1. Introduction
Childhood overweight and obesity are increasing globally, with adverse physical and psychosocial consequences on health and quality of life [
1]. Body image—or the perception and thoughts about one’s body—has been reported as one of the main factors related to psychological well-being that are more seriously compromised in obesity [
2,
3]. Moreover, body dissatisfaction seems to increase with overweight status [
2,
4]. Although weight loss and improved body appearance could represent primary motivators for Physical Activity (PA) in obese children [
5], it is still unclear whether weight loss might induce changes in the body representation of obese children [
3]. In particular, it is still unclear whether low levels of PA are determinants or consequences of body dissatisfaction [
6]. Furthermore, despite interactive effects seeming plausible [
6], most of the studies addressing the relationship between PA and body image have been conducted on adults, while the few studies among children have reported mixed results [
7,
8].
Cross-sectional studies have demonstrated an inverse relationship between Body Mass Index (BMI) and an individual’s perceptions of physical ability [
9]—mainly due to movement difficulties, including poor performance on weight-bearing tasks [
10]. In particular, an excessive body weight has adverse effects on actual motor skill competence [
11] and, consequently, on perceived physical competence [
4]. Of note, perceived competence has been linked with increased motivation towards and participation in PA [
12]. In fact, perceived physical competence is deemed to mediate individual’s emotional responses in PA [
13] and therefore, to influence PA behaviors [
14,
15]. Steiner [
16] argued that every action stimulates emotion, which in turn regulates our behavior [
17]. In fact, positive emotional experiences associated with PA sustain commitment toward actual levels of PA [
14,
18]. Thus, enhancing children’s self-perception of physical ability and pleasant emotions could ensure interest and adherence to PA [
12,
14]. Conversely, PA seems to increase body satisfaction through improvements in physical fitness and changes in perceived physical competence [
19,
20].
Notwithstanding these findings, a better understanding of associations over time between PA participation and the psychosocial determinants of active behavior in obese children is required [
21]. Research is needed to identify effective and sustainable ways to motivate overweight and obese children to increase their PA levels. As suggested by Stodden et al. [
22] and based on different theories of exercise motivation [
23,
24], a physically active lifestyle could be encouraged among obese children by enhancing their actual and perceived physical abilities and enjoyment. School Physical Education (PE) programs can greatly contribute to health promotion by providing a variety of enjoyable activities in which participants can monitor their progress and increase their performance satisfaction levels. A systematic review and two meta-analyses suggested that theory based, multi-component PA interventions are the most promising approaches in the school setting [
25,
26]. Therefore, it is important to ensure that school-based PA programs, especially around 11 years of age [
27], provide children with positive experiences to promote their adherence and improve motivation [
18]. However, scant research has been conducted on the relationship between psychosocial changes and weight loss, and the results are conflicting [
28,
29].
To the best of our knowledge, no studies have examined the relationship among changes in body composition, PA levels, motor performances, and psychosocial determinants of active behavior in obese children. Therefore, based on theories of motivation and behavioral change [
23,
24], we conducted a 6-month, multi-component pilot obesity program in school focused on nutritional education, skill-learning, enjoyable PA, and exercise training. Our study was part of a larger project, named “Healthy Lifestyles Intervention for Obese Schoolchildren (HeLIOS)”, which explored the impact of a school-based program on health and well-being among inactive overweight and obese children over time [
30]. Health can be broadly defined to include physical, psychological, and social well-being [
31]. Therefore, the current study was specifically designed to examine further aspects that were, for the most part, not considered in previous research, to broaden our understanding of modifiable factors that co-exist with pediatric obesity.
In particular, the aim of our study was to assess the feasibility of a school-based program, by examining changes in body composition, physical fitness, and self-reported measures of PA, perceived physical ability, body image, and pleasant/functional and unpleasant/dysfunctional psychobiosocial states in overweight boys and girls. A secondary purpose of this study was to investigate directional influences over time between PA and one of its key correlates (i.e., body image) by using a two-wave panel design with a 6-month time lag. In particular, the objective was to estimate reciprocal relationships between PA and body image across different time points (T
0 = baseline and T
1 = post intervention). Drawing on the causal priority approach [
32], we tested the hypotheses that baseline body image satisfaction would be positively associated with and predict higher levels of PA at T
1 (Hypothesis 1) or that PA levels at T
0 would be positively related to and precede changes in body image over a 6-month period (Hypothesis 2). Support of both hypotheses would indicate a reciprocal relationship between PA and body image.
