Introduction

The National Youth Orienteering Preparation Camp in China is a multi-week, high-intensity residential program designed to optimise both the physical and cognitive performance of adolescent athletes1. Orienteering combines sustained aerobic exertion with spatial navigation and rapid decision-making, exposing participants to overlapping domains of fatigue. Within this setting, psychological stress denotes perceived strain under competitive and environmental pressures2; physical fatigue reflects cumulative muscular and cardiovascular load3; and emotional states capture transient affective shifts such as anxiety, frustration, or positivity4. These processes interact continuously, influencing recovery, focus, and overall adaptation during intensive training.

This study focuses on three synchronised systems: autonomic regulation, indexed by heart-rate variability (HRV); endocrine adaptation, measured through salivary cortisol; and affective adjustment, assessed via self-report instruments. Their coordinated change provides a composite view of psychophysiological resilience in youth orienteers5. Such monitoring is valuable for performance management, offering quantifiable cues for adjusting workload, structuring recovery, and supporting emotional stability among adolescents.

Empirical evidence shows that autonomic suppression, hormonal fluctuation, and emotional dysregulation often occur together under cumulative stress. Cortisol levels typically rise with sustained cognitive and physical demand6, while HRV decreases during navigational errors and performance anxiety7. Concurrently, emotion-related attentional changes can impair spatial accuracy or broaden perceptual scope depending on valence8. These findings indicate that stress and fatigue in orienteering extend beyond muscular strain to include coordinated physiological–emotional responses.

High activity levels have been associated with stronger emotional intelligence and psychological resilience9, yet little is known about how these traits evolve under prolonged dual-task load. Existing research isolates stress, emotion, or physiology rather than tracing their joint development across a full training cycle. The present investigation addresses this gap by integrating HRV, cortisol, and affective indices within a longitudinal camp design to clarify how adaptive processes unfold in youth endurance-navigation contexts. It advances a dynamic view of resilience as the synchronisation of physiological regulation and emotional adjustment.

Objectives

This study aims to: (i) quantify temporal changes in psychological stress, physical fatigue, and emotional states during a national youth orienteering training camp in China; (ii) compare sex-specific trajectories in emotional and physiological responses; (iii) determine overtraining thresholds using a composite index of HRV suppression and cortisol elevation; and (iv) construct an integrated monitoring framework combining PANAS-C, HRV, and cortisol measures.

Hypotheses

H1: HRV and cortisol will change inversely across camp phases, reflecting coordinated autonomic–endocrine adjustment.

H2: Reductions in perceived stress will correspond with increases in positive affect.

H3: Athletes with higher baseline emotional intelligence will demonstrate greater physiological recovery and emotional stability.

Literature review

Stress and fatigue during multi-stage orienteering camps

The cumulative stress of orienteering training camps involves a complex interplay between mental fatigue, physical exhaustion, and incomplete recovery cycles that unfold over time. Lam et al.10 conducted a four-day field study with 11 national junior orienteers, recording moderate increases in mental fatigue (effect size = 1.06), physical fatigue (1.07), and mood disturbance post-training. Notably, mental fatigue remained elevated 48 h later, suggesting incomplete psychophysiological recovery. In a complementary race-based study, Lam et al.11 found that different orienteering formats produced distinct fatigue-recovery trajectories, with MF and PF increasing post-competition and varying by race type. However, both studies failed to incorporate physiological markers, such as HRV and hormonal indices, thereby limiting mechanistic understanding. The absence of real-time biomarker validation across these studies restricts their predictive utility in fatigue modelling.

Gorgulu et al.3 advanced this line by demonstrating parasympathetic withdrawal via HRV reduction (~ 23%) following checkpoint errors in elite orienteers, underscoring acute stress responses linked to cognitive failure. While the cardiac dimension is well documented, subjective stress ratings were not collected, weakening the connection to psychological domains. Together, these studies indicate that orienteering induces intertwined fatigue components that do not recover synchronously. Yet, a critical limitation is the lack of integrative models tracking psychological states and physiological outputs in parallel. The overall gap lies in the lack of longitudinal, multimodal protocols that synchronously measure perceived stress, mood, and physiological fatigue throughout camp duration to inform tailored recovery timelines.

Emotional regulation under prolonged training stress

Adolescent male and female orienteers exhibit distinct emotional regulation strategies and cognitive-affective adaptations when exposed to prolonged training stress, but camp-specific profiles remain underexplored. Chaplin and Aldao12 synthesised 555 effect sizes from over 21,000 participants, revealing that adolescent girls exhibit more internalising emotional responses (g = − 0.10) while boys exhibit more externalising responses (g = + 0.09), particularly under peer-related stress. Zhang et al.13 further demonstrated that girls possess greater expressive flexibility and emotional suppression capability, traits inversely correlated with depressive symptoms (r = − 0.34). These capacities likely buffer girls during prolonged training, but the non-sport context limits applicability. Abdulsamed et al.14 directly studied orienteering athletes and found that emotional regulation predicted tactical accuracy and decision efficiency, with males favouring cognitive reappraisal and females showing emotional suppression patterns. However, the study lacked longitudinal depth and age stratification, limiting insights into training-phase-specific changes.

Bao et al.15, using fNIRS neuroimaging, observed greater post-training neural efficiency in male orienteers, with higher gains in spatial memory (13.6% vs. 8.9% in females). These neural adaptations may underpin emotional resilience, but no emotional regulation metrics were recorded. Molina et al.16 linked cognitive reappraisal with faster recovery and better sleep quality, especially in females, while Türkmen and Biçer17 observed sex-specific gains in physical parameters post-orienteering. Yet, neither study examined psychological stress trajectories across a camp setting. Ferrari et al.18 found enhanced emotional restoration among girls in a school-based orienteering intervention, emphasising its psychosocial potential, though not under competitive stress. Across these works, a consistent shortfall is the absence of camp-specific, sex-stratified monitoring of emotional regulation evolution under prolonged stressors.

