[Skip to Navigation]
Sign In
Figure.  Scatterplot of Carbon Dioxide Content in Inhaled Air Under Filtering Facepiece Mask by Age
Scatterplot of Carbon Dioxide Content in Inhaled Air Under Filtering Facepiece Mask by Age

Linear regression line with locally weighted scatterplot smoothing.

Table.  Carbon Dioxide Values Under Various Conditions
Carbon Dioxide Values Under Various Conditions
Xiao  J, Shiu  EYC, Gao  H,  et al.  Nonpharmaceutical measures for pandemic influenza in nonhealthcare settings —personal protective and environmental measures.   Emerg Infect Dis. 2020;26(5):967-975. doi:10.3201/eid2605.190994 PubMedGoogle ScholarCrossref
Matuschek  C, Moll  F, Fangerau  H,  et al.  Face masks: benefits and risks during the COVID-19 crisis.   Eur J Med Res. 2020;25(1):32. doi:10.1186/s40001-020-00430-5PubMedGoogle ScholarCrossref
Schwarz  S, Jenetzky  E, Krafft  H, Maurer  T, Martin  D. Corona children studies “Co-Ki”: first results of a Germany-wide registry on mouth and nose covering (mask) in children. Published 2021. Accessed June 15, 2021. https://www.researchsquare.com/article/rs-124394/v1
Mitteilungen der Ad-hoc-Arbeitsgruppe Innenraumrichtwerte der Innenraumlufthygiene-Kommission des Umweltbundesamtes und der Obersten Landesgesundheitsbehörden.  [Health evaluation of carbon dioxide in indoor air].   Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz.  2008;51(11):1358-1369. doi:10.1007/s00103-008-0707-2PubMedGoogle ScholarCrossref
Walach  H, Weikl  R, Traindl  H,  et al. Is carbon dioxide content under nose-mouth covering in children without potential risks? a measurement study in healthy children. Published April 14, 2021. Accessed June 15, 2021. https://osf.io/yh97a/?view_only=df003592db5c4bd1ab183dad8a71834f
Kisielinski  K, Giboni  P, Prescher  A,  et al.  Is a mask that covers the mouth and nose free from undesirable side effects in everyday use and free of potential hazards?   Int J Environ Res Public Health. 2021;18(8):4344. doi:10.3390/ijerph18084344 PubMedGoogle ScholarCrossref
20 Comments for this article
Unsuitable CO2 meassurement device
Alejandro Keller, PhD in Natural Sciences | University of Applied Sciences and Arts Northwestern Switzerland
The authors use an CO2 incubator analyzer (0-20%) that has an accuracy of 1% of the range (i.e. 2000ppm) and a time response T^90<=20 seconds for CO2. This device is unsuitable for measuring the transient concentrations during the respiration processes. A normal respiration cycle has a duration of 3 to 4 seconds. Thus, it is impossible to separate the concentration of CO2 in inhaled and exhaled air using this device. This also explains why the authors measure average concentrations of around 2700ppm-CO2, way above the ambient value of 740ppm, even when no mask is present.

The authors refer to
the dead volume behind the mask as the main problem. The relevant data would be the comparison between this dead volume and the lung capacity and/or the volume of one respiration cycle. The lungs never collapse completely during respiration. Together with the rest of the respiratory airways, the respiratory system has also a dead volume that is much larger than the dead volume between the mask and the face. The comparison of these volume is of extreme importance for the discussion and may change the author's conclusions.
Simple Stratergies of Face Mask wear in Children
Harikrishnan Pandurangan, FDSRCS(Eng),FCrfOrth, PhD | Clinician Researcher, Craniofacial Orthodontist in Teeth"N"Jaws Center, Private practice in Chennai, India.
This paper brings out timely evidence on inhaled Carbon dioxide (CO2) in children due to face mask wearing in this Covid-19 pandemic.

The results show that there is a significant increase in the CO2 levels in inhaled air higher than advisable limits while wearing masks. This is true in adults as well and many feel tired or uncomfortable after wearing masks throughout the day or their work period.

We propose the following simple strategies for children wearing face masks to reduce inhaled CO2:

1. Use a single-layered medical or double-layered cloth face mask.
2. Increase the
air exchange ventilation in closed spaces like classrooms.
3.  Wear the face mask intermittently, like 30 minutes wear and 5-10 minutes off the mask with social distancing.
4. While off the mask, children can be encouraged to do minimal deep breathing exercises.
5. Limited time wearing of face mask in children who are asthmatic/respiratory problems, Nasal obstructions etc.
6. Avoid Double face mask or face mask and face shield in children.

We hope these simple cost-effective measures will be of benefit for children in this pandemic in all socio-economic settings.
Interested in the O2 data
Presten Witherspoon, DC | Private office
A review of the supplemental content appears to suggest that one of the measurements taken by researchers was the oxygen level of the blood. The concern with mask wearing among children is, realistically, two-fold: 1) that CO2 will be increased in the blood, and 2) that oxygen will be reduced in the blood.

