Competitive sprinting represents a domain of human performance where success is determined by fractions of a second. Athletes in disciplines such as the 100, 200, and 400 meters operate at the peak of their physiological capacity, an environment where psychological factors can be the ultimate arbiter of victory and defeat. Among the most pervasive of these factors is pre-competition anxiety, a complex emotional state characterized by feelings of apprehension and tension associated with an upcoming competition (Cheng et al., 2009). This anxiety manifests in two primary forms: cognitive anxiety, which involves negative thoughts, worries, and doubts about performance, and somatic anxiety, which encompasses the physiological arousal and symptoms like increased heart rate, muscle tension, and butterflies in the stomach (Morris et al., 1981; Steptoe and Kearsley, 1990). Research has consistently demonstrated a negative correlation between high levels of cognitive anxiety and athletic performance, as it can disrupt concentration and decision-making (Craft et al., 2003; Eysenck et al., 2007). Similarly, while a certain level of somatic arousal is necessary for peak performance, excessive physiological symptoms can lead to neuromuscular tightness and inefficient movement patterns, thereby impairing a sprinter's explosive power and coordination (Bandura, 1977).

In parallel with the need to manage anxiety, the cultivation of self-efficacy is paramount for athletic success. Defined by Bandura as an individual's belief in their capabilities to execute the courses of action required to produce given attainments, self-efficacy is a cornerstone of motivation and perseverance in sports (Carson and Collins, 2016). For a sprinter, this translates to a firm belief in their ability to execute a perfect start, maintain form under pressure, and push through the physiological and psychological pain of the final meters. High self-efficacy influences an athlete's choice of activities, effort expenditure, and persistence in the face of adversity (Moritz et al., 2000). It is a dynamic construct, susceptible to influence from past performances, vicarious experiences, verbal persuasion, and physiological states. Consequently, the debilitating effects of pre-competition anxiety can directly erode an athlete’s self-efficacy, creating a vicious cycle of doubt and underperformance. Therefore, interventions that can simultaneously mitigate anxiety and bolster self-efficacy are of immense value in sports psychology.

Traditionally, athletes have utilized psychological skills training (PST), incorporating techniques like goal-setting, imagery, and self-talk to manage their mental state (Birrer and Morgan, 2010). More recently, mindfulness-based interventions (MBIs) have gained significant traction within sports psychology as a powerful approach to enhancing mental well-being and performance (Gardner and Moore, 2017; Noetel et al., 2019). Mindfulness is defined as the practice of paying attention to the present moment, on purpose, and non-judgmentally (Kabat-Zinn, 2003). Rather than attempting to eliminate or control anxious thoughts, mindfulness teaches athletes to observe them as transient mental events without becoming entangled in their content. This decentering perspective can reduce the impact of cognitive anxiety and help regulate somatic arousal by fostering a greater awareness of bodily sensations (Shapiro et al., 2006). Studies have shown that MBIs can lead to significant reductions in anxiety symptoms and improvements in overall psychological well-being among athletes (Bühlmayer et al., 2017; Wang et al., 2023).

The proliferation of smartphone technology has created a paradigm shift in the delivery of mental health interventions. Smartphone applications offer an unprecedented opportunity to provide accessible, convenient, and scalable psychological support to athletes who often face significant time constraints and logistical challenges in accessing traditional face-to-face services (Wu et al., 2021; Linardon et al., 2024). These platforms can deliver guided mindfulness exercises, psychoeducational content, and progress tracking, allowing athletes to integrate mental training seamlessly into their daily routines (Gao et al., 2024). Commercially available apps like Headspace and Calm have been used in athletic contexts, demonstrating the potential of this delivery model (Huberty et al., 2021; Ren et al., 2022). This technological advancement makes it possible to provide consistent, evidence-based support that was previously difficult to implement on a large scale.