Cortisol–HRV biomarker models for early fatigue

While heart rate variability and cortisol are individually linked to fatigue, their joint predictive power in youth orienteering has not been empirically validated in real-time training contexts. Sundas et al.19 reviewed the use of AI-enhanced HRV analytics and RR interval modelling, concluding that HRV metrics can identify early signs of cardiac fatigue with up to 90% accuracy. However, both reviews focused on clinical or general populations, lacking sport-specific translation. Yang et al.20 conducted an RCT with treadmill-induced fatigue, finding that lnRMSSD declined ~ 40% at exhaustion, while sympathetic dominance increased. Pre-fatigue HRV also predicted total fatigue duration. Despite controlled protocols, the treadmill environment limits ecological validity for orienteering. Gronwald et al.21 reinforced that HRV-based daily tracking correlates with training adaptation, though data remains observational. These studies confirm HRV’s predictive utility, yet fail to integrate emotional or hormonal dimensions relevant to youth cognitive-endurance sports.

Coordinative tasks22 yielded cortisol kinetics different from those in endurance-only conditions, suggesting domain specificity in hormonal response. Bermejo et al.23 found cognitive performance improved in working memory despite elevated cortisol, showing differentiated impacts across brain domains. However, neither study included orienteering athletes or stress variability across a multi-day period. Others24 noted that HRV suppression precedes fatigue symptoms by ~ 3 days and tracks cognitive strain. Yet, no existing study has combined HRV and cortisol to detect early overtraining in navigation-intensive youth sports. The gap remains the absence of dual-biomarker models validated within orienteering-specific conditions.

Affective–cognitive depletion trajectories

The dual-task demands of orienteering, which combine spatial navigation and physical endurance, create unique trajectories of cognitive depletion and emotional strain, which are poorly mapped in camp settings. Waddington et al.25 demonstrated that high navigation plus physical stress in orienteering produced elevated lactate (6.1 mmol/L) and increased brain-derived neurotrophic factor (BDNF) by 38%, with corresponding memory gains (+ 18%). These synergistic effects of physical and cognitive load suggest enhanced neuroplastic adaptation, but the single-bout design did not explore cumulative fatigue. Waddington and Heisz26 reported that habitual orienteers showed superior spatial memory (+ 22%) and lower cognitive ageing, supporting long-term benefits of navigational engagement, though outside a camp context. Hohl et al.27 observed transient executive fatigue in adult orienteers post combined load, suggesting mid-session depletion even with cognitive gains. However, emotional states were not measured. These studies establish orienteering’s cognitive complexity but fail to track depletion across extended protocols.

Studies28,29 found that proactive navigation strategies reduce performance errors by 17% under fatigue, but data came from elite adults and lacked HRV or hormonal integration. Mishica et al.30 reported inverse HRV–cortisol correlations (~ r = − 0.55) in adolescent athletes, suggesting psychophysiological feedback loops in end-season fatigue states. Yet their study was not in orienteering and had a small sample size. Lundstrom et al.31 confirmed HRV drops as early fatigue indicators, but did not address navigation tasks. Together, these findings suggest that orienteering camps may produce unique affective–cognitive depletion arcs from cumulative dual-tasking, but this trajectory remains under-mapped. No study to date has integrated hormonal, emotional, and HRV data to model this combined stress arc in youth camp contexts.

Conceptual model

The Integrated Psychophysiological Fatigue–Emotion Synchronisation Model (IPFESM) organises the observed relationships among HRV, cortisol, and emotional states within a longitudinal design. It serves as an analytic framework for examining synchronised changes across these domains throughout the training cycle, providing a structured basis for identifying points of physiological and emotional strain (see Fig. 1). The model integrates real-time HRV metrics (RMSSD, LF/HF), salivary cortisol, and self-reported emotional state scales (PANAS-C) collected across three time points (T1–T3). By triangulating physiological strain, hormonal responses, and affective shifts, the IPFESM identifies high-risk fatigue patterns and points of emotional instability. It supports individualised training adaptation based on synchronised biomarker thresholds and emotion regulation markers, offering a practical framework for early intervention in endurance navigation sports in camp-based settings.

Fig. 1
figure 1

Tracks HRV, cortisol, and affect across phases to identify synchronised fatigue-emotion intervention points.

Full size image

Methods

Study design

This prospective, longitudinal, parallel-group study spanned 28 days and compared national camp participants with matched club-trained youth. Psychophysiological and emotional measures were collected at T1 (Day 1), T2 (Day 14), and T3 (Day 28) using identical instruments and procedures across groups. The design allowed for temporal tracking of intervention-specific effects within the IPFESM framework. References to “cumulative effects” therefore describe temporal changes across these phases rather than continuous physiological tracking. Data were collected at a centralised national youth camp in Jiangsu, under standardised indoor conditions, with all assessments and biomarker sampling conducted between 07:30–08:00 in a fasted, rested state. Participants refrained from structured exercise, caffeine, and food intake for at least twelve hours before each assessment. Compliance was verified through nightly logs and coach countersignatures to eliminate acute activity effects on physiological recordings.

The study employed a non-randomised allocation based on existing training enrollment structures. Camp participation was determined by national selection criteria, while matched controls were drawn from comparable club athletes who met identical eligibility standards. This approach preserved ecological validity and ensured continuity with real selection processes in Chinese youth orienteering.

