Post-Earthquake Well-Being, Trauma, and Physical Activity Motivations in Young Adults

1 Introduction

Natural disasters are extraordinary events that profoundly affect individuals’ lives physically, economically, and psychologically [1,2,3]. As a country that frequently experiences earthquakes due to its geological structure, Turkey faces recurring challenges that can adversely influence the psychological well-being of young people living or studying in affected regions [4,5,6]. Large-scale earthquakes not only cause material losses but also substantially disrupt individuals’ emotional balance, social relations, and overall life satisfaction [7,8,9]. Numerous studies have documented increases in anxiety, stress, motivational decline, and behavioral symptoms among students in the aftermath of such disasters [10,11,12].

University students represent a particularly relevant population in post-disaster contexts because they simultaneously navigate developmental, academic, and social transitions while maintaining performance expectations [13,14]. Consequently, disaster-related disruptions may have pronounced effects on student-specific outcomes, including psychological well-being, academic motivation, and behavioral adjustment [15,16]. In addition, educational interruptions such as university closures and disruptions to academic continuity may further increase stress and compromise academic engagement, underscoring the need for targeted psychosocial and educational support following natural disasters [17,18,19]. Within this context, students in sport sciences occupy a distinctive position. Sport sciences education encompasses not only physical performance but also psychological resilience, motivation, and self-regulation [14,20]. Psychological well-being includes dimensions such as meaning in life, self-acceptance, environmental mastery, positive relations, personal growth, and autonomy [21,22]. Disruptions in these domains can negatively affect students’ academic and athletic performance. Furthermore, motivational structures, particularly intrinsic and extrinsic motivation, are fundamental determinants of students’ participation in academic and sport-related activities [23]. Behavioral symptoms, as observable indicators of stress, anxiety, or depressive tendencies, provide crucial insights into an individual’s psychological state [24,25].

Positive emotions generated through physical activity, teamwork, and personal development may support sport sciences students’ psychological recovery and enhance their resilience following a disaster [26]. This perspective aligns with findings suggesting that regular physical exercise serves as an effective strategy for coping with trauma, facilitating both post-traumatic growth and psychological resilience [20,27]. Evidence on post-traumatic growth demonstrates that some individuals may achieve personal development through the challenges they experience, a process particularly meaningful for athletes [28]. The discipline and goal-oriented structure inherent in sports may offer a supportive framework for processing trauma and rebuilding psychological resources [26].

The present study adopts a resource–adaptation–motivation perspective in which disaster-related resource loss forms the contextual background, motivational processes represent the central explanatory mechanism, and psychological well-being constitutes the primary outcome.

Accordingly, this study aimed to examine the psychological well-being, post-earthquake trauma symptoms, recreational sport well-being, and sport motivation profiles of university students studying in Türkiye’s earthquake-affected region. In addition, the study investigated how demographic factors (gender, academic year, socioeconomic status, and prior earthquake exposure) shape these psychological and motivational outcomes.

In this regard, examining the psychological well-being, motivational structures, and behavioral symptom levels of sport sciences students studying in earthquake-affected regions is essential. Such research holds significant value both for developing post-disaster psychosocial support policies and for improving educational practices within sport sciences programs.

2 Conceptual Framework

University students living in disaster-affected regions constitute a vulnerable population, as they must simultaneously maintain academic responsibilities while coping with the psychological consequences of trauma [29,30]. The conceptual framework of the present study is organized around Self-Determination Theory (SDT) [23] as the primary explanatory perspective linking motivational processes with psychological well-being. Complementary theoretical perspectives are integrated to contextualize disaster-related resource loss, adaptation mechanisms, emotional processes, and demographic influences.

In disaster contexts, individuals frequently experience losses in personal, social, and material resources. Conservation of Resources (COR) Theory [31] provides the contextual foundation for understanding how such losses may weaken perceptions of control, competence, and security, thereby increasing vulnerability to psychological distress. Resource loss is particularly salient because its psychological impact is often stronger than resource gain [32,33]. Among university students, this process may manifest as reduced motivation and increased behavioral symptoms following earthquakes [6]. The ability to preserve or rebuild resources, therefore, plays a central role in maintaining psychological well-being in disaster-prone environments [20,32,34,35].

Within this context, adaptive recovery processes are conceptualized through Resilience Theory, which emphasizes individuals’ capacity for positive adaptation despite adversity [36]. Resilience reflects normative adaptive processes rather than extraordinary traits, allowing individuals to regain psychological stability following traumatic experiences [37]. For sport sciences students, participation in physical activity, teamwork, and athletic discipline may serve as important resilience resources that facilitate post-disaster recovery [4,38,39].

Beyond adaptation, positive emotional experiences may contribute to rebuilding psychological resources. According to Broaden-and-Build Theory [40,41], positive emotions expand individuals’ cognitive and behavioral repertoires, enabling the development of enduring psychological resources. This perspective supports the idea that emotional experiences related to sport participation may strengthen coping capacity and contribute to well-being following adversity.

Building on these contextual foundations, SDT [23] serves as the central explanatory mechanism of the present study. The theory conceptualizes motivation along a continuum ranging from amotivation to external and intrinsic motivation, proposing that more self-determined forms of motivation are associated with greater psychological well-being, persistence, and adaptive functioning. In this study, intrinsic motivation, integrated regulation, and identified regulation represent self-determined motivational orientations, whereas external regulation and amotivation represent less autonomous forms. The nature of students’ motivation toward sport participation in the post-disaster context may therefore play a key role in shaping psychological well-being [42,43].

