Digital conversion from https://books.google.pt/books?

id=KIQqAAAAMAAJ

Disclaimer: not reviewed, please double check with the source

Review of Studies on Flight Attendant

Health and Comfort in Airliner Cabins
NIREN L. NAGDA AND MICHAEL D. KOONTZ

NAGDA NL, KOONTZ MD. Review of studies on flight attendant health and toward standardization. Topics covered in one of the earliest
comfort in airliner cabins Aviat Space Environ Med 2003; 74:101-9.
major reviews of the subject by the National Research Council
Background: A number of studies have examined the effect of the airliner
cabin environment and other factors on the health and comfort of flight
(NRC; 20), were the environmental control systems used on
attendants (FAs), but no comprehensive review of such studies is available. commercial aircraft carrying passengers, air quality in emergency
Methods: This paper reviews studies conducted after 1980 that addressed situations, standards and regulations, exposures to cabin air
FA short-term health and comfort effects. Relevant literature was identified pollutants (including environmental aspects such as cosmic
using the National Institute of Health's PUBMED database. Results: Twenty-
radiation, relative humidity, and pressurization), and occupant
one studies were identified and classified into two types: in-flight surveys
and surveys of general flight experiences. Most studies used questionnaires
health effects associated with exposure. The report
to obtain perceptions of the cabin environment, comfort, and health-related recommended the elimination of smoking on domestic flights and
symptoms, but some included objective measurements. Only a few studies other actions to address health and safety problems and to obtain
used a random sample or control groups. Effects of confounding variables better data on cabin air quality.
generally have not been analyzed. Discussion: Most studies shared some
In 2000, the U.S. Congress (Wendell H. Ford Aviation
weaknesses such as poor response rate, significant response bias,
exclusive reliance on questionnaires, or limited analysis. Taken together, Investment and Reform Act of the 21st Century) directed the
the studies indicate that various complaints and symptoms reported by FAs Federal Aviation Administration (FAA) to ask the NRC to assess
appear to be associated with their job duties and with the cabin environment air contaminants in commercial aircraft and their health effects
Most notable are "dryness" symptoms attributable to low humidity and and to recommend approaches to improve cabin air quality. The
"fatigue".
recent NRC report (21) ranks air quality characteristics as high
symptoms associated with factors such as disruption of circadian rhythm.
Practically all symptoms are exacerbated by longer flight durations. Studies
concern, moderate concern, or low concern based on likelihood
citing problems of poor aircraft cabin air quality tend to be weak in design of exposure and potential severity of health effects. The
and have addressed only general flight experiences of FAs. Although certain characteristics of high concern are cabin pressure and ozone.
FA complaints are consistent with possible exposure to air pollutants, the Recommendations in the NRC report include taking effective
relationship has not been proven and such complaints also are consistent
measures to ensure that current regulations for ozone are being
with causes other than poor air quality. Keywords: airlines, cabin air quality,
environmental monitoring, commercial aircraft, occupational groups,
met, investigating the need and feasibility of installing air cleaning
surveys, signs and symptoms. equipment and establishing a research and surveillance program.
Standardization efforts for cabin air quality have been led by
the American Society of Heating, Refrigerating, and Air-

T HE HEALTH AND COMFORT of flight attendants (FAs) can

be affected by a variety of environmental factors in airliner
cabins, such as air quality, temperature, humidity, pressure,
conditioning Engineers (ASHRAE). Efforts are underway to
develop an aircraft cabin air quality standard under Special
Projects Committee 161P.
noise, vibration, and time- zone shifts. Human response to the Previous review articles on the topic of the cabin environment
environment can depend on one's current state of physical and and the associated health and comfort of occupants generally
mental well-being, as influenced by factors such as fitness, have focused on passengers. For example, Rayman (24) has
amount of rest, duty cycle, and job-related or personal- life discussed stressors such as barometric pressure changes,
stressors. Discerning the relative impacts of these various factors immobility, jet lag, noise, and vibration that are faced by travelers
on perceived or measured FA health and comfort is a complex on commercial airliners, and he has reviewed the related aspects
task; nonetheless, in the published literature there has been a of passenger safety, cabin air quality, and preventive med-
tendency to cite poor cabin air quality as a leading cause of From ENERGEN Consulting, Inc., Germantown, MD.
This manuscript was received for review in December 2001. It was
apparent flight- related symptoms. revised in October 2002. It was accepted for publication in November 2002.
Cabin air quality has been the focus of recent reviews and efforts Address reprint requests to: Niren L Nagda, Ph.D., who is Principal
Scientist, ENERGEN Consulting, Inc., 19900 Wild Cherry Lane,
Germantown, MD 20674; [email protected].
Aviation, Space, and Environmental Medicine • VOL. 74, No. 2 • February 2003
101
 FLIGHT ATTENDANT HEALTH—NAGDA & KOONTZ