While CO2 concentration is best assessed via bloodwork, which would be difficult to obtain from children in a timely manner after a short testing period, O2 concentration seems simpler data to collect. Current publication information does not list the O2 levels as measured by
the researchers. Adding that information to the study may provide further context for the effects of masks on blood chemistry beyond an observation that inhaled air increases in CO2 content underneath a child's mask.
What was measured doesn't matter
John Murphy, PhD CIH ROH | University of Toronto
The lack of methodological detail makes it unclear as to why the researchers concluded that they were measuring the concentration of inhaled CO2 as opposed to the concentration of exhaled CO2 or a mix of both. But the real question the author's don't clearly address is "Why does it matter?". The paper begins by stating "The question whether nose and mouth covering increases carbon dioxide in inhaled air is crucial.", and infer that it the reason is that the jurisdiction's regulatory limit for the concentration of CO2 in ambient indoor air is 2000 ppm, and that mask use results in overexposure in relation to that limit, which presumably is considered hazardous. Such a line of reasoning is mistaken. Indoor ambient CO2 limits are not set because CO2 itself is hazardous above those values, but because elevated CO2 correlates with elevated constellations of a range of indoor air pollutants that cause increasing occupant discomfort as levels rise. The CO2 indoor air standard is ultimately a ventilation effectiveness guideline, not a health-based exposure limit. It should also be recognized that any re-breathed CO2 is endogenous, and such air is qualitatively different from the ambient indoor air for which CO2 is solely a proxy indicator of composition, and to which the CO2 indoor air guidelines only apply.
Carbon dioxide and masking in healthy children
Sander Orent, M.D. | Physician consultant to fire/police services
I do toxicology work and came across an article years ago to which I wish I could reference that examined the effects of carbon dioxide levels on children in a middle school on cognition. They did this by generating carbon dioxide beneath one room to of about 500 ppm compared to atmospheric levels at the time were, as I recall, around 300 ppm. In any event they were able to show significant cognitive impairment in the classroom with the higher carbon dioxide levels. I think this finding would have direct bearing on your findings. I do however find myself conflicted about the balance of coronavirus infection risk to the child or others, especially with the emergence of the variants, with the risk of harm generated by elevated carbon dioxide levels in masked children. Especially important in light of some of the recent findings about the impact of coronavirus or the immune response to it on the brain. Thank you
Effects of increased CO2 concentrations
Michael Braun, PhD |
This study specifically for children plausibly follows the results of measurements for mask-waering adults by the " Landesagentur für Umwelt" in Bolzano, where a mean CO2 excess of about 4000 ppm was measured in the inhaled air :

The lower respiratory volume of children suggests a higher CO2 excess than observed at adults.

Several studies have investigated the effects of increased CO2 concentration.

In a study involving 10 subjects (Roberge et al., Physiological Impact of the N95 Filtering Facepiece Respirator on Healthcare Workers, Respiratory Care May 2010, 55 (5) 569-577, http://rc.rcjournal.com/content/55/5/569), it was found that mask-wearing
under light exertion led to a significant increase in blood CO2 levels in two of the 10 subjects (partial pressure of carbon dioxide of 50-52 mmHg instead of an average of about 42 mmHg in all subjects - values of 35-45 mmHg are considered normal).

A summary of the influence of carbon dioxide concentration in the atmosphere breathed in the range up to about 5,000 ppm can also be found in Azuma et al, Effects of low-level inhalation exposure to carbon dioxide in indoor environments: A short review on human health and psychomotor performance, Environment International, Volume 121, Part 1, December 2018, Pages 51-56, https://doi.org/10.1016/j.envint.2018.08.059

"A crossover experimental study conducted on 355 university students of four classrooms suggested that a 100 ppm increase in indoor CO2 concentration (range, 674-1450 ppm) was significantly associated with headache and this association was independent from other related indoor environmental factors, including temperature, relative humidity, and air exchange rate.... Exposures to CO2 at concentrations of 3000 ppm, including bioeffluents, significantly increased the intensity of reported headache, fatigue, and sleepiness."

An increased CO2 concentration, e.g. in indoor air, can therefore not only be regarded as a pure indicator for the presence of other harmful substances or germs. Harmful effects of increased CO2 concentrations themselves must also be taken into account.
Our Children Deserve Better Science
Eve Bloomgarden, MD | Northwestern University Chicago IL; Co-founder and COO of IMPACT (www.impact4hc.com)
Eve Bloomgarden,MD (Northwestern University, Co-founder IMPACT4hc), Elisabeth Marnik PhD (Husson University, author of Science Whiz Liz),
Alison Bernstein PhD (Michigan State University, Co-founder SciMoms and MommyPhD), Rebecca J. Heick, PhD (Augustana College, Author of Your Friendly Neighborhood Epidemiologist)

JAMA is a well respected peer reviewed journal. This research letter has extensive flaws and repercussions that may lead to very real harm for children. It is disappointing to see that this research letter was published in this journal.

