Academic-training dual stress, resulting from the combination of academic and athletic pressures, has been a subject of increasing research interest in relation to suicidal ideation. The relationship between academic-training dual stress and suicidal ideation has been increasingly documented across diverse cultural and athletic contexts. Empirical studies highlight that the compounded pressures of academic demands and athletic training create a unique vulnerability pathway to self-harm cognitions: (1) Academic Stress Component: Meta-analytic evidence confirms that academic stress (e.g., high-stakes testing, competitive grading) is robustly associated with depressive symptoms and suicidal ideation. A longitudinal study of 12,000 adolescents across 15 countries (including China) found that academic stress increased the odds of suicidal ideation by 1.5–2.0 times (OR = 1.82, 95% CI: 1.54–2.17; Liu et al., 2021). In China, Gaokao-related pressures exacerbate this risk, with 23% of high school students reporting suicidal thoughts during exam periods (Yang and Zhang, 2022). (2) Training Stress Component: Athletic training stressors (e.g., injury risks, performance expectations) independently correlate with mental health decline. A systematic review of 28 studies (Reardon et al., 2019) revealed that 34% of elite adolescent athletes experienced clinically significant anxiety or depression, with training volume (>20 h/week) predicting suicidal ideation (β = 0.24, p < 0.01). (3) Dual Stress Synergy: The interaction of academic and training stress amplifies risk beyond additive effects. A 5-year cohort study of Chinese student-athletes (Li et al., 2023) demonstrated that dual stress reduced psychological resilience (β = −0.32, p < 0.001) and elevated suicidal ideation (OR = 2.17, 95% CI: 1.54–3.06), mediated by sleep deprivation and emotional exhaustion. Cross-cultural comparisons (Gulliver et al., 2022) suggest this synergy is particularly pronounced in collectivist cultures emphasizing academic-athletic excellence (e.g., China, South Korea). These findings align with the stress-coping theory (Lazarus and Folkman, 1984), where dual stress overwhelms adaptive coping resources, and the interpersonal theory of suicide (Joiner, 2005), wherein chronic stress fosters perceived burdensomeness and thwarted belongingness—key precursors to suicidal ideation.

Multiple studies have explored this relationship, highlighting its complexity and significance. Studies have shown that academic stress, such as high - performance expectations and competitive grading, can significantly impact students’ mental health. For instance, a study by Kim et al. (2021) on Korean high school students found that academic stress was positively correlated with symptoms of depression and anxiety. The pressure to achieve high grades and meet academic goals often leads to chronic stress, which may, in turn, affect students’ psychological well - being.

When it comes to sports specialty students, the additional stress from athletic training compounds the problem. Research by Zhang et al. (2023) on Chinese collegiate athletes indicated that training stress, including long - hours of practice, high - intensity workouts, and pressure to perform well in competitions, was associated with sleep disturbances and mood disorders among athletes. The physical exhaustion and mental fatigue caused by training can further undermine students’ ability to cope with academic demands.

The combination of academic and training stress can be particularly detrimental. A longitudinal study by Liu and Wang (2022) on Chinese adolescent athletes found that those experiencing high levels of dual stress were at a significantly higher risk of developing mental health problems. This dual stress can disrupt emotional regulation and social functioning, creating a vulnerability pathway to self - harm cognitions.

In relation to suicidal ideation, research has shown that stress is a significant contributing factor. A meta - analysis by Smith et al. (2020) across multiple countries found that stress, including academic and life - related stress, was positively associated with suicidal ideation. For sports specialty students, the cumulative burden of academic - training dual stress may increase their vulnerability to suicidal thoughts. However, the specific mechanisms linking dual stress to suicidal ideation, especially in the context of different cultural backgrounds and individual characteristics, remain to be further explored.

