Relationship Between Carbon Dioxide Levels and Reported Headaches on the International Space Station

Article (PDF Available)inJournal of occupational and environmental medicine / American College of Occupational and Environmental Medicine 56(5):477-83 · May 2014with193 Reads
DOI: 10.1097/JOM.0000000000000158 · Source: PubMed
Abstract
Because of anecdotal reports of CO2-related symptoms onboard the International Space Station (ISS), the relationship between CO2 and in-flight headaches was analyzed. Headache reports and CO2 measurements were obtained, and arithmetic means and single-point maxima were determined for 24-hour and 7-day periods. Multiple imputation addressed missing data, and logistic regression modeled the relationship between CO2, headache probability, and covariates. CO2 level, age at launch, time in-flight, and data source were significantly associated with headache. For each 1-mm Hg increase in CO2, the odds of a crew member reporting a headache doubled. To keep the risk of headache below 1%, average 7-day CO2 would need to be maintained below 2.5 mm Hg (current ISS range: 1 to 9 mm Hg). Although headache incidence was not high, results suggest an increased susceptibility to physiological effects of CO2 in-flight.

Figures

FAST TRACK ARTICL E
Relationship Between Carbon Dioxide Levels and Reported
Headaches on the International Space Station
Jennifer Law, MD, MPH, Mary Van Baalen, MS, Millennia Foy, PhD, Sara S. Mason, BS, Claudia Mendez, MPH,
Mary L. Wear, PhD, Valerie E. Meyers, PhD, DABT, and David Alexander, MD
Objective: Because of anecdotal reports of CO2-related symptoms onboard
the International Space Station (ISS), the relationship between CO2and in-
flight headaches was analyzed. Methods: Headache reports and CO2mea-
surements were obtained, and arithmetic means and single-point maxima were
determined for 24-hour and 7-day periods. Multiple imputation addressed
missing data, and logistic regression modeled the relationship between CO2,
headache probability, and covariates. Results: CO2level, age at launch, time
in-flight, and data source were significantly associated with headache. For
each 1-mm Hg increase in CO2, the odds of a crew member reporting a
headache doubled. To keep the risk of headache below 1%, average 7-day
CO2would need to be maintained below 2.5 mm Hg (current ISS range: 1 to
9mmHg).Conclusions: Although headache incidence was not high, results
suggest an increased susceptibility to physiological effects of CO2in-flight.
On the Earth, carbon dioxide (CO2) is a trace constituent that
makes up 0.04% of the atmosphere, equating to a partial pres-
sure (ppCO2) of 0.3 mm Hg at standard pressure.1Nevertheless,
within a spacecraft, it has been impractical to control ppCO2to such
low levels because of mass constraints and consumable limitations.
CO2levels in spacecraft have typically been 2.3 to 5.3 mm Hg,
with large fluctuations occurring over hours and days.2The highest
ppCO2recorded in a US spacecraft was 14.9 mm Hg on Apollo 13.3
At the time of the study, the spacecraft maximum allowable concen-
trations (SMACs) for CO2were 15 mm Hg for 1 hour, 10 mm Hg
for 24 hours, 5.3 mm Hg for 7 to 180 days, and 3.8 mm Hg for 1000
days.4The SMACs provide the basis for the operational constraints
enacted by the flight rules, which are methods used by the National
Aeronautics and Space Administration (NASA) for planning and
executing operations.5
The effects of ambient CO2and exposure limits have been
well studied on the Earth. Physiologically, when blood CO2levels
rise, chemoreceptors in the carotid and aortic bodies quickly trigger
various centers in the medulla to send signals to the intercostal mus-
cles, diaphragm, and sinoatrial node to increase minute ventilation
and heart rate to enhance the body’s elimination of CO2.CO
2has
effects on the cerebral vascular tone, primarily driven by the changes
in extracellular pH. The lowering of the pH induces vasodilatation
mediated by nitric oxide, cyclic nucleotides, prostanoids, potassium
channels, and calcium ion exchange,6resulting in alteration of cere-
bral blood flow (CBF). Numerous studies reported an increase in
CBF of 1 to 2 mL/100 g/min for each 1-mm Hg increase in arte-
rial partial pressure of CO2(PaCO2),6–8 or 5.8% to 6.7% per 1-mm
Hg rise in ppCO2.9The increase in CBF results in an elevation of
intracranial pressure, presumably leading to headache, visual distur-
bance, impaired mental function, and other central nervous system
From the NASA Johnson Space Center (Dr Law, Ms Van Baalen, Dr Meyers,
and Dr Alexander), Wyle Science, Technology and Engineering (Drs Foy and
Wear), and MEI Technologies (Mrs Mason and Ms Mendez), Houston, Tex.
The authors declare no conflicts of interest.
Address correspondence to: Jennifer Law, MD, MPH, 2101 NASA Parkway, Mail
Code SD2, Houston, TX 77058 (jennifer.t.law@nasa.gov).