Psychological well-being itself is conceptualized within the Warwick–Edinburgh Psychological Well-Being Model [44], which integrates both hedonic and eudaimonic dimensions of well-being, including positive emotions, effective functioning, and meaningful social relationships. This framework provides a holistic conceptualization suitable for examining psychological adaptation among sport science students living in earthquake-affected regions.

Finally, demographic and contextual characteristics are incorporated as supporting explanatory dimensions. The Social Determinants of Health approach emphasizes socioeconomic factors such as income and living conditions as key influences on mental health and well-being [45]. Similarly, Gender Role Theory highlights differences in emotional expression and coping strategies [46], while Developmental Psychology and Self-Regulation frameworks suggest that advancing academic experience may strengthen self-regulatory capacities and adaptive functioning [47]. In sport contexts, gender roles, competitive expectations, and developmental factors may further shape motivational and emotional processes.

This integrated framework can position disaster-related resource loss and resilience as contextual foundations, positive emotional processes as supportive mechanisms, and self-determined motivation as the central pathway linking individual and contextual factors to psychological well-being among sport sciences students in earthquake-affected regions.

3 Methods

3.1 Study Design

This study employed a quantitative cross-sectional survey design to investigate the psychological well-being, motivational structures, and behavioral symptom levels of sport sciences students studying at two universities located in earthquake-affected provinces of Türkiye (Kahramanmaraş and Malatya). Data were collected using an online questionnaire administered during a single data-collection period.

The overall research framework was structured around three interrelated components: (a) demographic and contextual variables, including age, gender, marital status, academic year, socioeconomic indicators, active sport participation, earthquake-related experiences, and personal habits; (b) standardized psychometric instruments assessing psychological well-being, post-earthquake trauma symptoms, recreational sport-related well-being, and sport motivation; and (c) an analytical framework examining associations among demographic factors, motivational structures, behavioral symptoms, and psychological well-being within a post-disaster context.

Psychological well-being was assessed using the Warwick–Edinburgh Mental Well-Being Scale (WEMWBS), post-earthquake trauma symptoms were assessed with the Post-Earthquake Trauma Assessment Scale (PEAS), recreational sport-related well-being was measured using the Recreational Sport Well-Being Scale (RSWBS), and motivational orientations were measured using the Sport Motivation Scale-6 (SMS-6).

Fig. 1 presents a schematic overview of the study design, illustrating the measured constructs, data flow, and analytical structure used to examine relationships among the study variables.

Figure 1: Study design. Note: WEMWBS, Warwick-Edinburgh Mental Well-Being Scale; PETAS, Post-Earthquake Trauma Assessment Scale; RSWBS, Recreational Sports Well-Being Scale; SMS-6, Sport Motivation Scale.

3.2 Participants

Participants were recruited from the Faculties of Sport Sciences at two universities located in the earthquake-affected region of Türkiye: Kahramanmaraş Sutcu Imam University (Kahramanmaraş) and Inonu University (Malatya). Recruitment was conducted using a voluntary convenience sampling approach. Students were informed about the study during class sessions and invited to participate if they met the inclusion criteria. Participation was entirely voluntary, and no incentives were provided.

A total of 651 sport sciences students participated in the study. The mean age of the participants was 21 years, and the majority were male. Among the academic programs, the highest representation was from the Sports Management department. Most students reported a monthly family income between TRY 10,000 and 30,000, and more than half indicated that they were actively engaged in sport.

An a priori power analysis was conducted for the primary multiple regression model (k = 19 predictors) using an alpha level of 0.05, a statistical power of 0.95, and a medium effect size (Cohen’s f2 = 0.15) [48,49]. The analysis indicated a minimum required sample size of N = 218 participants. The final sample (N = 651) substantially exceeded this threshold, confirming adequate statistical power to detect medium effects within the primary regression model.

Additionally, a substantial proportion of students (41.32%) had previously experienced a major earthquake, while a smaller proportion (5.22%) reported losing a first-degree relative in the most recent disaster. Although the rate of receiving psychological support was relatively low, tobacco and alcohol use appeared to be at notable levels. These descriptive characteristics provide an important contextual foundation for interpreting the psychological well-being, motivational structures, and behavioral processes of young individuals living in the earthquake-affected region.

3.3 Data Collection

Participants’ demographic characteristics were assessed using questions covering gender, age, marital status, university of enrollment, academic department, year of study, family income, regular sport participation, earthquake-related experiences (including location during the earthquake and loss of a first-degree relative), prior earthquake exposure, receipt of psychological support within the past year, and cigarette and alcohol use.

Psychological well-being was assessed using the WEMWBS [44]. The scale consists of 14 items and is scored using a total score, with higher scores indicating better mental well-being. The Turkish adaptation was conducted by Demirtaş and Baytemir [50].

Post-earthquake trauma symptoms were assessed using the PEAS [51]. The scale includes subdimensions related to behavioral problems, emotional restriction, affective symptoms, cognitive restructuring, and sleep problems, with higher scores reflecting higher trauma severity.

Recreational sport–related well-being was assessed using the RSWBS, originally developed by Pi et al. [52] and later adapted into Turkish by Koç [53]. The scale includes subdimensions related to physical and mental health, life satisfaction, family relationship development, and positive emotion, with higher scores indicating greater sport-related well-being.