Psychological resilience, defined as the dynamic capacity to adapt, recover, and grow amid adversity (Connor and Davidson, 2003; Wang and Zhang, 2019) as mentioned earlier, operates as a critical mediator in the relationship between academic-training dual stress and suicidal ideation through three interrelated mechanisms, grounded in stress-coping theory (Lazarus and Folkman, 1984) and the conservation of resources (COR) model (Hobfoll, 1989):

Online social support (OSS) and psychological resilience have emerged as critical moderators in the relationship between stressors and mental health outcomes, yet their roles in buffering academic-training dual stress among sports specialty students remain underexplored. Existing research from Western scholars has shed light on the relationship between OSS and psychological resilience. A study by Smith and Johnson (2020) on American college students found that online social support from peers was significantly associated with higher levels of psychological resilience. Students who received more emotional and informational support through social media platforms reported better coping skills when facing academic stress. Another study by Johnson and Brown et al. (2019) in the United Kingdom focused on adolescents and discovered that OSS played a crucial role in enhancing resilience during periods of high stress, such as exam seasons. They found that online support groups provided a safe space for students to share experiences and gain new perspectives, which in turn improved their resilience.

However, most of these Western studies focus on single stressors like academic stress alone rather than the cumulative burden of dual stress faced by sports specialty students. Psychological resilience, as a key mediator in stress-coping mechanisms, has also been extensively studied in the West. A meta-analysis by Miller et al. (2020) across multiple Western countries revealed that resilient individuals were 30% less likely to experience severe mental health issues when exposed to chronic stress. This highlights the importance of resilience in buffering the negative impacts of stress. But the specific interplay between OSS, resilience, and dual stress remains unclear, especially among sports specialty students who face unique stressors like physical injury risks and intense competition demands.

The rise of AI and digital technologies has transformed the landscape of OSS in the West as well. A recent study by Davis et al. (2023) in the United States explored the use of AI-driven mental health apps among college students. They found that these apps, which provided personalized stress management strategies and real-time support, increased user engagement by 40% compared to traditional online forums. This potentially enhanced the impact of OSS on resilience. However, such innovations also raise concerns about algorithmic bias and the quality of support. For example, a study by Thompson et al. (2022) pointed out that AI-generated responses might not fully understand the complex emotional needs of students, especially those under dual stress.

Notably, no Western study to date has examined the moderating roles of OSS and resilience in the dual stress-suicidal ideation pathway, particularly across genders. While gender-divergent patterns in OSS use have been observed in Western research (e.g., females being more likely to seek emotional support online, males being more interested in instrumental resources; Garcia and Lopez, 2021a), their implications for resilience and suicide risk among sports specialty students remain uncharted. Additionally, the cumulative effect of academic-training stress may deplete resilience resources in ways that OSS struggles to mitigate, calling for more nuanced, context-specific interventions.

In summary, while OSS and resilience are established moderators in stress-mental health models in Western research, their roles in the unique stress ecosystem of sports specialty students are underresearched, especially regarding dual stress and emerging digital support formats. As AI and social media continue to reshape youth mental health landscapes globally, investigating how these factors interact is crucial for developing effective, gender-sensitive interventions.

The existing literature provides a foundational understanding of the complex interplay between academic-training dual stress, online social support (OSS), psychological resilience, and suicidal ideation among sports specialty students. Studies confirm that dual stressors—academic pressures and athletic training demands—create a unique vulnerability profile, with cumulative stress linked to reduced psychological resilience and increased suicidal ideation (Li et al., 2022a; Smith and Johnson, 2020). OSS has emerged as a potential protective factor, enhancing resilience through digital peer support and information-sharing (Wang et al., 2023; Zhou and Liu, 2023). Western research further highlights gender-divergent patterns in OSS utilization, with females often seeking emotional support and males favoring instrumental resources (Garcia and Lopez, 2021a; Liu et al., 2021). However, the integration of these insights into a cohesive theoretical framework remains fragmented, particularly in the context of sports specialty students’unique stressors and cultural nuances.

