The data used in this paper were from the sample survey data of the “Adolescent Health Theme Database” of the Chinese National Population Health Science Data Centre, obtained to investigate the health status and health-related behaviours of middle and high school students [18]. This is the first publicly available free database on the health and health-related behaviours of Chinese adolescents. For the data used in this paper, high school students were considered as the study population and the relevant data were processed by looking at specific information on all relevant variables; removing outliers for age, physical activity, total SCL-90 score, and control variables; and excluding samples with incomplete answers and missing values, resulting in a valid sample with a size of 3578.

Mental Health. The dependent variable in this paper is mental health, which is a continuous variable. This variable was derived from the Symptom Check-List-90 of the Chinese version (SCL-90) in the database [19]. The Cronbach coefficient alpha of this scale ranged from 0.78 to 0.90, and there was a high degree of correlation between the nine original sub-scales, with correlation coefficients ranging from 0.39 to 0.79 [20]. Each item was rated on a five-point Likert scale from 1 (not at all) to 5 (very severe). The SCL-90 scale results were processed by adding the individual scores of each of the 90 items in order to obtain the total score, then dividing the total score by 90 to obtain the total mean score (i.e., total mean score = total score/90). The total mean score indicates the reporter’s self-perception as a value between 1 and 5, representing the overall level of mental health with higher values indicating lower levels of mental health among adolescents [19,21].

Frequency of Physical Activity. The independent variable in this paper is the frequency of physical activity, which is a categorical variable. The relevant database question was “Do you participate in physical activity regularly?”, with the answers: “Never” = 1 (never), “Rarely” = 2 (rarely), “Often” = 3 (often), “Always” = 4 (always).

Life Satisfaction and Self-Rated Health. The mediating variables were life satisfaction and self-rated health, dichotomous variables. The relevant question for the former was “Are you satisfied with your life?” and the value assigned to life satisfaction was 1 = unsatisfied or 2 = satisfied. The relevant question for the latter was “Are you satisfied with your health?” and the value assigned to self-rated health was 1 = unsatisfied or 2 = satisfied. Questions about frequency of physical activity, life satisfaction, and self-rated health were taken from the Child and Adolescent Quality of Life Scale. The scale has good reliability with a Cronbach coefficient of 0.885, content validity of 0.63 (p < 0.01), and construct validity of 0.890 (p = 0.00) [22].

Furthermore, the following were chosen as control variables: Sex, age, whether the child was an only child, place of residence, and parental occupational classification (see Table 1).

(1) A general linear regression model was used to study the relationship of physical activity on the level of mental health in adolescents. The regression model is as follows: YMH=a+bXPaf+e

where XPaf denotes physical activity (the independent variable), and YMH  denotes mental health (the dependent variable).

(2) The Karlson–Holm–Breen (KHB) model was used to determine whether life satisfaction and self-rated health were significant, and the model also allowed for analysis of the relationship between life satisfaction, self-rated health and mental health levels and the contribution of the indirect effects of the two mediating variables.

The KHB mediation model is a method for estimating mediating effects in regression models. The method estimates the mediating effect by decomposing the total effect into a direct effect and an indirect effect [23]. The following equations are used in the KHB method:(1) Total Effect:YMH=b0+b1XPaf+b2MLS+eYMH=b3+b4XPaf+b5MSHR+e1 (2) Direct Effect: YMH=c0+c1XPaf+e2 (3) Indirect Effect:M=a0+a1XPaf+e3

where XPaf  denotes physical activity (the independent variable), YMH denotes mental health (the dependent variable), MLS  denotes the mediating variable of life satisfaction, MSHR  denotes the mediating variable of self-rated health, and M is the mediating effect.

The results indicated that the mean value of mental health for adolescents was 1.629. The level of mental health refers to the fact that, in general, the scores of the total symptom index ranged from 1.5 to 2.5, indicating the existence of mild symptoms and a low frequency of symptoms among Chinese adolescents. There were 885 adolescents with a mean SCL-90 total score greater than or equal to 2, meaning that about 24.73% of adolescents had varying degrees of mental health problems.