Sport motivation was assessed using the SMS-6, originally developed by Mallett et al. [54] and later adapted and revised for Turkish samples by Demir [55]. The scale measures amotivation, identified regulation, external regulation, integrated regulation, introjected regulation, and intrinsic motivation.

The internal consistency of the instruments was evaluated using Cronbach’s alpha coefficients calculated from the current sample (N = 651). The Cronbach’s alpha values were 0.866 for the WEMWBS, 0.933 for the PETAS, 0.948 for the RSWBS, and 0.935 for the SMS-6. These values indicated acceptable to good internal consistency for the measures used in this study.

3.4 Data Analysis

No missing data were identified in the dataset. All 651 participants provided complete responses across all variables; therefore, no missing data treatment or imputation procedures were required. Normality was examined using the Shapiro-Wilk test, and homogeneity of variances was assessed using Levene’s test. Descriptive statistics were reported as mean and standard deviation for quantitative variables meeting parametric assumptions, and as median and interquartile range (IQR) for those not meeting these assumptions. Categorical variables were summarized using frequencies and percentages.

For group comparisons, the Mann-Whitney U test was employed for two-group analyses, and the Kruskal-Wallis H test was used for comparisons involving three or more groups. When significant differences were detected, Bonferroni-corrected post hoc analyses were performed.

Correlation analyses were conducted using Spearman’s rho coefficient to visualize relational patterns among variables, and results were presented through heatmap correlation matrices. A significance level of p < 0.05 was considered statistically significant. All results were reported in accordance with APA formatting guidelines. Data analysis was conducted using SPSS version 28.0 (IBM Corp., Armonk, NY, USA) and Python version 3.10.

3.4.1 Supplementary Multivariate Analyses

In addition to the primary nonparametric analyses described above, three supplementary multivariate analyses were conducted to more rigorously address confounding variables and to examine indirect and conditional effects on participants’ Wellbeing Score. All supplementary analyses were performed using Python (version 3.10) with the statsmodels library (version 0.14).

3.4.2 Multivariate OLS Regression

A multivariate Ordinary Least Squares (OLS) regression model was estimated to examine the simultaneous effects of multiple predictor variables on Well-being Score. The model included the following predictor blocks: (a) demographic variables (gender, marital status, age, university name, department, class, family income, active sport participation, smoking, and alcohol use); (b) earthquake-related variables (province during earthquake, loss of a first-degree relative, prior earthquake experience, receipt of psychological support); (c) trauma composite scores derived from the PETAS (sleep problems, cognitive structure, behavioral problems, and emotional restriction); (d) post-earthquake appraisal items (increased attention to behavior/relationships, understood value of life more, need for help hurts honor, became emotional/cried easily); and (e) sport motivation subscales (external regulation, amotivation, introjected regulation, and intrinsic motivation). Categorical province variables were dummy-coded with a reference category. Model fit was evaluated using R-squared, Adjusted R-squared, and the overall F-statistic (significance level p < 0.05).

3.4.3 Mediation Analysis

A mediation analysis was conducted following the Baron and Kenny [56] causal steps approach to test whether Emotional Restriction Composite mediated the relationship between Lost Relative in Earthquake (independent variable) and Well-being Score (dependent variable). Three sequential OLS regression models were estimated: (a) regressing Well-being Score on Lost Relative in Earthquake to obtain the total effect (Path c); (b) regressing Emotional Restriction Composite on Lost Relative in Earthquake to obtain Path a; and (c) regressing Well-being Score simultaneously on Lost Relative in Earthquake and Emotional Restriction Composite to obtain Path b (mediator on outcome) and Path c’ (direct effect of the independent variable controlling for the mediator). The indirect effect was evaluated by inspecting the pattern of significance across these three models. Where a significant total effect was absent but individual paths (a and b) were significant, supplementary bootstrapping methods (e.g., Hayes’ PROCESS macro) are recommended to directly quantify the indirect effect and its confidence intervals.

3.4.4 Moderation Analysis

A moderation analysis was conducted to test whether Active Sport moderated the association between Behavioral Problems Composite (independent variable) and Well-being Score (dependent variable). An OLS regression model was estimated including: (a) the mean-centered Behavioral Problems Composite, (b) the binary Active Sport indicator, and (c) their product interaction term (Behavioral Problems × Active Sport). Mean-centering of the continuous predictor was applied prior to computing the interaction term to reduce multicollinearity. The interaction term coefficient and its statistical significance (p < 0.05) were interpreted as evidence of moderation. Significant main effects of both predictors were also reported to characterize the additive contributions of each variable to Well-being Score.

3.5 Use of Artificial Intelligence

Artificial intelligence (AI)-assisted tools (ChatGPT, OpenAI, San Francisco, CA, USA) were used exclusively for language polishing and improving grammar/fluency in the preparation of this manuscript. No AI tools were used for data collection, statistical analysis, or interpretation of results. The authors take full responsibility for the scientific content of the article.

3.6 Ethical Approval and Informed Consent

Ethical approval for this study was obtained from the Scientific Research and Publication Ethics Committee of Inonu University (Social and Human Sciences Ethics Committee) (Approval No.: E.587342, Decision No.: 28). The study was conducted in accordance with the principles of the Declaration of Helsinki. All participants were informed about the purpose and procedures of the study, and informed consent was obtained prior to participation. Participation was voluntary and anonymous, and participants had the right to withdraw from the study at any time without penalty.