icine. The Passenger Health Subcommittee within the Air manifested in a number of different ways, and a variety of factors
Transport Medicine Committee of the Aerospace Medical can alter how the environment may affect individual FAs.
Association has prepared a position paper on the subject of cabin Throughout this paper, we refer to aspects of environmental quality
air quality, covering aspects such as pressurization, ventilation, as environmental factors, to potential health/comfort effects as
contaminants, temperature, and humidity (31). In an earlier paper, outcomes, and to potential confounding variables as personal or
Kraus (13) con- ducted a review of epidemiological studies of flight-specific factors.
health effects for airliner pilots and flight attendants, but found the
published literature to be sparse and suggested that a broader
review be undertaken. However, as yet there has not been any RESULTS
comprehensive review with a specific focus on flight attendants. Scope and Design of Reviewed Studies
The primary purpose of this paper is to provide a review of prior The numbers of participating flights, airlines, aircraft, and FAs
studies concerning the airliner cabin environment, primarily as it are indicated for the reviewed studies in Table I. The studies, which
affects the health and comfort of FAs. Beyond the reasons cited vary considerably in scope and design, fall almost evenly into two
above, the experiences of FAs are particularly worthy of review groups: (i) those that have asked questions relating to experiences
because they have recurring exposures to cabin environments on a specific flight (i.e., in-flight studies), and (ii) those that have
under hypobaric conditions, with job requirements involving asked much broader questions (i.e., general flight experience).
physical activity. Although the workload of FAs is not commonly Most studies have focused just on the cabin crew, but a few
viewed as physically demanding, it has been noted in the literature (6,12,15,33) have examined both cockpit and cabin crews and one
that "... the cabin crew's working, day is a hard one due to the (23) included perceptions of passengers in addition to FAs. Flight
sustained nature of the work rather than tremendous energy conditions represented by all in-flight studies included in the review
expenditure at any one moment" (1). were "normal" as opposed to "emergency" or "episodic" conditions.
Several of the earlier studies (25,27) involved a relatively large
sample size, on the order of 1000 FAs or more, but these studies
METHODS asked only about their general flight experience rather than
The scope of the review in this paper includes studies involving
experiences that could be related to the characteristics of
collection of primary data on effects or outcomes related to FA
individual flights. A study by Erneling et al. (8) about a decade later
health or comfort that may be associated with the cabin
used a similar approach. The earlier studies distributed the
environment. The review to restricted to short-term effects such as
questionnaires with a monthly trade magazine for FAs. Only a few
discomfort, health-related symptoms, and acute illnesses
studies have used a truly random sample, and the earlier studies
(including exacerbation of chronic conditions); longer-term out-
did not include follow-up efforts to maximize response rates and
comes such as cancer or reproductive health effects are excluded.
thereby minimize potential non-response bias. In a study by
Studies involving either subjective measures (i.e., through
Galipault (9) where a limited set of FAs was selected "at random,"
questionnaires), including reported symptoms, or objective
the participants were chosen from a pool of volunteers. Although
assessments (i.e., through measurements and/or tests) of comfort
there were only 30 subjects, a noteworthy feature of this study is
or health are included. Also within the scope are prior studies with
its longitudinal aspect—participants were asked to make
direct measurements of the quality of the airliner cabin envi-
subjective assessments of sleepiness, fatigue, and stress over a
ronment (including climate, concentrations of air contaminants,
61-d period, in- ducting both duty and off-duty days.
lighting, noise, and motion), or with questions aimed at discerning
Investigations using broad questionnaires that were
FA perceptions of environmental quality. Among cabin
independent of specific flights tend to be the studies with the
environment factors, studies of cosmic radiation are excluded
largest number of respondents, but not necessarily the highest
because the focus of this review is on short-term effects and acute
response rates. For example, Smolensky et al. (27) had 3412
illnesses.
respondents, but a response rale of only 17%, whereas a later
The literature search was performed using the National Institute
study by Erneling et al. (8) had 908 respondents but a much higher
of Health's PUBMED database. Keywords and phrases used in the
response rate of 81%. In cases where onboard administration of
search included aircraft cabin environment, cabin crew health,
questionnaires was central to the approach, the number of flights,
cabin air quality, and flight attendants. The search was restricted
airlines, and aircraft were fairly limited, and in some cases these
to articles published since 1980. The database also was searched
numbers were not even reported. Study subjects typically were
for "related articles'' for the most relevant articles. The literature
predominantly female with an average age between 30 and 35 yr.
identified through the database search was supplemented by
The Enck et al. study (6) had equal numbers of males and females,
relevant articles and reports cited in selected articles.
but included both cockpit and cabin crew to achieve this balance.
Assessing the effects of the airliner cabin environment on FA
Among the in-flight studies, some (14,23) have used questions
health and comfort is a relatively complex process, because the
aimed at discerning outcomes such as perceptions
quality of the environment is multifaceted, potential effects can be

