Physical Activity and Body Image Perception in Adolescents: A Systematic Review

Adolescence represents a vulnerable phase of life for psychological health. The practice of physical activity (PA) appears to have a positive influence on adolescents, increasing self-esteem and producing a more positive body image. A systematic review of published articles over the past 10 years until June 2022 was conducted according to the PRISMA statement employing the electronic databases MEDLINE and Web of Science (639 records) to summarize the literature on the relationship between body image dissatisfaction (BID) and assessed by figural scales and practice of structured and unstructured PA in adolescents (10–18 years), taking into account BMI and/or weight status. All articles were independently reviewed using inclusion/exclusion criteria, retrieved data, and assessed quality with the adapted Newcastle-Ottawa Scale for observational studies. The main finding of interest that emerged from most of the 28 included studies is the negative association between BID and PA during adolescence: as PA increases, BID decreases. However, this updated systematic review also identified some flaws in the existing literature, highlighting the need for high-quality adolescent research using validated figural scales and objective PA assessments. In conclusion, the reviewed studies showed that PA involvement can be efficacious in protecting from body image perception concerns and enhancing body satisfaction. Future interventions should promote structured and unstructured PA during adolescence to improve self-esteem and body image.


Introduction
Overweight/obesity is a public health problem constituting an independent risk factor for several chronic non-communicable diseases and increased mortality [1,2]. Globally, there has been an increase in obesity and a decline in the underweight: the increase in obesity, including severe obesity, may be linked to increasing social disparities that limit healthy food access for people at the greatest risk of malnutrition [3]. According to WHO [4], the worldwide obesity prevalence nearly increased three-fold between 1975 and 2016, and more than 340 million children and adolescents (ages 5 to 19) were overweight or obese in 2016. During the recent COVID-19 pandemic, the increase in the rate of body mass index (BMI) almost doubled in comparison with the pre-pandemic period in persons aged 2-19 years [5].
There is scientific evidence that an unhealthy body image (BI) is linked to obesity and physical inactivity, exerting a key role in the emergence of eating disorders during adolescence. The image we have of our bodies is a dynamic feature that can change over a lifetime but the stage of adolescence is crucial for the development of a healthy BI due to the age-related transitions that occur during this period [6]. Sociocultural context greatly influences how adolescents perceive their bodies. The influence of body ideals interacts with the critical period of adolescence characterized by fast and diversified physical modifications that include changes in primary and secondary sexual characteristics, body shape, size, and composition. Key influences on BI include media, which target adolescents, and peers, which contribute to perceived body ideal beliefs. Especially the mass media can lead been carried out in this area by figural scales over the past decade; iii. to extend the analysis on BI or BID to PA contexts related to lifestyle -not just structured PA. In addition, finding possible gaps in the existing literature can direct future research on specific aspects of the relationship between PA and BID.

Materials and Methods
This review was carried out according to the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) [26]. The PRISMA checklist was reported in Supplementary Materials (Table S1). The protocol of this review was registered with PROSPERO (CRD42022342891).

Search Strategy and Selection Criteria
The systematic search was conducted in the PubMed and Web of Science databases on articles published in the last ten years until 13 June 2022 by the following combination of search terms: ("body image" or "body dissatisfaction" or "body satisfaction" or "body representation" or "silhouette scale") and (anthropometry or BMI or weight or "fat mass" or adiposity or "body composition") and ("sport activity" or "physical activity") and adolescen*. No geographical or cultural limitations were applied. The topic analysis was first discussed and agreed upon by all three reviewers. Articles resulting from database searches were then imported into Covidence software (Veritas Health Innovation Ltd, Melbourne, Australia), where duplicate records were removed. The search was then continued manually and independently by two reviewers (L.Z. and N.R.), discussing with the third reviewer (E.G.-R.) in case of discrepancies. Further eligible studies were included through consultation of the reference lists of previously selected articles.
All quantitative studies that investigated the possible association between BI or BID and PA, taking body composition into account, were included. Editorials and commentaries, case studies, protocol studies, conference proceedings, books and book chapters, and theses were excluded. In addition, excluded were studies related to the development and/or validation of BI screening tools, and studies in which BI assessment focused on specific features (e.g., muscularity) or parts of the body (e.g., breasts). We detailed the selection criteria used in Table 1. Table 1. Inclusion and exclusion criteria used in the articles screening.

