Psychological and Sleep Changes in Pre-Selected Members of the Polar Inland Expedition Team During Rapid Ascent High-Altitude Training

Summary

This study reveals that preliminary and backup Antarctic inland expedition team members demonstrated significantly elevated anxiety levels, prefrontal neuroplastic compensation (left anterior prefrontal cortex activation), and sleep fragmentation (Wake After Sleep Onset WASO) under high-altitude conditions. These findings provide an empirical basis for developing adaptive training protocols for extreme environments.

Abstract

This study established and implemented a multidimensional assessment protocol to systematically investigate neurobehavioral changes among 30 preliminary and backup members of the Antarctic inland expedition team undergoing rapid altitude training. The protocol integrates psychometric scales (BFI-44, DASS-21, PSQI), behavioral paradigms (N-back task), functional near-infrared spectroscopy (fNIRS), and wrist actigraphy to assess neurobehavioral adaptations across varying altitudes (0 m above sea level (a.s.l) / 3,700 m a.s.l / 4,300 m a.s.l) over an 8-day cycle. The core design comprises pre-ascent baseline testing; synchronized cognitive-fNIRS measurements at target altitudes; 24/8 sleep monitoring (including 8 h nocturnal actigraphy). This multimodal framework enables quantitative characterization of prefrontal compensatory activation, sleep fragmentation dynamics, and altitude-induced anxiety fluctuations. Findings reveal that, compared to Chinese normative data, participants exhibited significantly elevated agreeableness (t=3.940, P<0.001) and conscientiousness (t=9.736, P<0.001), alongside reduced neuroticism (t=-14.087, P<0.001). Acute hypoxia exacerbated anxiety (Z=-4.098, P<0.001), with greater severity in ethnic minority members (H=6.405, P=0.011). Actigraphy demonstrated altitude-dependent sleep fragmentation (WASO: 4,300 m a.s.l versus baseline, P=0.028). N-back tasks confirmed preserved working memory at 3,700 m a.s.l. (accuracy P=0.027) mediated by compensatory prefrontal activation (L-aPFC activation P=0.043). These data establish evidence-based selection criteria for polar science expeditions in high-altitude settings and delineate neuroplastic resilience thresholds under hypoxic stress.

Introduction

Polar expedition members must adapt to coexisting physiological and psychological challenges in extreme environments¹. The high-altitude conditions of Antarctica's interior (elevation > 4000 m a.s.l) can induce hypoxia-related cognitive impairment and sleep disruption. Although only Dome A (>4000 m a.s.l) is a permanent high-altitude Antarctic research station, traverses to inland sites such as Dome A and Dome F involve prolonged exposure (1-2 weeks) to altitudes exceeding 3000-4000 m a.s.l^2,3. Moreover, hypoxia is a key stressor encountered both during these traverses and at Dome A itself, reflecting the conditions simulated in this training protocol. Therefore, understanding neurobehavioral adaptation to hypoxia is essential for personnel selection and training optimization for inland traverse teams and Dome A operators, to ensure safety and performance in these logistically critical and scientifically valuable missions. The overall goal of this method is to establish a multimodal assessment protocol that systematically quantifies neurobehavioral adaptations during rapid altitude acclimatization, integrating psychological, cognitive, neurovascular, and sleep parameters into a unified framework. This technique was developed in response to critical limitations in existing research, which predominantly examines psychological or physiological parameters in isolation^4,5,6,7. The advantages of this multimodal technique over conventional methods include:

Ecological validity: Simultaneous fNIRS-behavioral coupling captures real-time neurocognitive adaptations^8,9, overcoming retrospective biases inherent in scale-only assessments⁷

Comprehensive profiling: Integrating actigraphy with validated scales (PSQI) provides objective verification of subjective sleep reports^10,11

Dynamic tracking: Repeated measures across altitudes reveal temporal adaptation patterns unattainable through single-point assessments⁴

Within the wider literature, this protocol addresses the escalating operational demands of Antarctic research¹² by providing the first standardized framework that: Quantifies prefrontal compensatory mechanisms by fNIRS during cognitive loading under hypoxia^8,13; correlates personality traits (BFI-44) with altitude-induced anxiety dynamics (DASS-21)^14,15; synchronizes objective sleep fragmentation (actigraphy) with neurobehavioral outcomes^16,17.