To examine the extent to which different frequencies of physical activity affected each factor of the SCL-90 scale, a general linear regression analysis was conducted between the nine elements and frequency of physical activity after controlling for all variables (Table 2). In these nine models, low-frequency physical activity was not correlated with obsessive–compulsive symptoms, partially correlated with somatization (p < 0.05), and correlated more significantly with interpersonal interactions and psychoticism (p < 0.01). Compared to low-frequency physical activity, medium- and high-frequency physical activity were significantly negatively correlated with each factor (p < 0.001). The more significant the relationship on each element as physical activity frequency increased. Of the nine models, high-frequency physical activity was most significantly correlated with anxiety, and low and medium-frequency physical activity was most significantly correlated with distress.

To analyze the relationship of between physical activity and mental health, a general linear regression was conducted with mental health as the dependent variable, incorporating the control and independent variables (frequency of physical activity) and mediating variables in turn (see Table 3).

First, the relationship of between physical activity and the level of mental health of adolescents was analyzed. Model 1 in Table 3 considers, with the inclusion of control variables, the effect of physical activity on mental health. The results indicate that the significance of the effect of physical activity on the level of mental health increases gradually as the frequency of physical activity increases (taking never participating in physical activity as the reference group); that is, as the frequency of physical activity increased, the level of mental health increased. The frequency of physical activity increased from low- to medium- and medium- to high-frequency by 0.168 and 0.1 units, respectively, and Hypothesis 1 was verified. The regression results for the control variables demonstrated that age, sex, and place of residence had significant effects on adolescent mental health. Both sex and age had a significant negative association with mental health (p < 0.001), with girls having 0.083 units higher mental health than boys (95% CI: −0.13, −0.04). Adolescent mental health increased by 0.041 units (95% CI: −0.07, −0.018) with increasing age. Rural adolescents had higher mental health, relative to urban adolescents (p < 0.05).

The mediating variables of life satisfaction and self-rated health were added into models 2 and 3. The correlation of physical activity on mental health gradually decreased, and the correlation of low-frequency physical activity on the level of mental health was no longer significant, while medium- and high-frequency physical activity still had significant correlations on the level of mental health. In addition, with the addition of the two mediating variables to model 3 at the same time, there were significant negative correlations (p < 0.001) between life satisfaction and self-rated health with mental health, with coefficients of −0.300 and −0.371, respectively. The regression coefficients of the mediating variables life satisfaction and self-rated health were both significantly negative, and the coefficient of self-rated health was smaller than that of life satisfaction, indicating that life satisfaction and self-rated health had significant contributions to the mental health level, where the contribution of self-rated health on mental health was greater than that of life satisfaction on mental health. From model 1 to model 3, the regression coefficients of low-, medium-, and high-frequency physical activity on mental health decreased from −0.171, −0.339, and −0.439 to −0.062, −0.126, and −0.194, respectively, indicating that the two mediating variables play a mediating role in the effect of physical activity on mental health. Therefore, on the basis of the general linear regression model, the KHB mediation model was utilized for the analysis of mediating effects [24].

The results of the mediated effects model indicated that the total effect of low-, medium-, and high-frequency physical activity on mental health was −0.171 (p < 0.01), −0.339 (p < 0.001), and −0.439 (p < 0.001), respectively, with mediated effects of −0.109, −0.213, and −0.246 (p < 0.001), respectively (see Table 4). In terms of direct effects, physical activity had a significant direct effect on mental health (p < 0.01). In terms of indirect effects, the indirect effects were negative, indicating partial mediating effects (p < 0.001) for both mediating variables (i.e., life satisfaction and self-rated health). Therefore, physical activity had an indirect positive effect on mental health through increasing life satisfaction and self-rated health and, so, Hypotheses 2 and 3 were verified. Further decomposition of the contributions of life satisfaction and self-rated health revealed that low-, medium-, and high-frequency physical activity acted on mental health through life satisfaction, with mediating effects of 46.15%, 41.67%, and 40.57%, respectively (see Table 5). Low-, medium-, and high-frequency physical activity also acted on mental health through self-rated health, with mediating effects of 53.85%, 58.33%, and 59.43%, respectively. The largest mediating effect for low-frequency physical activity on psychological well-being was through life satisfaction, while the largest mediating effect for high-frequency physical activity on psychological well-being was through self-rated health; therefore, Hypothesis 4 was rejected.