Inclusion Criteria Exclusion Criteria
-Original article in English -Assessment of body dissatisfaction only by questionnaires -Full-text availability -Assessment of satisfaction/dissatisfaction with specific body features or parts -Adolescents aged 10-18 years or mean age within this range -Development or validation of new BI screening tools -Assessment of body image perception -Studies of individuals affected by pathologies -Use of body silhouettes scale for the assessment of BI and BID -Duplicate studies * or studies with partially overlapping samples -Assessment of PA -Literature reviews -Assessment of indicators of overall and central obesity (at least BMI) -Editorials, commentaries, case and protocol studies, conference proceedings, books and book chapters, theses * In the case of additional data on the same sample, both studies were cited.
The relevant data were extracted from the selected articles as follows: article authors and year of publication; study design; geographical location; race/ethnicity; sample size; sex; age; BMI and/or weight status; PA assessment; silhouette rating scale used; BID and/or satisfaction; main outcomes according to the aims of this study. Due to the lack of quantitative homogeneity in the reviewed studies, we could not conduct a meta-analysis. The findings are therefore presented as a narrative analysis: all studies selected from the database searches were summarized in tabular form and organized in alphabetical order.

Quality Assessment
The quality assessment of selected studies was performed according to the adapted Newcastle-Ottawa Scale (NOS) [27] for observational studies. NOS is based on (1) clarity of the study objective, (2) sample selection (representativeness of the sample; sample size; response rate; ascertainment of exposure), (3) comparability (control of confounding factors; comparability of participants from different outcome groups), (4) outcome (assessment; statistical tests). Ascertainment of exposure (within criterion 2) was made with reference to the figural scale used (validated or not validated). The overall score for each study (range 0-16) is derived from the ratings of the single reported components. Based on Hillen et al. [27], studies that received 13 to 16 out of 16 possible points (scores > 75%) were considered of high quality, 9 to 12 points of moderate quality (scores > 50%), 8 and below of low quality (scores ≤ 50%). A low overall score is indicative of a higher risk of bias in the study. This assessment was independently carried out by two reviewers (N.R and E.G.-R.). Any disagreements between the two reviewers were discussed and resolved with the assistance of the third reviewer (L.Z.).

Study Selection and Description
A total of 639 articles were identified through the search conducted on the databases. Following the article selection procedure (Figure 1), 28 studies were included in the systematic review. After removing 139 duplicates, 500 articles were screened against title and abstract for eligibility for inclusion criteria. Studies excluded after the title and abstract screening were 217. Therefore, 283 studies were assessed for full-text eligibility; however, one study could not be retrieved. Thus, 282 studies were the number of articles assessed for full-text eligibility. Of these, 257 were excluded for several reasons: BI assessed by questionnaire (n = 117); conference proceedings, book or book chapters, case studies, commentary, protocol (n = 6); not written in English language (n = 1); non-compliance with other inclusion/exclusion criteria (n = 114); not related to the topic (n = 19). Therefore, 26 articles met the inclusion criteria. Two additional articles were included in the review from citation searching, so the articles included in the review were 28.