First, this is written by individuals with known bias against masks and other non-pharmaceutical interventions against
COVID, as well as vaccinations. The lead author is a psychologist with no training pertaining to this area of research.

Second, there are extensive methodological issues. They used a G100 CO2 incubator analyzer, which is made for measuring CO2 levels in incubators. There is no data supporting the use of G100 as a valid and accurate instrument for the type of measurement used in this study. They also did not address the amount of dead space within the mask that could be further making their measurements inconsistent or unreliable. It is also unclear how they could reliably distinguish between inhaled and exhaled air using the described methods.

Third, their results presented are incomplete and difficult to interpret as many previous commenters have noted. Their trial protocol included in supplement 2 outlines that they would also collect blood oxygenation, heart rate and breathing frequency. Yet none of this data was included in this letter. Their results also show almost the same CO2 readings for both surgical masks and respirator FFP2 masks. If their findings were accurate you would expect a difference, given the differing filter capabilities. It is also unclear why their measurement of CO2 in baseline inhaled air is different than in ambient air.

Fourth, there was also no discussion regarding whether these results are clinically meaningful. They’re using flawed measurements obtained with a device that was not designed for this purpose and stating that this demonstrates harm. There were no actual health metrics reported, like pO2 or pCO2 and no discussion of the actual data needed to confirm clinical significance. These would be more reliable indicators of potential harmful effects on children.

Overall, this paper misinterprets inappropriately collected data leading to incorrect, distorted, and dangerous conclusions.

Lastly, we have real life evidence from millions of children who have been successfully wearing masks every day for months. We have extensive evidence that masking is an essential protective strategy to slow and prevent the transmission of SARS-CoV-2. If this study were accurate and reflective of the real world we would see adverse health events reported worldwide. We do not. We also have data from other peer reviewed papers that has shown no harmful effects. This is a study searching for a mechanism for a non-existent problem. The subject of this letter, regardless of conclusion, should have prompted intense scientific scrutiny prior to publication. The bell cannot be unrung, as this research letter is already being used as “scientific proof” that masks are harming our children, but a retraction should be strongly considered.
Lack of Outcome Comparisons
Matthew Loxton, MA Knowledge Management | None
The method states that "carbon dioxide content in inhaled air" was measured, but this is not an accurate statement since the amount of gas in the mask region is a small fraction of tidal volume, and the method of measurement is sensitive to the respiration rate.

Secondly, to know if this measurement is of any clinical significance, both blood oxygenation and blood CO2 levels would need to be measured and compared to masked respiration and unmasked respiration.

Without these tests, any measurement of CO2 levels in the mask region is merely a curiosity, and not indicative of
a risk factor.
CO2 Dissolved in Exhaled Water
Morgan Robinson, BSc, MSc, PhD Candidate | University of Waterloo
One wonders how CO2 would be retained behind the mask if for instance other gases, such as O2, are not impeded by the mask as would be suggested by no change in O2 saturation.

It could be that the CO2 is dissolved in the exhaled water vapour. The solubility of CO2 in water is about 33x greater than that of oxygen. This water vapour does tend to build up in the mask.
Amy Houtrow, MD, PhD, MPH | University of Pittsburgh
As a member of the editorial boards for three journals I understand the challenges in evaluating the scientific rigor and validity of submitted manuscripts. In my experience it has become increasingly difficult to secure high quality peer reviews. Nonetheless, the fact that this manuscript made it through editorial screening, peer-review, and approval is concerning. The need for the study was not well justified, the premise biased, the methodology flawed and thus the resultant data unhelpful and the conclusions inappropriate. I fear that those with political incentive to demonize mask-wearing will use this study as ‘scientific evidence’ to undermine public health strategies.
Not a Randomized Trial
Robert Rosen, D.O. | Private Practice, Dermatology
The subtitle "Randomized Clinical Trial" implies comparison between clinical interventions. It is unclear what participants were randomized to, and no significant difference was found between 2 unidentified types of masks. The subtitle is bound to mislead casual (e.g. Twitter, Facebook) readers. Further the discussion references hypercapnia, defined as elevated CO2 concentration in blood, but they include no blood measurements of C02 nor any evidence their results could be extrapolated to that.
Flawed Methodology
Julius Cheng, MD | University of Rochester
Questions have been raised about the use of an environmental CO2 measuring device (PCE-CMM 10), and not a medically calibrated end-tidal CO2 device.

This is especially relevant since the headline-grabbing mean CO2 of "between 13 120 (384) and 13 910 (374) ppm" is higher than the stated measuring range of 400-5000 ppm by the device's manufacturer.

The authors did not measure the actual dead space behind any of the masks.