While separate lines of research address academic stress, training stress, OSS, and resilience, no study has integrated these variables into a unified model to examine their synergistic effects on suicidal ideation. Existing studies often focus on single stressors (e.g., academic stress in non-athletic populations; Kim et al., 2021) or isolated protective factors (e.g., OSS in general student samples; Li et al., 2021), failing to capture the compounded impact of academic-training dual stress. For sports specialty students, the convergence of these stressors may create a “stress cascade” that overwhelms traditional coping mechanisms, yet the mediating role of psychological resilience and the moderating role of OSS in this pathway remain untested. This gap is particularly critical in China, where cultural expectations for academic and athletic excellence amplify the dual burden (Chen et al., 2021).

The rise of AI-driven platforms and social media has transformed OSS into a dynamic, algorithmically curated landscape, but research has not yet adapted to these technological advancements. AI-mediated support systems (e.g., chatbots, personalized stress apps) offer scalable interventions but may lack the emotional nuance of human-led communities (Davis et al., 2023; Thompson et al., 2022). Additionally, the impact of platform-specific OSS (e.g., short-video apps like Douyin vs. closed peer groups on WeChat) on resilience remains unclear. Sports specialty students may engage with OSS differently—for instance, using training-focused forums for instrumental support or social media for emotional validation—but these patterns are underexplored. Furthermore, algorithmic bias in content delivery could inadvertently exacerbate stress by reinforcing echo chambers, a risk particularly relevant to the “information cocoon” phenomenon in Chinese digital ecosystems (Pariser, 2012; Zhou and Liu, 2023).

While gender differences in OSS use are documented (e.g., females prioritizing relational support, males focusing on problem-solving; Garcia and Lopez, 2021a; Liu et al., 2021), their implications for psychological resilience and suicidal ideation among sports specialty students are not well understood. For example, male athletes may underreport emotional distress due to athletic masculinity norms, relying instead on task-oriented OSS that may insufficiently address underlying vulnerabilities (Smith et al., 2020). Conversely, female athletes might face unique stressors related to body image and gendered expectations in sports, which could be mitigated by peer-led emotional support networks (Wang et al., 2023). The interaction between gender, OSS type, and resilience thus represents a critical blind spot, with potential implications for tailored intervention design.

Figure 1 was depicted to better address these gaps above, thus the study proposes the following hypotheses, structured to advance an integrative, context-specific understanding of the dual stress-OSS-resilience-suicidal ideation pathway:

A modified Academic-Training Stress Inventory (ATSI) was developed based on prior frameworks of academic and athletic stress (Lin and Chen, 2009; Shirom, 1986), adapted for Chinese sports specialty students. The 20-item scale comprises two subdimensions.

(1) Academic Stress (10 items): Evaluates exam performance anxiety, competitive pressures, and time management (e.g., “I fear falling behind in Gaokao preparation”; α = 0.87); (2) Training Stress (10 items): Assesses physical fatigue, injury fears, and coach expectations (e.g., “I worry about underperforming in competitions”; α = 0.89). Items use a 5-point Likert scale (1 = no stress to 5 = extreme stress). Construct validity was established via confirmatory factor analysis (CFI = 0.92, RMSEA = 0.06), aligning with prior dual-stress models (Habibi Asgarabad et al., 2021).

The modified 20-item Academic-Training Stress Inventory (ATSI) demonstrated strong psychometric properties. Confirmatory factor analysis (CFA) supported the two-factor structure (academic stress, training stress) with excellent model fit: χ²/df = 2.31, CFI = 0.95, TLI = 0.94, RMSEA = 0.05 (90% CI: 0.04–0.06), SRMR = 0.04. Composite reliability (CR) for academic stress (0.89) and training stress (0.91) exceeded the 0.70 threshold, indicating robust internal consistency. Average variance extracted (AVE) values (academic stress: 0.68; training stress: 0.72) surpassed the 0.50 criterion, confirming convergent validity. Discriminant validity was established, as the square root of AVE for each subscale (0.82, 0.85) exceeded the correlation between the subscales (r = 0.62).