3. Results
All children completed the intervention and the mean attendance for the exercise program was 82% (range 71–88%). Descriptive statistics and repeated-measures analyses of variance results for each variable by time of assessments (T
0 vs. T
1) are reported in
Table 1. The differences across measurements for body circumference and skinfold thickness mean values are represented in
Figure 1. Significant main effects by gender were found only for body fat percentage (
F1,16 = 6.00,
p = 0.026, η
p2 = 0.27) and hip circumference (
F1,16 = 5.75,
p = 0.029, η
p2 = 0.26), with females reporting higher values than males. Gender × time interaction effects did not emerge, suggesting that changes over time were similar for boys and girls.
From T0 to T1, height significantly increased, whereas BMI values and percent body fat declined. Arm and waist circumferences, and skinfold thickness of the biceps, subscapular and suprailiac regions also decreased over time. With regard to physical fitness, there were significant changes in all motor tests, with children showing better performances at T1 compared to T0. Significant time effects were shown for behavioral and psychosocial outcomes, with participants presenting lower body dissatisfaction and reporting higher PA and perceived physical ability scores after treatment.
For each time point, negative correlations were found between BMI and vertical jump (T
0:
r = −0.63; T
1:
r = −0.56), and positive correlations between speed times in the shuttle-run (T
0:
r = 0.59; T
1:
r = 0.57), and Harre circuit test (T
0:
r = 0.63; T
1:
r = 0.52). Similarly, negative correlations were found among percent body fat and vertical jump (T
0:
r = −0.60; T
1:
r = −0.58), shuttle-run (T
0:
r = 0.73; T
1:
r = 0.69), and Harre circuit test scores (T
0:
r = 0.55; T
1:
r = 0.48). It is interesting to note that BMI at T
0 was positively associated with body dissatisfaction (
r = 0.61), while PA was positively correlated with pleasant/functional psychobiosocial states (
r = 0.58) and negatively with unpleasant/dysfunctional states (
r = −0.49). At T
1, unpleasant/dysfunctional states correlated with the BMI (
r = 0.55), shuttle-run (
r = 0.54), and Harre circuit test scores (
r = 0.52). The magnitude of all these correlation coefficients ranged from moderate to moderately high [
49].
With regard to the cross-lagged analysis, PA levels and body image were not reciprocally related across time (
Figure 2). As shown in
Table 2, baseline body image was associated with levels of PA at T
1 (β = -0.48,
p = 0.02) beyond PA at T
0 (β = 0.66,
p = 0.01), and after controlling for the effects of covariates (i.e., gender and age). These findings suggest that the larger the body dissatisfaction of children before the intervention, the larger the increase in levels of PA six months later. The second hypothesis on the reverse relationship was not supported, given that PA at T
0 and body image at T
1 were not significantly related (β = 0.42,
p = 0.16).
4. Discussion
This was an uncontrolled, pilot study to assess the feasibility of implementing a school-based obesity program. Although limited by the small sample size and no control group, after the intervention we observed changes in obese boys and girls in almost all targeted variables. All the children who started the program completed it and showed improvements over six months in the measures of adiposity, actual and perceived physical abilities, PA levels, and body image. We did not find significant changes on the psychobiosocial state scores at T
1 compared to T
0, which was most likely due to initially high scores of pleasant/functional states and low scores of unpleasant/dysfunctional states reported by both girls and boys in the whole sample. These results are consistent with those of several previous studies conducted in the context of school PE [
48,
50,
51,
52], which suggest a general experience of a pleasant educational atmosphere.
