Smartwatch-Based Maximum Oxygen Consumption Measurement for Predicting Acute Mountain Sickness: Diagnostic Accuracy Evaluation Study (Preprint)

Abstract

Cardiorespiratory fitness plays an important role in coping with hypoxic stress at high altitudes. However, the association of cardiorespiratory fitness with the development of acute mountain sickness (AMS) has not yet been evaluated. Wearable technology devices provide a feasible assessment of cardiorespiratory fitness, which is quantifiable as maximum oxygen consumption (VO₂max) and may contribute to AMS prediction.

We aimed to determine the validity of VO₂max estimated by the smartwatch test (SWT), which can be self-administered, in order to overcome the limitations of clinical VO₂max measurements. We also aimed to evaluate the performance of a VO₂max-SWT–based model in predicting susceptibility to AMS.

Both SWT and cardiopulmonary exercise test (CPET) were performed for VO₂max measurements in 46 healthy participants at low altitude (300 m) and in 41 of them at high altitude (3900 m). The characteristics of the red blood cells and hemoglobin levels in all the participants were analyzed by routine blood examination before the exercise tests. The Bland-Altman method was used for bias and precision assessment. Multivariate logistic regression was performed to analyze the correlation between AMS and the candidate variables. A receiver operating characteristic curve was used to evaluate the efficacy of VO₂max in predicting AMS.

VO₂max decreased after acute high altitude exposure, as measured by CPET (25.20 [SD 6.46] vs 30.17 [SD 5.01] at low altitude; <i>P</i>&lt;.001) and SWT (26.17 [SD 6.71] vs 31.28 [SD 5.17] at low altitude; <i>P</i>&lt;.001). Both at low and high altitudes, VO₂max was slightly overestimated by SWT but had considerable accuracy as the mean absolute percentage error (&lt;7%) and mean absolute error (&lt;2 mL·kg^–1·min^–1), with a relatively small bias compared with VO₂max-CPET. Twenty of the 46 participants developed AMS at 3900 m, and their VO₂max was significantly lower than that of those without AMS (CPET: 27.80 [SD 4.55] vs 32.00 [SD 4.64], respectively; <i>P</i>=.004; SWT: 28.00 [IQR 25.25-32.00] vs 32.00 [IQR 30.00-37.00], respectively; <i>P</i>=.001). VO₂max-CPET, VO₂max-SWT, and red blood cell distribution width-coefficient of variation (RDW-CV) were found to be independent predictors of AMS. To increase the prediction accuracy, we used combination models. The combination of VO₂max-SWT and RDW-CV showed the largest area under the curve for all parameters and models, which increased the area under the curve from 0.785 for VO₂max-SWT alone to 0.839.

Our study demonstrates that the smartwatch device can be a feasible approach for estimating VO₂max. In both low and high altitudes, VO₂max-SWT showed a systematic bias toward a calibration point, slightly overestimating the proper VO₂max when investigated in healthy participants. The SWT-based VO₂max at low altitude is an effective indicator of AMS and helps to better identify susceptible individuals following acute high-altitude exposure, particularly by combining the RDW-CV at low altitude.

Chinese Clinical Trial Registry ChiCTR2200059900; https://www.chictr.org.cn/showproj.html?proj=170253