102 Aviation, Space, and Environmental Medicine • VoL 74, No. 2 • February 2003
FLIGHT ATTENDANT HEALTH—NAGDA & KOONTZ
TABLE I. PROFILE OF STUDY SAMPLES

Author, Year (Reference) General (G) # of Types of # of Flight Gender, % Age

or Flight- Airlines Aircraft Attendants
Specific (F) (Response F M Avg. ± SD Min Max
Rate)
1 2
Bassett & Spillane, 1987 (2) F (2) 1 — 26 (93%)3 61 39 — — —
Cone, 1984 (3) G 1 4 280 (28%)4 90 10 — <20 5 >60
de Ree et al., 2000 (5) F (50) 2 1 409 (68%) 73 27 34 — —
Enck et al., 1995 (6) G 1 — 131 (42%)6 527 48 32.1 ± 11.9 — —
Eng 1979 (7) G 18 — 774 (—) 95 5 — —8 —
Erneling et al., 1988 (8) G 1 4 908 (81%) 78 22 — <30 >50
Galipault, 1980 (9) G 9 — 30 (67%) 90 10 32.5 ± 3.6 24 43
Härmä et al., 1993 (11) F (3) 1 — 409 100 0 33.2 ± 7.3 21 50
Haugli et al., 1994 (12) G 1 4 565 (83%)10 76 24 — — —
Lee et al.,2000 (14) F (16) 1 3 185 — — — — —
Lindgren et al., 2000 (15) G 1 — 657 (81%)11 91 9 — — —
Marrison & Muir, 1988 (16) — 1 2 453 (51%) 83 17 — — —
Pelletier, 1998 (22) G 10 10+ 291 (3%)12 — — — — —
Pierce et al., 1999 (23) F (8) 1 1 27 (26%) — — — — —
Reed et al., 1980 (25) G 3 4 1330 (41%) 95 5 31 20 35+
Ross et al., 2000 (26) F (100)13 2 2 670 (46%) 62 38 35 20 58
Smolensky et al., 1982 (27) G 14 8 3412 (17%) 93 7 — 19 43
Suvanto et al., 1990 (28) F14 1 — 342 (—) 83 17 35 ± 5.6 23 51
Suvanto et al., 1993 (29) F(3) 1 — 4015 100 0 33 ± 6.9 21 50
Tashkin et al., 1983 (30) G 1 2 351 (8 to 55%) 81 19 — — —
Wieslander et al., 2000 (33) F (4) 1 1 3716 35 65 — — —

1
Number of flights in parentheses, where applicable.
2
Not reported.
3
Data for two flight attendants were excluded from the first of the two flights due to failure to collect sufficient number of urine samples.
4
A small group out of this sample (eight flight attendants) participated in a longitudinal study involving lung function measurements.
5
Exact range or minimum/maximum age not reported.
6
A total of 190 respondents from 450 cockpit crew and flight attendants included in the study; separate response rate for flight attendants not reported.
7
Statistics based on combination of cockpit and cabin crew.
8
The article stated that 70% of respondents were in the age range from 26 to 33.
9
Flight attendants selected from a total of 285 who had answered a questionnaire for the Suvanto et al., (28) study also reported here.
10
A total of 1240 respondents from 1464 cockpit crew and flight attendants included in the study; separate response rate for flight attendants not reported.
11
A total of 1513 respondents from 1857 cockpit crew and flight attendants included in the study; separate response rate or demographic information for flight
attendants not reported.
12
The low rate of response was attributed by the author to an observation that many respondents "accepted poor air quality as normal and not worth reporting."
13
Sixteen out of 100 flights included simultaneous air quality measurements.
14
Four flight segments included, but total number of flights was not reported.
15
Selected from 430 female flight attendants.
16
A total of 37 respondents from 40 cockpit crew and flight attendants included in the study; separate response rate for flight attendants was not reported.