In the regression model of the SCL-90 factors, high-frequency physical activity was most significantly correlated with low anxiety, while low and medium-frequency physical activity was most significantly correlated with reduced distress. Adolescents within this age group in China may experience psychological problems due to the excessive burden of college entrance exams, insufficient parental care, stringent discipline, and their own maladaptive habits [25]. Engaging in physical activity is beneficial for adolescents in reducing symptoms such as obsessive-compulsive symptoms and anxiety, fostering a healthy personality, and promoting emotional stability [26]. This substantiates the role of physical activity as an intervention for psychological problems.

It was determined that medium- and high-frequency physical activity were more significantly associated with mental health compared to low-frequency physical activity on mental health. This suggests that physical activity can promote mental health, and escalating the frequency of physical activity can effectively bolster adolescent mental health. Hamer also discovered that varying frequencies of physical activity resulted in distinct degrees of disparities in the mental health of students, demonstrating that mental health improved as the frequency of physical activity increased [27]. Moreover, in another study, the more weekly physical activities that adolescents participated in, the more pronounced the enhancement of their mental health [28]. Physical activity can ameliorate mental health and reduce depression, anxiety, and other negative emotions through physiological, psychological, and inflammatory mechanisms. As physical activity frequency increases, the body continually produces analgesic and euphoric endorphins, which improve adverse moods and modulate the central nervous system. Alternatively, physical activity may disperse negative emotions, augment self-esteem and self-worth, decrease chronic inflammation, and positively influence emotional states, thereby affecting mental health [29]. Consequently, the results of this study reaffirm the existing literature, indicating that adolescent mental health can be improved by augmenting the frequency of physical activity.

The results of this study demonstrated that adolescent physical activity exerts a positive influence on psychological well-being, with life satisfaction and self-rated health playing significant mediating roles between physical activity and psychological well-being.

An et al. concluded that physical activity among adolescents has a favorable impact on life satisfaction, which is in line with the findings of this study [30]. Some scholars have contended that physical activity not only mitigates psychological symptoms, but also enhances individual well-being and sustains elevated levels of life satisfaction [31]. Others have argued that physical activity bolsters individual life satisfaction by decelerating or restoring one’s physiological functions, enabling the attainment of information and the accomplishment of set goals [32].

The influence of life satisfaction on mental health has been substantiated. Life satisfaction should be regarded as a crucial aspect of positive mental health, as it has exhibited a moderate correlation with other mental health indicators within various populations [33]. Consequently, life satisfaction impacts adolescent mental health. Specifically, life satisfaction is also associated with psychiatric disorders, such as major depression, anxiety disorders, and suicidal tendencies, all of which profoundly affect mental health [34].

Moreover, studies investigating the effects of physical activity on self-rated health have corroborated that higher frequencies of physical activity correspond to improved self-rated health status. It has been observed that a high frequency of physical activity in adolescents is positively associated with self-rated ‘good health’ status, which concurs with the findings of this study [35,36]. An increased frequency of physical activity contributes to enhancing the perceived physical health level in adolescents, thus fostering positive self-rated health status [37].

Self-rated health is significantly correlated with mental disorders and the exacerbation of psychological symptoms [38,39]. Individuals with poor self-rated health tend to experience considerably more psychological distress [40]. Lower self-rated health exerts a negative impact on mental health and may be associated with emotions that emerge during illness. Research has indicated that pain constitutes a negative body-centered emotion, with more significant pain correlating with emotional stress [41,42]. Therefore, the discomfort and pain resulting from physical illness can lead to both a negative evaluation of one’s physical health and corresponding emotions, which can engender a certain degree of stress, subsequently making the connection with mental health.

This study discovered that low-frequency physical activity had the most substantial mediating effect on the mental health through life satisfaction. In contrast, the most significant mediating effect of high-frequency physical activity on mental health was through self-rated health. Furthermore, the mediating effects of self-rated health were consistently more significant than those of life satisfaction.

The impacts of varying physical activity frequency on life satisfaction are divergent. Adolescents with increased physical activity frequency and higher participation in active sports experience lower levels of depressive symptoms and higher life satisfaction [43,44]; however, in line with this study’s findings, high-intensity physical activity frequency was not associated with life satisfaction, while low- and moderate-intensity physical activity frequency improved life satisfaction [45]. Similarly, the study of Schmiedeberg did not reveal any increase in life satisfaction with heightened exercise frequency [46].