Ethics

The study was approved by the Ethics Review Committee of XX University (Approval No. AD2024-197). Written informed consent was obtained from all participants and legal guardians before enrollment. The study adhered fully to the ethical principles outlined in the Declaration of Helsinki (2013 revision), including voluntariness, anonymity, and the right to withdraw at any point without consequence.

Participants

Recruitment

Participants were recruited from the Chinese Orienteering Association through formal invitations to provincial teams. Coaches distributed study information in Mandarin or English. Screening included PAR-Q + interviews and an ACSM-based checklist32. Athletes meeting inclusion criteria and consenting in writing were enrolled by trained assistants, with procedures emphasising objectives, protocol transparency, and data confidentiality throughout recruitment. Power analysis estimated 64 participants for 85% power (α = 0.05); 72 were enrolled to accommodate subgroup comparisons and a 15% dropout rate, based on prior biomarker variability estimates.

Eligibility

Inclusion criteria were: (i) age 15–18 years confirmed via national ID or student registration card; (ii) active participation in officially recognised orienteering competitions for at least two years, verified through competition records and coach confirmation; (iii) written medical clearance issued by a certified sports physician indicating full readiness for high-intensity training. Exclusion criteria were: (i) current or recent use (within 30 days) of medications known to influence cortisol or cardiac autonomic regulation, including corticosteroids, beta blockers, SSRIs, SNRIs, and methylphenidate, verified via medication disclosure checklist adapted from the World Anti-Doping Agency (WADA) Prohibited List; (ii) diagnosis of any cardiovascular, endocrine (e.g., adrenal or thyroid disorders), or psychiatric condition (e.g., depression, anxiety disorder), confirmed via clinician review of prior medical documentation and responses to the General Health Questionnaire-28 (GHQ-28); (iii) presence of infection, injury, or inflammatory symptoms within 14 days before baseline, assessed via oral temperature (> 37.5 °C), limb mobility check, and symptom log; (iv) failure to adhere to pre-sampling behavior protocols including 8-hour overnight fasting, abstention from caffeine and exercise for 12 h prior, verified via participant log and coach countersignature.

Stratification and bias

Participants were stratified into intervention and control groups based on eligibility and consent, with post-enrollment matching on age (± 1 year), experience (± 1 year), and weekly training (± 1.5 h). The intervention group completed a 28-day national camp, while the control group maintained club-based training. Assessments at T1, T2, and T3 followed identical protocols. Misclassification and performance bias were minimised through coach-verified exposure logs, assessor blinding, and uniform testing conditions.

Intervention

Control

The control group continued their regular club-level training over the 28-day observation period. Each participant’s training log was reviewed and approved by their regional team coach and submitted weekly. The control protocol required athletes to maintain their average weekly training load within ± 15% of their 4-week pre-camp baseline, calculated from verified entries in the Orienteering China Athlete Training Registry.

Standard control group training consisted of four sessions per week: two running-based endurance workouts (45 to 60 min, moderate intensity; HR target: 60–70% HRmax), one terrain familiarisation session without timing or simulation drills (45 min, flat course), and one technical map review session (non-physical). Training heart rate was monitored using Polar OH1 optical sensors (Polar Electro, Finland), and session logs were submitted via timestamped Excel sheets, which were reviewed by a non-camp coach assigned to the study. No structured emotional regulation or fatigue-monitoring tools were used during the control period, but athletes adhered to the same assessment schedule at T1, T2, and T3.

Intervention

The intervention group underwent a structured 4-week national-level orienteering training camp designed to progressively increase cognitive and physical training loads. Individual maximal heart rate (HRmax) was derived using the validated age-adjusted formula (HRmax = 208 − 0.7 × age) and cross-checked against prior 3-minute field-test data provided by team coaches. Target heart-rate zones were expressed as percentages of each athlete’s verified HRmax.

The protocol followed a four-phase weekly progression model: Week 1 (Technical Load Initiation): Daily morning technical drills (6:30–8:00 AM), focused on compass bearing, terrain pacing, and punch control techniques. Drills were conducted in simulated environments using standardised 1:5,000-scale maps printed on waterproof Teslin paper. Each session included a pre-mapped 2.5 km circuit with 12–16 control points. Week 2 (Navigation Load Elevation): Terrain-based navigation under moderate physical exertion. Athletes completed five navigation sessions (60–75 min) on varied slope courses. GPS tracking was used for route validation (Garmin Forerunner 745, USA), with real-time splits verified post-run via QuickRoute software (Version 2.4). Heart rate targets were set at 70–80% HRmax. Errors in control sequencing were logged for each athlete and reviewed in nightly debrief sessions. Week 3 (Combined Load Simulation): Full-course simulations under time pressure, three sessions per week (90–105 min each), with terrain elevation ≥ 300 m and a minimum of 25 control points per course. Athletes wore chest-strap HR monitors (Polar H10, Polar Electro, Finland) and timing chips (Emit EKT, Norway). Decision-making stress was induced through pre-session route-option memorisation (60 s per map segment) and a no-pause rule at control points. Two sessions were conducted in heat-exposure settings (ambient temperature 28–30 °C) to elevate systemic stress. Week 4 (Recovery and Reinforcement): Tapered sessions including light map jogging (40 min) and technique reinforcement without timing. Psychological decompression routines included guided breathing (10 min post-session) and one-on-one feedback with assigned coaches trained in performance psychology (certified by the Chinese Athletics Psychological Services Board).

Hydration and nutritional timing were standardised across participants. Water intake was prescribed at 500 mL per hour during sessions. No caffeine or supplement use was permitted. Resting periods were enforced between 13:00 and 15:00 daily, and sleep duration was logged nightly using wrist-based actigraphy (Xiaomi Mi Band 6, China).