or symptoms at a single point in time (or pertaining to an Ross et al. (26) included domestic and international flights
entire flight), whereas others (5,26) have asked questions that were equally divided between Boeing 747 and 777
before, during, and after flights. Wieslander et al. (33) took aircraft Pierce et al. (23) restricted their study to Boeing
advantage of a smoking ban to survey the same set of 777 aircraft, Wieslander et al. (33) examined only Boeing
FAs, on the same set of four flights, just before and two 767 aircraft, Tashkin et al. (30) studied only flights on 747
weeks after the ban. Studies with questions administered and 747SP aircraft, and Lee et al. (14) chose to examine
at multiple flight stages (including before and after flights), only wide-body aircraft.
or using the same subjects across multiple flights, Some investigators have focused strictly on outcomes
represent a design whereby subjects effectively serve as and paid little attention to environmental factors (22^5).
their own "controls." An alternative approach for studies However, the more recent investigations generally have
with no longitudinal aspects has been the use of a included objective environmental measurements.
separate control group. For example, Enck et al. (6) used Because these measurements are relatively resource
administrative employees—matched with study subjects intensive, the sample size (number of aircraft) has been
on age and sex—as controls, and Bassett and Spillane (2) correspondingly limited—24 flights monitored by de Ree
matched on age, sex, and degree of manual labor. et al. (5), 16 by Lee et al. (14), 16 by Ross et al. (26), 8 by
Lindgren et al. (15) also used office workers as a control Pierce et al. (23), 6 by Lindgren et al. (15), and 4 by
group, as well as an external reference group for a subset Wieslander et al. (33).
of items taken from a standardized questionnaire. The use of objective outcome measurements has been
Some more recent studies have been built around
specific types of flights or aircraft. For example, de Ree et 103
al. (5) administered symptom questionnaires to FAs on
polar flights that were equally balanced in terms of aircraft
with/without humidifiers and ozone converters. A study by

Aviation, Space, and Environmental Medicine • Vol. 74, No. 2 • February 2003
 FLIGHT ATTENDANT HEALTH—NAGDA & KOONTZ

TABLE II. STUDY FACTORS REPORTED AND ANALYZED

Factors1
AUTHOR, YEAR, PERSONAL FLIGHT SPECIFIC ENVIRONMENTAL OUTCOMES/SYMPTOMS
(REFERENCE) DEMOGRAPHIC HEALTH
HISTORY

BESSETT & SPILLANE, O2, T (urinary cortisol excretion

1987 (2) rate)
CONE, 1984 (3) Q Q Q, T (peek expiratory flow rate,
PEFR)
DE REE ET AL., 2000 (5) O (O3 converter, M (ozone, humidity) Q
humidification)
ENCK ET AL., 1995 (6) Q Q (flight duration) Q
ENG, 1979 (7) Q Q Q3
ERNELING ET AL., 1988 (8) Q Q Q Q
GALIPAULT, 1980 (9) Q Q Q Q
HÄRMÄ ET AL., 1993 (11) Q, T (salivary hormones4)
HAUGH ET AL., 1994 (12) Q Q (flight duration) Q
LEE ET AL., 2000 (14) M5 Q
LINDGREN ET AL., 2000 (15) Q Q Q, M6
MARRISON & MUIR, 1988 Q Q O (operational.7 workload) Q
PELLETIER, 1998 (22) Q
8
PIERCE ET AL., 1999 (23) Q, M Q
REED ET AL., 1960 (25) Q Q Q (altitude, flight duration) Q

ROSS ET AL., 2000 (26) Q (ECS operation, altitude, Q, M9 Q, T (oximetry)

flight duration)
SMOLENSKY ET AL., 1982 Q Q Q
SUVANTO ET AL., 1990 (28) Q Q
SUVANTO ET AL., 1993 (29) Q, T (oral temperature, visual
search)
TASHKIN ET AL., 1963 (30) Q Q, T (pulmonary function tests)10

WIESLANDER ET AL., 2000 Q M11 Q, T (acoustic rhinometry, nasal

—
(33) lavage,12 tear-film break-up
time)

1
M — environmental measurements; Q — questionnaires; O — observations; T— medical tests. Only those factors that were reported and used in analyses are shown
here; bask demographic factors (e.g., age, sex) and monitoring factors (temperature, relative humidity) were reported in a number of studies, but are not included in this
table.
2
Diary of activities and mood ratings.
3
Factors influencing eve comfort including vision status (no correction, glasses, and hard and soft lenses).
4
Radioimmunoassay for salivary melatonin and cortisol
5
Carbon dioxide, carbon monoxide, humidity, noise, respirable particles,
6
Carbon dioxide, humidity, respirable particles.
7
Time and base of departure, length of duty period, and seven-day rostered period.
8
Acrolein, bacteria, carbon dioxide, carbon monoxide, formaldehyde, fungi, humidity, noise, ozone, pressure, respirable particles, and total volatile organic chemicals.

9
Air velocity, carbon dioxide, humidity, infrared thermography, microbiological measurements, ozone, volumetric flow, pressure, and VOCs
10
Battery of pulmonary function tests included timed respiratory and maximal expiratory flow volume curves, airways resistance, thoracic gas volume at functional
residual capacity, dosing volume and the slope of Phase II of the single-breath nitrogen washout curve.
11
Carbon monoxide, formaldehyde, humidity, microorganisms, ozone, respirable particles.
12
Analyzed for eosinophilic cationic protein, myeloperoxidase, lysozyme, and albumin.