Copyright C2014 by American College of Occupational and Environmental
Medicine
DOI: 10.1097/JOM.0000000000000158
symptoms. On longer exposures, Sliwka et al10 found that cerebral
blood flow velocity (CBFv) in the middle cerebral artery increased by
35% as detected by transcranial Doppler when subjects were exposed
to 23 days of CO2of 5.3 or 9.1 mm Hg; although CBFv responses
were similar for the two levels of exposure, headache complaints
were more frequent during the early days of exposure to the higher
level. Furthermore, CBFv increased at days 1 and 5 after discontin-
uation of hypercapnia. In addition, although CBF and cerebral blood
volume (CBV) change similarly during hypercapnia on the Earth,11
CBF and CBV may not have the same relationship in spaceflight
because of impaired venous drainage caused by the cephalad fluid
shift; therefore, increased flow may increase the volume.
Terrestrially, healthy males can tolerate CO2levels below
7.5 mm Hg indefinitely and up to 480 minutes at 11 mm Hg with-
out acute health effects. Individuals begin to experience headache
and dyspnea upon mild exertion after several hours of exposure to
15 mm Hg.12 Sweating and dyspnea at rest may be seen after ex-
posure to CO2of 23 mm Hg for 60 minutes. Dizziness, lethargy,
and uncomfortable dyspnea may develop within a few minutes of
exposure to CO2of 30 to 38 mm Hg. Still higher CO2concen-
trations will cause unconsciousness, muscle twitching, convulsions,
and eventually death.12,13
Since the early years of the International Space Station (ISS)
program, anecdotal reports have suggested that ISS crew members
develop CO2-related symptoms, such as headache, lethargy, malaise,
listlessness, and fatigue, at lower CO2levels than would be expected
terrestrially.2Headache was reported on two early occasions: once
while crew members were working inside a confined space having
reduced air flow, and the other when all of the crew members were
gathered in a single location. Also, these early ISS crew members
described their individual symptoms as similar to those they experi-
enced when they were intentionally exposed to excess CO2during
ground training.14 On later missions, there were reports of similar
symptoms when ppCO2rose above 4 mm Hg but remained under the
flight rule limit of 7.6 mm Hg. The crew noted that these symptoms
subsided within minutes of reducing ppCO2to the range of 2 mm
Hg or when they breathed 100% oxygen (O2)inanextravehicular
activity suit. Furthermore, the crew felt better and reported improved
performance when CO2levels were low. Similar crew observations
have been periodically noted since that time. These symptoms have
resulted in closer occupational surveillance and operational lowering
of the ISS CO2limits as more data are collected and flight rules are
changed.
Given the apparent increased sensitivity to CO2exposure dur-
ing spaceflight, it is important to understand the acute and chronic
effects of elevated CO2on orbit, particularly in light of symptoms
associated with the recently described spaceflight-induced visual im-
pairment/intracranial pressure (VIIP) syndrome.15, 16 The “VIIP syn-
drome” is a set of ocular structural and optic nerve changes thought
to be caused by events precipitated by the cephalad fluid shift crew
members experience during microgravity exposure. There is a subset
of crew members who experience visual performance decrements,
cotton wool spot formation, choroidal fold development, optic disc
edema, optic nerve sheath distention, and/or posterior globe flatten-
ing with varying degrees of severity and permanence. It is thought
that CO2exposure may contribute as a predisposing or exacerbating
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
JOEM rVolume 56, Number 5, May 2014 477
Law et al JOEM rVolume 56, Number 5, May 2014
factor because it affects intracranial pressure due to altered cerebral
blood flow and volume.
On the ISS, CO2levels are primarily managed by environ-
mental control and life-support systems (ECLSS) with regenera-
ble absorber technology: the Vozdukh in the Russian Segment and
the Carbon Dioxide Removal Assembly (CDRA) in the US Oper-
ating Segment (the USOS, composed of modules developed by the
NASA, Canadian Space Agency, European Space Agency, and Japan
Aerospace Exploration Agency). In addition, metal oxide and ex-
pendable lithium hydroxide canisters are available for extravehicular
activity or contingency. CO2monitoring is nominally performed us-
ing the Major Constituent Analyzer (MCA) in the USOS and the Gas
Analyzer in the Russian Segment. Portable Carbon Dioxide Moni-
tors can be deployed as needed. Although it is desirable to minimize
CO2exposure from a human health standpoint, it is not currently
practical to limit ppCO2to below 3 mm Hg with six crew members
using the suite of hardware.17
To study the effects of CO2on the ISS more rigorously, a
working group was formed as a multiyear collaboration between
the Astronaut Occupational Health Program, Space Toxicology Of-
fice, and Space and Clinical Operations at the NASA Johnson Space
Center. The objectives of this study were to comprehensively char-
acterize CO2levels since the beginning of the ISS program, iden-
tify in-flight headache events, and analyze the relationship between
ppCO2and headache.
METHODS
As part of the occupational surveillance program for astro-
nauts, relevant data were pulled on all USOS astronauts who com-
pleted an ISS mission between March 14, 2001, and May 31, 2012,
for inclusion in the study. Data collected before 2001 were excluded
because of limited environmental data, as CO2monitors did not
become operational until February 13, 2001.