Characteristics of Studies
Twenty-eight studies were included in this review ( Table 2)     BID 5% much heavier than ideal 29% somewhat heavier than ideal 50% same as ideal 15% thinner than ideal (BID: "Much heavier" indicates those whose self-image was >2 levels heavier than their ideal BI, "somewhat heavier" indicates those whose self-image was 1 level heavier than ideal, and "thinner" indicates those whose self-image was thinner than ideal) 7 silhouettes developed by Collins [34] BID was related to the BMI category: obese and OW students were more likely than NW students to identify as being much heavier or somewhat heavier than ideal. Students were less likely to become obese if they performed more vigorous exercise. Association of BI or BID with PA not assessed. The mean female desired body shape: 3.14 ± 0.82.
No BD in 18% of females and 21% of males. Desire for a thinner figure in two-thirds of girls and 44% of the boys. Desire for a bigger figure in 35% of males. 9 silhouettes developed by Stunkard et al. [29] More than 80% of adolescents reported being dissatisfied with their current body shape. Overall BID, when measured as an absolute value, was similar in male and female Thai adolescents but the majority of girls wanted to be thinner, while boys wanted to be either smaller or bigger. Among risk factors, lower socioeconomic status, higher BMI, and lower PA were found to be associated with higher body dissatisfaction explaining 20% of its variation. Participants from single-sex schools reported higher body dissatisfaction than those from mixed schools. In general, Thai adolescents reported a low PA.  (Using 9 figures, the image was assigned to one of four categories: images 1, 2, and 3 = UW, images 4 and 5 = NW, images 6 and 7 = OW, and images 8 and 9 = obesity) 9 silhouettes developed by Stunkard et al. [29] The level of engagement in perceived PA was not significantly associated with BID. In this study, 50% of the girls classified as obese (32%) regarded themselves as being NW. Obesity was not found to be associated with sedentary activities.  [29] Dissatisfaction with body size and ideal body size was not correlated with PA levels in the Ghanaian population. Dissatisfaction with body size in Ghanaians concerned both "too thin" and "too heavy" bodies.      [56] Low engagement in PA was found to be predicted by higher age, female sex, lower to middle SES, environments unfavorable to PA, non-practice of extracurricular sports, and BID. Lower levels of PA were found in students with OW/obesity and dissatisfaction with their BI. In the general sample, no significant correlation was found between PA level and BID in both genders. (Although unspecified, these were probably the silhouettes developed by Kakeshita et al. [48] for the adults) More than half of the participants were dissatisfied and had a distortion of their BI. Girls with an "inactive and sedentary" latent lifestyle were 1.71 times more likely to feel dissatisfied than those with an active and sedentary or inactive and non-sedentary lifestyle. BI disorders were associated with a decreased amount of MVPA. PA was negatively associated with BMI and BID; BMI and BID were positively associated. PA was significantly associated with BID among OW/obese adolescents. No sex differences were found in BID among boys and girls who were OW/obese, and sex-per PA interaction was not significant among these weight status categories. Adolescents, particularly NW boys, with higher levels of PA are less likely to be dissatisfied with their BI.   Desire to be thinner: BID score ≥ 1; desire to be heavier: BID score ≤ −1; satisfaction: BID score = 0.) 9 silhouettes developed by Stunkard et al. [29] About two-fifths of the total sample had a misperception of their weight status, and more than half of the total sample was dissatisfied with their BI. Boys tend to underestimate their weight status, while girls are more likely to overestimate it. In particular, females would be more sensitive to their weight status and BI due to external and internal pressures than males. Only the association between body perception and behaviors of dietary weight management was assessed. The number of subjects surveyed in the studies ranged from a minimum of 36 subjects (a study on selected athletes-male soccer players) to a maximum of 10,496 Finnish adolescents. In particular, looking at the number of subjects examined, only 1 study (3.6%) had fewer than 100 subjects, 14 studies (50.0%) had fewer than 1000 subjects, 11 studies (39.3%) had fewer than 5000 