sporadic. A 1983 study by Tashkin et al. (30) included pulmonary personal or flight-specific factors—that they have included or
function tests for 21 out of 351 FAs. In a 1984 cross-sectional addressed (Table 11). Age and gender have been the primary
survey with 280 respondents, Cone (3) took peak expiratory flow demographic factors used in the re- viewed studies; in some
rate (PEFR) measurements before, during, and after flight for a cases the investigators did not include such factors in the
small subset of eight respondents. Several years later, Bassett and analysis, even though the information was collected. Factors
Spillane (2) tested the urinary cortisol excretion rate for 28 cabin related to personal history have included personality, vision
crewmembers. Two studies reported in 1993 used objective status, smoking, job/health/personal-life satisfaction, and baseline
measurements — Härmä et al. (11) measured circadian variations health. The flight-specific factor common to most studies has been
in salivary melatonin and cortisol and Suvanto et al. (29) monitored flight duration, but in a few cases aspects such as altitude and
circadian rhythm of body temperature and alertness. More re- type of aircraft have been considered. in one case, noted
cently, Ross et al. (26) measured arterial oxygen saturation using previously, ozone converters and/or humidifiers were intentionally
pulse oximetry. included or excluded on study flights.
The prior studies also can be distinguished according to the Some investigations have examined a broad range of
types of factors — environmental, outcome, and symptoms whereas others have focused on specific out-

104 Aviation, Space, and Environmental Medicine • Vol. 74, No. 2 • February 2003
 FLIGHT ATTENDANT HEALTH—NAGDA & KOONTZ
comes such as respiratory symptoms (33), eye discomfort (5,7), ozone. Tashkin et al. (30) reported that symptoms thought to be
gastrointestinal problems (6), or disruption of circadian rhythm ozone-related were associated with high-altitude flights on aircraft
(11,28,29). Until recently, environmental factors were addressed without ozone converters, but further pulmonary function tests on
only through questionnaires aimed at obtaining occupants' a small part of the sample failed to reveal any abnormalities. The
perceptions of, or satisfaction with, environmental quality. study did not include any measurements of ozone, but the types
Investigations reported in 2000 by de Ree (5), Lee et al. (14), of flights monitored were assumed to be associated with higher
Lindgren et al. (15), Pierce et al. (23) and Wieslander et al. (33) ozone exposure. The studies by Cone (3) and Reed et al. (25)
have included one or more of the following measurements: also indicated that aircraft flying at higher altitude than other
temperature, humidity, ozone, carbon monoxide and carbon aircraft were associated with a higher frequency of respiratory
dioxide, respirable particles, microorganisms, formaldehyde, and
symptoms. However, measurements of ozone or other air
noise. As noted earlier, the use of objective outcome
contaminants were not conducted and the prevalence of ozone
measurements has been fairly limited (7 of out of 21 studies) and
converters was not reported in these studies (3,25,30). In addition,
there has been no apparent tendency toward increased use of
smoking—a potential confounding factor—was permitted on
such measurements in the more recent studies.
flights when these studies were conducted. Respiratory symptoms
were reported by about two-thirds of the flight crew in the Cone
Study Findings (3) study, especially on flights longer than five hours, and results
As noted above, the reviewed studies fall almost evenly into of pulmonary function tests on FAs indicated a small but
those concerning general flight experiences and those concerning
statistically significant drop between pre- and post-takeoff.
experience on a specific flight. Based on surveys concerning
However, the study response rate was low (28%), only eight
general flight experiences, the types of complaints or symptoms
subjects participated in the follow-up pulmonary function testing,
reported by flight attendants (not necessarily in order of
and several additional limitations were explicitly noted by the
importance) include respiratory symptoms, including colds,
blocked nasal passages, and difficulty breathing (3,8,12,22); dry- authors.
ness or irritation of skin, eyes, and throat (7,12,27); bloating of In a recent study (5) an identical type of aircraft (Boeing 747-
stomach, intestinal complaints, or nausea (6,8); swelling or aching 400) was flown over the same route by two different airlines, one
of legs (8); lower back pain (12); headache or earache (8,22); having aircraft equipped with ozone converters and the other
dizziness or faintness (22); and fatigue, sleep disorders, or without converters. Ozone levels and reported symptoms were
disruption of circadian rhythm due to time-zone displacement measured for 45 flights, and usable symptoms data were collected
(8,9,12,22,27). for 230 cabin crewmembers on these flights. The ozone
Environmental factors cited by FAs as possible causes of such converters were effective in reducing ozone levels but differences
problems include low humidity, tobacco smoke, dust, draft, stuffy in ozone-related symptoms (coughing, tightness of chest,
or dry air, noise, poor illumination, and poor air quality (which may shortness of breath, and "breathing hurts") in the two cases were
reflect air humidity/ movement in addition to contaminants/odors). small and not statistically significant Further, within the sets of
Such indications from general surveys may be instructive but still flights by the respective airlines, there was no correlation between
must be considered as subjective or speculative; the in-flight ozone levels and ozone-related symptoms.
surveys are more likely to offer insights on specific causative Perception of dryness or relative humidity has been included in
factors that may be related to re- ported outcomes. Potential many of the reviewed studies, but only a subset
causal agents or factors addressed in such studies, discussed (5,14,15,23,26,33) has reported measurements of cabin air
below, have included environmental tobacco smoke (ETS), humidity. The average measured humidity in the cabin was in the
relative humidity, ozone, other aspects of cabin air quality, range of 2 to 15% relative humidity (RH), with one study (14)
disruption of circadian rhythm, and daily activities or lifestyles of
reporting 22%. Complaints of dry air were common in all these
flight attendants.
studies and included dry and stuffy nose, irritation of eyes, dry
In-flight studies have confirmed previously established
skin, and dry lips. Such complaints increased with longer flight du-
relationships such as the effect of smoking on particle levels (17).
rations (26). These findings are consistent with the research on
Wieslander et al. (33) showed that a smoking ban caused a
reduction of more than 95% in particle concentrations and effects of low humidity conducted in laboratory chambers and in
reduced ocular symptoms, headache, and tiredness; ETS buildings (19). Further studies conducted in buildings have shown
exposure was associated with an increase in ocular and general that the low humidity also can contribute to "sick building
symptoms, decreased tear-film stability, and alteration in nasal pa- syndrome" (i.e., nonspecific symptoms such as irritation of eyes,
tency. Although the sample size was relatively small, the study skin, upper airways, headache, and fatigue) and that an increase
results are strengthened considerably by the before- and after- in RH of about 10% alleviated such symptoms (19).
intervention design coupled with objective outcome measures. Complaints concerning cabin air quality, including ETS, ozone,
Adverse health effects of ozone in the ambient atmosphere are relative humidity, and other environmental factors, have been
well established and the NRC (21) has designated ozone levels in reported in a number of studies, but
the cabin as of high concern.

