HRP. 0000000000000073 Larson, N. I., Neumark-Sztainer, D., and Story, M. (2009). Weight control behaviors and dietary intake among adolescents and young adults: longitudinal findings from project EAT. J. Am. Diet Assoc. 109, 1869–1877. doi: 10.1016/j. jada.2009.08.016 Lazzeri, G., Ahluwalia, N., Niclasen, B., Pammolli, A., Vereecken, C., Rasmussen, M., et al. (2016). Trends from 2002 to 2010 in daily breakfast consumption and its socio-demographic correlates in adolescents across 31 countries participating in the HBSC study. PLoS One 11:e0151052. doi: 10.1371/journal. pone.0151052 Lerner, R. M. (2018). Concepts and Theories of Human Development. New York, NY: Routledge. Marmot, M., and Bell, R. (2012). Fair society, healthy lives. Public Health 126, S4–S10. doi: 10.1016/j.puhe.2012.05.014 McArthur, L., Holbert, D., and Peña, M. (2005). An exploration of the attitudinal and perceptual dimensions of body image among male and female adolescents from six latin american cities. Adolescence 40, 801–816. McCabe, M. P., and Ricciardelli, L. A. (2004). Body image dissatisfaction among males across the lifespan: a review of past literature. J. Psychosom. Res. 56, 675–685. doi: 10.1016/S0022-3999(03)00129-6 McClintock, J. M. (2001). The underlying psychopathology of eating disorders and social phobia: a structural equation analysis. Eat. Behav. 2, 247–261. doi: 10.1016/S1471-0153(01)00032-0 McGuire, M. T., Wing, R. R., Klem, M. L., and Hill, J. O. (1999). The behavioral characteristics of individuals who lose weight unintentionally. Obes. Res. 7, 485–490. doi: 10.1002/j.1550-8528.1999.tb00437.x Mills, J. S., Roosen, K., and Vella-Zarb, R. (2011). “The psychology of body image: Understanding body image instability and distortion,” in Body Image: Perceptions, Interpretations, and Attitudes, ed. S. B. Greene, (New York, NY: Nova Science Publishers), 59–80. Mischner, I., van Schie, H., and Engels, R. (2013). Breaking the circle: challenging western sociocultural norms for appearance influences young women’s attention to appearance-related media. Body Image 10, 316–325. doi: 10.1016/j. bodyim.2013.02.005 Moreno, C., Ramos, P., Rivera, F., Jiménez-Iglesias, A., García-Moya, I., Sánchez- Queija, I., et al. (2016). Los adolescentes españoles: estilos de vida, salud, ajuste psicológico y relaciones en sus contextos de desarrollo. Resultados del Estudio HBSC-2014 en España. Madrid: Ministerio de Sanidad, Servicios Sociales e Igualdad. Moreno, C., Ramos, P., Sánchez-Queija, I., Rivera, F., Jiménez-Iglesias, A., García- Moya, I., et al. (2018). Informe técnico de la recogida de datos (cuestionario empleado y muestreo seguido) en el Estudio HBSC (Health Behaviour in School- Aged Children) - 2018. Madrid: Ministerio de Sanidad y Política Social. Moreno-Maldonado, C., Ramos, P., Moreno, C., and Rivera, F. (2018). How family socioeconomic status, peer behaviors, and school-based intervention on healthy habits influence adolescent eating behaviors. Sch. Psychol. Int. 39, 92–118. doi: 10.1177/0143034317749888 Neumark-Sztainer, D., Paxton, S. J., Hannan, P. J., Haines, J., and Story, M. (2006). Does body satisfaction matter? Five-year longitudinal associations between body satisfaction and health behaviors in adolescent females and males. J. Adolescent Health 39, 244–251. doi: 10.1016/j.jadohealth.2005.12.001 Frontiers in Psychology | www.frontiersin.org 18 August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 19 Ramos et al. Body Image and Internalizing