Monitoring and fidelity control

Protocol adherence in the intervention group was monitored through biometric logging, video recordings, and GPS validation. Coaches used HD camcorders (Sony FDR-AX53, Japan) to record sessions, while GPS files were cross-checked with course blueprints. Heart rate was continuously tracked, with < 90% compliance flagged. Daily compliance logs required dual signatures and timestamped photos. Staff completed a 2-day IPFESM training workshop. Control group fidelity was tracked using weekly digital diaries signed by coaches and reviewed by a blinded research coordinator.

Instruments

Primary outcomes

  1. (i)

    Heart Rate Variability (HRV): HRV was assessed as a physiological marker of autonomic nervous system regulation. Time-domain (RMSSD, SDNN) and frequency-domain (LF/HF ratio) indices were extracted using the Polar H10 chest strap sensor (Polar Electro, Finland) paired with the Kubios HRV Premium software, version 3.5 (Kubios Oy, Finland). Each measurement consisted of a 5-minute supine rest period under controlled ambient conditions. Data were recorded at 1,000 Hz and artefact-corrected using a medium filter setting.

  2. (ii)

    Salivary Cortisol: Cortisol was measured as an index of endocrine response to training stress. Samples were collected using Salivette Cortisol swabs (Sarstedt, Germany) under fasting, no-exercise, no-brushing conditions between 07:30 and 08:00 AM. Samples were frozen at − 20 °C and batch-analysed via enzyme-linked immunosorbent assay (ELISA) using the Salimetrics High Sensitivity Salivary Cortisol Kit (USA). The assay has a reported sensitivity of 0.007 µg/dL and intra-assay CV < 4.5%.

  3. (iii)

    Perceived Stress Scale (PSS-10): The 10-item Perceived Stress Scale was developed by Cohen et al.33 was used to assess perceived psychological stress over the past week. Each item is rated on a 5-point Likert scale (0 = never to 4 = very often). Total scores range from 0 to 40, with higher scores indicating greater stress. Reverse scoring is applied to items 4, 5, 7, and 8 before summing the total score. The Chinese version used in this study was validated, reporting a Cronbach’s alpha of 0.86 and test–retest reliability of r = 0.68 over 14 days in adolescent populations.

Secondary outcomes

  1. (i)

    Positive and Negative Affect Schedule – Child Form (PANAS-C): The PANAS-C is a 27-item self-report measure assessing transient emotional states. It comprises two subscales: Positive Affect (PA; 12 items) and Negative Affect (NA; 15 items), each rated on a 5-point scale (1 = very slightly to 5 = extremely). Scores are summed separately for PA and NA. This instrument was developed by Laurent et al.34, and validated in youth aged 9–18 years. The Chinese adaptation used in this study was validated with Cronbach’s alphas of 0.89 (PA) and 0.86 (NA), and a confirmatory factor analysis supported a two-factor structure (CFI = 0.93, RMSEA = 0.049).

  2. (ii)

    Subjective Fatigue Rating Scale – Borg RPE (6–20): Subjective physical fatigue during training was measured using the Borg Rating of Perceived Exertion Scale (6–20). Athletes rated their perceived exertion immediately post-session on a single-item Likert-type scale, where 6 indicates “no exertion” and 20 indicates “maximal exertion.” The scale was initially developed by35, and is widely used in exercise science. Validity in youth athletes was confirmed by correlation with %HRmax (r = 0.74–0.86) and by inter-session reliability coefficients exceeding 0.80.

  3. (iii)

    Sleep Duration and Quality Index (adapted from PSQI): Sleep was tracked daily using a 4-item abbreviated version of the Pittsburgh Sleep Quality Index (PSQI) covering sleep onset latency, total duration, perceived quality, and number of nighttime awakenings36. Athletes completed entries each morning via a structured log. This abbreviated version was validated in adolescent sport populations with adequate internal consistency (α = 0.79) and moderate convergent validity with actigraphy (r = 0.61).

Emotional profiling

Emotional profiling distinguished between stable traits and short-term emotional reactivity. Trait Emotional Profiling was conducted at baseline (T1) using the 24-item EISA-24, covering emotional awareness, self-regulation, utilisation, and empathy, scored 1–5, with total scores from 24 to 120. The scale showed α = 0.88 and r = 0.77. State Emotional Profiling was measured at T1–T3 using PANAS-C. Fatigue–Emotion Profiles were classified into Resilient, Reactive, or Dysregulated based on ± 1 SD changes in affect and HRV/cortisol, with blinded dual-analyst consensus ensuring classification reliability. Profiles were used for subgroup recovery analysis.

Statistical analysis

All analyses were performed using R statistical software (R Foundation for Statistical Computing, Vienna, Austria). Linear mixed-effects models were specified with random intercepts for participants and fixed effects for time, group, and their interaction. Random slopes were not included because each participant had three repeated measures. Missing observations (< 5%) were handled by restricted maximum likelihood estimation. Model residuals were checked for normality and homoscedasticity. Post-hoc pairwise contrasts were adjusted using the Holm–Bonferroni correction to control the family-wise error rate across multiple comparisons. Subgroup and moderation analyses examined interactions by sex, trait emotional intelligence, and baseline fitness levels using stratified models and interaction terms. Sensitivity analyses tested robustness by excluding participants with incomplete timepoint data and by applying nonparametric models to confirm consistency of findings. All statistical tests were two-tailed, with significance set at p < 0.05. Temporal effects were interpreted as phase-specific adaptations rather than direct causal outcomes, consistent with the longitudinal observational design.

Results

Preliminary assumption checks confirmed normality and homogeneity of residuals across all dependent measures. Linear mixed-effects analyses were conducted for time, group, and interaction terms, with all post-hoc comparisons Holm-Bonferroni-adjusted to maintain family-wise error control.