In the present study, significant gender differences were observed only in body fat percentage and hip circumference, with girls having higher values than boys at both T
0 and at T
1. These values may be an expression of maturational differences by gender. As was found in previous studies [
30,
53,
54], gender × time interaction effects did not emerge, suggesting that obese boys and girls may respond comparably to a non-competitive PA program at school and show improvements in almost all study variables after intervention. During treatment, all children performed a variety of non-competitive activities and enjoyable forms of individualized tasks in a supportive learning environment that enabled them to experience pleasant emotional states and improve. Although enjoyable skill-based activities may not conform to classical exercise training guidelines, a recent systematic review suggests that enjoyable interventions in overweight and obese children are needed to increase their actual and perceived physical abilities, which in turn, may lead to long-term adherence to PA participation [
55]. Children PA enjoyment depends, to a large extent, on their perceived physical ability [
15], which is thought to mediate the relationship between children’s motor competence and subsequent PA [
56]. In a previous longitudinal study, the perception of physical ability was found to be an antecedent to PA rather than an outcome [
53]. Therefore, to encourage overweight and obese children to adhere to a PA program, it is important to increase their perceived physical ability first [
53].
Thus, involving children in enjoyable experiences has the potential to increase PA levels, ensure continued interest and participation, and ultimately improve long-term health outcomes, hopefully reducing the prevalence of obesity [
14,
18,
55]. Conversely, fatigue, muscular strain, discomfort, or pain derived from PA may be related to low levels of perceived competence and discourage children’s participation [
21]. It is important to consider that obese children can experience more rapid fatigue on motor tasks requiring horizontal propulsion and vertical lifting of the body mass (e.g., walking, running, jumping) compared to their normal-weight peers [
4,
10]. In the current study, the negative associations found between BMI and body fat percentage with most health-related fitness tests (i.e., vertical jump, shuttle-run, and Harre) are in line with the contention that overweight and obesity can have adverse effects on motor performances [
4,
10,
11,
30]. Our post-intervention results also suggest that body-mass-dependent motor performances (i.e., shuttle-run and Harre circuit) were positively related to unpleasant/dysfunctional states. Therefore, especially in the early phases of an obesity intervention, weight-bearing tasks should be proposed gradually to enable successful performance, improve actual and perceived physical abilities, and thereby promote positive attitudes towards PA [
54].
Interestingly, findings in this study showed baseline BMI positively related to body dissatisfaction, indicating that obese children are dissatisfied with their body appearance. There is also evidence suggesting that body image may serve as a protective factor against BMI increase [
57] and may motivate exercise behavior [
6]. It seems plausible that children who are dissatisfied with their body appearance are less likely to engage in self-care behaviors that may enhance one’s ability to maintain a healthy weight over time. For instance, previous study findings found a direct relationship between body dissatisfaction and low PA levels [
58,
59]. As stated in our first hypothesis on the positive prediction of PA levels by body image, cross-lagged analysis results suggested that body image at T
0 was directly related to and predicted changes in PA levels at T
1, whereas the effect of PA to body image did not emerge. Therefore, in contrast to our second hypothesis, body image was found to be an antecedent to PA rather than a consequence of PA.
Based on the observed relationship between body satisfaction and engagement in PA, improving body image in obese children can be an important step to promote their participation in PA. A school-based health curriculum should simultaneously strive to improve body satisfaction and increase PA [
60] through programs focused on behavioral change and not just on weight per se. However, programs centered on body image changes should be implemented with caution, because raising awareness on body weight may enhance body dissatisfaction among obese children. Educators should provide a variety of enjoyable activities aimed at developing students’ actual and perceived physical competence and intrinsic motivation. Programs aimed at enhancing the perceptions of physical ability rather than physical appearance are expected to reduce body dissatisfaction [
19,
20] and result in increased PA levels.
Given the pilot nature of our study, there are some limitations that should be acknowledged. First, the small sample size limits the generalizability of the results, despite the fact that they represent valuable pilot data requiring confirmation in larger samples. Second, the study design involving only an experimental group does not allow us to establish the specific effects of the intervention. These limitations were also due to the public nature and complexity of the intervention that required the cooperation of parents and professionals (i.e., pediatricians, dietitians, and teachers) in a school setting. Beyond these limitations, a key strength of our intervention was the high attendance of the children, which suggests that they enjoyed the program. Another strength of the study is the use of self-reported measures of PA levels, perceived physical ability, body image, and psychobiosocial states together with objective assessments of physical fitness. Furthermore, body composition was assessed using multiple measures (i.e., waist circumference, skinfold thickness, and body fat) to provide a better indication of changes in adiposity than BMI alone [
61]. Finally, the pre-test post-test study design allowed us to address the issue of directionality between PA and body image in children treated for obesity.