And, the measured respiratory rate of the participants did not change. If the participants were getting clinically hypercapnic, you would think there would be some change in respiratory rate. />
Flawed Design, Unsafe Conclusions
Trisha Greenhalgh, FMedSci | University of Oxford
Walach et. al’s Research Letter [1] contains major flaws which cast doubt on the paper’s scientific integrity.

Most importantly, the main outcome cannot test the authors’ hypothesis. The authors used a surrogate outcome - carbon dioxide (CO2) levels inside the mask - as a proxy for CO2 levels inhaled by a child. This is actually a measure of ‘dead space’ CO2. Most of the air entering the child’s lungs will be inhaled from the ambient air outside the mask (in the same way a swimmer using a snorkel breathes in a small amount of ‘dead space’ air from
within the snorkel, but most air comes from outside it).

Second, CO2 levels were not correctly measured. The Geotech G100 is designed to be used with incubators. It has a response time of 20 seconds (T90 = 20 seconds, i.e. the instrument takes 20 seconds to reach 90% of a stable reading) [2]. In comparison, a capnograph [3], which would have been an appropriate instrument to use in this study, stabilizes in milliseconds. The study claims to have measured CO2 in inhaled air, but the authors provide no basis for distinguishing between inhaled and exhaled air, using the G100, at 1.5 cm from nose and face.

The authors report no meaningful physiological variables, such as blood oxygen saturation or arterial partial pressure of CO2 (pCO2). Measurement of transient increases in CO2 around the breathing zone (even if accurate) are not indicative of harm. Although Supplementary material reports that oximeters were used during the study and pulse rates appear to have been measured by oximeters, oxygen saturation level data are not reported.

The authors’ causal claims regarding physiological impact are wholly unsupported by the data.

The science of mask-wearing should be based on good data. Some of the authors have publicly associated themselves with, and taken funding from, non-science-based campaign groups [4]. We respectfully suggest that not only is this study scientifically flawed, but its publication puts the reputation of JAMA Pediatrics and its peer-review process at risk. 

Professor Trisha Greenhalgh FMedSci, University of Oxford
Megan L. Ranney MD MPH, Brown University
Asit Kumar Mishra PhD, Berkeley Education Alliance for Research in Singapore


1. Walach H, Weikl R, Prentice J, et al. Experimental Assessment of Carbon Dioxide Content in Inhaled Air With or Without Face Masks in Healthy Children: A Randomized Clinical Trial. JAMA Pediatr. Published online June 30, 2021. doi:10.1001/jamapediatrics.2021.2659

2. Geotech. G100 | QED. Published 2020. Accessed July 3, 2021. https://www.qedenv.com/products/g100/

3. Gravenstein N, Jaffe MB. 10 - Capnography. In: Ehrenwerth J, Eisenkraft JB, Berry JM, eds. Anesthesia Equipment (Third Edition). W.B. Saunders; 2021:239-252. doi:10.1016/B978-0-323-67279-5.00010-8

4. Hockertz S. Generation Maske: Corona: Angst und Herausforderung. Kopp, 2021.

Misrepresentation of Endogenously Produced CO2
Prof W Meier-Augenstein, Dr. rer. nat. | Robert Gordon University, Aberdeen
The values of 3.846 and 3.847% CO2 measured behind masks might actually represent CO2 endogenously produced as a result of the human body’s metabolism. In adults, body weight-adjusted endogenous CO2 production at rest is about 1.4 mmol/L which equates to a CO2 content of 5.03% in exhaled breath. The aforementioned values taken from the Table of this article are entirely consistent with what one would expect in breath exhaled by children into an enclosed yet not leak-tight space. Chances are that measurements of exhaled breath in a space formed by a hand cupped over the mouth and nose will have yielded values similar to those between mask and mouth and nose. Furthermore, to compare values of endogenous CO2 in exhaled breath as a result of the body’s energy expenditure at rest with threshold values for CO2 in enclosed spaces appears ill thought-out. It’s like comparing threshold concentrations for (e.g.) NOx in ambient air with NOx values measured directly at the exhaust of a car.
Bias Concerns
Steven Krager, MD, MPH | Clark County Public Health
I echo the other comments regarding the problems with the methods of this study and I'd also like to point out some ethical concerns.

• This study was funded by “Mediziner und Wissenschaftler für Gesundheit, Freiheit und Demokratie eV” ( Doctors and Scientists for Health, Freedom and Democracy, eV) which according to a Google search is an anti COVID-restriction advocacy organization that disseminates COVID-19 misinformation. The organization is also promoting the study on their website, without noting that they are the funder. Here is a quote from their article (translated by Google translate) “This clearly shows that when
wearing a mask, children regularly poison themselves through rebreathing of carbon dioxide and expose themselves to the highest health risks, which from a legal point of view clearly constitutes bodily harm.”

• The authors claim in the supplemental materials (Trial Protocol) that this study “will generate important knowledge that can be used for political action. This additional knowledge is central, as currently political decisions are made without the respective evidence base.”

