Recreation specialization refers to a continuum of behavior ranging from general to specialized, reflecting the equipment and skills utilized in leisure activities, as well as the preference for activity sites or environments (Bryan, 1977). This concept has been widely employed to assess the diverse behavioral characteristics exhibited by participants engaging in outdoor leisure activities (Liu and Lou, 2019; Tian, 2021). Previous research has established a three-dimensional framework for recreation specialization, comprising behavior, cognition, and affect (Song et al., 2018). Marathon runners typically undergo a progressive journey from novice to expert and demonstrate notable attributes associated with recreation specialization such as sport commitment, knowledge acquisition, and skill development (Tian, 2021).

Life satisfaction reflects individuals’ cognitive evaluation of their overall quality of life and is considered an important indicator of subjective well-being (Diener et al., 1999). It is commonly defined as “a positive assessment of one’s entire life based on subjective criteria” (Pavot and Diener, 1993). Although previous literature has overlooked the examination of the impact of recreation specialization on life satisfaction, indirect and reliable evidence has been provided. A case study conducted on elderly cyclists explored the relationship between personality, recreation specialization, and subjective well-being, revealing that recreation specialization can significantly enhance individuals’ subjective well-being (Tian et al., 2023). Another study confirmed that the affect dimension of recreation specialization had a significant positive influence on life satisfaction among long-distance runners (Sato et al., 2018). Furthermore, serious participants reported experiencing higher levels of life satisfaction and health perception (Heo et al., 2013; Kim et al., 2011; Long and Wang, 2013). Additionally, demographic variables such as age and marital status also play a significant role in the relationship between leisure sports participation and individuals’ subjective well-being (Tian et al., 2020).

Originating from Csikszentmihalyi’s (1975) study, flow experience has been frequently employed to elucidate individuals’ persistent engagement in certain activities devoid of material or economic gains. It denotes a psychological state wherein individuals immerse themselves in specific activities, strike a balance between challenges and skills, seamlessly integrate actions and consciousness, and derive intrinsic satisfaction (Csikszentmihalyi, 2002). Consequently, individuals develop disciplined exercise habits, accumulate extensive knowledge, skills, and experiences while exhibiting unwavering commitment towards engaging in leisure sports activities (Park et al., 2018). Recent research has demonstrated that leisure involvement significantly and positively impacts participants’ experiences (Cheng et al., 2016; Weng et al., 2020). Wu et al. (2013) further confirmed that individuals with higher levels of recreation specialization consistently encounter more frequent flow experience while recreation specialization moderates the influence of flow experience on addiction tendencies.

There exists a robust association between recreation specialization, flow experience, and life satisfaction. A higher level of recreation specialization indicates individuals’ persistent inclination towards their leisure choices (Bryan, 2000, 2001) and contributes to the attainment of enduring benefits such as self-image, self-satisfaction, and self-fulfillment (Tian et al., 2020). Previous research has suggested that recreation specialization significantly enhances individuals’ life satisfaction and subjective well-being (Heo et al., 2013; Kim et al., 2011; Tian et al., 2020). Drawing on data from an acrobatics show, a study identified flow experience as a crucial predictor and determinant of life satisfaction (Chen et al., 2010). Furthermore, recent findings have confirmed that flow experience plays a vital mediating role between demographics variables and life satisfaction among mountain bike tourists (Amatulli et al., 2021). In essence, it can be posited that recreation specialization enables individuals to attain heightened levels of flow experience while flow experience indirectly promotes the influence of recreation specialization on marathon runners’ life satisfaction.

Commitment refers to the process in which people willing to give their time or energy to leisure activity in order to express their needs (Buchanan, 1985). Psychological commitment reflects individuals’ inclination to actively engage in preferred activities despite the availability of alternative options (Lee and Kyle, 2014; Pritchard et al., 1999). It is considered a crucial factor in understanding why people choose specific leisure activities or revisit particular destinations (Backman and Crompton, 1991; Pritchard et al., 1999). The level of psychological commitment naturally increases with the progression of individuals’ involvement in recreational pursuits (Stebbins, 2007). By constructing a structural equation model, Ridinger et al. (2012) discovered that both leisure involvement and negotiation efficiency significantly predict marathon runners’ perceived constraint. Additionally, Kyle et al. (2004) confirmed that higher levels of leisure involvement are associated with stronger levels of perceived constraint on leisure activities or destinations.