Baseline features

Both groups demonstrated comparable baseline characteristics (see Fig. 2A). The control group reported a mean age of 17.8 ± 1.2 years, while the intervention group averaged 18.0 ± 1.1 years, with no significant difference (t(61) = − 0.683, p = 0.497, Cohen’s d = 0.17). Sex distribution was similar, with 59.4% male in the control group and 58.1% male in the intervention group (χ²(1) = 0.014, p = 0.905). Orienteering experience was 3.1 ± 1.6 years versus 3.4 ± 1.4 years, and weekly training loads were 6.8 ± 1.9 and 7.1 ± 2.1 h. Baseline physiological indices showed no significant between-group differences (see Fig. 2B). RMSSD was 42.3 ± 12.5 ms in the control group and 44.7 ± 13.1 ms in the intervention group (t(61) = − 0.752, p = 0.455, Cohen’s d = 0.19). SDNN values were 61.2 ± 14.8 ms and 64.1 ± 13.9 ms respectively (t(61) = − 0.869, p = 0.388, d = 0.22). LF/HF ratios were 1.48 ± 0.51 versus 1.52 ± 0.47 (t = − 0.314, p = 0.754), and morning salivary cortisol averaged 0.43 ± 0.16 and 0.46 ± 0.17 µg/dL.

Fig. 2
figure 2

(A) Baseline participant characteristics by study group, (B) Baseline heart rate variability and cortisol by study group, (C) Baseline perceived stress and trait emotional intelligence, and (D) Baseline sleep metrics and perceived fatigue.

Full size image

Baseline psychological indicators revealed minimal group differences (see Fig. 2C). PSS-10 scores averaged 19.1 ± 4.3 in the control group and 18.6 ± 4.5 in the intervention group (t(61) = 0.454, p = 0.652, d = 0.12). EISA-24 total scores were 92.4 ± 11.8 and 94.7 ± 12.1 respectively (t = − 0.756, p = 0.453, d = 0.19). Subscale means ranged from 21.7 ± 3.7 to 24.1 ± 3.6 in controls and 22.1 ± 3.5 to 24.5 ± 3.8 in the intervention group, with all p-values > 0.36 and d ≤ 0.23. Baseline sleep and fatigue indicators showed comparable distributions between groups (see Fig. 2D). Mean sleep duration was 7.2 ± 0.8 h for the control group and 7.3 ± 0.9 h for the intervention group (t(61) = − 0.446, p = 0.657, d = 0.11). Sleep latency averaged 17.3 ± 5.9 min and 18.1 ± 6.4 min, respectively (p = 0.586, d = 0.14). Other metrics, including sleep quality (3.6 ± 0.7 vs. 3.7 ± 0.8), awakenings (1.2 ± 0.5 vs. 1.3 ± 0.6), and Borg RPE (12.9 ± 2.1 vs. 13.2 ± 2.3), showed no significant differences (all p > 0.53).

Heart rate variability improved significantly over time in the intervention group compared with controls (see Fig. 3A). A linear mixed-effects analysis revealed a significant main effect of time, F(2, 122) = 12.37, p < 0.001, η² = 0.17, and group, F(1, 61) = 8.42, p = 0.005, η² = 0.12, together with a Group × Time interaction, F(2, 122) = 9.16, p = 0.004, η² = 0.13. Holm–Bonferroni-adjusted contrasts showed that RMSSD increased in the camp group (44.7 ± 13.1 → 54.6 ± 13.5 ms) but remained stable in controls (42.3 ± 12.5 → 40.7 ± 11.9 ms; p < 0.001, d = 1.12). At T1, RMSSD was 44.7 ± 13.1 ms in the intervention group and 42.3 ± 12.5 ms in controls (p = 0.455, d = 0.19). At T2, RMSSD increased to 50.2 ± 12.4 ms in the intervention group, significantly higher than 39.5 ± 11.7 ms in controls (t(61) = − 3.63, p = 0.001, d = 0.93). By T3, intervention RMSSD reached 54.6 ± 13.5 ms vs. 40.7 ± 11.9 ms in controls (p < 0.001, d = 1.12).

Fig. 3
figure 3

(A) Heart rate variability across timepoints by study group, (B) Salivary cortisol levels across timepoints by study group, and (C) Perceived stress scores over time.

Full size image

Salivary cortisol levels differed significantly between groups over time (see Fig. 3B). Cortisol showed significant main effects of time, F(2, 122) = 10.03, p = 0.002, η² = 0.14, and a Group × Time interaction, F(2, 122) = 6.85, p = 0.011, η² = 0.10. Adjusted contrasts indicated progressive declines in the camp group (0.46 ± 0.17 → 0.36 ± 0.13 µg/dL) compared with a slight increase in controls (0.43 ± 0.16 → 0.47 ± 0.17 µg/dL; p = 0.004, d = 0.76). At T1, levels were 0.43 ± 0.16 µg/dL in the control group and 0.46 ± 0.17 µg/dL in the intervention group (p = 0.465, d = 0.18). At T2, intervention group cortisol decreased to 0.39 ± 0.15 µg/dL, significantly lower than 0.51 ± 0.18 µg/dL in controls (t(61) = 2.79, p = 0.007, d = 0.71). At T3, intervention group levels further declined to 0.36 ± 0.13 µg/dL vs. 0.47 ± 0.17 µg/dL in controls (p = 0.004, d = 0.76). Perceived stress scores showed clear divergence across groups over time (see Fig. 3C). Perceived stress demonstrated significant effects of time, F(2, 122) = 16.48, p < 0.001, η² = 0.21, group, F(1, 61) = 14.22, p < 0.001, η² = 0.19, and their interaction, F(2, 122) = 11.71, p = 0.001, η² = 0.16. Follow-up comparisons showed that PSS-10 scores decreased in the camp group (18.6 ± 4.5 → 14.7 ± 4.0) but remained unchanged in controls (19.1 ± 4.3 → 19.3 ± 4.6; p < 0.001, d = 1.13). At baseline (T1), PSS-10 scores were 19.1 ± 4.3 in the control group and 18.6 ± 4.5 in the intervention group (t(61) = 0.45, p = 0.652, d = 0.12). At T2, intervention group scores dropped to 15.3 ± 4.2 compared to 19.7 ± 4.5 in controls (t(61) = 4.11, p < 0.001, d = 1.04). At T3, further reduction was observed in the intervention group (14.7 ± 4.0) versus 19.3 ± 4.6 in controls (t(61) = 4.46, p < 0.001, d = 1.13).