Despite the theoretical promise, the empirical evidence for smartphone-delivered mindfulness programs in athletic populations remains nascent and has yielded mixed results. While some meta-analyses and studies report favorable effects on anxiety (Blanck et al., 2018; Wang et al., 2023) other rigorous randomized controlled trials (RCTs) conducted with college athletes have found no significant reduction in anxiety levels compared to control conditions (Röthlin et al., 2020; Gao et al., 2024; Lee et al., 2024). These inconsistencies highlight a critical research gap and suggest that the design, content, and implementation of such digital interventions may be crucial determinants of their efficacy. Furthermore, there is a distinct lack of research focusing specifically on elite sprinters, a population with unique psychological demands. Moreover, few studies have simultaneously investigated the impact of a smartphone MBI on both pre-competition anxiety and sport-specific self-efficacy, leaving the interconnectedness of these constructs within the context of a digital intervention largely unexplored.

Therefore, the purpose of this study was to design and evaluate a bespoke, 8-week smartphone-delivered mindfulness program, named "MindSprint," for pre-competition anxiety and self-efficacy in competitive sprinters. We conducted a randomized controlled trial to test the following hypotheses: (H1) Sprinters in the MindSprint intervention group would demonstrate significantly greater reductions in both cognitive and somatic pre-competition anxiety from baseline to post-intervention compared with a waitlist control group. (H2) The MindSprint intervention group would show a significantly greater increase in sport-specific self-efficacy from baseline to post-intervention compared with the control group. (H3) As a planned mechanistic analysis, we further hypothesized that changes in dispositional mindfulness would mediate the intervention effect on cognitive anxiety, the anxiety component most directly aligned with worry, rumination, and decentering; associations with somatic anxiety and self-efficacy were examined as complementary exploratory relationships.

A total of 70 competitive sprinters (38 male, 32 female) were recruited from Wen Hua College and affiliated athletic clubs. Inclusion criteria required participants to be between 18 and 28 years of age (M = 22.4, SD = 2.8), actively competing in 100 m, 200 m, or 400 m events for at least three years, and owning an iOS or Android smartphone. Participants were excluded if they had an existing formal meditation practice (more than once per week) or were currently receiving psychological treatment for anxiety. The study received ethical approval from the University Institutional Review Board (IRB), and all participants provided digital informed consent prior to their enrollment, in accordance with established ethical guidelines (World Medical Association, 2013; Skelton et al., 2020).

This study employed a two-arm, parallel-group randomized controlled trial (RCT) design. Participants were randomly allocated, using a computer-generated sequence, to either the smartphone-delivered mindfulness intervention (IG; n = 35) or a waitlist control group (CG; n = 35). A researcher not involved in data collection managed the allocation concealment. Assessments were conducted at four time points: baseline (T0), mid-intervention (T1; 4 weeks), immediately post-intervention (T2; 8 weeks), and at a 4-week follow-up (T3; 12 weeks). To capture pre-competition psychological states while reflecting the realities of track-and-field scheduling, electronic assessments were targeted for the 24 hours before the nearest scheduled competition within each nominal assessment window. When a race did not fall exactly at week 4, week 8, or week 12, the nearest scheduled race within the relevant window was used. Competition level was recorded descriptively (club, collegiate/regional, or championship-level), but event importance was not entered as a statistical covariate.

The intervention group received access to "MindSprint," a custom-developed smartphone application designed specifically for this study. The 8-week program consisted of daily audio-guided mindfulness exercises, each lasting approximately 10-15 minutes (Huberty et al., 2019). The content was structured progressively. Weeks 1-2 focused on foundational mindfulness skills, including focused attention on the breath and body scan meditations (Kabat-Zinn, 2003). Weeks 3-4 introduced mindfulness of thoughts and emotions, teaching athletes to observe their mental content non-judgmentally. Weeks 5-6 centered on acceptance and self-compassion in relation to performance setbacks. Weeks 7-8 integrated sport-specific content, including mindful visualization of a race and mindfulness during movement exercises. A supplementary weekly outline describes the session themes, typical audio duration, notification schedule, adherence definition, and app navigation structure.

