Bayesian network meta-analysis of face masks' impact on human physiology

Several concerns regarding the safety of face masks use have been propounded in public opinion. The objective of this review is to examine if these concerns find support in the literature by providing a comprehensive overview of physiological responses to the use of face masks. We have performed a systematic review, pairwise and network meta-analyses to investigate physiological responses to the use of face masks. The study has been registered with PROSPERO (C RD42020224791). Obtained results were screened using our exclusion and inclusion criteria. Meta-analyses were performed using the GeMTC and meta R packages. We have identified 26 studies meeting our inclusion and exclusion criteria, encompassing 751 participants. The use of face masks was not associated with significant changes in pulsoxymetrically measured oxygen saturation, even during maximal-effort exercises. The only significant physiological responses to the use of face masks during low-intensity activities were a slight increase in heart rate, mildly elevated partial pressure of carbon dioxide (not meeting criteria for hypercarbia), increased temperature of facial skin covered by the mask, and subsequent increase of the score in the rating of heat perception, with N95 filtering facepiece respirators having a greater effect than surgical masks. In high-intensity conditions, the use of face masks was associated with decreased oxygen uptake, ventilation, and RR. Face masks are safe to use and do not cause significant alterations in human physiology. The increase in heart rate stems most likely from increased respiratory work required to overcome breathing resistance. The increase in carbon dioxide is too small to be clinically relevant. An increased rating of heat perception when using face masks results from higher temperature of facial skin covered by the mask.