Table 1 displays the demographic characteristics of the USOS
crew members included in this study. Astronauts are extensively
screened before selection into the corps and throughout their active
duty careers to ensure mission success and maintain crew health.18
As a result, astronauts at selection are free from many medical condi-
tions that are common in the general population. Notably, astronauts
are screened during selection for underlying cardiovascular and neu-
rovascular conditions that may predispose to headaches. Even so,
headache is one of the most common in-flight symptoms reported
by crew members, often occurring in the first few days of spaceflight
and associated with space adaptation syndrome.19 Headaches that
occur after the first week of spaceflight are presumed to have other
causes.
In-Flight Reports of Crew Symptoms
Cases were identified as crew member reports of in-flight
headache symptoms. Reports of these crew symptoms were cap-
tured from two primary archival sources: private medical conferences
TABLE 1. Characteristics of ISS USOS Crew Members
From Expeditions 2 to 31 (N=49)
Variables
Mean ±Standard
Deviation
Males, % 81.6
Females, % 18.4
Mean age at launch, yr 47.6 ±4.4
Mean mission duration, d 153.1 ±45.3
Mean number of observations per
crewmember
38.0 ±17.6
(PMCs) and Space Medicine Operations Team (SMOT) meeting
minutes. Private medical conferences are video teleconferences be-
tween individual crew members and their crew surgeons, held daily
for the first 5 days of flight and weekly thereafter, to discuss the
health of the crew member while on orbit. After each PMC, the
crew surgeons record their observations and impressions in the crew
members’ electronic medical records; before the introduction of the
electronic medical record, PMCs were recorded on paper notes or
audio/video tapes. The SMOT meeting is a multilateral, physicians-
only teleconference held weekly to review the health status of the
entire crew onboard the ISS. During each meeting, the crew sur-
geons report the condition of their crew members and any notable
observations or issues.
Any mention of headache symptoms in the PMC notes or
SMOT minutes was extracted by two reviewers with oversight by an
experienced epidemiologist. Private medical conferences or SMOT
notes that were not available electronically were converted from
paper notes or transcribed from audio or video tapes before data ex-
traction. Each PMC and SMOT note was categorized as the presence
or absence of a reported headache for each individual crew member.
Because of possible confounding by space adaptation syndrome,20
reports occurring within the first 7 days of flight were excluded from
the analysis.
All availableFlight Surgeon and Biomedical Engineer console
logs and the Crew Comments Database were also reviewed. There
was a complete agreement with the two primary sources (PMC and
SMOT compared with console logs and crew comments), so no new
information was added from these sources.
To combine the PMC and SMOT data sets and avoid dou-
ble reporting, the symptom’s timeframe, cause, and description were
used to reconcile the events that were reported in both sources. When
a symptom was reported from both sources, the PMC was chosen
to represent the report because PMCs are conducted before SMOT
meetings, therefore closer to the timing of the event. Questionable
reports were reviewed by the flight surgeon authors for final deter-
mination.
ISS CO2Levels
The MCA was the primary source of CO2monitoring used
in this study because its calibration and performance were well un-
derstood by the environmental control and life-support team. The
first MCA was activated during Expedition 2 on February 13, 2001;
CO2data were not available before this date. The second MCA was
subsequently added as the ISS grew in size and became the prime
CO2monitoring hardware after September 7, 2010. Between May
13, 2011, and January 28, 2012, backup sensors were used because
of MCA failures.
CO2measurements from these instruments were obtained by
the JSC Space Toxicology Office. Four different CO2measurements
were assessed: 7-day mean, 7-day peak, 24-hour mean, and 24-
hour peak. Because the weekly PMCs and SMOT meetings captured
crew symptoms for the entire preceding week and generally did not
provide a specific date of onset, the arithmetic mean and maximum
ppCO2were calculated for the 7-day period before each PMC or
SMOT date. Twenty-four-hour means and peaks were also calculated
in an effort to assess the more acute effects of CO2and reduce any
possible reporting bias that could occur if symptoms were more likely
to be reported, if they occurred immediately preceding the PMC or
SMOT than if they were earlier. These calculations were performed
uniformly on all PMC and SMOT dates and times without analyst
knowledge of whether crew symptoms were reported during each
period.
Statistical Analyses
All statistical analyses were performed using SAS 9.3 (SAS
Institute, Cary, NC). The relationship between headaches and CO2
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
478 C2014 American College of Occupational and Environmental Medicine
JOEM rVolume 56, Number 5, May 2014 CO2and Headache Reports on International Space Station
levels was analyzed using logistic regression modeling adjusted for
age at launch, mission elapsed time, whether the symptom was re-
ported in a PMC or a SMOT meeting (data source), and CO2level.
The following equation defined the logistic regression relationship
between the covariates and the probability of headaches (P):
log p
1p=β0+β2×age at launch +β3×missionelapsed time
+β4×I(PMC) +β5×CO2.
where I(PMC) was an indicator function whose value was 1 when
the data source was a PMC and 0 when the data came from a SMOT
meeting. Repeated observations within the same individual were
adjusted by using generalized estimating equations in fitting the un-
derlying logistic regression model. The exchangeable generalized
estimating equation covariance structure was chosen because it pro-
vided the best model fit statistics when applied to the complete case
data. Initially, sex was included in the model but no association was
found, so it was excluded from the model reported.
