Predicting severe high-altitude de-acclimatization syndrome from pre-return exposure dose and laboratory indicators: a prospective cohort study of 475 males
Abstract
High-altitude de-acclimatization syndrome (HADAs) poses significant health challenges to individuals returning from high-altitude residence, yet its de-acclimatization process of laboratory indicators and the extent to which it can be predicted before descent remain poorly characterized. We conducted a longitudinal cohort study involving 475 participants with repeated assessments at high-altitude residence (T0), 1 month after return to plain area (T1), and 3 months after return to plain area (T2). HADAs was diagnosed using established questionnaire criteria; outcomes were stratified by course (ever, persistent) and by severity. Cochran-Armitage trend tests were used to evaluate dose-response relationships of altitude and residence duration with HADAs incidence and persistence; Mann-Whitney U tests with Benjamini–Hochberg FDR correction were used to compare the 40 laboratory indicators between HADAs and non-HADAs participants; and machine-learning models (LightGBM, XGBoost, Random Forest, SVM, ensemble, TabPFN) were developed under a leakage-free protocol to predict severe HADAs from pre-return (T0) data. Altitude and residence duration showed robust dose-response associations with HADAs incidence and persistence (P for trend < 0.001), with over 85% incidence at altitudes above 5,000 m or residence for more than 18 months. After FDR correction, triglycerides, direct bilirubin and platelet distribution width differed between groups (q = 0.036). For severe HADAs, the TabPFN model achieved a held-out test AUC of 0.784 (95% CI 0.665–0.893). At the Youden-optimal threshold, its sensitivity and specificity were 0.739 and 0.750. Altitude, the altitude × duration interaction and pre-return erythrocyte indices were the dominant predictors. Cumulative exposure dose (altitude and residence duration) is the dominant determinant of HADAs incidence and persistence. Dysregulated lipid and platelet indices were core pathophysiological correlates of the de-acclimatization process. The pre-return model can identify the minority at risk of severe HADAs, supporting targeted monitoring while avoiding indiscriminate labeling in a population with high overall incidence.
Related
- Proteomics and network biology insights into inflammation suppression during prolonged high-altitude hypoxia acclimatization.
- Transcriptomic and epigenomic profiles analysis during human acclimatization to simulated high-altitude hypoxia
- Serum Cystatin 4 combined with routine clinical laboratory indicators: A support vector machine diagnostic model for early screening of gastrointestinal tumors
- Predicting visual impairment in type 2 diabetes using key performance indicators identified by the Black Widow and Whale Optimization Algorithms: A case study in Saudi Arabia
- Clinical and laboratory determinants of treatment response and rifampicin resistance among Xpert MTB/RIF-confirmed extrapulmonary tuberculosis patients: A prospective study from Central India
BibTeX
@article{liu2026predicting,
title = {{Predicting severe high-altitude de-acclimatization syndrome from pre-return exposure dose and laboratory indicators: a prospective cohort study of 475 males}},
author = {Ruohong Liu and Mu Yuan and LiYao Shui and Yufan Zhang and Dongxu Liao and Hongyin Liang and Yi Wen and Lijun Tang and Zhu Huang},
journal = {Scientific Reports},
year = {2026},
month = oct,
doi = {10.1038/s41598-026-73585-6},
url = {https://www.nature.com/articles/s41598-026-73585-6},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Predicting severe high-altitude de-acclimatization syndrome from pre-return exposure dose and laboratory indicators: a prospective cohort study of 475 males
AU - Ruohong Liu
AU - Mu Yuan
AU - LiYao Shui
AU - Yufan Zhang
AU - Dongxu Liao
AU - Hongyin Liang
AU - Yi Wen
AU - Lijun Tang
AU - Zhu Huang
PY - 2026
DA - 2026/10/04
JO - Scientific Reports
DO - 10.1038/s41598-026-73585-6
UR - https://www.nature.com/articles/s41598-026-73585-6
AB - High-altitude de-acclimatization syndrome (HADAs) poses significant health challenges to individuals returning from high-altitude residence, yet its de-acclimatization process of laboratory indicators and the extent to which it can be predicted before descent remain poorly characterized. We conducted a longitudinal cohort study involving 475 participants with repeated assessments at high-altitude residence (T0), 1 month after return to plain area (T1), and 3 months after return to plain area (T2). HADAs was diagnosed using established questionnaire criteria; outcomes were stratified by course (ever, persistent) and by severity. Cochran-Armitage trend tests were used to evaluate dose-response relationships of altitude and residence duration with HADAs incidence and persistence; Mann-Whitney U tests with Benjamini–Hochberg FDR correction were used to compare the 40 laboratory indicators between HADAs and non-HADAs participants; and machine-learning models (LightGBM, XGBoost, Random Forest, SVM, ensemble, TabPFN) were developed under a leakage-free protocol to predict severe HADAs from pre-return (T0) data. Altitude and residence duration showed robust dose-response associations with HADAs incidence and persistence (P for trend < 0.001), with over 85% incidence at altitudes above 5,000 m or residence for more than 18 months. After FDR correction, triglycerides, direct bilirubin and platelet distribution width differed between groups (q = 0.036). For severe HADAs, the TabPFN model achieved a held-out test AUC of 0.784 (95% CI 0.665–0.893). At the Youden-optimal threshold, its sensitivity and specificity were 0.739 and 0.750. Altitude, the altitude × duration interaction and pre-return erythrocyte indices were the dominant predictors. Cumulative exposure dose (altitude and residence duration) is the dominant determinant of HADAs incidence and persistence. Dysregulated lipid and platelet indices were core pathophysiological correlates of the de-acclimatization process. The pre-return model can identify the minority at risk of severe HADAs, supporting targeted monitoring while avoiding indiscriminate labeling in a population with high overall incidence.
SN - 2045-2322
ER -