However, among the reviewed FA studies, there are none that

convincingly demonstrate a cause-effect relationship for

Aviation, Space, and Environmental Medicine • Vol. 74, No. 2 • February 2003

105
 FLIGHT ATTENDANT HEALTH—NAGDA & KOONTZ
only a few such studies have included measurements of air quality complaints/symptoms generally increases with increasing age or
parameters. Erneling (8) reported that cabin air quality, specifically years of service. Such findings are reinforced by studies that have
tobacco smoke, dry air, and "stale" air, caused the greatest FA focused on the cock- pit crew. For example, Gander et at. (10)
discomfort among environmental factors; however, no found that, among crewmembers on long-haul flights, subjects of
measurements of air quality parameters were conducted. In a age 50 to 60 yr averaged 35 times more sleep loss per day than
study by Eng (7), over 95% of respondents reported some eye subjects of age 20 to 30 yr, consistent with the effect of
discomfort, with similar problems reported by those with and physiological aging.
without contact lenses. The most common problems were None of the reviewed FA studies were designed specifically to
conjunctival redness and dry eyes, and smoking was indicated to examine effects of cabin altitude. In one study with an even
be the most noticeable factor causing eye discomfort Both of these distribution of Boeing 747 and Boeing 777 aircraft by design, the
studies were con- ducted during the time when smoking was 777s also flew at a higher cruise altitude, on average, but there
permitted on aircraft and, thus, the effect of air quality factors were no notable differences between aircraft in terms of oxygen
independent of smoking cannot be discerned. Pelletier (22) saturation levels based on oximetry measurements. (An in-flight
reported large numbers of complaints of severe headache, study on arterial oxyhemoglobin saturation of airline pilots, using
difficulty breathing, nausea, dizziness, sudden fatigue, and other pulse oximeters and reported by Cottrell et al. (4), indicated that
symptoms; however, this was a casual survey of flight attendants the mean saturation level decreased from 97.0% preflight to 88.6%
and passengers the scientific basis for procedures used and at cruise altitude, with wide variations in individual levels.)
presentation/ analysis of results appear to be quite limited. Two studies involving control groups (6,15) have supported the
Recent studies, where smoking was not permitted and air notion that cabin crew are more prone to the types of symptoms
quality was measured in parallel with health symptoms or objective listed above than administrative or office workers. In addition, two
health measurements, include those by Pierce (23) and Ross (26). studies (6,12) have indicated that cabin crew are more likely than
In the Pierce study, no elevated cabin air concentrations were cockpit crew to report problems such as gastrointestinal, skin, eye,
reported, but the number of flights monitored was small due to the or musculoskeletal disorders, especially for long- distance flights.
pilot nature of the study. Measured concentrations of air In the Pierce et al. (23) study, in which FAs and passengers
contaminants also were low in the Ross study. The authors (26) responded to the same set of questions administered while in
noted that "the environmental measurements on 777 aircraft are flight, the FAs consistently rated aircraft and comfort factors lower
within generally recognized acceptable levels and therefore the than the passengers and were more likely to relate symptoms to
absolute levels of the individual parameters do not identify them as flying. Some of these differences may be related to the higher
causes of the reported symptoms.'' metabolic and respiration rates that on-duty flight attendants would
Circadian rhythm has been a focus of a number of these studies have as compared with seated passengers.
(2,11,28,29). A notable finding from two 1993 studies (11,29), for One FA lifestyle factor that may relate to certain symptoms such
4- to 5-d trips crossing 10 time zones, was progressive as gastrointestinal complaints is eating habits. Two surveys (8,16)
desynchronization or disruption of circadian rhythm during the trips conducted in the 1980s reported a relatively high frequency of
followed by a resynchronization or restitution time of 5 d. Thus, FAs upper GI symptoms such as bloating of stomach and other
who routinely work on such routes could be in a state of disruption stomach complaints among FAs. Another cross-sectional survey,
or restitution nearly half of the time during a typical work year. commissioned by FA organizations a few years earlier (17),
These findings related to disruption and restitution were based on reported that most FAs have poor eating and changed bowel
studies of different parameters (hormones, oral temperature, and habits. The earlier findings related to upper GI have been
visual search) that vary with circadian rhythm; additionally, mood confirmed in a recent study (6) with two groups, long-haul and
changes were documented both during and after flights relative to short-haul aircrew, along with matched controls from ground-
pre-flight states. In both studies (11,29), a repeated measures based staff from the same airline. Aircrews reported significantly
design was used whereby female FAs were chosen randomly from more dyspeptic (upper gastrointestinal) symptoms than ground
larger groups and subjects served as their own controls. The staff, mainly with long-distance flying, and cabin crew reported
studies also included both subjective and objective outcome significantly more intestinal com- plaints than cockpit crew. The
measures. The combination of careful design and multiple use of age- and sex- matched controls by Enck et al. (6) removed
outcome measurements in these two studies lends greater a potential confounding factor—menstruation—from their analysis.
credence to their results. The nature of FA work activities may partially ex- plain fatigue-
The collective studies consistently indicate that the frequency of related symptoms. FA complaints of fatigue or tiredness have been
many reported symptoms or problems, particularly those related to mentioned in most surveys of general flight experiences. One
circadian rhythm or fatigue, tends to increase with longer flights, survey (8) reported that tiredness after flights is more common
rapid changes in time zones, and early morning or late-night than tiredness before or during flights. A rare longitudinal study
departure times (2,6,8,16,28). Females tend to be more greatly
affected than males, and the frequency of certain types of