This method is appropriate for researchers and practitioners who require personnel selection systems for high-altitude polar operations, validation of neuroplastic resilience thresholds under hypoxic stress (>3,700 m a.s.l.), and objective biomarkers (e.g., L-aPFC activation) to predict task performance degradation. Implementation considerations: Recommended for adults aged 25-45 without cardiopulmonary comorbidities, requires 3-day acclimatization per altitude tier (3,700 m a.s.l to 4,300 m a.s.l), technical constraints: fNIRS optode placement limits concurrent headgear use. With increasing Antarctic research stations and diversified scientific missions¹², this protocol provides evidence-based solutions for optimizing adaptive training in extreme environments.

Protocol

1. Study participants

1. Participant selection
   1. Sampling method: A cluster sampling of 30 pre-selected and alternate team members from the Antarctic inland expedition team was taken as the study subjects.
   2. Study subjects: Include all pre-selected candidates and reserve members from the Chinese Antarctic Inland Expedition Team. Exclude individuals unable to complete the training regimen during the training period (e.g., those who failed to complete the training plan due to illness, physical discomfort, or other reasons).
2. Data collection
3. Demographic documentation: Record comprehensive sociodemographic data, including age, educational attainment, and previous polar expedition experience (Table 1).
4. Ensure that strict anonymization measures are enforced for all psychometric data, including: assignment of unique anonymized IDs (e.g., EXP##) to replace personal identifiers, removal of direct identifiers (e.g., names, addresses) from datasets, storage of identifiable data (e.g., consent forms) in password-protected, encrypted files separate from research data.

n=30	N	%
Ethnicity
Han Chinese	28	93.3
Ethnic Minority(Tibetan and Manchu)	2	6.67
Highest educational attainment
High school and below	1	3.3
College or Bachelor's Degree	11	36.7
Master's Degree	8	26.7
Doctoral Degree	10	33.3
Only Child
Yes	16	53.3
No	14	46.7
Marital Status
Unmarried	15	50
Married	15	50
Number of Scientific Expeditions Participated
Not Participated	24	80
1 Time	2	6.7
2 or More Times	4	13.3
Age: 32.72±7.73

Table 1: Demographic data. (A) This table presents the ethnic composition, educational level, only-child status, marital status, and frequency of research participation for the sample, with the last row showing age data (Mean ± SD). (B) N: Sample count; %: Percentage relative to total sample size (n=30), data shows Mean ± SD.