Symptoms Neumark-Sztainer, D., Story, M., Falkner, N., Beuhring, T., and Resnick, M. (1999). Sociodemographic and personal characteristics of adolescents engaged in weight loss and weight/muscle gain behaviors: who is doing what? Prev. Med. 28, 40–50. doi: 10.1006/pmed.1998.0373 Neziroglu, F., Khemlani-Patel, S., and Veale, D. (2008). Social learning theory and cognitive behavioral models of body dysmorphic disorder. Body Image 5, 28–38. doi: 10.1016/j.bodyim.2008.01.002 O’Dea, J. A., and Caputi, P. (2001). Association between socioeconomic status, weight, age and gender, and the body image and weight control practices of 6- to 19-year-old children and adolescents. Health Educ. Res. 16, 521–532. doi: 10.1093/her/16.5.521 Oliva, A., Parra, A., and Reina, M. C. (2015). Personal and contextual factors related to internalizing problems during adolescence. Child Youth Care For. 43, 505–520. doi: 10.1007/s10566-014-9250-5 Orbach, I., and Mikulincer, M. (1998). Body investment scale: construction and validation of a body experience scale. Psychol. Assess. 10, 415–425. doi: 10.1037/ 1040-3590.10.4.415 Overton, W. F. (2014). The process-relational paradigm and relationaldevelopmental-systems metamodel as context. Res. Hum. Dev. 11, 323–331. doi: 10.1080/15427609.2014.971549 Overton, W. F. (2015). “Process and relational developmental systems,” in Handbook of Child Psychology and Developmental Science: Theory and Method, eds W. F. Overton, and P. C. M. Molenaar, (Hoboken, NJ: Wiley), 9–62. Ozer, E. M., and Irwin, C. E. (2009). “Adolescent and Young adult health: from basic health status to clinical interventions,” in Handbook of Adolescent Psychology: Individual Bases of Adolescent Development, eds R. Lerner, and L. Steinberg, (Hoboken, NJ: Wiley), 618–641. Patalay, P., and Hardman, C. (2019). Comorbidity, codevelopment, and temporal associations between body mass index and internalizing symptoms from early childhood to adolescence. JAMA Psychiatry 76, 721–729. doi: 10.1001/ jamapsychiatry.2019.0169 Pereira, E. F., Graup, S., Lopes, A. S., Borgatto, A. F., and Daronco, L. S. (2009). Percepção da imagem corporal de crianças e adolescentes com diferentes níveis socioeconômicos na cidade de Florianópolis, Santa Catarina, Brasil. Rev. Bras. Saude Matern. Infant 9, 253–262. doi: 10.1590/S1519-382920090003 00004 Pereira, E. F., Teixeira, C. S., Gattiboni, B. D., Bevilacqua, L. A., Confortin, S. C., and Rodriguez de Silva, T. (2011). Adolescent body image perceptions and socioeconomic status: a systematic review. Rev. Paul. Pediatr. 29, 423–429. doi: 10.1590/S0103-05822011000300018 Prochaska, J. J., Sallis, J. F., and Long, B. (2001). A physical activity screening measure for use with adolescents in primary care. Arch. Pediatr. Adolesc. Med. 155, 554–559. doi: 10.1001/archpedi.155.5.554 Puccio, F., Fuller-Tyszkiewicz, M., Ong, D., and Krug, L. (2016). A systematic review and meta-analysis on the longitudinal relationship between eating pathology and depression. Int. J. Eat. Disord. 49, 439–454. doi: 10.1002/eat. 22506 Puhl, R. M., and Latner, J. D. (2007). Stigma, obesity, and the health of the nation’s children. Psychol. Bull. 133, 557–580. doi: 10.1037/0033-2909.133.4.557 Ramos, P., Brooks, F., García-Moya, I., Rivera, F., and Moreno, C. (2013). Eating habits and physical activity in dieter and non-dieter youth: a gender analysis of English and Spanish adolescents. Soc. Sci. J. 50, 575–582. doi: 10.1016/j.soscij. 