106 Aviation, Space, and Environmental Medicine • Vol. 74, No. 2 • February 2003
 FLIGHT ATTENDANT HEALTH—NAGDA & KOONTZ
The reviewed studies consistently indicate that FAs have
(9), spanning 61 consecutive days and using randomly selected reported various types of complaints and symptoms that appear to
FAs, indicated that end-of-day fatigue was related to fatigue at the be associated with their job duties and working environment In the
beginning of the day, length of the duty cycle, and amount of selected studies that have inducted comparison groups surveyed
walking. Disruption of circadian rhythm, discussed above, also in parallel with FAs, FAs tended to report more frequent or severe
contributes to fatigue. complaints/symptoms than both cockpit crew and passengers. The
collective evidence from these studies further points to certain
Commentary on Study Design and Analysis types of symptoms that are more prevalent among FAs than other
The relative simplicity of cross-sectional studies often has occupational groups, and suggests some likely causative factors.
resulted in a relatively large sample size, but many such studies Most notable in this regard are various "dryness" symptoms at-
have been prone to low response rates and consequent tributable to the low-humidity environment, "fatigue" symptoms
opportunities for non-response bias (e.g., a greater tendency to associated with time-zone changes, and symptoms such as
respond among FAs with complaints). The potential impacts of bloating and headache/earache that may be related to the
selection, recall or other types of biases could not be assessed hypobaric environment A very dear and consistent finding is that
based on the information provided in the articles and reports re- practically all symptoms are exacerbated by longer flight durations
viewed in this paper. Findings from both cross-sectional and in- and that at least some are related to early morning or late- night
flight studies are strengthened by design strategies such as use of flights as well.
a control group, as with Bassett and Spillane (2), Lindgren et al. Several of the reviewed studies that were conducted before the
(15), and Enck et al (6), or by pre-stratification of the study sample implementation of smoking bans have shown or confirmed that
by factors such as aircraft, route or equipment, as with de Ree et ETS exposure is irritating to FAs in addition to nonsmoking
al. (5), Reed et al. (25), and Ross et al. (26). passengers. One study in particular, with measurements before
Personal factors, although included in many studies, have not and after institution of a smoking ban on aircraft, demonstrated that
always been handled carefully in the data analysis to explore or FA exposure to ETS was associated with an increase in ocular and
adjust for their potential confounding effects. Two studies (15/25) general symptoms, decreased tear-film stability, and changes in
performed a more robust analysis by using logistic regression to nasal patency. Such findings serve to complement the body of
account for some of these factors. Results of the Reed et al. (25) evidence concerning deleterious effects of secondhand exposure
study indicated that flight duration was a significant factor for all of to tobacco smoke.
the 14 symptoms examined via logistic regression; Lindgren et al. Conclusions related to health effects of ozone are well
(15) found that females more often complained about low established in the criteria document published by the US.
temperature, dry air, and dust; and Enck et al. (6) noted a higher Environmental Protection Agency (32). However, four studies
frequency of gastrointestinal problems for females in both the reviewed here examined effects of ozone in terms of health-related
study and the control groups. Use of age- and sex-matched control symptoms for FA, and none of these demonstrated a cause-effect
groups (2,6,15), or subjects serving as their own "controls" through relationship. Direct measurements of ozone were included in only
repeated measures (e.g., before and after flights, as with a number one of the four studies; those measurements demonstrated the
of the in-flight studies), also strengthen the analysis by explicitly effect of ozone converters in reducing cabin ozone levels. Potential
addressing some of the personal factors. effects of altitude itself, other than the higher associated ozone
Due to the relatively small sample size, studies that have levels, have not been dearly established from the reviewed
included objective environmental or outcome measurements often studies.
have not added much clarity to the overall picture. One study (26) The findings of the reviewed studies relating to disruption of
with a fairly large sample size (100 flights) did not report any circadian rhythm for FAs are quite conclusive. Various objective
analysis relating environmental factors to outcomes. Similarly, little outcome measures that vary with circadian rhythm have been used
attention has been paid to how objective measurements correlate to demonstrate its disruption during flights and its subsequent
with subjective assessments of environmental conditions or health- restitution after flights, and mood changes have been documented
related symptoms. The limited cases with data to support such both during and after flights relative to preflight states. For 4-to 5-
analyses have not, as yet, established relationships between d trips where as many as 10 time zones are crossed, the restitution
subjective and objective measures. time is on the order of 5 d.
The collective findings also indicate that FAs are particularly
CONCLUSIONS prone to gastrointestinal problems and bloating of the stomach,
Most of the studies reviewed in this paper shared some more so titan both cockpit crew and ground-based airline staff, and
weaknesses such as poor response rate, significant response to fatigue. In addition to the nature of the working environment,
bias, exclusive reliance on questionnaires, or limited analysis. poor eating
Studies that focused on circadian rhythm generally were
exceptions. Most studies that did include objective measurements
had a relatively small sample size. Nonetheless, there are some
common findings from these studies that support certain types of
conclusions.