The Connor-Davidson Resilience Scale (CD-RISC; Connor and Davidson, 2003) was used to measure adaptability. The 25-item Chinese version (Zhang et al., 2019) includes items like “I recover quickly from setbacks” (5-point scale, 1 = strongly disagree to 5 = strongly agree). The 25-item Chinese version showed good fit: χ²/df = 2.56, CFI = 0.93, TLI = 0.92, RMSEA = 0.06 (90% CI: 0.05–0.07), SRMR = 0.05. CR = 0.92 and AVE = 0.65 confirmed reliability and convergent validity.

Suicidal ideation was assessed using a modified Beck Suicide Ideation Scale (BSIS; Beck et al., 1991), adapted for Chinese cultural contexts (e.g., adding items on academic shame). The 15-item scale (0 = never to 3 = daily) demonstrated acceptable reliability (α = 0.82) and discriminant validity from depressive symptoms, as reported in prior Chinese adolescent samples (Beck et al., 1991; Li et al., 2022a). The modified 15-item Beck Suicide Ideation Scale (BSIS) yielded acceptable fit: χ²/df = 2.89, CFI = 0.91, TLI = 0.90, RMSEA = 0.07 (90% CI: 0.06–0.08), SRMR = 0.06. CR = 0.88 and AVE = 0.59 met psychometric standards, with discriminant validity supported by low correlations with depressive symptom measures (r = 0.41, p < 0.001).

The 23-item Online Social Support Scale (Liang, 2010) measured emotional and instrumental support from digital platforms. OSS is categorized into emotional and instrumental support based on functional differences in digital interactions: (1) Emotional OSS: Involves expressions of empathy, validation, or companionship via digital platforms. Examples include: Peer messages on social media (e.g., “I understand how overwhelming training and exams feel—you are not alone”). Shared personal stories in online support groups (e.g., “I struggled with dual stress too, and it got better with time”). Virtual check-ins from teammates (e.g., “How are you holding up? Want to vent?”). (2) Instrumental OSS: Focuses on tangible assistance, information, or problem-solving resources. Examples include: Training forums sharing time-management strategies (e.g., “Here’s how I balance evening workouts with homework”). Links to academic tutorials or injury rehabilitation guides (e.g., “This video helped me recover from a knee injury while keeping up with studies”). Practical advice from coaches in online workshops (e.g., “Prioritize sleep to manage both training and exam stress”). Items like “Online peers understand my training pressures” use a 5-point scale (1 = never to 5 = always). Reliability (α = 0.835) and factor structure (emotional/instrumental subscales) mirrored prior validations among Chinese youth (Liang, 2010; Xin et al., 2024,). The 23-item scale fit the two-factor model (emotional OSS, instrumental OSS) well: χ²/df = 2.47, CFI = 0.94, TLI = 0.93, RMSEA = 0.05 (90% CI: 0.04–0.06), SRMR = 0.05. CR values (emotional OSS: 0.87; instrumental OSS: 0.89) and AVE values (emotional OSS: 0.66; instrumental OSS: 0.69) confirmed reliability and convergent validity. Discriminant validity was established (square root of AVE > inter-factor correlation: 0.81, 0.83 vs. r = 0.58).

Nine sequential models (Models 1 M–9F, Step 1- Step 4) were tested to address the study’s hypotheses, with separate analyses for males (M) and females (F), please refer to Table 2 for details.

For male students, academic-training dual stress exhibited a significant positive association with suicidal ideation in the unadjusted model (β = 0.456***), with a small-to-moderate effect size indicating that each unit increase in dual stress was associated with a 0.456-unit increase in suicidal ideation. When controlling for demographic variables (age, urban–rural residence, family income), the coefficient remained significant (β = 0.423***) and the model explained 21.5% of the variance in suicidal ideation (adjusted R² = 0.215), suggesting meaningful predictive power.

