Abstract
Medical personnel need practical guidelines on how to construct high altitude ascents to induce altitude acclimatization and avoid acute mountain sickness (AMS) following the first night of sleep at high altitude. Using multiple logistic regression and a comprehensive database, we developed a quantitative prediction model using ascent profile as the independent variable and altitude acclimatization status as the dependent variable from 188 volunteers (147 men, 41 women) who underwent various ascent profiles to 4 km. The accumulated altitude exposure (AAE), a new metric of hypoxic dose, was defined as the ascent profile and was calculated by multiplying the altitude elevation (km) by the number of days (d) at that altitude prior to ascent to 4 km. Altitude acclimatization status was defined as the likely presence or absence of AMS after ~24 h of exposure at 4 km. AMS was assessed using the Cerebral Factor Score (AMS-C) from the Environmental Symptoms Questionnaire and deemed present if AMS-C was ≥0.7. Other predictor variables included in the model were age and body mass index (BMI). Sex, race, and smoking status were considered in model development but eliminated due to inadequate numbers in each of the ascent profiles. The AAE (km·d) significantly (P < 0.0001) predicted AMS in the model. For every 1 km·d increase in AAE, the odds of getting sick decreased by 41.3%. Equivalently, for every 1 km·d decrease in AAE, the odds of getting sick increased by 70.4%. Age and BMI were not significant predictors. The model demonstrated excellent discrimination (AUC = 0.83 (95% CI = 0.79-0.91) and calibration (Hosmer-Lemeshow = 0.11). The model provides a priori estimates of altitude acclimatization status resulting from the use of various rapid, staged, and graded ascent profiles.