4 Results

The study sample consisted of 651 students (age: Mean = 21, SD = 3), including 364 males (55.91%) and 287 females (44.09%) (Table 1). Most participants were single (94.62%). The majority were enrolled at İnönü University (83.87%), whereas 16.13% were studying at Kahramanmaraş Sütçü İmam University.

Regarding academic departments, 42.86% were students in the Sports Management Department, followed by Physical Education and Sports for the Disabled (19.35%), Physical Education and Sports Teaching (19.05%), and Coaching Education (18.74%). Year of study distribution indicated that third-year students constituted the largest group (31.95%), followed by first-year (26.57%), second-year (25.35%), and fourth-year (16.13%) students.

Monthly family income was most frequently reported as TRY 10,000–20,000 (32.10%), followed by TRY 20,000–30,000 (20.43%). A smaller proportion reported income levels above TRY 60,001 (10.29%). Regular participation in sports was reported by 57.60% of the participants.

Regarding earthquake-related variables, 5.22% reported losing a first-degree relative in the earthquake, and 41.32% reported previous experience of a major earthquake. Psychological support within the past year was reported by 7.99% of participants. Smoking was reported by 32.87% of participants, while alcohol use was reported by 11.67%.

Gender-based comparisons were conducted using the Mann–Whitney U test, and effect sizes were reported using the r coefficient (Table 2). Analysis of WEMWBS scores revealed a significant gender difference, with male participants demonstrating higher psychological well-being scores than females (U = 57,086, p = 0.047, r = 0.077). The effect size indicated a small effect.

Regarding PETAS subdimensions, female participants scored significantly higher than males on sleep problems (U = 43,813, p < 0.001, r = 0.140), cognitive restructuring difficulties (U = 37,854, p < 0.001, r = 0.238), behavioral problems (U = 43,150, p < 0.001, r = 0.151), affective symptoms (U = 40,496, p < 0.001, r = 0.195), and emotional restriction (U = 44,347, p < 0.001, r = 0.132). Effect sizes ranged from small to moderate, with the largest effect observed in cognitive restructuring difficulties.

No significant gender differences were observed in the RSWBS subdimensions, including physical and mental health (U = 48,954, p = 0.134, r = 0.056), life satisfaction (U = 50,752, p = 0.461, r = 0.027), family relationship development (U = 52,186, p = 0.891, r = 0.003), and positive emotion (U = 52,592, p = 0.968, r = 0.004). Effect sizes were negligible across these dimensions.

Within the motivation subscales, only external regulation showed a significant gender difference, with female participants scoring higher than males (U = 58,384, p = 0.013, r = 0.099), reflecting a small effect size. No significant gender differences were found for integrated regulation (U = 54,955, p = 0.299, r = 0.042), amotivation (U = 49,321, p = 0.191, r = 0.050), identified regulation (U = 54,110, p = 0.497, r = 0.028), introjected regulation (U = 51,956, p = 0.817, r = 0.007), or intrinsic motivation (U = 54,670, p = 0.360, r = 0.038). Effect sizes for these comparisons were negligible to small.

Comparisons based on marital status were conducted using the Kruskal-Wallis H test, and effect sizes were reported using η2 coefficients (Table 3). Age differed significantly across marital status groups (H = 53.012, p < 0.001, η2 = 0.079), with married participants showing higher median age scores than single and divorced participants, indicating a moderate effect size.

WEMWBS scores did not differ significantly according to marital status (H = 0.129, p = 0.938, η2 = 0.000), suggesting a negligible effect.

For PETAS, a significant difference was observed only for cognitive restructuring (H = 6.884, p = 0.030, η2 = 0.008), with married participants demonstrating higher median scores compared to the other groups, reflecting a small effect size. No significant differences were found for sleep problems (H = 0.875, p = 0.641, η2 = 0.000), behavioral problems (H = 5.444, p = 0.063, η2 = 0.005), affective symptoms (H = 1.533, p = 0.456, η2 = 0.000), or emotional restriction (H = 0.158, p = 0.919, η2 = 0.000).

RSWBS scores did not show significant marital status differences for physical and mental health (H = 2.685, p = 0.261, η2 = 0.001), life satisfaction (H = 2.141, p = 0.346, η2 = 0.000), family relationship development (H = 0.822, p = 0.655, η2 = 0.000), or positive emotion (H = 0.532, p = 0.771, η2 = 0.000). Effect sizes across RSWBS dimensions were negligible.

SMS-6 subdimensions also showed no significant differences according to marital status, including integrated regulation (H = 3.511, p = 0.175, η2 = 0.002), external regulation (H = 2.161, p = 0.343, η2 = 0.000), amotivation (H = 0.091, p = 0.955, η2 = 0.000), identified regulation (H = 2.143, p = 0.345, η2 = 0.000), introjected regulation (H = 1.574, p = 0.460, η2 = 0.000), and intrinsic motivation (H = 4.087, p = 0.131, η2 = 0.003). Effect sizes ranged from negligible to small.

Overall, marital status was associated only with age and cognitive restructuring, whereas WEMWBS, most PETAS subdimensions, RSWBS, and SMS-6 scores did not show meaningful differences, with effect sizes remaining small or negligible.

Comparisons across academic year levels were conducted using the Kruskal-Wallis H test, and effect sizes were reported using η2 coefficients (Table 4).

WEMWBS scores did not differ significantly across year levels (H = 3.423, p = 0.332, η2 = 0.000), indicating a negligible effect.