2. Preparation of psychological and sleep assessment instruments

1. Standardized psychometric scales: Carry out a researcher-administered questionnaire-based assessment in a standardized interview format to ensure comprehension and minimize self-reporting bias. Guide participants verbally through each item, with clarifications provided as needed.
   1. Big five inventory-44 (BFI-44)¹⁴
      1. Select the 44-item BFI-44 questionnaire to assess five personality dimensions. Instruct participants to rate statements using a 5-point Likert scale (1=Strongly disagree to 5=Strongly agree).
      2. Calculate domain scores for Extraversion, Agreeableness, Neuroticism, Openness, and Conscientiousness. Higher scores indicate stronger expression of corresponding traits. The scale includes 28 positively keyed items.
   2. Depression anxiety stress scales-21 (DASS-21)¹⁵
      1. Administer the 21-item DASS-21 with a 4-point Likert scale (1= Did not apply to me at all to 4=Applied to me very much). Record subscale scores for Depression, Anxiety, and Stress. Elevated scores reflect greater severity of emotional disturbances.
   3. Pittsburgh Sleep Quality Index (PSQI)¹⁰
      1. Distribute 18-item PSQI questionnaires to participants. Instruct completion within 10 min using 0-3 scoring per domain: Subjective sleep quality, Sleep latency, Sleep disturbances, hypnotic medication use, daytime dysfunction. Calculate global scores (0-21 scale). Higher total scores on the PSQI indicate poorer overall sleep quality.
2. Sleep monitoring
   NOTE: The apparatus used here for wrist actigraphy was ActiGraph wGT3X-BT¹⁶.
   1. Initialize devices using the software with standardized timestamps.
   2. Launch ActiLife v6.15.0. Click on Device > Initialize > Set. Set the sampling rate: 60 Hz, Start time: YYYY-MM-DD HH:MM:SS (synchronized with atomic clock), Subject ID: EXP## (anonymized format).
   3. Set a checkpoint to verify that the initialization log displays Initialization Successful.
   4. Calibrate the software using manufacturer protocols after placement on the non-dominant wrist. Insert the device into the USB Dock. Click Calibrate and then follow the on-screen gravity vector test protocol.
   5. Collect 24/8 activity data at a 60 Hz sampling rate. Perform a visual checkpoint and confirm the green status LED blinks every 15 s (normal operation).
   6. Carry out data acquisition. Post-deployment, insert the device into the dock. Select ActiLife. Click on Download > RAW .gt3x format > Process using Cole-Kripke algorithm. Use Cole-Kripke algorithm for sleep-wake classification. Validate device-derived sleep parameters via participant sleep diaries¹⁷.
   7. Perform sampling with continuous 24/8 recording at 60 Hz (ensure space between value/unit: 60 Hz)

3. Cognitive assessment and fNIRS acquisition

1. Letter N-back task^8,9
   1. Using E-Prime 2.0 software, uppercase letters are randomly presented in the center of a 16-inch LCD monitor. Participants are required to make judgments: press the F key if the current letter matches the one presented N positions earlier or press the J key if it does not match (N=1, 1-back; N=2, 2-back; N=3, 3-back).
   2. Record accuracy (ACC) and reaction time (RT) metrics for 50 trials per difficulty level. 
   3. In the E-Prime output settings, configure the DataFile.Name as "EXP##Nback.edat2". For the stimulus onset time, select the variables: ACC (Correct = 1 / Incorrect = 0) and RT (in milliseconds). Click Export > Excel format.
   4. Click on Analyze > General Linear Model > Repeated MeasuresWithin-Subject Factors: N-level(1,2,3). Set checkpoints at Raw data: 50 trials/level = 150 trials total.
   5. For the practice phase, in E-Prime, click on BlockProc > Practice Properties. Set Max trials: 10, Feedback Screen: Correct! (green) / Incorrect (red), and Advancement: Press the 'Q' key to repeat the practice phase, press the 'P' key to proceed to the formal experiment. A checkpoint is set, limiting the practice phase to a maximum of 3 repetitions (Figure 1).

Figure 1: Letter N-back paradigm. This schematic illustrates the letter n-back task paradigm across three difficulty levels (1-back, 2-back, and 3-back). Please click here to view a larger version of this figure.

2. fNIRS data acquisition
   NOTE: The NirScan-500C, Huichuang, Danyang was used here.
   1. Carry out optode configuration. Position a 48-channel fNIRS headpiece according to the 10-20 international system.
   2. Set spatial arrangement with source-detector distance at 30 mm, coverage: Prefrontal cortex (PFC) and bilateral temporal regions, and optode matrix: 15 source optodes + 16 detector optodes (Table 2, Table 3, Figure 2).
   3. Signal Acquisition Parameters
      1. Set signal acquisition parameters using the NirScan-500C software. Select Dual-wavelength near-infrared light: Set wavelengths to 730 nm ± 5 nm and 850 nm ± 5 nm.
      2. Configure Sampling rate: Set to 11 Hz by clicking Settings > Acquisition Rate. Enable Task-event synchronization: Activate hardware triggers by selecting Sync > External Trigger.
      3. Record raw optical intensity changes for subsequent conversion to oxygenated/deoxygenated hemoglobin concentrations using the modified Beer-Lambert law. Verify parameter settings in the software log (e.g., Wavelengths: 730/850 nm, Rate: 11 Hz).