Missing Data
CO2data gaps were identified and divided into the following
categories for disposition:
rQuestionable data: Single data point or multiple data points with
identical values were recorded over a long period. The data in
this category were discarded and the observations were treated as
missing because it was questionable whether the hardware was
working.
rLimited data: Less than 50% of the data points were available,
but values varied in an expected manner. Because each period
contained thousands of data points in 10-minute intervals, the
data in this category were retained and used to calculate averages.
rNo data: Data were completely missing during the period. Data
from the secondary sensors were used to fill in the gaps as avail-
able.
To address the missing CO2data in an unbiased manner,
regression-based multiple imputation was performed using date,
time, CO2levels, symptoms, sex, number of crew members onboard,
and habitable volume of the ISS at the time to inform the missing
data. The multiple imputation process determined that date, time,
and other CO2measures were most important in predicting missing
CO2levels. Distributions of each missing value were determined on
the basis of the observed informing variables, and 20 random data
sets were generated with the missing data imputed with random ob-
servations from the underlying distributions. The logistic regression
analysis was then repeated for each imputed data set, and the results
from each were then combined to produce the final result.
RESULTS
In-Flight Reports of Crew Symptoms
There was limited overlap between the PMC and SMOT
records, with only 17 events reported in both sources. As described in
the Methods, only the PMC observations were included when over-
lap was observed to avoid double-counting single events. A total of
46 reported headaches and 1670 nonreports were observed.
ISS CO2Levels
Initial evaluation of CO2levels using the primary sensors
described previously resulted in a data set that was 68% complete
for 24-hour data and 60% complete for 7-day data. As a result,
CO2levels were collected from secondary monitoring sources, re-
sulting in 24-hour/7-day CO2data that were categorized as follows:
87.2/86.8% complete, 5.31/3.92% questionable, 1.83/4.29% limited,
and 5.64/4.99% no data.
Table 2 summarizes the descriptive statistics for CO2lev-
els calculated using the multiple imputation procedure. Data clas-
sified as limited (<50% observed) were calculated on the basis of
large numbers of observations available, and were therefore consid-
ered complete for analysis purposes. Questionable data, resulting
from large stretches of time where levels were nearly identical, were
deemed unreliable and were treated as missing data in the analysis.
Figure 1 shows the CO2levels over time during the study period.
Average 7-day CO2levels over the ISS program were used as an
example, because it is most reflective of the medical event-reporting
period.
Management of CO2levels on the ISS has changed over time
as seen in 7-day averages (Fig. 1). Current ISS flight rules require
consultation with the flight surgeon when CO2levels average higher
than 5.3 mm Hg over 5 days or 6.0 mm Hg over 24 hours, and mea-
sures must be taken to lower the ppCO2if levels reach or exceed
7.6 mm Hg. This action is required even in the absence of CO2-
related symptoms. The presence of CO2-related symptoms below
these levels has led to flight rule waivers, known as “chits,” to man-
age CO2levels even lower on a mission-to-mission basis until the
flight rules can be changed permanently. Updates to the flight rules
are currently in work. The periods during which chits are in place
are indicated by color bands in Fig. 1. Early in the ISS program,
chits controlled ppCO2to 5.3 mm Hg or levels that were contingent
on the duration and magnitude of the excursion or timing relative
to complex operations. Since Expedition 23 in 2010, chits have rou-
tinely controlled CO2levels to 4 mm Hg or less. The incidence of
headache per week in a crew member, estimated through logistic
regression adjusted for repeated measures and missing values, was
3.28% before Expedition 23 and decreased to 1.6% since controls
were put in place to keep ppCO2to 4 mm Hg or less (pre-Expedition
23, mean 7-day CO23.60 mm Hg; post-Expedition 23, mean 7-day
CO22.54 mm Hg; P=0.068).
Association of Crew Symptoms and CO2
CO2level, age at launch, time in-flight, and data source were
all significantly associated with headache for 24-hour and 7-day av-
erages and peaks. Table 3 shows the results for 7-day CO2averages
as a representative example. Seven-day CO2averages are most rep-
resentative of the medical event-reporting period for both PMCs and
SMOT notes. We found that for each 1-mm Hg increase in CO2,the
odds of a crew member reporting a headache almost doubled; for
each year older a crew member was at launch, the odds of a headache
TABLE 2. CO2Levels for Expeditions 2 to 31
Variables
Average 24-hr CO2,
mm Hg
24-hr Peak CO2,
mm Hg
Ave r a ge 7-d C O 2,
mm Hg
7-d Peak CO2,
mm Hg
Mean ±SD 3.41 ±0.93 3.91 ±1.08 3.39 ±0.83 4.50 ±1.15
Median 3.50 3.98 3.41 4.51
Range 0.972–5.78 1.22–8.32 1.17–5.56 1.63–8.32
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
C2014 American College of Occupational and Environmental Medicine 479
Law et al JOEM rVolume 56, Number 5, May 2014
0
1
2
3
4
5
6
7-Day Aver age CO 2 Level (mmHg)
Date
ppCO2 controlled to 5 mmHg
Conditional
ppCO2 controlled to 4 mmHg
FIGURE 1. International Space Station CO2levels over time. Time bands indicate ISS expeditions for which CO2levels were
controlled by chits.