Positive and negative affect scores were assessed at three timepoints across groups (see Fig. 4A). A mixed-effects model revealed significant Group × Time interactions for both affect dimensions: positive affect F(2, 122) = 9.54, p = 0.002, η² = 0.14; negative affect F(2, 122) = 7.86, p = 0.005, η² = 0.11. Post-hoc contrasts showed increases in positive affect within the camp group (25.2 ± 5.4 → 30.1 ± 5.5) and reductions in negative affect (− 3.1 ± 2.4 points) relative to controls (p < 0.01 for both). At T1, positive affect was 24.8 ± 5.1 in controls and 25.2 ± 5.4 in the intervention group (t(61) = 0.382, p = 0.704, d = 0.10). At T2, intervention scores increased to 29.4 ± 5.3 versus 23.9 ± 5.0 in controls (t(61) = 4.574, p = 0.016, d = 1.21). At T3, positive affect further rose to 30.1 ± 5.5 in the intervention group, exceeding the control group’s 24.5 ± 5.2 (t(61) = 5.213, p = 0.003, d = 1.36).

Fig. 4
figure 4

(A) Positive and negative affect scores across time, (B) Sleep metrics across time by study group, and (C) Borg RPE weekly averages by study group.

Full size image

Sleep outcomes were analysed using mixed models with time and group as fixed effects. A significant Group × Time interaction emerged, F(2, 122) = 6.27, p = 0.013, η² = 0.09. Self-reported sleep duration increased in the camp group (6.78 ± 0.76 → 7.35 ± 0.78 h) and remained unchanged in controls (6.72 ± 0.81 → 6.60 ± 0.82 h). Sleep quality also improved significantly (F(2, 122) = 5.61, p = 0.020), based on (abbreviated PSQI: duration, quality, latency, awakenings) at T1–T3; wrist actigraphy was used to verify diary adherence only (see Fig. 4B). At T1, sleep duration was 6.72 ± 0.81 h in controls and 6.78 ± 0.76 in the intervention group (t(61) = 0.307, p = 0.760, d = 0.09). At T2, self-reported sleep duration was longer in the intervention group (7.28 ± 0.74 h vs. 6.55 ± 0.83 h), accompanied by higher perceived sleep-quality scores (4.1 ± 0.7 vs. 3.0 ± 0.8) and fewer self-reported awakenings (1.2 ± 0.5 vs. 2.0 ± 0.6; adjusted p < 0.05). At T3, these self-reported differences remained consistent, with higher perceived sleep quality and longer duration in the intervention group (adjusted p < 0.01). Perceived exertion was recorded weekly using Borg RPE mean and peak scores across both groups (see Fig. 4C). RPE scores declined significantly over time in the camp group, F(3, 183) = 8.92, p < 0.001, η² = 0.13, whereas they remained stable in controls, yielding a Group × Time interaction, F(3, 183) = 7.34, p = 0.002. Mean RPE decreased from 12.6 ± 1.5 to 11.1 ± 1.4 versus 13.2 ± 1.6 in controls (p = 0.007, d = 1.33). In Week 1, mean RPE was 12.3 ± 1.4 in controls and 12.6 ± 1.5 in the intervention group (t(61) = 0.726, p = 0.470, d = 0.17). From Week 2 onward, intervention participants showed consistently lower exertion, with Week 2 RPE at 11.7 ± 1.4 vs. 13.1 ± 1.6 (t(61) = 3.641, p = 0.002, d = 0.95), Week 3 at 11.4 ± 1.3 vs. 13.3 ± 1.5 (p = 0.005, d = 1.28), and Week 4 at 11.1 ± 1.4 vs. 13.2 ± 1.6 (p = 0.007, d = 1.33).

Fig. 5
figure 5

(A) Baseline EISA-24 subscale scores by study group, (B) PANAS-C score change from T1 to T3, and (C) Emotional profile classification frequencies.

Full size image

Trait emotional intelligence was assessed using EISA-24 subscales at baseline for both groups (see Fig. 5A). Trait emotional intelligence scores differed by group, F(1, 61) = 5.71, p = 0.020, η² = 0.09, driven by higher self-regulation and emotional utilisation in the camp group. Emotional state profiles derived from PANAS-C changes also showed a Group × Time interaction, F(2, 122) = 8.38, p = 0.004, η² = 0.12, indicating greater positive affect retention and reduced reactivity in camp participants. Emotional awareness averaged 21.5 ± 3.2 in controls and 22.8 ± 3.0 in the intervention group (t(61) = 1.753, p = 0.085, d = 0.44). Self-regulation scores were markedly higher in the intervention group at 24.3 ± 3.1 compared to 20.1 ± 2.9 (t(61) = 5.711, p = 0.001, d = 1.46). Emotional utilization also differed significantly, 23.6 ± 3.2 vs. 19.7 ± 3.0 (t(61) = 5.173, p = 0.003, d = 1.32). Empathy scores showed a moderate difference, 23.9 ± 3.5 vs. 22.3 ± 3.4 (p = 0.059, d = 0.50).