Previous studies have identified the crucial role of psychological commitment in mediating the relationship between leisure participation and sports experience. Cheng et al. (2016) confirmed a positive association between higher levels of leisure involvement and stronger perceptions of flow experience and psychological commitment, with psychological commitment found to mediate the link between leisure involvement and flow experience. Scholars have also established that dimensions of psychological commitment can indirectly influence individuals’ behavioral loyalty through their engagement in leisure activities (Iwasaki and Havitz, 2004). Furthermore, Sato et al. (2015, 2016, 2018) demonstrated that participating in marathon events leads to positive experiences and a sense of achievement, which contribute to enhancing runners’ life satisfaction. Tian et al. (2022b) provided evidence for the mediating effect of psychological commitment on the relationship between different dimensions of recreation specialization (i.e., behavior, cognition, and affect) and life satisfaction. Therefore, it is plausible that recreation specialization and flow experience positively impact runners’ psychological commitment; moreover, enhanced psychological commitment may facilitate improvements in life satisfaction while also serving as a mediator between recreation specialization and life satisfaction.

In general, the objectives of this study are twofold. The first is to evaluate the direct effects among recreation specialization and life satisfaction. The second is to investigate the mediating effects of flow experience and psychological commitment on the role of recreation specialization on life satisfaction. Based on the aforementioned discussion, the hypotheses of this study are as follows:

The study comprised a total of 404 Chinese marathon runners. Within the sample, the majority of participants were male (269 or 66.6%). A significant proportion fell within the age range of 18 to 44 years old (368 or 91.1%). Most of the respondents possessed a college or university level education (284 or 70.3%), while a smaller fraction had completed high school or below (27 or 6.7%). The majority of runners were unmarried (251 or 62.1%), with over 30% being married individuals. Nearly half of the respondents reported an annual income exceeding US $3,001 (198 or 49%). Moreover, most participants were members of a running group (277or 68.6%) (See Table 1).

Recreation specialization was assessed using a modified version of a scale from previous studies (Iwasaki and Havitz, 2004; McIntyre and Pigram, 1992; Song et al., 2018), comprising nine items that encompassed three dimensions: behavior (three items), cognition (two items), and affect (four items). This measure aimed to evaluate the level of participation and engagement among marathon runners. Participants rated each item on a five-point Likert scale ranging from “strongly disagree” (scored 1) to “strongly agree” (scored 5). The recreation specialization statement used in this study was: “Running is important to me.” Higher scores indicated a greater degree of recreation specialization among marathon runners. Internal consistency analysis yielded Cronbach’s α coefficients ranging from 0.75 to 0.94 for the different dimensions of recreation specialization in this study. Confirmatory factor analysis results showed χ² = 73.30, df = 24, RMSEA = 0.07, RMR = 0.04, CFI = 0.98, TLI = 0.96.

The Satisfaction with Life Scale (SWLS), developed by Diener et al. (1999), was employed to assess the overall quality of life among marathon runners. SWLS comprises five items, which are rated on a seven-point Likert scale ranging from “1” indicating “strongly disagree” to “7” representing “strongly agree.” The statement for life satisfaction was as follow: “In most ways my life is close to my ideal”. A higher total score indicates a greater level of life satisfaction among runners. In this study, the life satisfaction sub-scale demonstrated excellent internal consistency with a Cronbach’s α coefficient of 0.92.

Flow Short Scale (FSS), adapted from previous literature (Engeser and Rheinberg, 2008; Rheinberg et al., 2003), was employed to assess the flow state of marathon runners. The FSS comprises 10 items, encompassing absorption in activity (four items) and fluency of performance (six items). Participants rated these items on a seven-point Likert scale ranging from “1” indicating “not at all” to “7” representing “very much.” The statement for flow experience was as follow: “When running, I am totally absorbed in what I am doing”. Higher total scores indicate a stronger level of flow experience. The Cronbach’s α coefficients for flow experience dimensions were found to be 0.87 and 0.95, respectively. Confirmatory factor analysis yielded χ² = 101.50, df = 34, RMSEA = 0.08, RMR = 0.05, CFI = 0.95, TLI = 0.93.