TABLE 3. Parameter Estimates in the Logistic Regression*
Parameter Description Estimate (SE) P
Multiplicative
Effect eβ(CI)
β0Intercept 1.06 (2.28) 0.64
β2Age at launch, yr 0.099 (0.047) 0.036 0.905 (0.825–0.994)
β3Mission elapsed time, days 0.006 (0.002) 0.004 0.994 (0.989–0.998)
β4Data source was a private
medical conference
0.897 (0.397) 0.024 2.78 (1.13–5.35)
β5Average 7-d CO2level 0.632 (0.197) 0.002 1.88 (1.27–2.78)
*For every 1 unit increase in the covariate of interest, the predicted odds of a headache are multiplied by Exp(β).
CI; confidence interval; SE, standard error.
decreased by almost 10%; for each day spent in-flight, the odds of a
headache decreased by 0.6%; and headaches were more likely to be
reported during PMCs than SMOTs. This last effect was likely influ-
enced by the decision to use PMC reports when the same headache
event was reported on both the PMC and the SMOT. When sex was
included in the model, no significant effects were found (main effect
P=0.71; interaction effect P=0.49), so it was eliminated from the
final model.
From the parameter estimates using the median value for the
other covariates, a plot was generated to predict the probability of
headache on the basis of average 7-day CO2(see Fig. 2). To keep
the risk of headache below 1%, the average 7-day CO2would have
to be below 2.5 mm Hg.
DISCUSSION
Compared with terrestrial ambient ppCO2,CO
2levels on-
board the ISS are chronically elevated but within the historical
SMACs, which were based on terrestrial data. These levels were
previously thought to not cause detrimental physiological effects,21
but on the basis of the current analysis the higher CO2levels present
on the ISS have a statistically significant association with the inci-
dence of reported headaches in a healthy astronaut cohort. Although
the incidence is not high, the concern of the space medical commu-
nity is that headaches may be an indicator of underlying increased
intracranial pressure likely from the synergy between CO2-induced
cerebral vasodilatation and decreased venous drainage because of
the loss of the hydrostatic gradient in microgravity.
With recent interest in VIIP, we performed the same analysis
on vision changes to evaluate their association with CO2. We did not
find a relationship between vision changes during spaceflight and
CO2levels in this study, but we believe the lack of association was
primarily due to the method of data collection and analysis not being
appropriate for nonspecific vision outcomes. Vision changes tend
to be chronic in nature, occurring over weeks to months, whereas
our CO2data set focused on acute changes occurring within 7 days.
Whereas headaches are more readily recognizable, a crew member
experiencing an insidious vision change may gradually adjust read-
ing distances or attribute it to presbyopia and may not report visual
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
480 C2014 American College of Occupational and Environmental Medicine
JOEM rVolume 56, Number 5, May 2014 CO2and Headache Reports on International Space Station
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
01234567
Predicted Probability of Headache
7-Day Average ppCO2(mmHg)
Predicted
95% Confidence Limits
FIGURE 2. Predicted probability of headache on the basis of average 7-day CO2levels.
changes as a discrete event or symptom until prompted. Furthermore,
we noted that vision changes were inconsistently reported. For ex-
ample, a crew member’s first event reported was described as the
same visual disturbance experienced for weeks. It is clear that vision
data need to be collected and analyzed differently than headaches.
In our data extraction, we noted other suspected CO2
symptoms. These crew comments suggest effects of CO2beyond
headaches and vision change. The following anecdotes were partic-
ularly interesting:
r“Last week, the crew reported that they were not aware of the high
CO2levels. After further discussion, however, they reported that
they noticed a big difference when the CDRA was turned back on
and the levels went down. [Two crew members] stated that levels
as high as 6.0 [mm Hg] were incompatible with performing on-
board activities. The crew members reported sleep difficulties and
increased irritability, which was noticed by ground controllers.”
rOn another occasion, the CDRA was turned off during an as-
sessment of the Vozdukh, which led to elevation of ppCO2above
6.2 mm Hg. Although the crew reported that they were unaware of
the elevated levels, the ground controllers reported that the crew
seemed irritable during that time.
rFatigue was reported by a crew member, who described it as atypi-
cal when compared to regular workday fatigue. The crew member
reported feeling better the morning the CDRA was activated, as
well as during the following days.
These examples suggest a possible association between CO2
and sleep difficulties, irritability, and fatigue. It is also worth not-
ing that some of the symptoms were recognized by the ground sup-
port team. Epidemiological and interventional research suggests that
higher levels of CO2in indoor areas are correlated with headache,
mucosal irritation, decreased work performance, and perceptions
of poor ventilation.22–27 A recent study by Satish et al28 demon-
strates decrements in decision-making performance after exposure to
ppCO2up to 1.9 mm Hg for 2.5 hours. These reductions in cognitive
decision-making after CO2exposure can have critical consequences,
especially in closed cabin environments during contingency opera-
tions. How these results relate to responses of astronauts to acute
spikes in CO2after chronic exposure in the ISS environment is yet
to be determined. This is a focus area of NASA studies currently
being planned.