Core weekly modules were released sequentially to preserve the intended progression, whereas short practice components (e.g., breathing space and body scan exercises) remained available for optional review after first release. The app sent one daily push-notification reminder and displayed a personal dashboard with completed sessions and total practice time (Yardley et al., 2016; Volgemute et al., 2025). Adherence was defined descriptively as the proportion of the 56 recommended daily sessions completed across the intervention period. No separate formal usability trial was conducted before the efficacy study; app feasibility was monitored through backend completion logs, attrition, and participant use patterns.

The waitlist control group received no mindfulness intervention during the 12-week study period and continued with regular training and competition schedules. At baseline, waitlist participants were told that they would receive access to MindSprint after completing the T3 follow-up assessment. They received the same electronic assessment reminders as the intervention group but did not receive app content, mindfulness prompts, or practice feedback during the trial.

Pre-Competition Anxiety: The Competitive State Anxiety Inventory-2 Revised (CSAI-2R) was used to measure pre-competition anxiety (Cox et al., 2003; Andrade Fernández et al., 2007). This validated instrument consists of three 9-item subscales assessing cognitive state anxiety, somatic state anxiety, and state self-confidence. Participants rated items on a 4-point Likert scale, with subscale scores ranging from 9 to 36. Higher scores indicate greater levels of the respective construct. The CSAI-2R has demonstrated strong psychometric properties in athletic populations.

Self-Efficacy: Sport-specific self-efficacy was measured using the Athlete Self-Efficacy Scale (ASES) (Jackson et al., 2011). This 15-item scale assesses an athlete's belief in their ability to perform well under pressure, focus effectively, and overcome challenges in their sport. Items are rated on a 10-point Likert scale. The ASES has shown good validity and high internal consistency (Cronbach's α = .88) in previous research with adult athletes (Stephen et al., 2022; Evans et al., 2024).

Mindfulness: Dispositional mindfulness was assessed using the 39-item Five Facet Mindfulness Questionnaire (FFMQ), which measures five facets of mindfulness: Observing, Describing, Acting with Awareness, Non-judging of inner experience, and Non-reactivity to inner experience. It is a widely used and validated measure of mindfulness skills.

Engagement: Adherence and engagement with the MindSprint app were measured objectively via the application's backend server. Data collected included the number of mindfulness sessions completed, the total duration of practice in minutes, and the frequency of app opens (Perski et al., 2017; Miller et al., 2019).

Analyses were conducted with IBM SPSS Statistics, version 29.0 (IBM Corp., Armonk, NY, USA). Two-sided tests used an alpha level of.05. Baseline group comparisons used independent-samples t tests for continuous variables and chi-square tests for categorical variables. Primary and secondary outcomes were analyzed with 2 × 4 mixed-model repeated-measures ANOVAs, with group (intervention vs. control) as the between-participant factor and time (T0, T1, T2, and T3) as the repeated factor. Mauchly's test was used to assess sphericity; Greenhouse-Geisser corrections were applied when the assumption was violated. Significant interactions were followed by Bonferroni-adjusted simple-effects comparisons.

The intention-to-treat dataset retained all 70 randomized participants. For the eight participants who withdrew after randomization, missing post-randomization questionnaire scores were imputed by the last-observation-carried-forward method before repeated-measures analysis. Effect sizes for ANOVA were reported as partial eta squared (ηp²). Between-group Cohen's d values were calculated from pooled standard deviations at each time point, with positive values denoting lower anxiety in the intervention group and negative values denoting higher self-efficacy or mindfulness in the intervention group. Confidence intervals for indirect effects were estimated with 5, 000 bias-corrected bootstrap samples.