• The primary author was the lead author of a paper retracted from the journal Vaccines titled “The Safety of COVID-19 Vaccinations—We Should Rethink the Policy” which made this unfounded claim- “For three deaths prevented by vaccination we have to accept two inflicted by vaccination.” The data in their paper in no way supports this conclusion, and to make such a claim while knowing the limitations of the data is further evidence that the author is not concerned about advancing science, but rather their own agenda.
Weak Methodology, No Standards
Jennifer McDonald, MD, FRCPC, PM&R | University of Ottawa
In their research letter on masks in children [1] Walach et. al failed to prove their findings are of physiological importance. Furthermore, their methodology rejects, without justification, a widely-accepted and long-established standard [2] for how and when to measure CO2 levels in relation to masks.

The authors conclude that the reported elevated CO2 levels will lead “to impairments attributable to hypercapnia” without measuring the partial pressure of carbon dioxide (PaCO2). PaCO2 can be estimated non-invasively using a readily-available transcutaneous CO2 monitor or by measuring end-tidal CO2 via capnography [3]. Oxygen saturation data were also omitted despite apparently being

CO2 levels vary widely during a breathing cycle and are best studied as a capnograph waveform with a fit-for-purpose CO2 sensor. The authors purport to measure CO2 content in the “inhaled air” of children wearing masks. Inexplicably, they used an inappropriate sensor to measure CO2 concentration of mask dead space, which represents the air between the mask and the face, can be thought of as an extension of the anatomical dead space (nose, tracheal, bronchi), and represents a small fraction of the child’s tidal volume. If the dead space were large, this would lead to increased respiratory effort to maintain the same volume of respirable air, which would be noted physiologically as decreased breathability rather than silent CO2 accumulation [4].

Many of the facepieces studied are covered under EN-149:2001+A1:2009 [2], and those with “CE” are certified by an EN Notified body through laboratory testing. The authors could have used the well-recognised dead space CO2 measurement method therein – employing a respiration simulator, and acceptance criterion of ≤10,000 ppm average CO2 inhalation concentration. The European Norm for surgical masks (EN 14683:2014) does not include a requirement for CO2 concentration, indicating that the international expert committee creating it did not consider this a factor of concern for wearer safety.

Products with different internal volumes and quality of fit to face were studied – naturally expected to form different inhaled CO2 concentrations depending on retention, leakage and dilution by inhaled air. Yet very similar concentrations were reported across all facepiece types – and the authors fail to justify this, despite the wide variation in contributing factors.

We respectfully suggest that not only is this study irredeemably flawed, but the reputation of the journal and its peer-review process is now at risk.

Simon Smith, PhD. ARCS CChem FRSC(UK), Member, PPE Committee, NASEM
Ralph Wunderlin
Jennifer McDonald, MD, University of Ottawa


1. Walach H, Weikl R, Prentice J, et al. Experimental Assessment of Carbon Dioxide Content in Inhaled Air With or Without Face Masks in Healthy Children: A Randomized Clinical Trial. JAMA Pediatr. Published online June 30, 2021.

2. CEN. EN 149:2009-08, Respiratory Protective Devices - Filtering Half Masks to Protect against Particles - Requirements, Testing, Marking. Beuth Verlag GmbH; 2009.

3. Hoffmann U, Essfeld D, Stegemann J. Comparison of arterial, end-tidal and transcutaneous PCO2 during moderate exercise and external CO2 loading in humans. Eur J Appl Physiol Occup Physiol. 1990;61(1-2):1-4.

4. Hopkins S, Dominelli P, Davis C, et al. Face Masks and the Cardiorespiratory Response to Physical Activity in Health and Disease. AnnalsATS 2021;18(3).

Are the Measurements Correct?
Herbert Renz-Polster, MD | Mannheim Institute for Public Health, University of Heidelberg
Several commentators have pointed out that the technical measurement device used in this study may not be appropriate because of a high latency of the analysis. Therefore, the measurements taken during the inspiratory phase may in part include CO2 concentration measurements inadvertently taken during the expiratory phase - which may explain the reported high CO2 concentrations during inspiration in children wearing a mask.

Such measurement errors would become immediately obvious if the researchers had performed due baseline measurements, i.e. measurements without masks during the inspiratory phase only and the expiratory phase only. The authors have not performed such
baseline measurements (personal communication with the lead author). Instead, they report what they call a "baseline" CO2 concentration. As it turns out, this is the average of measurements taken over a 3 minute period at 15 second intervals (personal communication with the lead author) - a completely nonsensical value from a physiological or pathophysiological standpoint.

Also, even if the point measurements during inspiration reported were accurate reflections of the CO2-concentrations under the mask, they would only reflect the concentration at the given time point. CO2-concentrations in air inspired from a mask with a certain dead space, however, are bound to vary greatly between the beginning of the inspiration and later phases of the inspiration when more and more ambient air gets entrained from the room environment. Therefore, it may not be correct to state that the inspired CO2 concentrations measured in children wearing masks are 6-fold higher than the limit allowed by German regulatory authorities for ambient air at the working place of 0,2% (which clearly is a concentration present during the whole inspiratory cycle).