Interestingly, our results showed that a crew member’s age at
launch and time in flight were inversely related to headache reports. A
possible explanation is that older crew members may have less com-
pliant vasculature and therefore be less susceptible to CO2-mediated
vasodilation. There may also be an adaptation to chronic exposure
to CO2. Of note, although men and women have been reported to
have different responses to hypercapnia hypothesized to be related
to different prostaglandin levels in men and women,9this study did
not find a statistically significant sex effect in predicting headaches.
Nevertheless, the focus of this study was not on determining whether
there were sex-differentiated effects.
Methodologically, we created data sets from operational data
collected for other purposes to conduct occupational surveillance.
This was the first application of multiple imputation to astronaut
occupational surveillance data to handle missing data that could
confound finding the association of interest. Our results were based
on more complete data sets than prior work, but there were still
many limitations to this analysis. We assumed that environmental
CO2measurements were representative of the individual exposures,
and we did not account for potential localized pockets of elevated
CO2. Despite our best effort to collect crew symptoms, PMC reports
were missing for several expeditions and many console logs were not
available to fill in the gaps. Underrepor ting of symptoms is also likely
for a number of reasons: a crew member might not have reported a
symptom unless it was sufficiently severe; symptoms that occurred
earlier in a reporting period might be less likely recalled by a crew
member by the time of the next PMC; a persistent symptom might
be reported only once and then not updated in further reports; and
in the early expeditions, crew surgeons might not have recorded all
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
C2014 American College of Occupational and Environmental Medicine 481
Law et al JOEM rVolume 56, Number 5, May 2014
crew-reported symptoms in PMCs if they were deemed to have “no
mission impact.” Once the effects of CO2became better known,
crew members might have reported symptoms or attributed them to
elevated CO2more often.
Furthermore, headaches are multifactorial, and it is clear from
the crew reports that the symptoms varied in type and severity. Pos-
sible confounders include nasal congestion secondary to increased
dust in some recent missions; space adaptation syndrome and other
factors that may increase vascular congestion, thereby increasing
crew susceptibility to headaches; exposure to CO2during transport
to the ISS on the Soyuz, which has been reported to exceed 20 mm
Hg; and differences in individual susceptibility.
It is also possible that some crew members do not experience
headache as a CO2symptom. ISS crew members receive CO2ex-
posure training before each mission, so they can identify their own
CO2symptoms. The crew members breathe into an anesthesia bag
containing 100% oxygen and replace the oxygen with the CO2they
exhale until 8% CO2is reached or they desire to stop the test. The in-
structor helps participants complete a symptom documentation form
to record the symptoms they develop and the percent CO2level at
which each symptom is first experienced. Although the CO2levels in
this ground training cannot be compared with on-orbit exposure be-
cause of differences in exposure time and O2prebreathe, the presence
or absence of headache development during ground training may be
an indicator of individual susceptibility to CO2-induced headache.
Operationally, the first ISS flight rules limited CO2exposure
to 7.6 mm Hg, on the basis of earlier SMAC values. Recognition of
crew symptoms at lower levels led to the implementation of chits to
control ppCO2to 5.3 mm Hg initially, then 4 mm Hg more recently.
Our results demonstrated that the chits have successfully resulted
in lower ppCO2and lower rates of reported headaches, supporting
ongoing efforts to update the flight rule limit to 4 mm Hg or below.
According to our analysis, average 7-day CO2concentrations
would need to be maintained below 1.97 mm Hg to keep the risk
of headache below 1%, a standard threshold used in toxicology and
aerospace medicine. Nevertheless, our results alone are not sufficient
to define a new SMAC. The toxicology community will need to weigh
other studies and health effects beyond headaches to set new SMAC
values.
More work is needed to elucidate the effects of CO2in the
setting of spaceflight. To better evaluate vision changes, Astronaut
Occupational Health Program is beginning to correlate VIIP classifi-
cations and individual findings within the classes with aggregate CO2
data for each mission. The NASA Human Research Program has be-
gun a perspective “Ocular Health Study” to collect eye and vascular
data on crew members, and the data will provide objective measures
for correlation with CO2. Additional occupational surveillance with
emphasis on behavioral, cognitive, and performance effects is also
planned, and a NASA study has been proposed to repeat Satish’s
decision-making study28 and collect additional relevant endpoints.
CONCLUSIONS
This study was the first comprehensive effort to analyze the
relationship of headache reports and exposure to elevated CO2on
the ISS. We found that headache reports were positively associated
with CO2levels, which would have to be further lowered than current
operational limits to reduce the risk of headache below 1%. Efforts
are ongoing to study visual and cognitive effects of acute and chronic
exposure to CO2in the setting of spaceflight to inform SMACs and
new hardware requirements.
ACKNOWLEDGMENTS
We thank the JSC ECLSS sustaining engineering team for
providing CO2data for our analysis, and the following individuals
who assisted with data processing: Deborah Eudy, Kevin Keenan,
MD, Selina Zalesak, Sara Perry, Zane Conrad, and David Reyes,
MD, MPH.