Figure 2: Channel arrangement, position, and configuration of the fNIRS probe. (A) Red circles represent light sources, blue squares represent detectors, and numbered areas represent the nearest source-detector pairs (channels) to measure the brain activities. (B) Coregistered positions of the optodes on a standard brain atlas. The anatomical position of each channel on the brain atlas is reported in detail in Table 1. Please click here to view a larger version of this figure.

Channel	Brodmann area	Anatomical area	Coverage percentage
CH1	21	Middle Temporal gyrus	97.44%
CH2	22	Superior Temporal Gyrus	43.21%
CH3	38	Temporopolar area	51.84%
CH4	47	Inferior prefrontal gyrus	100.00%
CH5	47	Inferior prefrontal gyrus	56.35%
CH6	10	Frontopolar area	62.75%
CH7	10	Frontopolar area	51.96%
CH8	10	Frontopolar area	100.00%
CH9	11	Orbitofrontal area	59.66%
CH10	11	Orbitofrontal area	46.67%
CH11	10	Frontopolar area	98.66%
CH12	38	Temporopolar area	57.25%
CH13	21	Middle Temporal gyrus	98.63%
CH14	22	Superior Temporal Gyrus	43.46%
CH15	21	Middle Temporal gyrus	92.79%
CH16	21	Middle Temporal gyrus	75.85%
CH17	22	Superior Temporal Gyrus	62.67%
CH18	43	Subcentral area	36.79%
CH19	44	pars opercularis_ part of Broca's area	57.05%
CH20	6	Pre-Motor and Supplementary Motor Cortex	73.08%
CH21	10	Frontopolar area	49.10%
CH22	46	Dorsolateral prefrontal cortex	78.62%
CH23	10	Frontopolar area	93.20%
CH24	9,46	Dorsolateral prefrontal cortex	77.95%
CH25	10	Frontopolar area	87.63%
CH26	10	Frontopolar area	100.00%
CH27	10	Frontopolar area	100.00%
CH28	10	Frontopolar area	99.65%
CH29	47	Inferior prefrontal gyrus	58.95%
CH30	10	Frontopolar area	92.46%
CH31	45	pars triangularis Broca's area	64.19%
CH32	46	Dorsolateral prefrontal cortex	86.52%
CH33	21	Middle Temporal gyrus	56.73%
CH34	22	Superior Temporal Gyrus	53.70%
CH35	42	Primary and Auditory Association Cortex	64.69%
CH36	43	Subcentral area	46.23%
CH37	9,46	Dorsolateral prefrontal cortex	60.54%
CH38	9	Dorsolateral prefrontal cortex	52.09%
CH39	9,46	Dorsolateral prefrontal cortex	92.77%
CH40	10	Frontopolar area	89.88%
CH41	9,46	Dorsolateral prefrontal cortex	91.08%
CH42	9	Dorsolateral prefrontal cortex	84.34%
CH43	10	Frontopolar area	92.46%
CH44	9	Dorsolateral prefrontal cortex	83.67%
CH45	9,46	Dorsolateral prefrontal cortex	90.37%
CH46	45	pars triangularis Broca's area	39.27%
CH47	9,46	Dorsolateral prefrontal cortex	98.16%
CH48	6	Pre-Motor and Supplementary Motor Cortex	80.97%

Table 2: Brain region distribution of fNIRS measurement channel after calibration by 3D locator. (A) Spatial distribution characteristics of 48 fNIRS channels calibrated by a 3D positioning system across Brodmann areas and corresponding anatomical brain regions. Coverage percentage indicates the effective signal acquisition proportion from target brain regions. (B) Channel: fNIRS optical measurement channel number (CH1-CH48); Brodmann area: Brodmann area number (e.g., 22 = Auditory Association Cortex); Anatomical area: Standard terminology per Terminologia Anatomica (1998); Coverage %: Effective coverage percentage of the optode pairs for the target brain region.