The mediation model was specified a priori for the cognitive-anxiety change score because cognitive worry was the outcome most directly linked to mindfulness-based decentering. The model used the full randomized sample, intervention assignment as the independent variable, change in FFMQ score from T0 to T2 as the mediator, and change in cognitive anxiety from T0 to T2 as the dependent variable. Baseline values were represented through change-score construction and were not entered as separate covariates. Correlations between engagement, mindfulness, somatic anxiety, and self-efficacy were treated as exploratory association analyses.

Of the 98 sprinters assessed for eligibility, 70 met the inclusion criteria and were randomized into the intervention group (IG; n = 35) or the waitlist control group (CG; n = 35). Over the course of the 12-week study, five participants (14.3%) from the IG and three (8.6%) from the CG withdrew, citing time constraints or injury. As described in the Statistical Analysis subsection, the intention-to-treat analysis retained the full randomized sample of 70 participants. The flow of participants through each stage of the trial is detailed in the CONSORT diagram (Figure 1). Baseline demographic and athletic characteristics are presented in Table 1. No statistically significant differences were found between the two groups at baseline on demographic, athletic, or primary outcome variables, indicating successful randomization.

Engagement data, collected automatically from the MindSprint application, are summarized in Table 2. Participants in the intervention group completed an average of 5.8 sessions per week (SD = 1.3) out of the 7 recommended sessions, representing an adherence rate of 82.9%. The average duration of a completed session was 11.2 minutes (SD = 2.1). Over the 8-week intervention, the mean total practice time was 515.2 minutes (SD = 120.4). This level of engagement is considered robust for a mobile health intervention (Giovanazzi et al., 2022).

The mean scores and standard deviations for cognitive and somatic anxiety at each time point are presented in Table 3.

For cognitive anxiety, the 2x4 mixed-model ANOVA revealed a significant group-by-time interaction effect, F(3, 204) = 18.33, p < .001, ηp² = .212, indicating that the changes in cognitive anxiety over time differed significantly between the two groups. Simple effects analysis showed a significant decrease in cognitive anxiety for the intervention group from T0 to T2 (p < .001), which was maintained at T3 (p < .001). No significant changes were observed in the control group across any time points (p > .05). Between-group comparisons confirmed that the intervention group had significantly lower cognitive anxiety than the control group at T1, T2, and T3 (all p < .001).

Similarly, for somatic anxiety, a significant group-by-time interaction was found, F(3, 204) = 22.71, p < .001, ηp² = .250. The intervention group experienced a significant reduction in somatic anxiety from T0 to T2 (p < .001), an effect that was sustained at the T3 follow-up (p < .001). The control group's scores remained stable over time (p > .05). The intervention group reported significantly lower somatic anxiety than the control group at T1, T2, and T3 (all p < .001). These longitudinal changes are visualized in Figure 2, and the distributional shifts from baseline to post-intervention is illustrated in Figure 3.

Mean scores for the Athlete Self-Efficacy Scale (ASES) are presented in Table 4. The mixed-model ANOVA revealed a significant group-by-time interaction effect for self-efficacy, F(3, 204) = 15.96, p < .001, ηp² = .190. Simple effects analysis indicated a significant increase in self-efficacy for the intervention group from T0 to T2 (p < .001), which was sustained at T3 (p < .001). The control group showed no significant change in self-efficacy. Between-group analyses showed the intervention group had significantly higher self-efficacy scores than the control group at T1, T2, and T3 (all p < .001). The group differences at each time point are displayed in Figure 4. An exploratory bivariate correlation in the intervention group showed that total mindfulness practice time was positively associated with the change in self-efficacy from T0 to T2 (r = .58, p = .002; Figure 5). This association was not adjusted for baseline ASES scores and is therefore interpreted as descriptive evidence of an engagement-response pattern rather than as an independent causal estimate.

As shown in Table 5, the intervention produced a significant increase in dispositional mindfulness. The ANOVA for the FFMQ total score yielded a significant group-by-time interaction, F(3, 204) = 25.11, p < .001, ηp² = .269. The intervention group's FFMQ scores increased significantly from T0 to T2 (p < .001) and were maintained at T3.