REFERENCES
1. National Oceanic and Atmospheric Administration. Trends in atmospheric
carbon dioxide. Earth Syst Res Lab. 2009. Available at http://cdiac.ornl.gov/
trends/co2/sio-keel.html. Accessed October 22, 2013.
2. James J. The Headache of Carbon Dioxide Exposures. SAE Technical Paper.
Chicago, IL; 2007. Available at http://papers.sae.org/2007-01- 3218/
3. Michel E, Waligora J, Horrigan D. Shumate. Environmental factors. Biomed-
ical Results of Apollo. Washington, DC: NASA SP-368; 1975.
4. James J. Carbon dioxide. Spacecraft Maximum Allowable Concentrations for
Selected Airborne Contaminants: Vol 5. Washington, DC: National Academy
Press; 2008:112–124.
5. Keyser L. Apollo Experience Report: The Role of Flight Mission Rules in
Mission Preparation and Conduct. NASA Johnson Space Center; 1974.
6. Brian JE Jr. Carbon dioxide and the cerebral circulation. Anesthesiology.
1998;88:1365–1386.
7. Iadecola C, Zhang F. Permissive and obligatoryroles of NO in cerebrovascular
responses to hypercapnia and acetylcholine. Am J Physiol. 1996;271:R990–
R1001.
8. Wang Q, Pelligrino D, Koenig HM, Albrecht RF. The role of endothelium,
and nitric oxide in rat pial arteriolar dilatory responses to CO2in vivo. J
Cereb Blood Flow Metab. 1994;14:944–951.
9. Vavilala MS, Lee LA, Lam AM. Cerebral blood flow and vascular physiology.
Anesthesiol Clin N Am. 2002;20:247–264.
10. Sliwka U, Krasney JA, Simon SG, SchmidtP, Noth J. Effects of sustained low-
level elevationsof carbon dioxide on cerebral blood flow and autoregulation of
the intracerebral arteries in humans. Aviat Space Environ Med. 1998;69:299–
306.
11. Ito H, Kanno I, Ibaraki M, Hatazawa J, Miura S. Changes in human cerebral
blood flow and cerebral blood volume during hypercapnia and hypocapnia
measured by positron emission tomography. J Cereb Blood Flow Metab.
2003;23:665–670.
12. Appendix B—Overview of Acute Health Effects. 2013. Available at http://
www.epa.gov/ozone/snap/fire/co2/appendixb.pdf. Accessed August 8, 2013.
13. Wong KL. Carbon dioxide. Spacecraft Maximum Allowable Concentra-
tions for Selected Airborne Contaminants: Vol 2. Washington, DC: National
Academy Press; 1996:105–188.
14. Interim Report Meeting Minutes, 9 July 2009, Houston, TX. Available at
http://iss-www.jsc.nasa.gov/nwo/ppco/cbp_miocb/bbt_docs/bbtcal/Minutes.6
.09-Jul-2009.htm. Accessed October 1, 2009.
15. Marshall-Bowman K, Barratt MR, Gibson CR. Ophthalmic changes and in-
creased intracranial pressure associated with long duration spaceflight: an
emerging understanding. Acta Astronaut. 2013;87:77–87.
16. Mader TH, Gibson CR, Pass AF, et al. Optic disc edema, globe flatten-
ing, choroidal folds, and hyperopic shifts observed in astronauts after long-
duration space flight. Ophthalmology. 2011;118:2058–2069.
17. James J, Meyers V, Sipes W, Scully R, Matty C. Crew health and performance
improvements with reduced carbon dioxide levels and the resource impact to
accomplish those reductions. Proceedings of the 41st International Conference
on Environmental Systems. Portland; 2011. AIAA 2011–5047.
18. Gray G, Johnston SL. Medical evaluations and standards. In: Barratt MR,
Pool SL, eds. Principles of Clinical Medicine for Space Flight. NY: Springer;
2008;59–68.
19. Marshburn TH. Acute care. In: Barratt MR, Pool SL, eds. Principles of
Clinical Medicine for Space Flight. NY: Springer; 2008;101–122.
20. Ortega H, Harm D. Space and entry motion sickness. In: Barratt MR, Pool
SL, eds. Principles of Clinical Medicine for Space Flight. NY: Springer;
2008;211–222.
21. Bacal K, Beck G, Barratt MR. Hypoxia, hypercarbia, and atmospheric control.
In: Barratt MR, Pool SL, eds. Principles of Clinical Medicine in Space Flight.
NY: Springer; 2008;445–474.
22. Erdmann CA, Apte MG. Mucous membrane and lower respiratory building
related symptoms in relation to indoor carbon dioxide concentrations in the
100-building BASE dataset. Indoor Air. 2004;14(suppl 8):127–134.
23. Federspiel CC, Fisk WJ, Price PN, et al. Worker performance and ventilation
in a call center: analyses of work performance data for registered nurses.
Indoor Air. 2004;14(suppl 8):41–50.
24. Milton DK, Glencross PM, Walters MD. Risk of sick leave associated with
outdoor air supply rate, humidification, and occupant complaints. Indoor Air.
2000;10:212–221.