Affective state changes from T1 to T3 were evaluated using PANAS-C scores (see Fig. 5B). Positive affect increased in the intervention group by + 4.9 ± 2.7 compared to a slight decline of − 0.3 ± 2.1 in the control group. Negative affect decreased by − 3.1 ± 2.4 in the intervention group, whereas the control group showed a minor increase of + 0.2 ± 2.0. The between-group difference was statistically significant (t(61) = 6.289, p = 0.001), yielding a large effect size (Cohen’s d = 1.59), indicating marked emotional improvement.

Emotional profile classification frequencies across groups were examined using categorical distribution metrics (see Fig. 5C). In the control group, 8 participants (26.7%) were classified as Resilient, 14 (46.7%) as Reactive, and 8 (26.7%) as Dysregulated. In contrast, the intervention group showed 21 Resilient (70.0%), 6 Reactive (20.0%), and 3 Dysregulated (10.0%) profiles. A chi-square test indicated a significant difference in the distribution of emotional profiles between groups, χ²(2) = 14.527, p = 0.002, highlighting a higher prevalence of adaptive emotional regulation in the intervention group.

Fig. 6
figure 6

(A) HRV and PSS by high vs. low trait EI (intervention group only), and (B) Key outcome metrics by sex (intervention group only).

Full size image

Comparative psychophysiological metrics by trait emotional intelligence were analysed within the intervention group only (see Fig. 6A). Within-group comparisons revealed significant sex effects on HRV, F(1, 28) = 6.12, p = 0.019, η² = 0.18, and negative affect, F(1, 28) = 5.03, p = 0.032, η² = 0.15. Trait emotional intelligence moderated stress and recovery indices, showing an EI × Time interaction, F(2, 56) = 7.41, p = 0.008, η² = 0.21, where high-EI athletes demonstrated larger RMSSD gains (+ 9.3 ms) and lower cortisol (− 0.042 µg/dL) across camp phases. Participants with high trait EI (n = 15) showed higher RMSSD at T3 (64.2 ± 8.7 ms) and lower cortisol levels (0.181 ± 0.031 µg/dL) compared to those with low EI (48.7 ± 6.9 ms; 0.223 ± 0.028 µg/dL). Perceived stress scores were lower in the high EI group (12.3 ± 3.4) versus the low EI group (17.5 ± 3.9). Positive affect scores were also higher (31.8 ± 4.1 vs. 24.6 ± 3.7), with a significant group difference, t(28) = 4.237, p = 0.002, Cohen’s d = 1.55. Key psychophysiological and recovery-related outcomes were compared by sex within the intervention group (see Fig. 6B). Males (n = 15) recorded higher RMSSD at T3 (58.5 ± 7.4 ms) and lower cortisol levels (0.195 ± 0.029 µg/dL) than females (50.7 ± 6.8 ms; 0.214 ± 0.026 µg/dL). Negative affect scores were lower in males (13.1 ± 3.2) than in females (15.8 ± 3.5), while sleep quality ratings were higher (4.1 ± 0.6 vs. 3.6 ± 0.7). The group difference was significant, t(28) = 2.638, p = 0.013, with a Cohen’s d of 0.97. Outcome changes were stratified by baseline weekly training hours within the intervention group. Participants in the low training group (< 5.0 h) showed a mean HRV increase of + 4.7 ± 3.2 ms, PSS reduction of − 2.4 ± 2.1, cortisol decrease of − 0.014 ± 0.010 µg/dL, and PA gain of + 1.3 ± 2.0. Moderate group (5.1–7.5 h) recorded + 7.2 ± 3.5 ms, − 3.6 ± 1.8, − 0.023 ± 0.009, and + 3.9 ± 2.6, respectively. The high training group (> 7.5 h) achieved + 10.4 ± 3.7 ms, − 5.1 ± 2.2 ms, − 0.031 ± 0.011 ms, and + 5.8 ± 2.9 ms. Between-group comparison was significant, F(2, 27) = 5.261, p = 0.011, η² = 0.28.37,38,39,40,41

Discussion

Participants in the intervention group showed a marked increase in parasympathetic activity, with RMSSD rising from 44.7 ± 13.1 ms at baseline to 54.6 ± 13.5 ms at the end of the program, indicating substantial autonomic recovery. In contrast, previous findings reported HRV suppression in similar cohorts under cognitive–endurance stress. Gorgulu et al.3 observed a ~ 23% drop in RMSSD following navigational errors, particularly in female orienteers, highlighting stress-linked vagal withdrawal. Yang et al.20 further reported a ~ 40% decline in lnRMSSD after fatigue induction, with sympathetic dominance dominating recovery. The study’s contrasting results reflect successful vagal restoration driven by cumulative neuromodulatory and regulatory inputs during the intervention. The observed autonomic rebound is consistent with prior models of prefrontal–vagal regulation and baroreceptor adaptation, but such processes were not directly measured in this study. Therefore, these interpretations remain inferential, highlighting potential mechanisms rather than confirmed physiological pathways. The combined HRV, cortisol, and affective results reinforce a multilevel model of resilience in which physiological recovery, emotional stability, and perceived control operate in synchrony. This synthesis supports the emerging view that resilience is not the product of any single domain but rather a dynamic coordination of autonomic, endocrine, and affective systems during adaptation to complex stressors.

These integrated outcomes align directly with the conceptual logic of the Integrated Psychophysiological Fatigue–Emotion Synchronisation Model (IPFESM). Within this framework, adaptive resilience is theorised to arise from synchronised adjustments across three systems: autonomic regulation, endocrine modulation, and affective adaptation. The present findings empirically support this synchronisation pattern: enhanced HRV and reduced cortisol corresponded with improved affective balance, indicating coordinated recovery rather than isolated physiological or emotional shifts.

