The study was conducted at the Department of Cardiology, University of Leipzig. The 12 active and healthy male volunteers were recruited from medical staff. Subjects with cardiac, pulmonary or inflammatory diseases or any other medical contraindications were not included. The characteristics of the participants are shown in Table 1. The study was conducted in accordance with the latest revision of the Declaration of Helsinki and was approved by the Ethical Committee of the Medical Faculty, University of Leipzig (reference number 088/18-ek). Written informed consent was obtained from all the participants.

Medical history was taken using a questionnaire. Subjects received physical examination and vital parameters, body measurements and a resting electrocardiogram (ECG). Each subject performed three incremental exertion tests (IET), one “no mask” (nm), one with surgical mask (sm) and one with FFP2/N95 mask (ffpm). The order of the masks worn was randomly assigned using the GraphPad Quickcalcs online randomization tool [13]. Tests were performed at the same time of day with a minimum of 48 h between two tests. To assess baseline respiratory function, spirometry for each setting (nm, sm, ffpm) was performed. The participants were blinded with regard to their respective test results to avoid influence by an anticipation bias. Statistical analysis was performed by an independent and fully blinded scientist who was not involved in the conduction of the tests.

IET were performed on a semi-recumbent ergometer (GE eBike, GE Healthcare GmbH, Solingen, Germany, Germany) at a constant speed of 60–70 revolutions per minute (rpm). The test began at a workload of 50 W with an increase of 50 W within 3 min (as a ramp) until voluntary exhaustion occurred. Each subject continued for an additional 10-min recovery period at a workload of 25 W.

We used typical and widely used disposable FFP2/N95 protective face masks (Shaoguan Taijie Protection Technology Co., Ltd., Gao Jie, China) and surgical masks (Suavel^® Protec Plus, Meditrade, Kiefersfelden, Germany), both with earloops.

The spirometry mask was placed over the fm and fixed with head straps in a leak-proof manner (see Fig. 1A1, B1). After fitting the spirometry mask, subjects performed (a) inspiration and (b) expiration with maximal force. During both maneuvers, the valve of the mask was closed leading to abrupt stop of the air flow (see Fig. 1A2, B2). The fitting was carefully checked for the absence of any acoustic, sensory or visual indication of leakage (e.g., lifting of the mask, whistling or lateral airflow) by the investigators and the test person. The correct fitting and leak tightness were confirmed before each test was started.

Cardiac output (CO), stroke volume (SV) (measured by impedance cardiography; Physioflow, Manatec Biomedical, Macheren, France), heart rate (HR) (GE-Cardiosoft, GE Healthcare GmbH, Solingen, Germany), maximum oxygen consumption (VO_2max) and minute ventilation (VE) were monitored continuously at rest, during IET and during recovery. Lung function and spirometry data were collected through a digital spirometer (Vyntus™ CPX, Vyaire Germany, Hoechberg, Germany). For each modality (nm, sm, ffpm), data of three expiratory maneuvers with 1‐min intervals were collected using the best values obtained for maximum forced vital capacity (FVC), forced expiratory volume in 1st second (FEV1), peak expiratory flow (PEF) and Tiffeneau index (TIFF). The arterio-venous oxygen difference was computed using Fick’s principle with avDO₂ = VO₂/CO. Cardiac work (CW) was measured in joules (J) and calculated according to the formula CW = SV (in m³) × SBP (in Pa). Capillary blood samples (55 µl) were taken from the earlobe at baseline and immediately after cessation of maximum load and analyzed (ABL90 FLEX blood gas analyzer, Radiometer GmbH, Krefeld, Germany). Blood pressures (BP) was observed at rest, every 3 min during the IET and after the first 5 min of recovery period.

We used a published questionnaire published by [14] to quantify the following ten domains of comfort/discomfort of wearing a mask: humidity, heat, breathing resistance, itchiness, tightness, saltiness, feeling unfit, odor, fatigue, and overall discomfort. The participants were asked 10 min after each IET how they perceived the comfort in the test.

All values are expressed as means and standard deviations unless otherwise stated, and the significance level was defined as p < 0.05. Data were analyzed using Microsoft Office Excel^® 2010 for Windows (Microsoft Corporation, Redmond, Washington, USA) and GraphPad Prism 8 (GraphPad Software Inc., California, USA). For distribution analysis, the D’Agostino–Pearson normality test was used. For normal distribution, comparisons were made using one-way repeated measures ANOVA with Turkey’s post hoc test for multiple comparisons. Otherwise, the Friedman non-parametric test and Dunn’s post hoc test were used. The study was powered to detect a difference of 10% in VO_2max/kg between nm and ffpm.

The results of the pulmonary function tests are shown in Table 2. Both sm and ffpm significantly reduce the dynamic lung parameters. The average reduction of FVC was −8.8 ± 6.0% with sm and −12.6 ± 6.5% with ffpm. FEV1 was −7.6 ± 5.0% lower with sm and −13.0 ± 9.0% with ffpm compared to no mask. The peak flow measurement showed that both sm and ffpm significantly reduced the PEF (−9.7 ± 11.2% and −21.3 ± 12.4%, respectively).

The results of IET under different conditions are depicted in Table 3. None of the masks had impact on the examined parameters under resting condition. The average duration of IET compared to the test without mask was slightly decreased by −29 ± 40 s with sm (p = 0.07) and significantly decreased by −52 ± 45 s with ffpm (p = 0.005). Under maximum load, there was a large reduction of the performance measures Pmax and VO_2max, especially with ffpm (Fig. 2). Furthermore, these parameters were significantly reduced in ffpm compared to sm.

