Association between Physical Fitness, Physical Activity Level and Sense of Coherence in Swedish Adolescents; An Analysis of Age and Sex Differences

Introduction: The aim of this study was to investigate the association between physical fitness, physical activity, and sense of coherence (SOC) in Swedish adolescents (n = 2028 males, n = 1287 females) aged 14 to 18 using a cross-sectional design. Methods: Using the Swedish Physical power Mental harmony and Social capacity (FMS) student profile, participants performed physical tests measuring their cardiovascular ability (CV) and muscular strength. Questionnaires were used to measure physical activity levels (PA), the participation in organized physical training, sedentary behavior (SB), screen time, and SOC value. Multiple linear regression analyses were used to analyze the association between SOC and independent variables. Results: The regression analyses explained a small, shared variance between SOC and the independent variables in boys (4.3%) and girls (3.3%). SB showed a positive association with SOC both in girls (β = 0.114, p = 0.002) and boys (β = 0.109, p = 0.013). Abdominal strength was positively associated, while VO2max was inversely associated, with SOC (β = 0.113, p = 0.022; β = −0.109, p = 0.026, respectively) in boys. Girls had poorer fitness than boys did across all age groups except at age 18. PA levels decreased from age 14 to 18 in girls and boys, but without differences between sexes. Abdominal strength decreased from age 14 to 18 in girls and boys. In general, girls had lower SOC than boys across all age groups. Conclusions: Poor sedentary behavior was significantly associated with weaker SOC for both genders, indicating overall physical activity as the most important factor for stronger SOC. However, emotional support in vulnerable environments may have a bigger impact than physical activity or sedentary behavior on the SOC value for adolescents.


Introduction
Adolescence is an intensive and sensitive period of physical and psychological changes where social relationships and emotional and academic abilities are developed simultaneously [1,2]. It is often described as a period of "storm and stress" [3] due to the onset of many stress-induced mental health problems (e.g., depression, anxiety, and post-traumatic stress) that increase in childhood and peak in adolescence [4][5][6][7]. Mental health problems play a big role in the global burden of disease and affect approximately 10-20% of children and adolescents worldwide [7], while the number of adolescents with emotional and mental problems has increased during the recent decade, particularly in central and northern Europe [8]. One of the major concerns with adolescents' mental health is that it affects the mental state when they become adults [8,9]. Not surprisingly, the problem of mental health of adolescents is among the most important issues in public health nowadays [10,11]. adolescents with a higher activity level, physical fitness, excessive screen time, and poorer sedentary behavior display a stronger SOC.

Participants
Fourteen-to eighteen-year-old Swedish high-school students (2028 girls; 61% and 1287 boys; 39%) from 45 schools located in three urban regions in the middle and south of Sweden, including communities with smaller (<50,000) and bigger (>50,000) population, were invited and accepted to participate in the study. They were divided into several age groups: girls (age 14, n = 40; age15, n = 239; age 16, n = 1312; age 17, n = 384; and age 18, n = 53) and boys (age 14, n = 40; age 15, n = 137; age 16, n = 773; age 17, n = 260; age 18, n = 77), respectively ( Table 1). All participants included in this study met the requirements for inclusion, which were being a Swedish school student, 14-18 years old, attending a school using the FMS instrument, having a complete set of data, e.g., had undergone all the physical tests and answered all the questionnaires included in the FMS instrument during 2014-2020. The FMS profile is a tool designed by the Swedish Institute of Physical, Mental, and Social capacity for adolescents aged 12-19 years old and includes a questionnaire (e.g., PA level and sedentary behavior), fitness tests (e.g., cardiovascular fitness, muscular strength, and flexibility) and anthropometric measurements (e.g., body height and mass) [26]. Parents/guardians and the participants were all informed about the fitness test and the FMS questionnaire. A written consent was signed by the parents/guardians and the students to participate in the study. The present investigation was conducted in accordance with the Declaration of Helsinki for human studies and was approved for retrospective and prospective data by the Swedish Ethical Review Authority (Dnr 2019-05076).