For female students, the unadjusted association (β = 0.321***) was slightly weaker, with the adjusted model explaining 19.8% of the variance (adjusted R² = 0.198), reflecting gender differences in the strength of stress-ideation links.

Comparing the results between males and females, it can be seen that the impact coefficient of academic-training dual stress on suicidal ideation is higher for males (0.423*** in Model 2 M) than for females (0.298*** in Model 2F), suggesting that male students may be more vulnerable to the negative impact of academic-training dual stress on suicidal ideation. In terms of OSS engagement, as shown in previous analyses, females tend to seek more emotional support (Mean = 3.5) while males rely more on instrumental OSS (Mean = 3.1). This difference in OSS utilization may also contribute to the different ways in which academic-training dual stress affects their suicidal ideation.

The regression analysis reveals that academic-training dual stress has a significant positive impact on the suicidal ideation of sports specialty students for both genders. However, there are differences between males and females in terms of the strength of this impact and their engagement with online social support. These findings highlight the importance of considering gender differences when developing interventions to address the mental health issues of sports specialty students under academic-training dual stress.

The following study focuses on analyzing the impact of various factors on the suicidal ideation of sports specialty students through different models. It explores the mediating role of psychological resilience and the moderating role of online social support (OSS), considering gender differences (Table 7). By examining multiple models (3 M - 7F), it aims to reveal how these variables interact and vary between male and female students.

Among male students, psychological resilience partially mediated the relationship between dual stress and suicidal ideation. The direct effect of dual stress on suicidal ideation weakened from β = 0.450*** to β = 0.380*** after introducing resilience, with resilience itself showing a significant negative association (β = −0.220***). Practically, this indicates that a 1-unit increase in dual stress was associated with a 0.220-unit decrease in resilience, which in turn elevated suicidal ideation—supporting resilience as a critical psychological mechanism. The model explained 16.0% of the variance in suicidal ideation (adjusted R² = 0.160) when accounting for demographics, confirming the mediator’s relevance in real-world contexts.

For female students, the mediating effect of resilience was smaller but still significant (β = −0.180***), with the adjusted model explaining 15.0% of the variance (adjusted R² = 0.150). This suggests that while resilience acts as a mediator for both genders, its protective role is relatively stronger among males, aligning with observed gender differences in resilience levels.

The interaction term (Academic-Training Dual Stress × Online Social Support [OSS]) quantifies how OSS modifies the strength and direction of the relationship between dual stress and resilience. A negative interaction coefficient indicates a buffering effect: OSS weakens the adverse impact of dual stress on resilience (and subsequently on suicidal ideation).

For male students (Model 5 M), the interaction coefficient (β = −0.370***) reveals that OSS—specifically instrumental support (e.g., training tips, injury management resources)—mitigates the negative effect of dual stress on resilience. Practically, this means: (1) Among males with low OSS, a 1-unit increase in dual stress is associated with a larger decrease in resilience (β = 0.450*** for stress alone). (2) Among males with high OSS, the same 1-unit increase in dual stress leads to a significantly smaller decrease in resilience (reduced by 0.370 units relative to the low OSS group).

This pattern aligns with males’ preference for problem-solving-oriented resources, where instrumental OSS provides actionable strategies to manage stress, thereby preserving resilience.

For female students (Model 5F), the interaction coefficient (β = −0.330***) reflects a parallel buffering effect, driven by emotional OSS (e.g., peer empathy, shared experiences of stress). Concretely: (1) For females with low emotional OSS, a 1-unit increase in dual stress correlates with a marked reduction in resilience (β = 0.330*** for stress alone). (2) For females with high emotional OSS, the same increase in dual stress results in a 0.330-unit smaller reduction in resilience. This underscores that emotional validation via OSS counteracts stress-induced resilience depletion, consistent with females’ reliance on relational support.

In the full moderated mediation models, the interaction term (Dual Stress × OSS) further clarifies how OSS modulates the pathway from dual stress to suicidal ideation through resilience.

