Effects of Wearing Facemasks During Brisk Walks: A COVID-19 Dilemma

Ophir Bar-On, MD; Yulia Gendler, PhD; Patrick Stafler, MD; Hagit Levine, MD; Guy Steuer, MD; Einat Shmueli, MD; Dario Prais, MD; Meir Mei-Zahav, MD


J Am Board Fam Med. 2021;34(4):798-801. 

In This Article


Healthy adult volunteers were recruited. Vital signs, oxygenation, and ventilation were monitored under 3 different conditions: (a) at rest, (b) during a slow walk (4 km/h), and (c) during a brisk walk (7 km/h), with the latter 2 on a standard treadmill at 0° inclination. Each segment lasted 5 minutes, and each was performed twice—in random order—once without, and once with, a standard disposable surgical facemask (Non-Woven 3-ply FaceMask, LeJian Protective Equipment, China). A 5-minute recovery period was allocated between segments. The room settings were 25°C with 48% humidity.

Using a Welch Allyn 300 Vital Signs Monitor, we documented heart rate and oxygen saturation continuously; carbon dioxide was measured end-tidally (EtCO2) using LifeSense II (Nonin Medical Inc., Minneapolis, USA) via nasal prongs. This device also measured respiratory rate. Values were registered at 0 minutes, 2 minutes, 4 minutes, 5 minutes, and at recovery end.

Normal CO2 was defined as 35–45 mmHg; Normal oxygen saturation was defined as ≥97%.[5,6]

Participants rated their subjective difficulty using Borg's Exertional Scale,[7] an established numeric tool (range, 1 to 10), measuring the combined feeling of effort and exertion, breathlessness, and fatigue during physical work. Furthermore, participants graded their individual sensations, comparing the same activity, with versus without a mask, namely: difficulty breathing, shortness of breath, "choking" feeling, headache, dizziness, and weakness.

Data were analyzed using SPSS, version 25 (SPSS Inc; Armonk, NY). Results are presented as mean±standard deviation (SD) or median+range according to variable distribution. For comparison of parameters measured with versus without a mask, paired-samples t -test was used when normal distribution was assumed; otherwise, a Wilcoxon rank-sum test was used. A general linear model was used for repeated measurements analysis. Sample size was calculated using power analysis (α=0.05, 1-β=0.8). The study was approved by the organization's Institutional Research Board (RMC0325-20), and participants signed informed consent before entering the study.