Variables and Measurement
The variables in this study included independent variables (age, sex, BMI, upperand lower-body strength, abdominal strength, VO 2 max, everyday physical activity, organized physical training in a school/club setting, sedentary behavior, and screen time) and a dependent variable (sense of coherence).
The FMS profile includes anthropometric measurements, questionnaires (including SOC questionnaire), and the fitness tests follow the standardized protocols [26]. Anthropometrics included body mass and height measured by the school nurses. Body mass index (BMI) was calculated for each student [body mass (kg)/body height squared (m 2 )]. Physical education (PE) teachers conducted the fitness tests during the PE sessions.

FMS Questionnaire
Four single-item questions from the FMS student profile questionnaire were used to measure the prevalence of physical training, everyday PA, sedentary behavior style, and screen time among the participants. (1) Physical training: How often (times a week) do you train in a club, so that you become at least 30 min breathless (e.g., running, playing ball sports, etc.)? (answers included: never, once a week, two times, three times, at least four times). The self-reported questionnaires consisted of a Likert-type scale ranging from 1 to 5, and it has previously been validated in large sample studies [26,42,43]. Sedentary behavior and screen time were inversely coded with a higher score, meaning lower sedentary behavior and less screen time.

Sense of Coherence Questionnaire
The present study used the shortened 13-item SOC questionnaire. Every question option was ranked 1-7. The item response in the questionnaire was counted as 1 for the lowest rank and 7 for the highest rank. Four questions were counted in a reverse score setting in the 13-item questionnaire. An example of a reverse counted question is "Have people you trust made you disappointed?"; ranking 1 means it "never happens", and ranking 7 means it "happens a lot." A high total score in the questionnaire represents a high value of SOC [44]. The dependent SOC variable in the data analysis consisted of the total score of the questionnaire (ranged 13-91). The 13-item questionnaire has shown to be a reliable [44][45][46] and valid [47] instrument, measuring how individuals manage stressful situations and stay well [15].

Fitness Measurements
Muscular strength was assessed using the following tests: sit-ups (number of repetitions), upper body strength (arm lift with 5 kg for females and 5 or 10 kg for males (number of repetitions), and lower body strength (static wall squat, i.e., holding position "sitting chair" for as long as possible, measured in seconds). All tests are described elsewhere in more detail [26].

Cardiovascular Fitness
To test cardiovascular fitness, PE teachers could choose one of several maximal and submaximal tests, depending on the available resources, equipment, and the pupils' characteristics (e.g., overall fitness level, familiarity with the tests, etc.). A Cooper test was most frequently used. It estimates the participant's maximal oxygen consumption based on the distance covered in 12 min of running [48]. It has shown to be a valid and reliable test for estimating VO 2 max [49]. Alternatives to the Cooper test were the 1-Mile Walk test [50], Step test [51], and Åstrand cycle ergometer test [51,52], all of which were considered valid and reliable.

Statistical Analyses
The Shapiro-Wilk test was used to analyze the normality of the distribution for all variables. Continuous variables were presented by mean and standard deviation (SD), while one-way analysis of variance (ANOVA) with consecutive Bonferroni post hoc test and the independent t-test were used to analyze the differences across the age groups and differences between girls and boys, respectively. For ordinal variables, the Kruskal-Wallis Test and Mann-Whitney Test were used. Standard multiple linear regression analyses were used to analyze the association between the dependent variable (i.e., SOC value) and the independent variables separately for girls and boys. Independent variables consisted of ordinal (i.e., everyday physical activity, physical training, sedentary behavior, and screen time) and continuous variables: anthropometry (i.e., BMI), physical fitness (i.e., VO 2 max, abdominal strength, upper body strength, and lower body strength), and age. Before the calculation of multiple regression, the collinearity between the predictors was checked. The variance inflation factor (VIF) for all predictors was below 2.5, showing low multicollinearity. Therefore, all predictors were included in the multiple regression calculations. Significance level was set at p < 0.05. Statistical analyses were performed using IBM SPSS Statistics 27 (IBM, Armonk, New York, NY, USA).