For PETAS, no significant differences were observed among academic year groups for sleep problems (H = 0.546, p = 0.909, η2 = 0.000), cognitive restructuring (H = 2.356, p = 0.475, η2 = 0.000), behavioral problems (H = 1.299, p = 0.712, η2 = 0.000), affective symptoms (H = 1.424, p = 0.650, η2 = 0.000), or emotional restriction (H = 0.602, p = 0.869, η2 = 0.000), with effect sizes remaining negligible.

RSWBS scores showed significant differences across year levels in physical and mental health (H = 19.217, p = 0.001, η2 = 0.025), life satisfaction (H = 14.778, p = 0.001, η2 = 0.018), family relationship development (H = 14.986, p = 0.001, η2 = 0.018), and positive emotion (H = 12.256, p = 0.005, η2 = 0.014). Post-hoc comparisons indicated that fourth-year students demonstrated higher median scores compared with lower academic year groups. Effect sizes for these differences were small.

SMS-6 subdimensions did not show significant differences across academic year levels, including integrated regulation (H = 1.293, p = 0.677, η2 = 0.000), external regulation (H = 2.365, p = 0.479, η2 = 0.000), amotivation (H = 3.352, p = 0.321, η2 = 0.000), identified regulation (H = 0.778, p = 0.839, η2 = 0.000), introjected regulation (H = 1.698, p = 0.593, η2 = 0.000), and intrinsic motivation (H = 2.413, p = 0.444, η2 = 0.000). All effect sizes were negligible.

Overall, differences across academic year levels were limited to RSWBS dimensions, where fourth-year students exhibited higher well-being-related outcomes; however, effect sizes remained small, suggesting modest practical differences despite statistical significance.

Comparisons across family income levels were conducted using the Kruskal-Wallis H test, and effect sizes were reported using η2 coefficients (Table 5).

WEMWBS scores differed significantly across income groups (H = 23.222, p = 0.001, η2 = 0.028), with higher-income groups demonstrating higher median psychological well-being scores. Although statistically significant, the effect size indicated a small effect.

For PETAS, no significant differences were observed across income levels for sleep problems (H = 5.620, p = 0.335, η2 = 0.000), cognitive restructuring (H = 4.140, p = 0.517, η2 = 0.000), behavioral problems (H = 6.839, p = 0.248, η2 = 0.003), affective symptoms (H = 1.998, p = 0.867, η2 = 0.000, or emotional restriction (H = 6.571, p = 0.281, η2 = 0.002). Effect sizes across PETAS dimensions were negligible.

RSWBS scores showed significant differences across income levels in physical and mental health (H = 14.563, p = 0.011, η2 = 0.015), family relationship development (H = 12.994, p = 0.021, η2 = 0.012), and positive emotion (H = 16.633, p = 0.004, η2 = 0.018), with higher-income groups generally displaying higher median scores. Life satisfaction did not differ significantly across income categories (H = 9.399, p = 0.083, η2 = 0.007). Effect sizes for significant RSWBS differences were small.

SMS-6 subdimensions revealed significant differences for integrated regulation (H = 13.987, p = 0.013, η2 = 0.014), introjected regulation (H = 13.091, p = 0.018, η2 = 0.013), and intrinsic motivation (H = 12.784, p = 0.021, η2 = 0.012), indicating higher median scores among students in higher income groups. No significant differences were observed for external regulation (H = 8.636, p = 0.112, η2 = 0.006), amotivation (H = 1.831, p = 0.867, η2 = 0.000), or identified regulation (H = 8.750, p = 0.103, η2 = 0.006). Effect sizes across SMS-6 dimensions were small.

Overall, income level was associated with higher WEMWBS scores, several RSWBS subdimensions, and selected SMS-6 motivational components; however, effect sizes remained small, indicating modest practical differences despite statistical significance.

Fig. 2 presents Spearman correlation coefficients (r) and significance levels among all study variables.

Positive correlations were observed among well-being–related measures, including WEMWBS, RSWBS subdimensions (physical and mental health, life satisfaction, family relationship, and positive emotion), and related indicators (r range = 0.43–0.63, p < 0.001).

PETAS subdimensions (sleep problems, cognitive difficulties, behavioral problems, affective symptoms, and emotional limitation) showed positive correlations with one another (r range = 0.36–0.74, p < 0.001). Negative correlations were observed between PETAS variables and several well-being indicators (e.g., WEMWBS and emotional limitation: r = −0.35, p < 0.001).

Positive correlations were also found among SMS-6 motivation subdimensions, including integrated regulation, identified regulation, introjected regulation, intrinsic motivation, and external regulation (r range = 0.58–0.88, p < 0.001).

All coefficients represent bivariate associations and should be interpreted as correlational findings.

Figure 2: Spearman correlation heatmap for all participants.

Fig. 3 presents Spearman correlation coefficients (r) and significance levels among study variables for female participants.

Positive correlations were observed among well-being–related variables, including WEMWBS and RSWBS subdimensions (physical and mental health, life satisfaction, family relationship, and positive emotion) (r range = 0.45–0.68, p < 0.001). For example, life satisfaction was positively correlated with positive emotion (r = 0.68, p < 0.001), and WEMWBS was positively correlated with life satisfaction (r = 0.60, p < 0.001).

PETAS subdimensions (affective, cognitive, behavioral, and emotional limitation symptoms) showed positive correlations with one another and negative correlations with several well-being indicators (e.g., WEMWBS and emotional limitation: r = −0.30, p < 0.001).