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
482 C2014 American College of Occupational and Environmental Medicine
JOEM rVolume 56, Number 5, May 2014 CO2and Headache Reports on International Space Station
25. Sepp¨
anen OA, FiskWJ, Mendell MJ. Association of ventilation rates and CO2
concentrations with health and other responses in commercial and institutional
buildings. Indoor Air. 1999;9:226–252.
26. Shendell DG, Prill R, Fisk WJ, Apte MG, Blake D, Faulkner D. Associations
between classroom CO2concentrations and student attendance in Washington
and Idaho. Indoor Air. 2004;14:333–341.
27. WargockiP, Wyon DP, Sundell J, Clausen G, Fanger P. The effects of outdoor
air supply rate in an office on perceived air quality, sick building syndrome
(SBS) symptoms and productivity. Indoor Air. 2000;10:222–236.
28. Satish U, Mendell MJ, Shekhar K, et al. Is CO2an indoor pollutant? Direct
effects of low-to-moderate CO2 concentrations on human decision-making
performance. Environ Health Perspect. 2012;120:1671.
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
C2014 American College of Occupational and Environmental Medicine 483
  • [Show abstract] [Hide abstract] ABSTRACT: Twenty-five subjects were exposed to different levels of carbon dioxide (CO2 ) and bioeffluents. The ventilation rate was set high enough to create a reference condition of 500 ppm CO2 with subjects present; additional CO2 was then added to supply air to reach levels of 1,000 or 3,000 ppm, or the ventilation rate was reduced to allow metabolically-generated CO2 to reach the same two levels (bioeffluents increased as well). Heart rate, blood pressure, End-Tidal CO2 (ETCO2 ), oxygen saturation of blood (SPO2 ), respiration rate, nasal peak flow and forced expiration were monitored and the levels of salivary α-amylase and cortisol were analyzed. The subjects performed a number of mental tasks during exposures and assessed their levels of comfort and the intensity of their acute health symptoms. During exposure to CO2 at 3,000 ppm, when CO2 was added or ventilation was restricted, ETCO2 increased more and heart rate decreased less than the changes that occurred in the reference condition. Exposure to bioeffluents, when metabolically-generated CO2 was at 3,000 ppm, significantly increased diastolic blood pressure and salivary α-amylase level compared with pre-exposure levels, and reduced the performance of a cue-utilisation test: these effects may suggest higher arousal/stress. A model is proposed describing how mental performance is affected by exposure to bioeffluents. This article is protected by copyright. All rights reserved.
    Full-text · Article · Feb 2016
  • [Show abstract] [Hide abstract] ABSTRACT: This article briefly reviews the fidelity of ground-based methods used to simulate human existence in weightlessness (spaceflight). These methods include horizontal bed rest (BR), head-down tilt bed rest (HDT), head-out water immersion (WI), and head-out dry immersion (DI; immersion with an impermeable elastic cloth barrier between subject and water). Among these, HDT has become by far the most commonly used method, especially for longer studies. DI is less common but well-accepted for long-duration studies. Very few studies exist that attempt to validate a specific simulation mode against actual microgravity. Many fundamental physical and thus physiological differences exist between microgravity and our methods to simulate it, and between the different methods. Also, although weightlessness is the salient feature of spaceflight, several ancillary factors of space travel complicate Earth-based simulation. In spite of these discrepancies and complications, the analogs duplicate many responses to 0 G reasonably well. As we learn more about responses to microgravity and spaceflight, investigators will continue to fine-tune simulation methods to optimize accuracy and applicability.
    Article · Feb 2016
  • [Show abstract] [Hide abstract] ABSTRACT: In order to gain insights into microgravity-induced ophthalmic changes (Microgravity Ocular Syndrome), we investigated the effect of acute hypercapnia following 10-d bed rest and hypoxia on posterior eye structures. Female subjects (N=7) completed three 10-day experimental interventions: i) normoxic bedrest (NBR; PIO2=132.9+0.3 mmHg) ii) hypoxic ambulatory confinement (HAMB; PIO2=90.4+0.3mmHg), and iii) hypoxic bedrest (HBR; n=12; PIO2=90.4+0.3mmHg). Before, and on the last day of each intervention, optical coherence tomography (OCT) of the optic disc was performed and the thicknesses of the retinal nerve fibre layer (RNFL), retina and choroid were measured. OCT examinations were conducted with the subjects breathing the prevailing normocapnic breathing mixture (either normoxic or hypoxic) and then following a 10-min period of breathing the same breathing mixture, but with the addition of 1% CO2 Choroidal thickness was greater during both bedrest conditions (NBR and HBR) compared the ambulatory (HAMB) condition (ANOVA, p<0.001). Increases in RNFL thickness compared to baseline were observed in the hypoxic trials (HBR, p<0.001; and HAMB, p=0.021), but not the normoxic trial (NBR). A further increase in RNFL thickness (p=0.019) was observed after the 10-min hypercapnic trial in the NBR condition only. The fact that choroidal thickness was not affected by PO2and PCO2, but increased by bedrest, suggests a hydrostatic rather than a vasoactive effect. The increements in RNFL thickness were most likely associated with local hypoxia and hypercapnia-induced dilatation of the retinal blood vessels.
    Article · Mar 2016
Show more

We use cookies to give you the best possible experience on ResearchGate. Read our cookies policy to learn more.