Results
Descriptive statistics (mean ± SD) and differences across age and sex groups in anthropometric, physical fitness, and SOC variables are presented in Table 1. Differences in physical training frequency, PA level, sedentary behavior, and screen time across age and sex groups are presented in Table 2. Legend: Data are presented at 1-5 point scale, with higher scores indicating a higher level of PT and PA, and lower level of SB and ST; p-value of Kruskal-Wallis test indicates significant differences between the age groups; * Indicates significant differences between boys and girls in accompanied age group at p < 0.05.

Anthropometrics
Boys were significantly taller (p < 0.005) and heavier (p < 0.005) than girls across all age groups (14-18 years old). Body mass increased across all age groups in boys (f = 6.07, p = 0.001). In contrast, there were significant differences between age groups (f = 4.67, p = 0.001) in girls but without significant increase across the groups. There were no differences in body height across age groups, either in boys or in girls (f = 1.23, p = 0.30, f = 1.78, p = 0.13, respectively). BMI showed no significant differences across age and gender groups. However, girls had significantly higher BMI in ages 16, 17, and 18 compared with age 14.

Physical Fitness
Boys had higher VO 2 max than girls across all age groups. There was a difference in both girls (f = 12.85, p = 0.00) and boys (f = 9.85, p = 0.001) in each age group, with a tendency to decrease with age from 57.3 ± 8.9 mL/kg/min at age 14 to 48.6 ± 13.1 mL/kg/min at age 18 in boys and from 51.1 ± 5.3 mL/kg/min at age 14 to 44.4 ± 7.2 mL/kg/min at age 18 in girls.
Lower-body strength was higher in boys than girls in age groups 16 and 17, with a tendency to decrease with age in both groups (f = 2.92, p = 0.02; f = 4.38, p = 0.00, respectively). On the other hand, upper-body strength was better in boys at age 14, but poorer at age 18. Boys had higher abdominal strength (i.e., sit-ups) than girls in all age groups except at age 14. Moreover, it declined with age, both in girls and boys (f = 4.61, p = 0.00; f = 3.90, p = 0.00, respectively).

Sense of Coherence
Boys had a stronger SOC than girls in age groups 15, 16, and 17. There were no differences in the SOC level between age groups in girls (f = 0.38, p = 0.82). On the other hand, boys in age group 15 had stronger SOC than those from age group 18.

Factors Associated with a Sense of Coherence
The multiple regression analyses showed a significant association of the independent variables with SOC both in girls (p = 0.003) and boys (p = 0.005) ( Table 3). However, the used predictors explained only 3.3% (R 2 = 0.033) variability of SOC in girls and 4.3% (R 2 = 0.043) in boys. Sedentary behavior showed a significant positive association with SOC both in girls (β = 0.114, p = 0.002) and boys (β = 0.109, p = 0.013), meaning the better sedentary behavior, the stronger SOC. Abdominal strength was positively associated, while VO 2 max was inversely associated with SOC (β = 0.113, p = 0.022; β = −0.109, p = 0.026, respectively), in boys. There were no other significant associations between independent variables and SOC.

Discussion
The main finding in the present study was that adolescents with less sedentary behavior had stronger SOC. Moreover, abdominal strength showed positive association, while VO 2 max had a negative association with SOC in boys. They also had a higher value of SOC and VO 2 max across all age groups compared to girls. Differences between girls and boys in everyday physical activity, sedentary lifestyle, and screen time were small. However, there was a significantly higher prevalence of physical training frequency for boys compared to girls in age groups 15-16.