Anatomical area	Channel
L-aPFC	11,25,27,28,30,43
R-aPFC	6,7,8,21,23,26,40
L-DLPFC	32,44,45,47
R-DLPFC	22,24,37,38,39,41,42

Table 3: Areas of region of interest. (A) Display the channel configuration distribution for specific functional brain regions in the fNIRS measurement system, including the anterior Prefrontal Cortex (aPFC) and Dorsolateral Prefrontal Cortex (DLPFC) in both left and right hemispheres. (B) Abbreviations: L = Left hemisphere; R = Right hemisphere; aPFC = anterior Prefrontal Cortex; DLPFC = Dorsolateral Prefrontal Cortex.

4. Data acquisition

1. Psychological assessment implementation
   NOTE: The team members departed from sea level (Shanghai, 0 m above sea level) and took a commercial flight (approximately 8 h) to Lhasa, located at 3,700 m above sea level, where they conducted a 72 h (3-day) scientific research skills training. Subsequently, the team transferred y ground transportation (approximately 2 h) to Yangbajain at 4,300 m above sea level to continue another 72 h (3-day) scientific research skills training. Data collection was carried out immediately upon the team’s arrival at each of the three aforementioned altitudes (Figure 3).
   1. Conduct standardized administration: Launch pre-assessment group briefing sessions. Administer questionnaires with a 30 min time constraint.
   2. Mitigate bias: Read standardized instructions from scripted protocols. Eliminate suggestive phrasing by following pre-approved text.
   3. Synchronize multimodal data: Enable synchronization between cognitive tasks and fNIRS acquisition. Time-lock stimuli to fNIRS triggers by selecting E-Prime > Trigger Settings > Sync to fNIRS.
   4. Ensure all sessions adhere to identical scripts and timing constraints.
2. Sleep parameter monitoring
   1. Implement continuous actigraphy protocol: Collect 24 h/8 h wrist actigraphy data. Concurrently record sleep diary.
   2. Ensure data integrity: Perform daily battery checks at 07:00-08:00 by clicking Device > Status > Battery. Validate data continuity by clicking ActiLife > Tools > Data Integrity Check.  

Figure 3: Psychological assessment flowchart. (A) This flowchart illustrates a three-stage psychometric assessment protocol across graded altitude exposures: Low altitude (0-2 m a.s.l.): Baseline measurements in Shanghai (Day 1-2), Mid altitude (3650 m a.s.l.): Acclimatization-phase measurements in Lhasa (Day 3-5), High altitude (4300 m a.s.l.): High-exposure measurements in Yangbajing (Day 6-8). (B) Identical assessment modules implemented at each stage: Cognitive Task: Letter N-back (3 difficulty levels); Psychological Scales: Depression Anxiety Stress Scale (DASS-21), Pittsburgh Sleep Quality Index (PSQI); Sleep Monitoring: Subjective (Sleep diary), Objective (Actigraphy). Please click here to view a larger version of this figure.