subjects, and only two studies (7.1%) had more than 5000 subjects. Studies with the smallest sample sizes preferably involved only one sex and longitudinal or repeated measures design. Most samples included both genders, 5 studies (17.8%) included only females, and one (3.6%) only males. All the studies were conducted on samples of students except two studies on athletes, two studies on Chinese samples [46,69], that employed a secondary-analyses of data from the China Health and Nutrition Survey (CHNS), and the Indonesian household survey [59]. The majority of studies had a cross-sectional design, and only three studies had a longitudinal or repeated measures design [38,42,63].
For what concerns BMI and weight status, 12 studies (42.9%) reported the percentage of weight status categories (the majority determined from the cut-off proposed by Cole et al. [70,71] adjusted by age and sex, and two by the BMI percentiles (Underweight: <5th BMI-P; Normal weight: 5th to 85th; Overweight: 85th to 95th; Obese: ≥95th), while two others used Z-score of BMI for age (BMIZ) (thinness was defined as BMIZ less than −2 and overweight or obesity above 1), 7 (25%) reported mean values of BMI, and nine (32.1%) reported both mean values of BMI and the percentages of weight status categories.
The mean BMI of the participants included in the samples retrieved in the present review ranged from 16.9 kg/m 2 of the youngest female sample of rhythmic gymnasts aged 10.7 years [53] to 22.5 kg/m 2 of Spanish male sedentary students aged 12-17 years [50]. The female adolescents of Kuwait reported the highest percentage of OW/obesity (61.4%), of which 44.3% of obese. In addition, the USA samples reported a high percentage of overweight/obesity: 50.7% in the girls [45], of which 31.6% of obese, and 45% in the mixed sample in the longitudinal study of Schuster et al. (survey 1) [38]. The lowest percentage of OW/obese was reported in the Indonesian sample [59], with 14% of OW/obese for boys and 11% OW/obese for girls, but a high percentage of stunting and thinness (19% and 10% in males; 23% and 4% in females, respectively) was also reported. Africans [44] and Israelians [35] showed a low percentage of OW/obese (19.5% in the former, 20.4% (for males), and 16.9% (for females) in the latter).
PA was assessed with different instruments: the majority with indirect methods such as questionnaires, validated (IPAQ, PAQ, 24 h Physical Activity Recall, Baecke questionnaire) or not validated (a single question on h/week of sports training or questions on PA) to divide the samples in active and sedentary or in categories according to PA levels (low, moderate or high PA). Only three studies have directly assessed PA: one by the pedometer, the other by the pedometer plus the 24 h PA recall, and the last one by the accelerometer to determine the sedentary time, light PA, moderate PA, and vigorous PA (min/h) plus a self-reported questionnaire. In epidemiological studies, self-reporting is usually the most feasible method of assessing PA, especially when a high number of participants are interviewed due to the relative inexpensiveness, the minimal inconvenience to the participant, and the relative ease of administration.
According to the inclusion criteria, we chose only studies performed using the body silhouette scales for the assessment of BI and BID. Eleven studies assessed BI using the Stunkard Rating Scale with nine silhouettes, six studies used the seven silhouettes developed by Collins for preadolescents, three studies used the Contour Drawing Rating Scale proposed by Thompson and Gray, three studies on Brazilian samples used the silhouettes developed by Kakeshita et al. [48], one study each the BIDA questionnaire, the Body Silhouette Chart developed by Sanchez-Villegas et al., and the 13 silhouettes developed by Gardner et al. adapted for Spanish by Rodriguez et al. [52]. Zach et al. [35] used five silhouettes (not reported) drawn by an artist, and Min et al. [46] used nine new silhouettes reported in the paper. From the current (also called actual or feel) and ideal figures, the researchers calculated some indices of dissatisfaction, which, under different names, generally assessed the discrepancy between feel and ideal BI, thus coming to define the percentage of satisfied/dissatisfied (without discrepancy between current and ideal figures or with the discrepancy, respectively) subjects or mean value of dissatisfaction (negative values mean a desire of a bigger body and positive values mean a desire of a thinner body).