Differences in life satisfaction levels across physical activity frequencies could be attributed to the relationship between life satisfaction and academics in adolescents [47]. Engagement in academic activities indirectly affects the life satisfaction of students, with positive academic attitudes and involvement effectively promoting their life satisfaction [48]. The lives of Chinese high school students are dominated by studying in a concentrated learning atmosphere. The competitive environment and high expectations can increase pressure on adolescents, affecting their perceived quality of life [49]. Consequently, higher physical activity frequency may detract from time dedicated to studying, socializing, and resting, thus impacting academic performance, interpersonal interactions, and rest quality, subsequently reducing life satisfaction.

Health perception levels in adolescents are influenced by their physical activity engagement. Embodied cognitive theory [50] posits that cognitive processes are deeply rooted in the body–world interaction, and the body’s state directly affects cognitive functions, with physical behavior influencing mental activity and cognitive perception. Increased physical activity frequency leads to positive health perception changes in individuals and promotes better self-rated health outcomes through good exercise behavior. Thus, the level of self-rated health increases with the frequency of physical activity.

Comparing the two mediating variables across physical activity frequencies, the mediating effect of self-rated health was notably higher than that of life satisfaction. Self-rated health pertains to a single health indicator focused on one’s health status and is more individual-centric. The health perceptions of adolescents depend on their physical health status, with health, disability, higher BMI, stress, sleep quality, and poor health behaviors associated with lower self-rated health [16,51]. Consequently, higher physical activity frequency, as positive health behavior, will be directly associated with higher self-rated health. Simultaneously, increased physical activity participation by adolescents decreases the likelihood of engaging in health-risk behaviors and enhances health behaviors, thus improving self-rated health [52]. Physical activity frequency, duration, and intensity are directly related to physiology [53], and physical health will positively contribute to favorable self-rated health evaluations [54]. Therefore, physical activity can be made to drive mental health through self-assessment of health.

Conversely, life satisfaction is a subjective assessment of the overall quality of life [55]. Factors such as the environment, social desirability effects, personality, interpersonal relationships, community, and life events all influence the life satisfaction of adolescents [55,56]. Although physical activity frequency is closely related to life satisfaction, its effect on life satisfaction is relatively small compared to other influencing factors and can be affected by different life aspects. For instance, physical activity frequency is influenced by intrinsic motivation in buffering the adverse effects of stress on life satisfaction in adolescents [57]. As a result, the relationship of physical activity frequency and life satisfaction may be confounded by other factors.

Higher physical activity frequencies—as optimistic, healthy behaviors—were more significantly associated with self-rated health. In contrast, less correlation between the frequency of physical activity and life satisfaction, as life satisfaction is more profoundly and complexly affected by other individuals or the external environment. Therefore, the mediating effect of self-rated health was found to be higher than that of life satisfaction.

This study offers valuable insights into the effects of physical activity frequency on adolescents’ mental health, as well as the mediating roles of life satisfaction and self-rated health. The main findings were as follows: (1) Physical activity was significantly correlated with mental health in adolescents, mental health levels increase with frequency of physical activity, and (2) Life satisfaction and self-rated health played significant mediating roles in the relationship between physical activity frequency and mental health. Specifically, low-frequency physical activity has the most pronounced mediating effect on mental health through life satisfaction, while high-frequency physical activity exerts the most significant mediating effect on mental health through.

Despite the valuable contributions of this study, there are some limitations to consider. First, the use of the SCL-90 scale, which reflects mental health status over one week, may not fully capture the long-term effects of physical activity frequency on mental health. Future research could employ longitudinal designs to better understand the causal relationships between physical activity frequency, life satisfaction, self-rated health, and mental health over time.

Second, the sample consists of Chinese high school students, which limits the generalizability of the findings to other cultural contexts and age groups. Future studies could examine the relationships of interest in diverse populations, considering factors such as cultural differences, socio-economic status, and age.

Third, the present study only focuses on the mediating roles of life satisfaction and self-rated health. There may be other potential mediators, such as social connectedness, academic performance, and sleep quality, which could be explored in future research to provide a more comprehensive understanding of the relationship between physical activity frequency and mental health.

Lastly, the study relies on self-reported measures, which may be subject to reporting biases. Future research could incorporate objective assessments of physical activity and mental health, as well as qualitative data, to gain a richer understanding of the phenomena under investigation.