Analyses of the FFMQ subscales revealed significant interactions for Acting with Awareness, Non-judging, and Non-reactivity (all p < .001), but not for Observing or Describing. Figure 6 provides a visual characterization of the change profile across the five mindfulness facets for the intervention group.

The correlation matrix (Table 6) summarizes change-score associations. For outcomes measured in the full sample, changes in mindfulness (ΔFFMQ) were strongly and negatively correlated with changes in cognitive anxiety (r = -.65, p < .001) and somatic anxiety (r = -.61, p < .001), and positively correlated with changes in self-efficacy (r = .59, p < .001). Engagement correlations used only intervention-group participants because app-use data were available only for those assigned to MindSprint (n = 35). Table 7 summarizes between-group effect sizes at each time point. The planned mediation analysis (Table 8) supported the hypothesized cognitive-anxiety pathway, indicating that the increase in mindfulness skills mediated the relationship between intervention assignment and reduction in cognitive anxiety (indirect effect = -3.87, 95% CI [-5.62, -2.19]). Somatic anxiety and self-efficacy were not modeled as mediation outcomes in this trial because the planned mechanistic test focused on cognitive worry, the construct most directly linked to mindfulness-based decentering.

To provide a more granular characterization of the intervention's impact, Figure 7 presents weekly psychological profile heatmaps for four intervention-group participants selected to illustrate distinct adherence-response patterns (rapid improvement with consistent use, gradual improvement, fluctuating response, and lower/variable engagement). These maps visualize the week-by-week interplay between anxiety levels, self-efficacy, and app engagement, highlighting the heterogeneity of individual responses rather than identifying statistically representative cases. Figure 8 offers a component analysis of the intervention material itself, illustrating the distribution of completed MindSprint practice components among intervention-group users. The "Daily Mindful Moment" and the "Pre-Sleep Body Scan" accounted for the largest shares of completed sessions, suggesting athletes favored shorter, easily integrated practices.

The primary aim of this study was to evaluate a custom-designed, smartphone-delivered mindfulness program, "MindSprint," for reducing self-reported pre-competition anxiety and enhancing self-reported sport-specific self-efficacy in competitive sprinters. The findings of this randomized controlled trial support the study hypotheses. Participants in the 8-week MindSprint program demonstrated significant reductions in both cognitive and somatic anxiety, alongside increases in sport-specific self-efficacy, when compared with a waitlist control group. These changes were maintained at the 4-week follow-up, suggesting that the intervention may help athletes develop durable psychological skills.

The marked reduction in pre-competition anxiety aligns with a substantial body of literature demonstrating the anxiolytic effects of mindfulness-based interventions in athletic populations (Mehrsafar et al., 2019). The large effect sizes observed for both cognitive anxiety (d = 2.24) and somatic anxiety (d = 2.22) at post-intervention are particularly noteworthy and compare favorably to, or exceed, the moderate-to-large effects reported in meta-analyses of traditional face-to-face mindfulness programs (Bühlmayer et al., 2017). Our positive results stand in contrast to some previous studies of smartphone-delivered mindfulness for athletes that found null effects (Gao et al., 2024). Several factors may explain this discrepancy. First, the "MindSprint" application was not a generic, one-size-fits-all program; its content, language, and examples were specifically tailored to the experiences of competitive athletes, which may have enhanced engagement and relevance. Second, the 8-week duration provided sufficient "dosage" for skills to be learned and consolidated, which may not be the case in shorter interventions. Third, the high adherence rate (82.9% of recommended sessions) suggests the program was successfully integrated into the athletes' routines, a common barrier in digital health studies (Kelders et al., 2012).