Association between Weight Status, PA, and Body Dissatisfaction
Several studies underlined the association between BI and BMI, especially in overweight and obesity categories [38,41,43,46,57,59,60,63,65,67]. In almost all the studies that reported this association, overweight and obese adolescents had higher body dissatisfaction and less body perception consistency, evaluating themselves as slimmer than they were. In contrast, Zach et al. [35] reported that a large percentage of overweight and obese adolescents were satisfied with their body shape and size, regardless of sex. Moreover, Michels et al. [44] did not find any significant association between the level of body dissatisfaction and BMI.
Six studies that analyzed the association between PA and BID found no significant association or correlation between these two variables [28,35,43,44,54,55], and nine studies reported a negative association between PA and BID or body perception [39,41,42,45,53,57,58,62,68], whereas one study reported an increase in body dissatisfaction with an increase in PA level [47].

Quality of Studies
Half of the examined studies showed a high level of quality with an overall score ≥ 13 (Table 2), and the remaining 14 were found to be studies of moderate quality (8 < score < 13).
No study received an overall score of <9 ( Table 2, Table S2).
In more detail, 15 studies (53.6%) provided an adequate description of the random sampling process, 16 studies (57%) provided an adequate description of the study sample, 25 studies (89.3%) used a validated measurement of BI, and 21 studies (75%) controlled for potential confounders.