Determinants of Sense of Coherence (SOC)
The hypothesis that adolescents with a higher PA level, physical fitness, and poorer sedentary behavior would display a stronger SOC irrespective of gender was partially supported. The present study showed less sedentary behavior being associated with a stronger SOC in adolescents, irrespective of the sex and age. This is consistent with a recent study showing a weak SOC being associated with insufficient PA [53]. Moreover, the finding is in line with several studies showing higher sedentary behavior being associated with worse mental health [20,22,54,55]. Furthermore, Iannotti et al. (2009) suggest the sedentary behavior and mental health may be affected by time spent watching television, which, in turn, reduces social interaction, personal problem-solving situations, and does not activate and challenge one's cognitive and physical capacities [56]. It seems that screen time and usage of social media increase not only sedentary time but also lead to social pressure and body dissatisfaction among adolescents [57,58]. However, the social media and screen time can provide some sort of social connection and interaction among adolescents, which could explain the non-significant association between screen time and SOC in the present study.
Unexpectedly, there were no significant associations between SOC and physical training (PT) and everyday PA, which is in line with some previous studies [13,19], but opposing to others that showed a positive association between PA and mental health [59,60]. The use of a single item (e.g., question) to assess adolescent PA and PT might be a possible explanation for the lack of association between PA, PT, and SOC.
Furthermore, BMI was not significantly associated with SOC. The present findings support previous studies which have suggested that BMI and mental health are not associated with adolescents younger than 20 years [61], or when controlling for fitness variables [62,63].
The only strength variable positively associated with SOC was boys' abdominal strength that was tested in the same way across all age groups and both sexes. The result is in line to previous research where strength is positively associated with mental health outcomes, such as self-esteem and self-perceptions [64], physical self-perception [65], perceived health status and life satisfaction [66], and lower depression [67]. However, in the current study there were no observed association between strength of other muscle groups (e.g., upper-and lower-body) and SOC. Possible explanation can be that boys aged 14-16 had the option of choosing either a 5 kg or a 10 kg dumbbell, compared to boys aged 17-18 who had to use the 10 kg dumbbell, which in turn could equalize their number of repetitions and in turn reduce the power of the regression analysis to detect potential association to SOC.
Interestingly, the present study showed an inverse association between SOC and cardiovascular (CV) fitness (i.e., VO 2 max) only in boys. It seems that increased CV fitness might be a product of external stressors (e.g., high training volume and pressure), which, in turn, could reduce the coping capacity to deal with everyday stress [68]. The observed association only in boys might be connected to a higher prevalence of physical training at age 15-18, compared to girls in the present study. Specifically, sports participation on a high-level puts many demands and stress on the participants. A recently published study demonstrated that more than 50% of the athlete students frequently reported studyrelated stress as a predictor of general well-being and general risk of depression, while 21% showed impaired well-being [68]. In line with that, Demirel (2016) suggested reasons such as physical or somatic stress associated with vigorous exercise programs (e.g., overuse injuries) and fear of occupational careers as possible factors associated with stress for athletic students [69].
In the present study, 47% of male adolescents reported that they trained in sports at a moderate to high-intensity level at least 4 times per week, revealing their active sports participation. In accordance with that, Mayer and Thiel (2014) reported a slight, but significantly lower SOC score in elite athletes than in the general population, indicating a high demanding sports participation as a risk factor that might weaken SOC [70]. In brief, the social context of sports participation is considered to be an extremely demanding environment that includes constant competitive pressure to perform at a high level which might increase injury risk, absence from the participation and competition, cause overtraining syndrome, and reduce psychological resilience to stress, which all together may lead to a decreased overall health and SOC [70,71].

Differences between Age and Sex in SOC
Generally, boys had a stronger SOC in every age group compared to girls. This is in accordance with previous studies showing boys having a stronger SOC [72][73][74]. The difference was significant at ages 15, 16, and 17, which is congruent with previous research by Moksnes et al. (2012) where boys had a significantly higher value at ages 15-16, compared to within age groups 13-14 and 17-18 [72]. The differences between the sexes in SOC in the present study might be explained by differently perceived stress in boys and girls. In brief, boys have usually reported higher levels of performance stress [73,75] and higher levels of achievement and self-relevant stressors [76], while girls intend to blame themselves in relationship problems [77] and are more affected by social relationship [78]. On the other hand, the SOC value showed small variability across age groups in both sexes, which is contrary to Antonovsky's (1987) thesis, indicating that SOC was developing by age [12]. A possible explanation may be that adolescents between 14-18 are in the same school stage (i.e., high school) where social, emotional, and academic demands are similar across the grades.