However, there is one set of values reported that does not fit well with the assumption that these measurements are all unreliable because of the latency of the measurements and thus represent admixture of point measurements from the next respiratory phase. And this is the measurements reported for exhalation in children with masks. These reported numbers are a) pretty similar between the 2 types of masks used and b) compatible with the normal values of CO2-concentration reported for exhaled CO2 in the literature (about 4%).

So the question would be: if the inspiratory values are to be considered unreliable for methodological reasons - why are the expiratory values clearly within the expected range?

Also, the authors report a measurement range in participants wearing a mask from between 1783 ppm and 5240 ppm CO2 (see table). This is physiologically not plausible as the CO2 concentration in exhaled air does not fluctuate between individuals that much (1).

However, what they report as an average concentration for "exhaled air with FFP2", is 3.846 ppm. And what they report for "exhaled air with surgical mask" is 3.847 ppm. I.e., both values are around 3,8% - which is pretty much what a quick Google search gives you as average CO2 concentration in exhaled air.

The only values in this study which you don´t have to take at face value but that you can actually verify fall within the generally reported normal range - in spite of them allegedly being an average of values that physiologically don´t make sense. A remarkable coincidence. References

1. https://www.physiology.org/doi/pdf/10.1152/jappl.1998.85.2.642

Protocol Mismatch, Weird Measures
Mark Dublin, sweng | None
As others have pointed out, why the mismatch between the protocol vs the actually implemented trial? The latter provided just a fraction of the intended measures.

The rationale for the 2-mask-type sequence is unclear.

The difference between the 'baseline', the (supposed) averages, and the 'joint inhaled/exhaled' CO2 data is unclear. While the ~13k ppm data fits the G100 range, I see no reference in the datasheet to the purported 1 samples per second sampling rate. If that's actually a capability of the G100 as datalogger, I wonder why they did 3 seconds (3') baseline, 3' joint, 3' inhaled,
and 3' exhaled? I'd rather see measures for 3' baseline, 3' mask, and reported the 3x60 baseline datapoints and 3x60 mask datapoints from which some signal analysis could be done to infer the CO2 concentration at hose entry, assuming constant pump flow throughout the tests.

The added 'effective' dead space is a dynamic quantity that depends also on breathing rate, bigger than the geometric mask-face space due to the modulation of the CO2 concentration gradient at airways-ambient interface.

Additional Methodologic and Physiologic Questions
William Bennett, PhD | Center for Environmental Medicine, Asthma, and Lung Biology University of North Carolina-Chapel Hill, NC 27599
The authors' premise was not totally without merit but the actual data they present is not physically or physiologically possible suggesting a flaw in the CO2 measurement technique. Wearing a mask may increase dead space slightly – though because the masks are gas permeable, exhaled CO2 rapidly diffuses and mixes with the ambient CO2 concentration within the mask space (i.e. is not like increased dead space associated with putting a CPAP mask on or breathing through a snorkel). As a worst case, physical mask volume is about 25-50 ml of a 300 ml tidal inhalation. In their supplement the authors said they would measure the volume of the masks used but do not report this important detail. The end-exhaled concentration CO2 may have a transient peak of 2-4% concentration of CO2 that is re-inhaled from the mask space but again only in the first 25-50 ml of the inhalation of a 300 ml tidal volume. The remaining inhaled volume for each breath will be at ambient levels (0.04%). So the results they publish in the table and figure can only be possible as peak CO2 concentrations, probably at end-exhalation/start of inhalation, measured close to the nostrils/upper lip; it is not possible that the data are mean CO2 concentrations for the entire breath. They should have shown how/where/when the sampling occurred in the breath (e.g. provide a sample tracing of CO2 concentration during breathing).

Finally, their implication that the effect of increasing dead space with masks leads to health effects is very overstated from their single measure of peak CO2 inside the mask over a 9 minute sampling period and anecdotal reporting from some of the children (e.g. irritability). In their experimental design (supplement) they said they would measure O2 sats and breathing frequency/tidal volume, none of which was included in the results. A simple measure of breathing frequency with and without mask would tell them (and us) whether a slightly increased mean inhaled CO2 resulted in a physiological response. Again no data on this endpoint was provided.
Concerned for Biases
Angela Duke, MD, MPH, FACOG | Private Practice
We write with concern that the literature forming the basis for Walach et al's research letter was not fairly interpreted.

