Background

The coronavirus-19 (COVID-19) pandemic has had a long-term and profound effect on people’s lives, studies, work, and social activities around the world. Wearing face masks, which prevent the transmission of infectious respiratory pathogens, has become a major strategy to prevent its spread [1,2].

Some clinical trials have investigated the effects of wearing face masks on subjects’ performance during exercise. Fikenzer et al conducted repetitive cardiopulmonary exercise tests on 12 healthy male subjects in a randomized order: no mask, surgical mask, and an FFP2/N95 mask; the forced expiratory volume and peak expiratory flow of the subjects decreased during the lung function tests, while peak power and ventilation also decreased during the cardiopulmonary exercise test [3]. Shaw et al showed that wearing a face mask during vigorous exercise had no discernable detrimental effect on blood or muscle oxygenation or exercise performance in young, healthy participants [4]. The conclusions of these studies were based on comparisons of maximum exercise capacity, which does not necessarily represent the actual exercise conditions during a routine exercise.

However, there is a lack of studies on continuous exercise capability (anaerobic threshold), which we believe is the crux of the issue. Any physical activity that exceeds the anaerobic threshold can lead to anaerobic metabolism and unsustainable exercise. Continuous physical activity is only possible if the exercise intensity does not exceed the anaerobic threshold. According to previous studies, the endurance threshold for an 8-h shift of heavy manual occupational labor is 33%–40% of an individual’s maximum oxygen uptake [5]. Even in physically highly demanding occupational tasks, the ventilatory threshold of an individual was only exceeded for short periods at 35–69% of maximal oxygen uptake [6]. To study the harm caused by wearing a mask during exercise, it is important to determine an individual’s ventilatory threshold. However, there are limited studies on the effect of exercising with a mask on the ventilatory anaerobic threshold.

Two separate incidents of the sudden death of students wearing face masks in China have raised concerns about whether wearing face masks during exercise results in physiological changes that endanger human life. Considering that wearing masks is thought to prevent oxygen inhalation, increase repetitive inhalation of carbon dioxide, and increase work of breathing [7], we hypothesized that wearing face masks might advance both the peak and ventilatory anaerobic threshold (VAT) during cardiopulmonary exercise testing. This study aimed to investigate the effect of wearing face masks on exercise performance and the VAT in healthy young subjects through a cardiopulmonary exercise test (CPET), as well as the mechanisms involved.

Material and Methods

SUBJECTS:

A total of 34 young, healthy volunteers were recruited from the medical staff of our Rehabilitation Department. The inclusion criteria were: 1) 18–30 years of age, 2) healthy and without cardiopulmonary diseases, 3) not using drugs that affect heart rate, and 4) signed the informed consent form. The exclusion criteria were: 1) any signs of hemodynamic instability during exercise, such as pale complexion and decreased blood pressure, or 2) unable to complete the cycling task for any reason.

STUDY DESIGN:

This study was registered at http://www.chictr.org.cn/ (ChiCTR2000034457) and approved by the Ethics Committee of Nanjing First Hospital (KY20210125-02). The study was conducted as a randomized, counterbalanced, cross-over trial. Study participants were randomized to perform an incremental cycle ergometer exercise test wearing a surgical face mask or not on 2 separate occasions, with at least 48 h between 2 separate occasions. Randomization was performed using a computerized random number generator. Subjects with odd number were first subjected to cardiopulmonary exercise test without mask, and subjects with even number were first subjected to cardiopulmonary exercise test with mask. All participants wore the same brand and model of surgical face masks.

CARDIOPULMONARY EXERCISE TEST (CPET) PROTOCOL: All CPETs were performed by the same operator, who has been a CPET operator at the Rehabilitation Department for >10 years. The fit of the surgical face mask beneath the Schiller mask was standardized; the correct fit was confirmed by exhaling with maximal force before each test to identify air leakage. At the beginning of each exercise test, subjects rested for 3 minutes, then performed warm-up exercise for 3 minutes (unload cycling), followed by a ramp incremental exercise stage until voluntary exhaustion occurred. Each subject continued an additional 5-min recovery period at a work rate of 20 W. The increase in the work rate varied from 10 to 25 W/min according to the estimated exercise capability of each subject and ensured that the subject reached the peak within 10 minutes. The power increasing scheme for each subject’s 2 exercise tests was the same so that the detection of exercise peak and the anaerobic threshold were not affected. Gas exchange was measured using a breath-by-breath instrument (Schiller AT-104) with a mask with 80 to 100 mL of dead space. All values were averaged across 10-second epochs. At the end of each CPET test, the physician estimated the location of the subject’s VAT. At the end of all tests, the location of the VAT for each exercise test was re-verified by another physician. The VAT point of each test was determined using the V-slope method [8] and other ventilatory parameters, including the VE/VCO2 nadir, PETCO2 response, and RER transition point.