Positive correlations were also observed among SMS-6 motivation subdimensions, including integrated regulation, identified regulation, introjected regulation, and intrinsic motivation (r range = 0.65–0.75, p < 0.001).

All coefficients represent bivariate associations and should be interpreted as correlational findings.

Figure 3: Spearman correlation heatmap for female participants.

Fig. 4 presents Spearman correlation coefficients (r) and significance levels among study variables for male participants.

Positive correlations were observed among well-being–related variables, including WEMWBS and RSWBS subdimensions (physical and mental health, life satisfaction, family relationship, and positive emotion) (r range = 0.40–0.60, p < 0.001). For example, life satisfaction was positively correlated with positive emotion (r = 0.64, p < 0.001), and WEMWBS was positively correlated with life satisfaction (r = 0.54, p < 0.001).

PETAS subdimensions (sleep problems, cognitive difficulties, behavioral problems, affective symptoms, and emotional limitation) showed positive correlations with one another and negative correlations with several well-being indicators. For instance, WEMWBS was negatively correlated with emotional limitation (r = −0.36, p < 0.001).

Positive correlations were also observed among SMS-6 motivation subdimensions, including integrated regulation, identified regulation, introjected regulation, and intrinsic motivation (r range = 0.70–0.83, p < 0.001).

All coefficients represent bivariate associations and should be interpreted as correlational findings.

Figure 4: Spearman correlation heatmap for male participants.

4.1 Multivariate Regression Analysis of Wellbeing

This section reports the results of the multivariate OLS regression analysis examining associations between demographic variables, earthquake-related factors, resilience indicators, motivational variables, and Wellbeing Score.

The overall model showed a significant fit to the data (R2 = 0.356, adjusted R2 = 0.296, F = 5.887, p < 0.001), indicating that the included predictors explained 35.6% of the variance in Wellbeing Score (Table 6).

As shown in Table 7, significant predictors (p < 0.05) included Province During Earthquake 5 (β = 11.972, p = 0.016), Province During Earthquake 20 (β = −3.265, p = 0.029), Province During Earthquake 26 (β = −5.109, p = 0.043), Family Income (β = 0.349, p = 0.004), Active Sport (β = 1.334, p = 0.002), Behavioral Problems Composite (β = −0.419, p = 0.001), Emotional Restriction Composite (β = −0.505, p < 0.001), Increased Attention Behavior Relationships (β = 0.518, p = 0.031), and Understood Value of Life More (β = 0.806, p < 0.001).

Gender, marital status, age, loss of a relative during the earthquake, sleep problems, cognitive structure, and motivational subdimensions (external regulation, amotivation, introjected regulation, and intrinsic motivation) were not significantly associated with Well-being Score within the multivariate model (p > 0.05).

Table 6 presents overall model fit statistics across analyses, and Table 7 presents the regression coefficients for Wellbeing Score. All estimates represent associations within the fitted model.

4.2 Mediation Analysis: Lost Relative, Emotional Restriction, and Well-Being

This section reports the mediation analysis examining associations among Lost Relative in Earthquake, Emotional Restriction Composite, and Wellbeing Score using regression-based models following the Baron and Kenny approach [56].

In the first model (Path c; total effect), Wellbeing Score was regressed on Lost Relative in Earthquake. The total effect was not statistically significant (β = −0.886, p = 0.380).

In the second model (Path a), Emotional Restriction Composite was regressed on Lost Relative in Earthquake. The predictor showed a significant positive association with the mediator (β = 3.730, p < 0.001).

In the final model (Paths b and c’), Wellbeing Score was regressed simultaneously on Lost Relative in Earthquake and Emotional Restriction Composite. Emotional Restriction Composite was significantly associated with Wellbeing Score (β = −0.327, p < 0.001), whereas Lost Relative in Earthquake was not statistically significant (β = 0.335, p = 0.732).

Table 8 presents the regression coefficients for the mediation analysis models. All estimates represent associations within the fitted models.

4.3 Moderation Analysis: Behavioral Problems, Active Sport, and Well-Being

This section reports the moderation analysis examining whether Active Sport moderates the association between Behavioral Problems Composite and Well-being Score.

The overall moderation model was statistically significant (R2 = 0.082, adjusted R2 = 0.077, F = 19.20, p < 0.001).

As shown in Table 9, the interaction term Behavioral Problems by Active Sport was not statistically significant (β = −0.008, Std. Err. = 0.111, t = −0.074, p = 0.941).

Behavioral Problems Composite showed a significant negative association with Well-being Score (β = −0.281, p = 0.001), whereas Active Sport showed a significant positive association (β = 2.399, p = 0.033).

Table 9 presents the regression coefficients for the moderation analysis model. All estimates represent associations within the fitted model.

5 Discussion

The present study offers a multidimensional perspective on university students’ psychological, motivational, and recreational well-being following a major earthquake. Crucially, the findings revealed a closely interconnected psychosocial pattern broadly consistent with existing trauma, well-being, and motivation literature. Trauma symptoms, psychological well-being, motivational regulation, and recreational well-being appeared closely associated, with these relationships unfolding alongside sociodemographic characteristics such as gender, academic year, and income level.