Psychological Commitment Scale (PCS), adapted from previous studies (Kyle et al., 2004; Pritchard et al., 1999), was utilized to assess the level of commitment among marathon participants. PCS comprised 10 items encompassing four dimensions: positional involvement (three items), resistance to change (three items), information complexity (two items), and volitional choice (two items). These items were rated on a seven-point Likert scale ranging from “strongly disagree” (score 1) to “strongly agree” (score 7). The statement for psychological commitment was as follows: “My preference to participation in running will not willingly change”. A higher total score indicates a stronger level of psychological commitment. The Cronbach’s α coefficient for the factors of psychological commitment ranged from 0.75 to 0.90. The confirmatory factor analysis yielded the following results: χ² = 89.90, df = 29, RMSEA = 0.07, RMR = 0.06, CFI = 0.96, TLI = 0.94.

In accordance with previous research (Tian et al., 2020), a demographic variables questionnaire comprising six items, namely gender, age, marital status, education level, income, and membership of running group was employed by the researchers.

The data for this study were collected at the 2022 Shaoxing International Marathon Event, held on November 12 in Shaoxing, a city in Zhejiang province, China. This event has been officially recognized as a nationally influential event by the China Athletic Federation and consistently attracts over 15,000 runners annually since 2017. Based on the size of the event, the data collection process involved random sampling (i.e., every 25th runner) through Questionnaire Star, a web-based questionnaire platform deployed near the marathon finish area. The studies involving human participants were reviewed and approved by the Shaoxing University Ethics committee. The participants provided their written informed consent to participate in this study. A total of 432 questionnaires were initially collected; However, 28 samples were excluded from the data analysis according to Bryan’s (1977) definition of recreation specialization due to a low running frequency per week (e.g., running less than once per week). Consequently, a final sample size of 404 questionnaires was used to examine the study hypotheses.

The data in this study were analyzed using software of Amos 22.0 and JASP 0.16.1. Initially, confirmatory factor analysis was employed to assess the validity of all adapted scales. Subsequently, JASP was utilized to provide descriptive statistics and evaluate the internal consistency reliability of all variables, as well as test research Hypotheses 1 through Hypotheses 3.

To control for common method variance, this study employed Harman’s single—factor test to examine common method variance. The results indicated that the first unrotated factor accounted for only 35.60% of the total variance, which was less than 40% of the total explained variance. According to the suggestion of Tang and Wen (2020), this study does not have a serious common method bias.

The mean, standard deviation, and correlation coefficient of each variable are presented in Table 2. Marathon runners demonstrate higher levels of recreation specialization (M = 3.323; SD = 0.811), life satisfaction (M = 5.104; SD = 1.189), flow experience (M = 5.369; SD = 1.093), and psychological commitment (M = 5.369; SD = 1.164). The study findings reveal significant correlations among recreation specialization, life satisfaction, flow experience, and psychological commitment (p < 0.01), with correlation coefficients ranging from 0.489 to 0.851.

Table 2 also shows the composite reliability, and square roots of the average variances extracted of recreation specialization, flow experience, psychological commitment, and life satisfaction. Previous research suggests that a recommended acceptable standard of an AVE value higher than 0.50 and a CR value higher than 0.70 (Fornell and Larcker, 1981). Overall, the composite reliability value ranged from 0.783 to 0.927; the average variances extracted ranged from 0.646 to 0.764. In addition, the square roots of the average variance extracted values of all the subscales were higher than the correlations with other constructs, demonstrating the high discriminant validity of all the subscales. Therefore, the constructs in the hypothesized model demonstrated acceptable reliability and validity.

The statistical software JASP 0.16.1 was utilized to analyze Hypotheses 1 through 3, with recreation specialization serving as the predictor variable, flow experience and psychological commitment as mediators, life satisfaction as the outcome variable, and demographic variables included as control variables. As shown in Table 3, the results indicate that a positive relationship was found between recreation specialization (β = 0.528, p < 0.001) and life satisfaction in the model without mediators, which supports H1 (see Figure 1).

The mediating effects were assessed and the 95% confidence intervals were computed, as presented in Table 4. The study findings provide confirmation that both flow experience and psychological commitment play a significant mediating role in the relationship between recreation specialization and life satisfaction, as indicated by the exclusion of zero from all confidence intervals. Therefore, H2 (a*b = 0.491, 95% CI = 0.271–0.659) and H3 (a*b = 0.093, 95% CI = 0.030–0.164) are supported, highlighting that flow experience partially explains its influence on life satisfaction through psychological commitment (See Table 4). However, the direct effect between recreation specialization and life satisfaction becomes non-significant (95% CI = −0.117–0.666) in the model with mediators. Therefore, it can be concluded that flow experience and psychological commitment fully mediate the impact of recreation specialization on life satisfaction.