Differences between Age and Sexes in Physical Activity and Fitness Level
Expectedly, boys were taller and heavier across all age groups compared to girls. BMI showed a tendency of increasing with age. There was no difference in BMI between sexes, except in age 15, with girls having higher values. This small difference between the sexes can be explained by a low power of BMI to evaluate body composition (i.e., body fat and muscle mass). Thus, a higher BMI in girls might be a product of higher adiposity mass, while in boys it might be a result of bigger muscle mass, which is reasonable to assume when results showed higher activity level and participation rate in organized sport activities in boys than in girls. Moreover, girls had lower levels of everyday PA at age 14-17 compared to boys, which is in line with previous studies [23,79]. There were no significant differences in sedentary behavior and screen time across the age groups in either of the sexes. In general, the PA level decreased with age for both sexes, which is in line with previous studies [25].

Physical Fitness
Fitness results showed that boys had a higher cardiovascular capacity (i.e., VO 2 max) in all age groups compared to girls, which is in line with previous studies [25,80,81]. A higher PA level, cardiac size and function, as well as greater muscle mass and mechanical efficiency (e.g., larger levers) may explain the male superiority in aerobic capacity compared to female adolescents [25,82]. Additionally, boys were significantly stronger in the lower body at age 16-17, and in abdominal strength at age 15-17 compared to girls, which might explain the higher results in the cardiovascular tests too. Interestingly, the 14-year-old adolescents in the present study showed significantly higher values of VO 2 max compared to other groups. A possible explanation is that they were significantly lighter than older adolescents, which may have positively affected their relative VO 2 max. In addition, both girls and boys of this age showed to be the strongest in the abdominal and lower body tests. Moreover, they showed to have the highest everyday PA, which, in turn, could contribute to increased VO 2 max [37,83]. Results in the upper body strength were inconsistent between the sexes. However, a bias in the present study might be that boys aged 14-16 had the option of choosing either a 5 kg or a 10 kg dumbbell, compared to boys aged 17-18 who had to use the 10 kg dumbbell.

Strengths and Limitations
The strength of the present study was a relatively large sample size, representing the urban population of students aged 14-18 years from the middle and south of Sweden. However, the study has several limitations, which must be acknowledged. First, the cross-sectional character of the study design does not allow for the establishment of firm causal links between the variables, but only describes the relationship between them. Consequently, a longitudinal follow-up research design would be more appropriate to reveal variables that affect SOC. Second, the levels of PA, sedentary behavior and screen time were assessed by a subjective measure, including only a single question per studied measure, which could reduce the validity of the assessed construct. Third, the study did not include some potentially confounding variables (e.g., school factors, interpersonal relationships, nutritional behavior, health issues, substance use, etc.) that could affect the SOC strength. Fourth, the fact that boys aged 14-16 had the option of choosing either a 5 kg or a 10 kg dumbbell, compared to boys aged 17-18 who had to use the 10 kg dumbbell, could affect the strength score (e.g., number of repetitions), which, in turn, could reduce the power of the measure to discriminate between the age groups. Additionally, the number of 16 year old girls is quite high to other age-groups and the boys, which in turn could affect the analyses. Moreover, even though the used cardiovascular fitness tests were previously validated to estimate VO 2 max, the fact that the students could choose one of 3 different tests could affect the accuracy of the data.

Conclusions
In conclusion, even though the amount of explained variance in the regression model was low, the study revealed that adolescents with less sedentary behavior, irrespective of sex and age, had a stronger SOC and thus might have a better ability to cope with everyday stress. The results emphasize the importance of positive PA behavior in the development of a stronger SOC in adolescents. Furthermore, the findings indicate a stronger SOC and higher PA level in boys, which might affect the development of health-related strategies that target boys and girls separately. The strategies should also take into account the fact that PA consistently reduced with age. Interestingly, increased VO 2 max was associated with a weaker SOC in boys. This might be related to the reported increased training frequency, which in turn could lead to increased mental and physical burden and thus weakened SOC. This implies a need for well-designed training plans and emotional support for the participants in sport clubs that could potentially be socially vulnerable environments. Informed Consent Statement: Informed consent was obtained from all individuals included in the study. We confirm that neither the manuscript nor any parts of its content are currently under consideration or published in another journal.

Data Availability Statement:
The dataset supporting the conclusion of this article is available from the authors upon reasonable request and the completion of a data transfer agreement.