Consistent with prior research, female students reported higher levels of trauma-related symptoms, including cognitive restructuring difficulties, affective disturbances, sleep problems, and behavioral symptoms. Importantly, this pattern aligns with evidence suggesting heightened psychological vulnerability among women following disasters [57,58]. Similar observations have been reported among students directly exposed to earthquakes [59]. In parallel, the higher psychological well-being scores observed among male students correspond with studies documenting stronger trauma-related symptom severity among women in post-earthquake contexts [60,61,62,63]. These findings reinforce the presence of gender-related variations in post-disaster psychological responses.

The inverse association observed between trauma symptoms and psychological well-being converges with a well-established body of literature. Research employing the WEMWBS [64,65,66,67,68] consistently highlights income level as a relevant correlate of well-being, underscoring the broader role of social determinants in post-disaster adjustment. Likewise, lower income has repeatedly been associated with increased vulnerability to trauma-related responses [69,70]. Viewed collectively, these findings emphasize the importance of accounting for socioeconomic and gender-related inequalities when considering psychosocial support strategies after disasters [71,72,73].

A progressive increase in psychological well-being across higher academic year levels was also observed, aligning with developmental perspectives emphasizing growing maturity, self-efficacy, and coping capacity. Comparable tendencies have been reported in both post-pandemic and post-earthquake contexts [68,73]. Within this framework, traditional gender-role expectations emphasizing caretaking and emotional responsibility may be linked to stronger distress responses among women during crisis situations [74]. In cultural settings such as Türkiye, where social and economic resources may vary, these dynamics may become more pronounced in relation to psychological outcomes [69,71].

Findings from the motivational dimension indicated that motivational regulation was closely linked with both individual and contextual resources. Higher income levels corresponded with higher intrinsic motivation, integrated regulation, and introjected regulation, suggesting that motivational orientations may coexist with structural opportunities as well as personal characteristics. Furthermore, the higher external regulation scores observed among female students align with literature describing reduced motivational autonomy under trauma-related conditions [75,76]. Similarly, prior trauma exposure was associated with lower psychological functioning and weaker motivational profiles, echoing studies linking socioeconomic disadvantage and trauma burden to reduced resilience [58,77,78,79].

With regard to recreational well-being, higher scores were observed among students with advanced academic year levels and higher income, consistent with evidence demonstrating positive associations between recreational engagement and well-being [80]. At the same time, its negative association with trauma symptoms and positive association with psychological well-being suggest that recreational participation may coexist with more adaptive functioning in post-disaster contexts. The absence of gender or marital-status differences further indicates that recreational well-being may be more strongly related to participation intensity and environmental opportunities than to biological factors [79]. This interpretation is also consistent with literature emphasizing the roles of social support, motivation, and access to facilities in shaping sport participation [81].

When considered holistically, findings across all four scales suggest a strongly interconnected psychosocial structure in which trauma symptoms, well-being, motivation, and recreational functioning co-occur in meaningful ways. Higher trauma levels were associated with lower well-being and reduced autonomous motivation, whereas higher well-being and recreational engagement tended to coincide with lower trauma symptom levels and stronger intrinsic motivation. Importantly, the consistent associations observed with academic year and income further highlight the relevance of social determinants in shaping post-disaster adjustment.

From a theoretical standpoint, these patterns align closely with the conceptual frameworks introduced earlier in the study. Social Comparison Theory provides a useful lens for contextualizing gender-related differences in trauma responses, while SDT helps interpret the co-occurrence of stronger intrinsic motivation and higher well-being. In addition, Positive Psychology perspectives suggest that higher academic year and income may correspond with accumulated psychological resources, whereas Self-Regulation Theory frames recreational participation as part of adaptive regulatory processes. Collectively, these converging interpretations indicate that the findings remain theoretically coherent while fully consistent with the associative nature of a cross-sectional design.

5.1 Interpreting Multivariate Analysis

Multivariate analysis refined the psychosocial landscape of post-disaster wellbeing by isolating the unique contributions of interrelated factors. After accounting for shared variance, emotional restriction and behavioral problems emerged as the strongest negative correlates, suggesting these trauma-related processes are centrally implicated in maladaptation. In contrast, Province 5 showed a positive association with psychological well-being, whereas Provinces 20 and 26 were associated with lower well-being levels. Conversely, family income and active sport participation demonstrated independent positive associations, highlighting the concurrent salience of socioeconomic resources and behavioral engagement. Provincial differences further underscored the role of contextual conditions, supporting an ecological perspective wherein wellbeing reflects individual-environment interplay [58].

These findings align with prior research indicating that while direct effects of academic stress may not always be statistically evident, its influence likely operates indirectly through emotional and behavioral pathways [82]. Within this framework, psychological resilience appears as a positive correlate, potentially coexisting with adaptive functioning among trauma-exposed students [82,83]. The independent role of sports participation suggests behavioral activation may facilitate stress regulation and resilience, consistent with evidence linking physical activity to reduced distress [84,85].

Socioeconomic gradients further contextualize these patterns. Higher income and education likely confer cumulative advantage through differential resource access [86], while demographic variations in gender and residential context indicate that post-disaster adjustment differs meaningfully across social groups [87]. Collectively, these results support a multidimensional interpretation wherein emotional regulation, contextual resources, and behavioral engagement concurrently shape well-being, while remaining mindful of the cross-sectional design, which precludes causal inference.