Discussion
The principal aim of this systematic review was to investigate the association between body image dissatisfaction (BI and BID) and PA practice in adolescents, taking BMI or weight status into account. The hurdles we addressed primarily involved the assessment of dissatisfaction with BI, which was reported as an index of the discrepancy between the actual and ideal figure in some studies, while others reported the percentage of satisfaction or dissatisfaction. The type of BI rating scale used differed in the studies reviewed, although the use of Stunkard's scale [29] prevailed. Second, we found a variety of methods for assessing PA in the studies reviewed. Because of the different measures of BI and BID, and PA, a narrative review of the 28 included studies was conducted.
Despite the difficulties faced, this review provides an important update of the literature related to PA and BID in adolescents over the past decade, highlighting a slightly higher interest in the female sex (5 out of 28 studies; while 1 out of 28 studies involved only the male sex and 4 were mixed-sex studies). Although the limited number of reviewed studies on adolescents and the different survey methodologies applied to preclude a decisive view of the relationship between BI/BID and PA, the main findings of this systematic review indicate mostly (56.3% of studies) a negative association (moderate-medium) between PA and BID in adolescents, coming to highlight a better body perception in the physically active adolescents [50] and an association between BI disorders and reduced moderate-to-vigorous physical activity (MVPA) [58]. However, due to differences among samples in PA practiced (amount and type of activity), weight status, and sex of participants, this relationship is complex, and the results contradictory, so much so that 37.5% of reviewed studies detected no association between these variables and 6.3% showed a positive association. In contrast, generally, higher body satisfaction results in active adolescents or those with high PA levels, and lower body dissatisfaction in sedentary or physically inactive girls was also reported [47]. In the field of sports, Fernandez-Bustos et al. [50] found greater dissatisfaction in girls who practiced aesthetic/light PA or were inactive than in those who practiced other types of PA, confirming the importance of the type of PA practiced. Furthermore, Zaccagni et al. [53], analyzing a sample of rhythmic gymnasts, pointed to a possible source of misunderstanding in the ideal BI, showing that the ideal figure in sport does not coincide with the ideal figure in everyday life-the dissatisfaction between perceived and sport ideal figure was greater than the dissatisfaction between perceived and general ideal figure.
Diverse physical, cultural, social, and psychological changes that characterize early and medium adolescence interact with BI and BID between ages 10 and 18. Restricting to physical changes during this period, sex-specific differences occur that result in an increase in body fat mass and a decrease in lean mass in the female sex compared to the male, with consequent changes in body shape and size. The eventual practice of PA and sports may also interact, leading to eventual improvements in the perception of BI [15]. However, a decline in PA practice during the period of adolescence is well known [72,73], leading especially girls to be less active than boys and not following the WHO recommendations of at least 60 min of MVPA per day [74], as confirmed by this review with a percentage of inactivity reaching almost three times as high in females as in males in Spain [50] and exceeding 80% inactivity in Brazilian girls [47] with less than 10% of girls participating in the recommended 60 min/day of MVPA for seven days in the USA [45]. Inactivity in China was high (about 60%) but similar in the two sexes [69], while in Indonesia, girls met the daily PA recommendations slightly less than boys [59]. The only exception is the high MVPA averages (over 5 h per day, value not disaggregated by sex) reported for Ghanaian adolescents [44].
Because body weight dissatisfaction is reported as a factor associated with lack of PA, physical inactivity can result, as a consequence, of the interaction between weight status and BI, according to Duarte et al. [73]. An alternative hypothesis is that, instead of discouraging, BI may motivate PA and sports participation [15]. In general, adolescent girls of all ethnic groups showed greater body dissatisfaction than boys. However, in addition to Malaysia, dissatisfaction appears lower in adolescents from African countries and African Americans than in adolescents from Western countries, in agreement with the literature [75]. In Ghanaian adolescents, in particular, the greatest body dissatisfaction involved a body that was too thin [44], unlike the other studies reviewed. This result is consistent with the known tendency of non-White populations (particularly Africans and African Americans) toward a more favorable BI than Whites and a preference for a heavier body [25,76] according to a different beauty ideal. Despite the greater dissatisfaction that generally characterizes Western countries, Finnish adolescents show a high percentage of satisfaction with their BI (about 60%), with small differences between genders [68]. This pattern may at least partly be explained by a possible larger presence of preadolescents in the sample characterized by a lower age range (9-12 years) than in most of the studies considered. Indeed, it is believed that the physical, cultural, and diminished self-esteem changes that occur during adolescence can contribute to an increased awareness of BI, resulting in greater concerns about weight (particularly in females) and musculature (in males) [77,78].
Other previous analyses conducted on adults had shown small/medium associations between PA and BI [17,25], pointing to a stronger PA impact of PA on BID among young people, according to Bassett-Gunter et al. [17]. In particular, the number of hours spent in sports activities was found to be a predictor of body dissatisfaction with a decrease in BID as the number of hours spent in sports activities increases [19]. In this review, a positive effect of sports training on BID was found in a study [42] carried out by a repeated measure design: a decrease in BID and changes in anthropometric measurements were evident after 12 weeks of soccer training.
An important aspect of the relationship between PA and BI/BID may be the intensity of PA. Although it is plausible that this factor, as well as multiple others, is involved in the association, it is not possible to explore it in depth because of the limited data provided by the studies reviewed and the heterogeneity in the methods used to assess PA (from different types of questionnaires to accelerometers). Taking into account the accuracy and reliability of accelerometry [79], we emphasize the need for adequate studies designed to allow an understanding of the actual role of PA on BI/BID. According to Gualdi-Russo et al. [80], a decrease in body adiposity occurs in children and adolescents as MVPA increases and sedentary time decreases. If PA affects adiposity and, in particular, weight status, this could reverberate on BI/BID. Several anthropometric parameters (BMI, waist circumference, waist-to-hip ratio, waist-to-height ratio) are possible indicators of overall and central obesity and predictors of weight status [81,82]. However, the vast majority of examined studies have only considered BMI, classifying adolescents into weight categories. According to NCD-RisC [83], BMI can be regarded as "an imperfect measure of the extent and distribution of fat in the body, but it has the major advantage of having consistent and comparable data in many populationbased surveys". The relationship between weight status and BI/BID is complicated and could significantly affect the health behaviors of adolescents. Specifically, negative health consequences that may result from the interaction between weight status and BI include eating disorders, physical inactivity, and dysfunctional exercise [6]. Adolescent girls, in particular, will be at higher risk of developing clinical eating disorders because of intense BID and peer pressure to look thinner. At the same time, although PA is usually encouraged as a behavior that promotes health [79,84], compulsive exercise is a commonly used purging strategy to balance caloric intake or modify weight, body size, or shape [85]. This is probably the source of some apparent inconsistencies in the examined literature indicating greater PA practice in overweight/obese adolescents or those who perceived themselves as fat [35], although some examined studies show little awareness of the importance of PA among overweight/obese adolescents [46,59]. Lastly, further inconsistencies in the analysis of relationships between BID and BMI may arise from the study of sex-mixed samples (n = 4) due to the different characteristic patterns of the two genders (linear relationship in females, curvilinear relationship in males).