A key finding of this study is the improvement in athlete self-efficacy. This supports the theoretical proposition that mindfulness can bolster self-efficacy by altering an athlete's relationship with internal experiences. By learning to observe self-critical thoughts and feelings of doubt non-judgmentally, athletes may be less likely to identify with them, thereby preserving their belief in their capabilities (Gardner and Moore, 2004). The mindfulness practice of returning attention to the body and breath may also reduce distracting cognitive anxiety, allowing athletes to focus on the task at hand and draw confidence from physical readiness. The positive association between total practice time and gains in self-efficacy (Figure 5) is consistent with an engagement-response pattern, although this unadjusted association cannot establish that engagement independently caused the observed change. This finding extends the literature by suggesting that a mobile-delivered MBI may support psychological strengths such as self-efficacy as well as anxiety management.

Our investigation into the mechanisms of change provides further insight into how the intervention may have achieved its effects. As hypothesized, the MindSprint program was accompanied by increases in dispositional mindfulness, particularly in the facets of Acting with Awareness, Non-judging, and Non-reactivity (Figure 6). The planned mediation analysis (Table 8) provided statistical support for the cognitive-anxiety pathway, showing that increases in mindfulness skills accounted for part of the relationship between intervention assignment and reductions in cognitive anxiety. This pattern is consistent with the view that athletes learned to notice pre-race thoughts, respond less judgmentally, and reduce escalating worry.

The use of a smartphone-based delivery model is a central feature of this study. The high engagement rates and low attrition (14.3% in the IG) relative to typical digital interventions, which can see dropout rates exceeding 50% (Eysenbach, 2005; Geraghty et al., 2013) underscore the feasibility of this approach for dedicated athletic populations. The convenience of the app allowed sprinters to practice mindfulness at times that suited their demanding training schedules, such as during travel to competitions or before sleep. The objective engagement data collected from the app (Table 2, Figure 8) is a significant methodological strength, providing a more accurate measure of adherence than self-report diaries (Lee et al., 2017). The component analysis (Figure 8) offers practical insights for future app development, indicating a preference for shorter, highly practical exercises that can be easily integrated into a daily routine. This suggests that micro-interventions may be a particularly effective strategy for athletes.

Despite the strengths of this research, including its randomized design, use of validated measures, longitudinal follow-up, and focus on an understudied population, several limitations must be acknowledged. First, the sample size was based on the available eligible sprinter cohort rather than a documented a priori power calculation, and prospective trial registration was not documented in the study materials. Second, the use of a waitlist control group means we cannot rule out placebo effects, expectancy effects, or the non-specific benefits of receiving attention as part of a study; moreover, waitlist participants knew they would receive app access after follow-up, which may have influenced expectations or motivation. Third, pre-competition assessments were anchored to scheduled races, but competition timing and event importance could not be fully standardized across athletes, and event level was not modeled as a covariate. Fourth, outcome measures were based exclusively on self-report questionnaires. Future studies would be strengthened by including objective performance data (e.g., reaction times or race times), physiological markers of anxiety such as heart rate variability or salivary cortisol, and active-control conditions. Finally, engagement-response analyses were exploratory and did not adjust for baseline self-efficacy; these associations should be interpreted as descriptive rather than causal.

The practical implications of these findings are relevant for athletes, coaches, and sport psychologists. This study suggests that a sport-specific smartphone application can be a feasible and accessible tool for supporting psychological preparation in competitive athletes. Such apps may serve as a standalone resource or as a supplement to traditional sport psychology support. Coaches can encourage athletes to use such tools as part of a regular mental-skills routine, and sport psychologists may use mobile applications to support remote monitoring and practice continuity.

In conclusion, this randomized controlled trial showed that an 8-week, smartphone-delivered mindfulness program, MindSprint, was associated with reductions in self-reported pre-competition cognitive and somatic anxiety and increases in sport-specific self-efficacy among competitive sprinters. The outcomes were sustained at a 4-week follow-up, and the cognitive-anxiety pathway was partly mediated by changes in mindfulness skills. These findings support continued investigation of tailored digital mindfulness interventions in sport, with future trials incorporating active controls, objective performance and physiological outcomes, prospective registration, and stronger baseline-adjusted analyses.