5.2 Interpreting Mediation and Moderation Analysis

Mediation analysis revealed a theoretically significant indirect pathway from trauma exposure to well-being through emotional restriction. Although the total effect was non-significant, the presence of constituent paths suggests that post-disaster well-being is more proximally tied to emotion dysregulation than to the objective burden of loss itself. This pattern aligns with contemporary process-oriented models wherein indirect effects may exist independently of total effects, particularly in complex psychosocial ecologies where exposure exerts influence through intervening mechanisms rather than direct linear impact [88]. The application of bootstrapped confidence intervals strengthens confidence in this indirect pathway by circumventing normality assumptions [5,89].

Moderation analyses clarified the functional role of behavioral engagement. Active sport participation demonstrated a positive main effect on wellbeing, yet failed to moderate the relationship between behavioral problems and wellbeing. This differential pattern suggests that sport participation functions as a generalized positive correlate rather than a targeted buffer against specific maladaptive processes. Such findings resonate with evidence that physical activity confers broad mental health benefits without necessarily attenuating condition-specific risks [85,90].

Synthesized with the main multivariate findings, these analyses converge on a coherent psychosocial architecture: emotional regulation processes emerge as the more proximal determinant of post-disaster wellbeing, while active sport participation and contextual resources operate as independent positive correlates rather than conditional modifiers. This underscores that adaptive recovery hinges primarily on the capacity to process trauma-related emotions, with behavioral engagement and socioeconomic resources providing concurrent but distinct contributions. The absence of moderation effects further suggests these domains function additively rather than synergistically within this post-disaster context.

6 Study Limitations

Despite the strengths of this study, several limitations should be acknowledged. First, although the sample size was statistically adequate as confirmed by an a priori power analysis, participants were recruited through convenience sampling from two universities located in earthquake-affected regions; therefore, the findings cannot be generalized to all university populations or different cultural and regional contexts.

Second, all variables were assessed using self-report measures, which may introduce social desirability bias and subjective interpretation. In addition, the cross-sectional design precludes causal inference; findings should therefore be interpreted as associative rather than predictive or causal relationships.

Third, although supplementary multivariate analyses were conducted to control for potential confounders, several contextual variables were not directly measured. In particular, post-earthquake environmental conditions such as displacement status, housing instability, and variability in living arrangements may have influenced psychological outcomes but were not examined explicitly.

Finally, while validated quantitative scales enabled systematic comparisons, they do not fully capture the complexity of individual post-disaster experiences. Future research using longitudinal and mixed-method designs would provide a more comprehensive understanding of recovery trajectories and contextual adaptation processes.

7 Theoretical and Practical Implications

The findings contribute to a more integrated theoretical understanding of post-disaster wellbeing by showing that emotional and behavioral processes are closely associated with psychological adjustment alongside contextual resources. The consistent associations of emotional restriction and behavioral problems with lower well-being support trauma-informed and self-regulation perspectives emphasizing emotional processing as a central dimension of adaptation. At the same time, the independent positive associations of active sport participation and family income highlight the complementary roles of behavioral engagement and socioeconomic resources, suggesting that post-disaster well-being reflects a multidimensional psychosocial structure rather than a single pathway. In addition, the mediation pattern, characterized by significant indirect paths despite a non-significant total effect, points toward more complex relational mechanisms and supports the consideration of indirect and context-sensitive models in future theoretical work.

From a practical perspective, these findings suggest that interventions targeting emotional regulation and behavioral difficulties may be particularly relevant for students exposed to disaster-related stress. Active sport participation appears to be associated with higher well-being as a general supportive resource, indicating its potential value within broader well-being promotion strategies. At the same time, observed contextual differences, including provincial variation and socioeconomic disparities, highlight the importance of resource-sensitive and context-aware support approaches. Overall, post-disaster interventions may benefit from combining psychological support, opportunities for structured physical activity, and policies addressing socioeconomic inequalities in order to support more comprehensive recovery processes.

8 Conclusions

This study advances understanding of post-disaster psychosocial functioning by empirically integrating trauma symptomatology, psychological well-being, motivational regulation, and recreational engagement within a single analytical framework applied to university students affected by the 2023 Kahramanmaraş earthquakes. The findings collectively indicate that post-disaster adjustment is associated with a multidimensional structure wherein trauma-related processes, intrapersonal resources, and contextual conditions interrelate. Importantly, female students reported elevated trauma symptoms, while variations in well-being and motivational indicators across academic year and income strata suggest that pre-existing developmental and socioeconomic disparities may be more pronounced in disaster contexts.

Trauma symptoms were inversely associated with psychological well-being, whereas recreational engagement and motivational factors exhibited consistent positive associations. These patterns imply that adaptive recovery may involve more than symptom attenuation, potentially encompassing the preservation of motivational structures and social-behavioral engagement. Multivariate extensions refined this picture, revealing emotional restriction and behavioral problems as the most robust negative correlates, while family income and active sport participation contributed independently and positively. Mediation analyses further identified emotional processing as a key relational link, whereas moderation tests indicated that sport participation is associated with generalized benefits rather than condition-specific buffering.

Synthesized, these findings suggest that post-disaster recovery among university students may be conceptualized as a process involving resource maintenance across psychological, motivational, and behavioral domains, not merely clinical symptom reduction. From an intervention standpoint, this implies that effective support architectures may benefit from integrating trauma-informed psychological care with structured physical activity, motivational reinforcement, and equity-oriented policies that address socioeconomic vulnerabilities.

By offering an integrated, contextually grounded portrayal of post-disaster functioning, this study reinforces the relevance of holistic, resilience-oriented perspectives that acknowledge the complexity of student well-being in the aftermath of collective trauma.