Strengths, Limitations, and Directions for the Future
This systematic review provided an overview of the literature of the last 10 years and identified eventual gaps in the literature to improve future studies. Among the strengths of this review is the systematic search of articles from the literature by evaluating a wide variety of evidence. A specific strength relates to including only studies that have assessed BID through figure scales. A limitation of this review is that we screened only two major databases, and we purposely decided to exclude gray literature. Moreover, other limitations reflect the limitations of the studies analyzed. First, we chose to consider only studies that used figural stimuli, but there is evident heterogeneity in the figural scales used, as well as in the methods employed to assess PA. With particular reference to BI, some studies have relied on unclearly defined or non-validated body image rating scales or inappropriate BI rating scales (e.g., scales developed for preadolescents or adults but applied to adolescents) with results of weak or unknown value. The use of small convenience samples in several studies is an additional limiting factor in assessing an association between BID and PA. Concerning PA, the limited information on this variable did not allow us to detect any abnormal compulsive exercise behavior. Accurate measurements of the PA level would be needed to assess the real impact on BI. In future studies, we suggest further investigation of the quality and quantity of exercise performed to discriminate behaviors attributable to dysfunctional exercise. The same assessment of overall and central obesity, limited in this review to BMI because it was found to be the most widely used, was obtained from directly measured anthropometric measurements in some studies and in other studies from self-reported height and weight values resulting in a possibly incorrect estimate. Secondly, the studies reviewed were generally cross-sectional, the limit of which is often ascribed to the impossibility of establishing causal relations. However, Wunsch et al. [86] think that this possibility is rather related to "whether or not our modeling strategy is structural". Therefore, more robust study designs and/or modeling strategies should be employed in the future to facilitate the understanding of causal relations between PA and BI/BID. Given the significant differences between genders in BID (dissatisfaction more pronounced with a clear preference toward slimmer figures in girls) and PA (lower amount in girls that are also less interested in vigorous PA), it is recommended that statistical analyses be conducted separately by reporting data by sex. Moreover, there is a need for future research to improve and standardize the methodologies used, especially in PA and BI assessments.
Finally, the exclusion of non-English articles certainly resulted in a loss of information, but we felt this choice was in line with current international systematic syntheses.

Conclusions
Up-to-date local data on adolescents' BI perception and PA practice are indispensable for planning health strategies and supporting non-communicable disease management. Monitoring changes in ideal BI during adolescence should be a primary health objective to guarantee the psychological and physical well-being of adolescents: health interventions should include strategies to promote health-related benefits and positive BI through PA among adolescents. The results highlighted in this review show a ubiquitous prevalence of unsatisfactory BI, although more pronounced in Western countries, with a concurrent trend toward inactivity and overweight in adolescents. This picture implies a high risk of the spread of nutritional disorders. Considering social and behavioral differences among subgroups based on sex, BMI, and PA, a unique approach may be inadequate: specific recommendations tailored to the needs of adolescents should be provided by researchers and health professionals.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijerph192013190/s1; Table S1: Prisma 2020 Checklist; Table S2: NOS scores for all included studies in alphabetical order (range: 0-16 stars, with higher scores indicating better quality research). The range of each item was reported.