Walach et al state as their main principle that “the evidence base for [compulsory face masks for schoolchildren] is weak”, but both studies that the authors cited were in favour of such coverings (1,2). Xiao et al. said this about cloth masks: “[i]n theory, transmission should be reduced the most if both infected members and other contacts wear masks” (1). Similarly, Matuschek et al., who looked at studies that examined various types of masks from FFP3 to
cloth, included discussion of a study of cloth masks that found they reduced small aerosols by 10% to 60% depending on material and manufacture (2). This is quite substantial, given the limitations to cloth mask.

Walach et al also cite a non-peer-reviewed preprint for the idea that masks have adverse effects on children. This reference is now marked with an editorial warning that it was based on a selective survey, had multiple methodological limitations, and could not demonstrate a causal link between mask wearing and negative effects. This last point is exactly why it was cited in the present letter (3).

The journal Vaccines last week fully retracted another piece written by the present lead author that had questioned the benefit-harm balance of COVID-19 vaccines. The editors said they had serious concerns about misinterpretation of data by authors. This is eerily similar to our concerns.about the present letter.

Finally, the authors of the present letter appear to be associated or involved with many websites taking a rather strong stance on masks, casting the impartiality of this study into question. For example it was funded by Mediziner und Wissenschaftler fur Gesundheit, Freiheit und Demokratie (MWGFD). Three authors are members. It promotes itself saying “[d]uring the corona crisis, we came together in our criticism of the excessive restrictions” (4). The primary named author appears on a video discussing that masking children is a “crime” and one of the other authors is the CEO of the company whose website posts an anti-masking video (5).

We are concerned that the letter, misinterpreting the literature, and in light of the surrounding circumstances, is not a fair and unbiased presentation. Other commenters have pointed out its flawed methodology.

Angela Duke, MD, MPH, FACOG
Jonathan Mesiano-Crookston, BScH Biochem, JD, private practice (lawyer)

1-3 numbered are consistent with primary article's references (due to character limitations).

4. Mediziner und Wissenschaftler fur Gesundheit, Freiheit und Demokratie Physicians and Scientists for Health,Freedom and Democracy (MWGFD) www.mwgfd.de

5. tpi consult: toxikologie pharmakologie immunologie GmbH. CEO Stefan Hockertz
https://tpi-consult.de/deu/, accessed July 1, 2021.

Research Letter
June 30, 2021

Experimental Assessment of Carbon Dioxide Content in Inhaled Air With or Without Face Masks in Healthy Children: A Randomized Clinical Trial

Author Affiliations
  • 1Poznan University of the Medical Sciences, Pediatric Clinic, Poznań, Poland
  • 2Obstetric, Gynecological, and General Practice, Passau, Germany
  • 3Psychotherapeutic Practice, Müllheim, Germany
  • 4General Practice, Gernsbach, Germany
  • 5Traindl Consult, Vienna, Austria
  • 6Psychotherapeutic Practice for Children and Youth, Müllheim, Germany
  • 7tpi consult GmbH, Bollschweil, Germany
JAMA Pediatr. Published online June 30, 2021. doi:10.1001/jamapediatrics.2021.2659

Many governments have made nose and mouth covering or face masks compulsory for schoolchildren. The evidence base for this is weak.1,2 The question whether nose and mouth covering increases carbon dioxide in inhaled air is crucial. A large-scale survey3 in Germany of adverse effects in parents and children using data of 25 930 children has shown that 68% of the participating children had problems when wearing nose and mouth coverings.

The normal content of carbon dioxide in the open is about 0.04% by volume (ie, 400 ppm). A level of 0.2% by volume or 2000 ppm is the limit for closed rooms according to the German Federal Environmental Office, and everything beyond this level is unacceptable.4


We measured carbon dioxide content in inhaled air with and without 2 types of nose and mouth coverings in a well-controlled, counterbalanced, short-term experimental study in volunteer children in good health (details are in the eMethods in Supplement 1). The study was conducted according to the Declaration of Helsinki and submitted to the ethics committee of the University Witten/Herdecke. All children gave written informed consent, and parents also gave written informed consent for children younger than 16 years. A 3-minute continuous measurement was taken for baseline carbon dioxide levels without a face mask. A 9-minute measurement for each type of mask was allowed: 3 minutes for measuring the carbon dioxide content in joint inhaled and exhaled air, 3 minutes for measuring the carbon dioxide content during inhalation, and 3 minutes for measuring the carbon dioxide content during exhalation. The carbon dioxide content of ambient air was always kept well under 0.1% by volume through multiple ventilations. The sequence of masks was randomized, and randomization was blinded and stratified by age of children. We analyzed data using a linear model for repeated measurements with P < .05 as the significance threshold. The measurement protocol (trial protocol in Supplement 2) is available online.5 Data were collected on April 9 and 10, 2021, and analyzed using Statistica version 13.3 (TIBCO).


