“You can leave your mask on”: effects on cardiopulmonary parameters of different airway protective masks at rest and during maximal exercise

Discussion

In this experimental study on healthy subjects, we demonstrated how the use of protective masks (both surgical and FFP2 masks) is associated with a significant worsening of FEV₁ and FVC and of cardiorespiratory parameters both at rest and at peak exercise. At rest, V˙_O2, V˙_CO2 and V˙_E decreased, the latter due to a RR reduction which was paralleled by an increase in t_I. At peak exercise, an increase in dyspnoea and a reduction in peak V˙_O2 measured during a standardised maximal effort at CPET were observed. The effect was predominantly driven by a reduction in V˙_E. Our data suggest that V˙_E is reduced owing to a decrease in both RR and V_T along the three conditions, with a parallel increase in t_I.

During respiratory virus outbreaks, such as the current COVID-19 pandemic, protective masks, together with social distancing, have proven to be essential devices for the containing infection, both in everyday life and in hospitals, where health workers most often use FFP2 masks [8, 17, 18]. Breathing discomfort in members of the general population, who are not accustomed to daily mask use, has been frequently reported by word of mouth and social networks [19], even becoming a potentially dangerous political statement for some [20]. Several peer-reviewed publications have demonstrated the minimal clinical impact of wearing protective masks, both in healthy subjects and in patients with respiratory diseases (e.g. chronic obstructive pulmonary disease), although in the absence of accurate cardiorespiratory parameters [21–23]. Regardless, in the public opinion, the long-term practicability and tolerability of protective masks are still questioned, and, in spite of their mandatory use, they are frequently worn below the nose and mouth, rendering them useless.

In the present study, we selected 12 healthy subjects who were already familiar with CPET in our laboratory. The subjects were representative of the general population; indeed, their observed peak exercise V˙_O2 with no mask was 101% of the predicted value. Masks slightly affected the breathing pattern even at rest, as indicated by the light breathing discomfort experienced by some subjects (figure 2). A similar observation was made in two previous reports [24, 25]. We performed spirometry with a commercial CPET mask that had been demonstrated as adequate in previous studies [26, 27]. In the present study, spirometry with different masks showed a reduction in FEV₁ and FVC that was paralleled by a longer t_I [25]. This datum is consistent with an increased resistance to airflow during inspiration [25, 28]. Of note, some voluntary hyperventilation at rest was present in all conditions as shown by the relatively high RR and P_ETO2 values. However, a progressive reduction of V˙_E, V˙_O2, V˙_CO2 and P_ETO2 was observed from no mask to surgical mask to FFP2 mask, revealing an involuntary adjustment of V˙_E to the variable flow resistance. Given the rapidity of this ventilatory adaptation (the masks were worn just before starting the CPET), these changes could be explained by a rapid adaptation of the chemoreceptor to tolerate higher arterial CO₂ and lower O₂ values. To summarise the data at rest: 1) subjects spontaneously adapted to the increased airflow resistance by reducing V˙_E, P_ETO2, V˙_O2 and V˙_CO2 with an unchanged RER, suggesting rapid chemoreceptor response; 2) although RER was in the normal range for resting condition and sufficient resting time was allowed, some voluntary hyperventilation was present in all conditions as shown by RR and P_ETO2; 3) the involuntarily altered breathing pattern was hampered by increased airflow resistance. Indeed, subjects responded to this resistance by self-adjusting t_I and t_E, likely to minimise their respiratory effort. In other words, given the increased cost of breathing while wearing protective masks, healthy subjects trigger an innate mechanism by maintaining their V˙_E to a lower set point.

Protective masks did not affect gas exchange kinetics pattern, given that the V˙_E/V˙_CO2 slope and V˙_O2/work relationships were unchanged. However, a slight anticipation of the threshold metabolism was observed with protective masks. It must be underlined that we, as well as others [24, 25, 28], studied the effects on respiration of various types of surgical masks when worn under the standard CPET silicon mask both during spirometry and CPET. This was necessary to allow respiratory gas measurements. It is possible, but unknown, that the silicone mask per se influenced the respiratory function, albeit minimally.

The reported index of dyspnoea at peak exercise showed a clear worsening with mask wearing (figure 2). At peak exercise, with the different types of masks V˙_E showed a greater decrease than V˙_O2 and V˙_CO2, a datum accompanied by an increased t_I and again suggestive of increased resistance to airflow. Of note, true ventilatory limitation, as assessed by applying the standards for exercise limitation during CPET, was not observed [29]. Indeed, peak exercise breathing reserve, as measured by (FEV₁×35)–observed V˙_E [30] was always >20 L·min⁻¹, being 45.5±25.9 L·min⁻¹ with no mask, 61.6±19.5 L·min⁻¹ with a surgical mask and 66.1±22.9 L·min⁻¹ with an FFP2 mask. Increased resistance of the masks was also shown by the spirometry data. The reduced V˙_E at peak exercise could be due to an increased airway opening resistance reducing ventilatory capacity, which would then lead to dyspnoea and reduced performance. The reduction of S_aO2, V˙_E and V_T and the increase in t_I altogether support a V˙_E-mediated effect of masks on exercise performance, which was clearly reduced as shown by the lower peak V˙_O2 and workload achieved. We did not observe signs of respiratory fatigue, as shown by an unchanged peak exercise MIP and MEP, or ventilatory limitation to exercise. Unfortunately, flow/volume curves during exercise were not performed to avoid any possible interference of these respiratory manoeuvres with peak exercise performance. Because our results were obtained in a population of middle-aged healthy subjects, more studies are needed to assess the cardiorespiratory effects of various protective masks on exercise performance in older subjects or in patients with proven exercise limitation. Finally, we analysed the effects of masks using a maximal workload incremental protocol aimed at achieving peak exercise in ∼10 min performed in a temperature and humidity controlled laboratory located at sea level. This is the gold standard for maximal exercise performance evaluation [12, 31]. However, efforts performed with different exercise protocols (e.g. as for during daily life activities [32]), or in different ambient conditions (i.e. temperature, humidity or at altitude) may produce different results. Indeed, in these conditions the effects of protective masks are unknown.

In conclusion, the use of protective masks in healthy middle-aged subjects 1) slightly influences cardiorespiratory variables at rest and during exercise, and 2) reduces peak V˙_O2 by ∼10% due to an increase in airflow resistance, although V˙_E limitation was far from being reached. Accordingly, the general population should be aware that the use of protective masks in healthy subjects is associated with modest respiratory discomfort but their use is safe even during maximal exercise, albeit with a slight reduction in performance.