Morning salivary cortisol decreased notably from 0.46 ± 0.17 µg/dL to 0.36 ± 0.13 µg/dL in the intervention group, reflecting an adaptive endocrine profile. This pattern aligns with prior research showing that moderate, skill-based physical engagement is associated with lower basal cortisol and enhanced stress recovery capacity. Budde et al.22 reported greater cortisol elevations following endurance than coordinative tasks, highlighting cortisol’s sensitivity to cumulative physical stress and recovery kinetics. The study findings align with these results but uniquely demonstrate that multi-day integrative training can suppress baseline cortisol, suggesting not only acute recovery but also chronic downregulation of HPA hyperactivity. While the observed reduction may indicate improved regulatory balance, the present data cannot isolate underlying neuroendocrine mechanisms. From the perspective of the IPFESM framework, this endocrine stabilisation represents the hormonal component of fatigue–emotion synchronisation, in which cortisol modulation parallels autonomic recovery. The observed coupling between decreasing cortisol and rising HRV supports the model’s premise that physiological recalibration and emotional regulation unfold in tandem, driven by shared adaptive control loops rather than independent systems.

Participants who trained more than 7.5 h weekly showed the greatest increases in RMSSD (+ 10.4 ± 3.7 ms), perceived stress reductions (− 5.1 ± 2.2), and positive affect gains (+ 5.8 ± 2.9), indicating that both training volume and emotional adaptability contributed to enhanced resilience. Within the IPFESM framework, trait emotional intelligence can be interpreted as a moderating factor that enhances the synchronisation efficiency between the physiological and affective subsystems. High-EI participants likely maintained tighter alignment between autonomic recovery and emotional appraisal, facilitating faster stabilisation of stress indicators throughout the training cycle. Melguizo Ibáñez et al.9 similarly found that trainees exceeding 150 min of weekly activity demonstrated superior emotional intelligence and psychological well-being, with emotional intelligence mediating 28% of the activity–resilience link. This reinforces the study finding that trait EI moderated physiological and affective outcomes. Additionally, Ferrari et al.18 highlighted orienteering’s unique restorative potential, with a 27% stress reduction and 34% improvement in nature connectedness, especially in females. In the study, emotional restoration was significantly more pronounced in female participants, supporting sex-specific affective pathways. These findings suggest that emotionally intelligent individuals leverage both physical exertion and environmental affordances such as sensory immersion and terrain variability to regulate stress. The observed buffering may arise from combined neurocognitive mechanisms of attentional refocusing, interoceptive attunement, and limbic modulation activated through structured movement in naturalistic settings.

Participants in the intervention group exhibited significant improvements in positive affect, rising from 25.2 ± 5.4 to 30.1 ± 5.5, with concurrent reductions in negative affect and fatigue indicators. These affective enhancements were particularly pronounced among individuals with higher baseline training volumes and stronger emotional regulation capacities. Liu et al.5 found that physical activity enhanced positive affect (β = 0.25), with resilience and emotional self-efficacy acting as mediators. Participants with higher activity levels reported 19% greater emotional regulation. The structured orienteering protocol in the study likely replicated these dynamics by offering rhythm, challenge, and reflective intervals that activated both cognitive and emotional self-regulatory pathways. These associative trends suggest that structured, emotionally engaging physical activity may foster adaptive affective regulation; however, causal relationships should be interpreted cautiously, as analyses were correlational rather than experimental.

Theoretical and practical implications

This study provides initial empirical support for the Integrated Psychophysiological Fatigue–Emotion Synchronisation Model (IPFESM), demonstrating that resilience under dual cognitive–physical load can be understood as a product of multi-system coordination. Theoretical refinement of this model may guide future predictive frameworks linking physiological regulation with emotional adaptation. Practically, it supports the use of structured orienteering as a scalable intervention to enhance heart rate variability, stress regulation, and emotional intelligence. Practically, the results support the feasibility of integrating wearable monitoring and emotional-profiling approaches within athletic and youth-training contexts similar to those studied here. While such frameworks may also hold promise for other cognitively demanding environments (e.g., academic or pre-military programs), these applications require targeted validation before broad implementation.

Limitations and future research

One limitation of this study is its short intervention duration, which limits inferences about sustained emotional or physiological change. Another limitation is the absence of objective sleep tracking, which could have complemented self-reported data. Although power analysis indicated adequate sensitivity, the modest sample size and single-cohort design may limit generalizability and the detection of smaller group effects, particularly for sex or training-volume comparisons. In future work, the study will address these by implementing longitudinal tracking over six months and incorporating actigraphy-based sleep monitoring to verify physiological recovery patterns.

Conclusion

The study identified significant associations between participation in the orienteering-based program and improvements in heart rate variability, cortisol regulation, sleep quality, and emotional resilience. Participants showed marked improvements in both physiological recovery and psychological flexibility, especially among those with higher baseline fitness or emotional regulation capacity. Emotional profiling confirmed a shift toward resilient patterns post-intervention, with trait emotional intelligence moderating these effects. Overall, the findings provide preliminary empirical grounding for the IPFESM by illustrating how synchronised autonomic, endocrine, and emotional adjustments co-occur during structured endurance–navigation tasks. These outcomes highlight the potential of navigation-based physical activity combined with emotion-regulation strategies as a feasible approach to strengthen psychophysiological resilience. Future work should test this framework in broader youth and sport populations before extending it to other professional or academic settings. As a real-world application, this protocol offers a cost-effective, scalable model for resilience-building programs in educational, military, and therapeutic policy domains.