The mean (SD) age of the children was 10.7 (2.6) years (range, 6-17 years), and there were 20 girls and 25 boys. Measurement results are presented in the Table. We checked potential associations with outcome. Only age was associated with carbon dioxide content in inhaled air (y = 1.9867 – 0.0555 × x; r = –0.39; P = .008; Figure). Hence, we added age as a continuous covariate to the model. This revealed an association (partial η2 = 0.43; P < .001). Contrasts showed that this was attributable to the difference between the baseline value and the values of both masks jointly. Contrasts between the 2 types of masks were not significant. We measured means (SDs) between 13 120 (384) and 13 910 (374) ppm of carbon dioxide in inhaled air under surgical and filtering facepiece 2 (FFP2) masks, which is higher than what is already deemed unacceptable by the German Federal Environmental Office by a factor of 6. This was a value reached after 3 minutes of measurement. Children under normal conditions in schools wear such masks for a mean of 270 (interquartile range, 120-390) minutes.3 The Figure shows that the value of the child with the lowest carbon dioxide level was 3-fold greater than the limit of 0.2 % by volume.4 The youngest children had the highest values, with one 7-year-old child’s carbon dioxide level measured at 25 000 ppm.


The limitations of the study were its short-term nature in a laboratory-like setting and the fact that children were not occupied during measurements and might have been apprehensive. Most of the complaints reported by children3 can be understood as consequences of elevated carbon dioxide levels in inhaled air. This is because of the dead-space volume of the masks, which collects exhaled carbon dioxide quickly after a short time. This carbon dioxide mixes with fresh air and elevates the carbon dioxide content of inhaled air under the mask, and this was more pronounced in this study for younger children.

This leads in turn to impairments attributable to hypercapnia. A recent review6 concluded that there was ample evidence for adverse effects of wearing such masks. We suggest that decision-makers weigh the hard evidence produced by these experimental measurements accordingly, which suggest that children should not be forced to wear face masks.

Back to top
Article Information

Accepted for Publication: June 7, 2021.

Published Online: June 30, 2021. doi:10.1001/jamapediatrics.2021.2659

Retraction: A notice of retraction was published on July 16, 2021.

Corresponding Author: Harald Walach, PhD, Poznan University of the Medical Sciences, Pediatric Clinic, ul. Szpitalna 27/33, PL-60-572 Poznań, Poland (harald.walach@uni-wh.de).

Author Contributions: Dr Walach (principal investigator) had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: All authors.

Acquisition, analysis, or interpretation of data: Walach, Weikl, Diemer, Traindl, Kappes, Hockertz.

Drafting of the manuscript: Walach, Traindl.

Critical revision of the manuscript for important intellectual content: Walach, Weikl, Prentice, Diemer, Kappes, Hockertz.

Statistical analysis: Walach.

Administrative, technical, or material support: Weikl, Prentice, Diemer, Traindl, Kappes, Hockertz.

Supervision: Weikl, Diemer, Traindl, Kappes, Hockertz.

Other–liaising with all other authors: Walach.

Conflict of Interest Disclosures: None reported.

Funding/Support: Mediziner und Wissenschaftler für Gesundheit, Freiheit und Demokratie eV, a public charity, has organized this study and covered only essential expenses, such as travel.

Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 3.

Xiao  J, Shiu  EYC, Gao  H,  et al.  Nonpharmaceutical measures for pandemic influenza in nonhealthcare settings —personal protective and environmental measures.   Emerg Infect Dis. 2020;26(5):967-975. doi:10.3201/eid2605.190994 PubMedGoogle ScholarCrossref
Matuschek  C, Moll  F, Fangerau  H,  et al.  Face masks: benefits and risks during the COVID-19 crisis.   Eur J Med Res. 2020;25(1):32. doi:10.1186/s40001-020-00430-5PubMedGoogle ScholarCrossref
Schwarz  S, Jenetzky  E, Krafft  H, Maurer  T, Martin  D. Corona children studies “Co-Ki”: first results of a Germany-wide registry on mouth and nose covering (mask) in children. Published 2021. Accessed June 15, 2021. https://www.researchsquare.com/article/rs-124394/v1
Mitteilungen der Ad-hoc-Arbeitsgruppe Innenraumrichtwerte der Innenraumlufthygiene-Kommission des Umweltbundesamtes und der Obersten Landesgesundheitsbehörden.  [Health evaluation of carbon dioxide in indoor air].   Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz.  2008;51(11):1358-1369. doi:10.1007/s00103-008-0707-2PubMedGoogle ScholarCrossref
Walach  H, Weikl  R, Traindl  H,  et al. Is carbon dioxide content under nose-mouth covering in children without potential risks? a measurement study in healthy children. Published April 14, 2021. Accessed June 15, 2021. https://osf.io/yh97a/?view_only=df003592db5c4bd1ab183dad8a71834f
Kisielinski  K, Giboni  P, Prescher  A,  et al.  Is a mask that covers the mouth and nose free from undesirable side effects in everyday use and free of potential hazards?   Int J Environ Res Public Health. 2021;18(8):4344. doi:10.3390/ijerph18084344 PubMedGoogle ScholarCrossref