2013.09.017 Ramos, P., Rivera, F., and Moreno, C. (2012). “Key variables in the development of eating disorders during the adolescence: implications for sex differences,” in Relevant Topics in Eating Disorders, ed. I. Jauregui, (Rijeka: InTech), 31–44. Roberts, C., Freeman, J., Samdal, O., Schnohr, C., de Looze, M. E., NicGabhainn, S., et al. (2009). The Health Behaviour in School-aged Children (HBSC) study: methodological developments and current tensions. Int. J. Public Health 54, 140–150. doi: 10.1007/s00038-009-5405-9 Rotenberg, K. J., Taylor, D., and Davis, R. (2004). Selective moodinduced body image disparagement and enhancement effects: are they due to cognitive priming or subjective mood? Int. J. Eat. Disord. 35, 317–332. doi: 10.1002/eat. 10263 Rutter, M., Caspi, A., and Moffitt, T. E. (2003). Using sex differences in psychopathology to study causal mechanisms: unifying issues and research strategies. J. Child Psychol. Psychiatry 44, 1092–1115. doi: 10.1111/1469-7610. 00194 Schnohr, C. W., Makransky, G., Kreiner, S., Torsheim, T., Hofmann, F., De Clercq, B., et al. (2013). Item response drift in the family affluence scale: a study on three consecutive surveys of the Health Behaviour in School-aged Children (HBSC) survey. Measurement 46, 3119–3126. doi: 10.1016/j.measurement.2013. 06.016 Shagar, P. S., Harris, N., Boddy, J., and Donovan, C. L. (2017). The relationship between body image concerns and weight-related behaviours of adolescents and emerging adults: a systematic review. Behav. Change 34, 208–252. doi: 10.1017/bec.2018.3 Shomaker, L. B., and Furman, W. (2010). A prospective investigation of interpersonal influences on the pursuit of muscularity in late adolescent boys and girls. J. Health Psychol. 15, 391–404. doi: 10.1177/1359105309350514 Siegel, J. M. (2002). Body image change and adolescent depressive symptoms. J. Adolescent Res. 17, 27–41. doi: 10.1177/0743558402171002 Sobrino-Bazaga, A., and Rabito-Alcoón, M. F. (2018). Gender differences existing in the general population in relation to body dissatisfaction associated with the ideal of beauty: a systematic review. Stud. Psychol. 39, 548–587. doi: 10.1080/ 02109395.2018.1517860 Stalsberg, R., and Pedersen, A. V. (2010). Effects of socioeconomic status on the physical activity in adolescents: a systematic review of the evidence. Scand. J. Med. Sci. Sports 20, 368–383. doi: 10.1111/j.1600-0838.2009.01047.x Stice, E., Trost, A., and Chase, A. (2003). Healthy weight control and dissonancebased eating disorder prevention programs: results from a controlled trial. Int. J. Eat. Disord. 33, 10–21. doi: 10.1002/eat.10109 Story, M., French, S. A., Resnick, M. D., and Blum, R. W. (1995). Ethnic/racial and socioeconomic differences in dieting behaviors and body image perceptions in adolescents. Int. J. Eat. Disord. 18, 173–179. doi: 10.1002/1098-108x(199509) 18:2<173::aid-eat2260180210>3.0.co;2-q Strauss, R. S. (1999). Comparison of measured and self-reported weight and height in a cross-sectional sample of young adolescents. Int. J. Obes. 23, 904–908. doi: 10.1038/sj.ijo.0800971 Ter Bogt, T., van Dorsselaer, S., Monshouwer, K., Verdurmen, J., Engels, R., and Vollebergh, W. (2006). Body mass index and body weight perception as risk factors for internalizing and externalizing problem behavior among adolescents. J. Adolescent Health 39, 27–34. doi: 10.1016/j.jadohealth.2005.09. 