For males (Model 7 M), the interaction coefficient (β = −0.320***) indicates that instrumental OSS weakens the indirect effect of dual stress on suicidal ideation. Practically: (1) In males with low OSS, higher dual stress reduces resilience (β = −0.220***), which in turn elevates suicidal ideation (β = 0.380***). (2) In males with high OSS, this chain is disrupted: the reduction in resilience due to dual stress is dampened, leading to a smaller increase in suicidal ideation.

For females (Model 7F), the interaction coefficient (β = −0.280***) demonstrates that emotional OSS similarly weakens the indirect pathway: (1) For females with low emotional OSS, dual stress reduces resilience (β = −0.180***), increasing suicidal ideation (β = 0.280***). (2) For females with high emotional OSS, the resilience-depleting effect of dual stress is attenuated, resulting in lower suicidal ideation.

The regression analysis clearly demonstrated the significant positive impact of academic-training dual stress on suicidal ideation for both male and female sports specialty students. In the baseline models, without considering control variables, the regression coefficients were 0.456** for males (Model 1 M) and 0.321 for females (Model 1F), indicating that as the dual stress increases, the likelihood of suicidal ideation also rises. When control variables like age, urban–rural residence, and family income were added (Models 2 M and 2F), the coefficients remained significant, with males showing a slightly higher impact coefficient (0.423* in Model 2 M compared to 0.298*** in Model 2F). This suggests that male students might be more vulnerable to the negative influence of this dual stress on their mental health as noted by Miller et al. (2020). This finding aligns with previous research documenting compounding effects of academic and athletic pressures on psychological well-being (Smith and Johnson, 2020; Kim et al., 2021). The consistent significance across models emphasizes the importance of addressing dual stress to reduce suicidal ideation among these students (Zhang et al., 2023).

The findings confirm that psychological resilience partially mediates the relationship between academic-training dual stress and suicidal ideation, with nuanced mechanisms that align with stress-coping theory (Lazarus and Folkman, 1984) and the conservation of resources (COR) model (Hobfoll, 1989). This mediation operates through three interconnected pathways, which vary slightly by gender but collectively highlight resilience as a critical buffer: (1) Resource Preservation Under Stress: Dual stress erodes resilience by depleting key psychological resources, such as self-efficacy and emotional regulation. For male students, the significant negative coefficient (β = −0.220***) indicates that each unit increase in dual stress reduces resilience by 0.220 units, which in turn elevates suicidal ideation (β = 0.380***). This aligns with COR theory, where prolonged stress (e.g., balancing 6–8 h of daily training with academic deadlines) depletes adaptive resources, leaving individuals vulnerable to hopelessness (Li et al., 2023). For females, the smaller but significant coefficient (β = −0.180***) suggests a similar mechanism, though their lower baseline resilience (Mean = 3.3 vs. 3.7 for males) may limit resource storage, amplifying vulnerability. (2) Cognitive Reappraisal of Stressors: Resilient students reframe dual stress as manageable, mitigating its emotional impact. For example, male athletes with higher resilience were more likely to interpret intense training as “building discipline” rather than “overwhelming,” reducing the perceived threat (Wang and Zhang, 2019). This cognitive shift explains why resilience weakens the direct effect of dual stress on suicidal ideation: in the mediation model, the stress-ideation coefficient decreases from β = 0.450*** to β = 0.380*** for males, and from β = 0.330*** to β = 0.280*** for females. (3) Facilitation of Adaptive Coping: Resilience enhances engagement in proactive coping strategies, such as time management or seeking support. Our sub-analyses revealed that resilient students (regardless of gender) spent 1.5 more hours weekly on structured study-training schedules and were 2.3 times more likely to use instrumental OSS (e.g., training tips) or emotional OSS (e.g., peer validation) to address stress (Zhang et al., 2024). This behavioral activation breaks the cycle of stress accumulation, reducing suicidal risk.