COLLECTED INDICATORS:

General information about each subject was collected, and all indicators from the 2 exercise tests were recorded, with or without wearing a face mask, including work rate (WR), oxygen uptake (VO 2 ), heart rate (HR), ventilation per minute (VE), tidal volume (VT), breath frequency (BF), dead space ratio (VD/VT), carbon dioxide ventilation equivalent (VE/VCO 2 ), and end-tidal CO 2 pressure (PETCO 2 ). The VAT was denoted by work rates and oxygen uptake, abbreviated as [email protected] and VO 2 @VAT, respectively.

STATISTICAL ANALYSIS:

R version 4.0.3 (R version 4.0.3 (2020-10-10) – “Bunny-Wunnies Freak Out” Copyright (C) 2020 The R Foundation for Statistical Computing Platform: https://www.r-project.org/ ; x86_64-w64-mingw32/x64 (64-bit)) was used for statistical analysis. Continuous variables were represented as mean ± standard deviation. Categorical variables were expressed as frequencies and percentages. The paired t test was used to compare data from 2 exercise tests with and without masks. For correlation analysis, Pearson correlation analysis was used if both variables had normal distribution; otherwise, the Spearman correlation analysis was used. After grouping the participants, the group t test was used to compare the 2 groups. The cor.test() function in R was used for correlation analysis and test. The glm() function was used to fit the logistic regression model. The step() function was used for variable screening, and the model was automatically selected according to the Akaike information criterion (AIC) (stepwise regression method). P <0.05 was considered statistically significant. The graphs were drawn using GraphPad PRISM 5.01 for Windows (GraphPad Software, Inc., San Diego, CA, USA).

Results

DEMOGRAPHIC AND BASELINE DATA: A total of 34 young, healthy volunteers were enrolled, consisting of 18 men and 16 women (Figure 1). All 34 participants completed the 2 CPETs. The main characteristics were an age of 23.4±3.6 years, a height of 170.3±9.7 cm, a weight of 63.3±13.5 kg, and a BMI of 21.7±3.7 kg·m−2. The physiological indexes at rest included heart rate of 87.7±12.5 bpm, systolic blood pressure of 114.8±17.8 mmHg, and diastolic blood pressure of 77.0±9.6 mmHg. All pulmonary function test indexes were good (Table 1).

COMPARISON OF PEAK INDEXES AND VAT INDEXES BETWEEN 2 EXERCISE TESTS WITH AND WITHOUT A MASK: As shown in Table 2, the peak indexes of the cardiopulmonary exercise test while wearing face masks, such as WR, VO2/kg, HR, VE, and VE/VCO2, were significantly lower than those in the cardiopulmonary exercise test without a face mask (all P<0.05). However, the PETCO2 was significantly higher than those in the cardiopulmonary exercise test without a face mask (P=0.004). The VAT indexes of the cardiopulmonary exercise test while wearing face masks, such as WR, HR, VE, BF, VD/VT, and VE/VCO2, were significantly lower than those in the cardiopulmonary exercise test without a face mask (all P<0.05).

: There was a positive linear correlation between δ[email protected] and δ[email protected] (r=0.495, P=0.003, Figure 2).

SUBGROUP ANALYSIS: According to the aggregation characteristics of scattered points, δ[email protected]=−6 watts at the X-axis position was selected as the vertical dividing line, and the 34 subjects were divided into 2 groups: the stable VAT group (VS group) and the advanced VAT group (VA group). The subgroup analysis of the VAT indicators showed that all subjects in the VS group and VA group had similar VAT indicators in the exercise test without a mask. Compared with the VS group, WR, VO2/kg, and VE were significantly decreased in the VA group after wearing masks (all P<0.05) (Table 3).

As shown in Table 4, subjects in the VA group demonstrated a significant difference in δHR, δVE, δVT, δVE/VCO2, and δPETCO2 compared with subjects in the VS group (all P<0.05). All demographics and physiological characteristics were not significantly different between the 2 groups. Multivariable logistic regression analysis revealed that the factors independently associated with decreased VAT were δVE, δBF, and δVE/VCO2 (Table 5).

Discussion

LIMITATIONS:

Although the tightness test of the mask was assessed in a calm state, we could not guarantee the absence of gas leakage during the exercise test. Due to many missing values, systolic blood pressure data could not be provided, which is important because it may be closely related to the risk of wearing masks. However, we did detect a decrease in systolic blood pressure in many subjects, whether at peak or VAT. When portable finger pulse oximetry was added, SpO 2 did not significantly decrease during the exercise test while wearing a mask, and in some cases, it even increased. Since SpO 2 values cannot be recorded synchronously in real-time, we could not provide SpO 2 values at each observation time point. Blinding was not possible for the participant nor any study personnel in the same room as the participant. Finally, no body composition tests were performed, and follow-up studies will be conducted.

Conclusions

Our data show that wearing surgical masks causes ventilatory anaerobic threshold to occur earlier in some healthy young subjects during cardiopulmonary exercise test. Whether VAT occurs in advance is related to the changes in VE, BF, and VE/VCO2.