007 Thompson, J. K., Coovert, M. D., Richards, D. J., Johnson, S., and Cattarin, J. (1995). Development of body image, eating disturbance, and general psychological functioning in female adolescents: covariance structure modelling and longitudinal investigations. Int. J. Eat. Disord. 18, 221–236. doi: 10.1002/1098-108x(199511)18:3<221::aid-eat2260180304>3.0.co;2-d Tiggemann, M., Gardiner, M., and Slater, A. (2000). “I would rather be size 10 than have straight A’s”: a focus group study of adolescent girls’ wish to be thinner. J. Adolesc. 23, 645–659. doi: 10.1006/jado.2000.0350 Timlin, M. T., Pereira, M. A., Story, M., and Neumark-Sztainer, D. (2008). Breakfast eating and weight change in a 5-year prospective analysis of adolescents: project EAT (eating among teens). Pediatrics 121, 638–645. doi: 10.1542/peds.2007-1035 Turrell, G., Hewitt, B., Patterson, C., and Oldenburg, B. (2003). Measuring socio-economic position in dietary research: is choice of socio-economic indicator important? Public Health Nutrit. 6, 191–200. doi: 10.1079/PHN200 2416 Vereecken, C., Pedersen, T. P., Ojala, K., Krølner, R., Dzielska, A., Ahluwalia, N., et al. (2015). Fruit and vegetable consumption trends among adolescents from 2002 to 2010 in 33 countries. Eur. J. Public Health 25, 16–19. doi: 10.1093/ eurpub/ckv012 Viner, R. M., Ozer, E. M., Denny, S., Marmot, M., Resnick, M., Fatusi, A., et al. (2012). Adolescence and the social determinants of health. Lancet 379, 1641– 1652. doi: 10.1016/s0140-6736(12)60149-4 Wadsworth, M. E., and Achenbach, T. M. (2005). Explaining the link between low socioeconomic status and psychopathology: testing two mechanisms of the social causation hypothesis. J. Consult. Clin. Psychol. 73, 1146–1153. doi: 10.1037/0022-006X.73.6.1146 Walby, S. (2011). “Globalization and multiple inequalities,” in Analyzing Gender, Intersectionality, and Multiple Inequalities: Global, Transnational and Local Contexts: Advances in Gender Research, Vol. 15, eds E. Ngan-Ling Chow, and Frontiers in Psychology | www.frontiersin.org 19 August 2019 | Volume 10 | Article 1952 fpsyg-10-01952 August 21, 2019 Time: 17:31 # 20 Ramos et al. Body Image and Internalizing Symptoms M. Texler Segal, (Bingley: Emerald Group Publishing Limited), 17–33. doi: 10.1108/s1529-2126(2011)0000015006 Wang, Z., Byrne, N. M., Kenardy, J. A., and Hills, A. P. (2005). Influences of ethnicity and socioeconomic status on the body dissatisfaction and eating behaviour of Australian children and adolescents. Eat. Behav. 6, 23–33. doi: 10.1016/j.eatbeh.2004.05.001 Ward, P. R., Verity, F., Carter, P., Tsourtos, G., Coveney, J., and Wong, K. C. (2013). Food stress in adelaide: the relationship between low income and the affordability of healthy food. J. Environ. Public Health 10:968078. doi: 10.1155/ 2013/968078 Yannakoulia, M., Lykou, A., Kastorini, C. M., Papasaranti, E. S., Petralias, A., Veloudaki, A., et al. (2016). Socio-economic and lifestyle parameters associated with diet quality of children and adolescents using classification and regression tree analysis: the DIATROFI study. Public Health Nutr. 19, 339–347. doi: 10. 1017/S136898001500110X Zeller, M. H., Saelens, B. E., Roehrig, H., Kirk, S., and Daniels, S. (2004). Psychological adjustment of obese youth presenting for weight management treatment. Obes. Res. 12, 1576–1586. doi: 10.1038/oby.2004.197 Conflict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Copyright © 2019 Ramos, Moreno-Maldonado, Moreno and Rivera. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Frontiers in Psychology | www.frontiersin.org 20 August 2019 | Volume 10 | Article 1952