Assessment of the hemodynamic parameters (Table 3) showed that ffpm decreased avDO₂ by 16.7 ± 11.2% compared to nm. Stroke volume and cardiac output and cardiac work did not differ significantly (nm: 4.3 ± 0.8 J, sm: 4.7 ± 1.4 J, ffpm: 4.6 ± 0.9 J; p = 0.29).

The masks showed a marked effect on pulmonary parameters: VE for both sm and ffpm was significantly reduced by −12.0 ± 12.6% and −23.1 ± 13.6%, respectively, compared to nm (see Table 3; Fig. 1). Compared to nm, tests with ffpm showed a significant reduction in breathing frequency with an additional decrease in tidal volume (−9.9 ± 11.3% and −14.4 ± 13.0%, respectively). At the same time, a longer inhalation time was observed (sm: 12 ± 15%, p = 0.043; ffpm: 19 ± 16%, p = 0.005). There were no differences in exhalation time.

Measurements of the metabolic parameters pH, PCO₂, PO₂ and lactate and the heart rate recovery did not differ significantly between the three tests (Table 3).

Subjective ratings for different sensations and overall discomfort for sm and ffpm compared to nm are depicted in Table 4. In general, the negative ratings for all items of discomfort increased consistently and significantly from sm to ffpm. There were several-fold negative reports for the ffpm compared to nm and sm for breathing resistance. The relative aggravation in overall discomfort compared to the standard procedure for spiroergometric tests is shown in Fig. 2.

Spirometry showed reduced FVC, FEV1 and PEF with the surgical mask and even greater impairments with the FFP2/N95 mask. Wearing the FFP2/N95 mask resulted in a reduction of VO_2max by 13% and of ventilation by 23%. These changes are consistent with an increased airway resistance [15]. Studies testing increased upper airway obstruction induced by added resistance at the mouth report similar effects on the lung functions parameter with increased breathing resistance [16]. The reduction in ventilation resulted from a lower breathing frequency with corresponding changes of the inhaling and exhaling time and a reduced tidal volume. This is in agreement with the effects of respiratory protective devices or additional external breathing resistance [16, 17]. The increased breathing resistance, which is likely higher during stress, leads to an elevated breathing work and a limitation of the ventilation. The data of this study are obtained in healthy young volunteers, the impairment is likely to be significantly greater, e.g., in patients with obstructive pulmonary diseases [18]. From our data, we conclude that wearing a medical face mask has a significant impact on pulmonary parameters both at rest and during maximal exercise in healthy adults.

Increased breathing resistance in ffpm and sm requires more work of the respiratory muscles compared to nm leading to higher oxygen consumption. Additionally, a significant proportion of cardiac output is directed via different mechanisms, e.g., sympathetically induced vasoconstriction, to the respiratory musculature [19]. Furthermore, the increased breathing resistance may augment and prolong inspiratory activity leading to more negative intrathoracic pressure (ITP) for longer durations. This assumption is supported by the findings on inspiration times which were higher while wearing a fm. Prolonged and more negative ITP increases the cardiac preload and may lead to higher SV at the one hand which is consistent with our results showing a statistical trend towards higher SV while wearing ffpm or sm [20, 21]. In addition, cardiac afterload increases because of an increased transmural left-ventricular pressure resulting in enhanced myocardial oxygen consumption [22]. In these healthy volunteers, functional cardiac parameters do not differ significantly at baseline, at maximal load and during recovery. However, there is a non-significant trend towards a higher cardiac work (Joule) compared to the test without mask. This is of relevance since significantly less watts (−5%) was achieved in the tests with masks. The relation of cardiac power to the total power is approximately 10% lower with ffpm. These data suggest a myocardial compensation for the pulmonary limitation in the healthy volunteers. In patients with impaired myocardial function, this compensation may not be possible.

The measurements show that surgical masks, and to a greater extent FFP2/N95 masks, reduce the maximum power. P_max (Watt) depends on energy consumption and the maximum oxygen uptake (VO_2max). The effect of the masks was most pronounced on VO_2max. Since the cardiac output was similar between the conditions, the reduction of P_max was primarily driven by the observed reduction of the arterio-venous oxygen content (avDO₂). Therefore, the primary effect of the face masks on physical performance in healthy individuals is driven by the reduction of pulmonary function. In addition, the auxiliary breathing muscles have been described to induce an additional afferent drive which can contribute to an increase of the fatigue effect [23‐25].

The performance of several different populations may be significantly affected by face masks. For athletes the use of fm will reduce physical performance. Less pronounced but mechanistically similar effects have been observed for mouthguards [26‐28]. The increased breathing resistance is especially problematic for patients with chronic obstructive pulmonary diseases. Patients with diffusion disorders have reduced capacity to compensate due to the reduced tidal volume. Another example of a population at risk is patients with heart failure. The observed mechanisms may lead to more severe symptoms in individuals with impaired capacity for myocardial compensation.

Health care professionals and others are faced with significant psychological distress during viral outbreaks [29]. Measures to maintain the quality of life both during emergency situations and long term care are increasingly important. Adequate personal protective equipment and adequate rest are considered keys to reduce the risk of adverse psychological outcomes [29]. Our sample primarily consisted of physicians working at a university hospital who are very familiar with medical masks and have a positive attitude towards personal protection. Our data show that FM leads to severe subjective discomfort during exercise. FFP2/N95 masks are perceived as more uncomfortable than sm. In particular, breathing resistance, heat, tightness and overall discomfort are the items with the greatest influence on subjective perception. This finding is in agreement with the literature [14, 30]. Wearing of fm is perceived as subjectively disturbing and is accompanied by an increased perception of exertion. It is likely that the masks negatively impact on the dynamics of perception especially at the limit of exercise tolerance [31, 32]. In addition to the severe impact on ventilation, the data suggest the associated discomfort as a second important reason for the observed impairment of physical performance.