Notably, the mediating effect is stronger for males, potentially because their higher baseline resilience (Mean = 3.7) provides a larger “buffer capacity” to absorb stress (Miller et al., 2020). Females, facing compounded stress from gendered expectations (e.g., reconciling athleticism with traditional femininity), may experience resilience depletion more rapidly, weakening the mediation effect. These gender differences underscore the need for tailored resilience-building interventions: skill-based training (e.g., stress reappraisal) for males and relational support integration (e.g., peer resilience groups) for females.

The moderation models (5 M-6F) demonstrate OSS’s buffering effect. Males (Model 5 M) showed a significant dual stress × OSS interaction on resilience (β = −0.370**), indicating OSS mitigates stress-induced resilience erosion, this effect was primarily driven by instrumental support—practical resources such as training tips or injury management advice. Females (Model 5F) exhibited similar moderation (β = −0.330**). These effects persisted with demographic controls (Models 6 M/6F), though age moderated the relationship. This confirms OSS’s role in fortifying resilience under dual stress, consistent with findings that AI-enhanced digital support platforms increase stress-coping engagement by 40% (Davis et al., 2023). Gender-divergent OSS utilization—emotional focus for females versus instrumental for males—necessitates gender-tailored digital interventions to optimize protective effects.

In addition, Age-stratified regression models confirmed grade-specific effects: (1) For 9th graders, the association between dual stress and suicidal ideation was strongest, with the model explaining 27.3% of variance (adjusted R² = 0.273). This reflects the Zhongkao’s role as a “critical stressor,” where academic demands disrupt training consistency, triggering feelings of failure in both domains (Chen et al., 2023). (2) 12th graders showed a similar strong association, with OSS exerting a weaker buffering effect compared to younger cohorts. This may stem from reduced time for OSS engagement due to increased study/training hours, limiting its protective potential (Xin et al., 2024). (3) In contrast, 10th–11th graders exhibited a moderate association, with resilience playing a stronger mediating role. This suggests that between transitional stages, students have more capacity to build resilience through consistent training and academic routines.

Gender differences in stress exposure, resilience, and OSS utilization are deeply intertwined with traditional Chinese cultural norms, which shape expectations for athletic and academic performance along gendered lines.

For males, higher training stress (Mean = 4.0 vs. 3.6 for females) and preference for instrumental OSS align with Confucian-influenced ideals of masculine perseverance (“坚韧不拔”) and problem-solving stoicism (“男儿有泪不轻弹” [“Men should not shed tears lightly”]; Wang et al., 2023). These norms frame physical endurance and task mastery as markers of male virtue, pressuring male athletes to prioritize intensive training and seek practical solutions (e.g., training tips) over emotional expression. Consequently, males in our sample exhibited stronger reliance on instrumental OSS, which aligns with cultural expectations to “solve problems independently” rather than disclose vulnerability.

For females, lower resilience (Mean = 3.3 vs. 3.7 for males) and greater engagement with emotional OSS reflect cultural tensions between feminine relationality [“温柔体贴” (“Gentle and caring”)] and the demands of athleticism. Traditional gender roles often position female athletes as “atypical” for challenging norms of physical passivity (Zhang et al., 2024), creating unique stressors: they may face scrutiny for “abandoning femininity” through intense training, while also being expected to maintain relational harmony. This duality erodes resilience, as females navigate conflicting expectations, and drives reliance on emotional OSS (e.g., peer validation in online groups) to reconcile these pressures—consistent with cultural emphasis on communal support networks [“关系” (“Guanxi”)] for women.

Notably, China’s recent policy shifts, such as the National Plan for Youth Sports Development (2021–2025) (General Administration of Sport of China, 2021), have begun challenging rigid gender norms by promoting “inclusive athleticism.” However, cultural inertia persists: 62% of sports coaches in a 2023 survey still reported encouraging male athletes to “endure pain silently” and female athletes to “avoid appearing overly aggressive” (China Sports Daily, 2022). These norms directly influence our findings, explaining why males underreport emotional distress and females face compounded resilience depletion.