5. Data analysis and statistics

1. For data preprocessing in a spreadsheet, apply exclusion criteria to remove incomplete datasets with >10% missing responses by selecting Data > Filter > Missing Values. Filter cognitive data to discard RT outliers (<150 ms or> mean ±2.5 SD) using Data > Sort & Filter > Advanced, and calculate the HitRT (in milliseconds) and rejRT (in milliseconds). Handle missing data to implement multiple imputation in SPSS 29.0 by clicking Transform > Multiple Imputation (5 iterations).
2. fNIRS signal processing
   NOTE: This was done using NirSpark v1.8.8, Huichuang, Danyang.
   1. Convert raw signals to optical density (OD). Click File > Import Raw > Select .nirs file. Select Convert to Optical Density and apply algorithm: OD = -log10(I/I₀). Verify OD range 0.1-1.2 AU in View > Channel Metrics. Reject channels with OD >1.5(poor contact) by selecting Tools > Reject Channels.
   2. Calculate hemodynamic changes. Navigate to Analysis > Hemodynamic > Parameters. Set Law as Modified Beer-Lambert, DPF as 66, and Image Mode as adult_head_template. Execute calculation as [HbO, HbR] = mBLL(OD, [730,850], [6,6]) and HbT = HbO + HbR. Confirm DPF value matches participant age (e.g., 6.0 for 25-45 years).
   3. Denoise signals by clicking on Preprocessing > Filter > Settings. Select Bandpass Butterworth 4th-order and set the range to 0-0.2 Hz. Perform motion correction via the automatic detection (std_thr=6, amp_thr=0.5) and spline interpolation method and set artifact rejection threshold to SD >±6. Verify SNR improvement >10 dBin by clicking View > Pre/Post Comparison.
   4. Quantify activation as follows. Define baseline as Mean HbO during the entire post-stimulus task period. Calculate task response as Mean HbO during 2-20 s post-stimulus. Generate ROI averages by selecting Analysis > ROI Masks > Anatomical Templates.
3. Sleep efficiency calculation
   1. Import and preprocess data by clicking File > Import > Select .gt3x file. Set the Epoch to 60s and validate the wear time by selecting Tools > Wear Time Validation > Choi algorithm. To perform actigraphy and sleep diary synchronization, click on Tools > Sleep > Detect Sleep Periods. Then, calibrate the actigraphy data against the sleep diary. Ensure sleep diary timestamps are within ± 5 min of actigraphy data.
   2. Extract parameters as follows. Navigate to Reports > Sleep Report. Select parameters: Latency, TIB, TST, WASO. Compute Sleep Efficiency = (TST / TIB) x 100 by selecting Options > Formulas. Export results by clicking on Export > Excel .xlsx > Filename: EXP##_SleepEfficiency.xlsx. Confirm actogram displays ≥3 clear sleep/wake cycles.
4. Statistical testing
   1. Perform descriptive statistics. Report normally distributed data as mean ± SD and non-normal data as median [IQR].
   2. Execute parametric tests. Run one-sample t-tests by clicking Analyze > Compare Means > One-Sample T Test. Conduct RM-ANOVA by clicking Analyze > General Linear Model > Repeated Measures. Apply Greenhouse-Geisser correction for sphericity violations.
   3. Apply nonparametric alternatives as needed. Use Nonparametric Tests on Related Samples for Wilcoxon signed-rank tests. Control covariates by selecting Analyze > General Linear Model > ANCOVA. Set significance threshold by applying α=0.05 (two-tailed) with Bonferroni correction for multiple comparisons.

Results

Personality traits

Personality characteristics of the pre-selected polar inland expedition members were evaluated using the Big Five Inventory (BFI). Compared to Chinese undergraduate norms¹⁸, candidates demonstrated (Table 4) significantly higher scores in Agreeableness (t=3.940, P<0.001) and Conscientiousness (t=9.736, P<0.001); a markedly lower score in Neuroticism (t=−14.078, P<0.001)¹⁸. No ...

Discussion

Personality traits of pre-selected polar inland expedition members differ significantly from Chinese norms

As measured by the Big Five Inventory-44 (BFI-44), candidates exhibited elevated agreeableness and conscientiousness alongside reduced neuroticism compared to undergraduate norms (Table 3). This triad-heightened social cooperativeness, task-oriented rigor, and emotional stability-aligns with the psychosocial demands of Antarctic expeditions.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Actigraph	Actigraph	wGT3X-BT	America
ActiLife	Actigraph	6.15.0	America
E-Prime	Psychology Software Tools	2.0	America
IBM SPSS Statistics	IBM	29	America
Microsoft Office	Microsoft Corporation	2021	America
NirScan	Huichuang Medical Equipment	500C	Danyang, China
NirSpark	Huichuang Medical Equipment	1.8.8	Danyang, China