Aviation, Space, and Environmental Medicine • Vol. 74, No. 2 • February 2003 107
 FLIGHT ATTENDANT HEALTH—NAGDA & KOONTZ

TABLE III. FACTORS AND OUTCOMES WARRANTING

habits among FAs have been cited as a potential causative factor Consideration in future research.
for intestinal problems. Factors contributing to end-of-day fatigue
for FAs include fatigue at the beginning of the day, length of the Major Areas Specific Factors/Symptoms
duty cycle, and disruption of circadian rhythm. Another symptom
consistently experienced by FAs is skin/eye/throat dryness or Demographic Factors Age, gender, mortal status
irritation. The pervasive low humidity in the airliner cabin Height, weight
Years of experience
environment is a likely causative factor here.
Homebase
Personal History Baseline physical health (e.g., health status,
Factors vision status, chronic and acute conditions)
Among the air quality contaminants that have been
Baseline state of mind (e.g., attitudes, stress)
measured on aircraft in the reviewed studies and in other studies
of cabin air quality, only for ETS (prior to smoking bans) is there Work schedule, aircraft usually flown
evidence of an association with acute health outcomes. Measured Orcadian rhythm, sleeping pattern
levels of airborne contaminants, other than ozone and ETS Eating smoking habits
tracers, have been relatively low in aircraft cabins (also see, for Professional or union affiliation
Hight-Specific Aircraft type and size
example, 18). Factors Flight duration, time zones
Time of day
From the results of studies using questionnaires, it is
Flight route
apparent that FAs have real symptoms that they attribute to poor Environmental Air movement, ventilation
cabin air quality. However, with the possible exceptions of ETS, Factors Temperature, humidity
ozone, and low humidity, there is no direct evidence from there Airborne contaminants
Odon
studies to support such a claim. One particularly intriguing
Altitude
observation is that FAs tend to rate cabin air quality as poor Lighting
whereas objective measurements indicate that contaminant con- Noise
centrations in cabin air typically are well below levels of concern. Vibration, motion
Studies (e.g., 22) that cite problems of "poor aircraft cabin air Outcomes Fatigue
Aching / swelling / bloating
quality" have addressed only the general flight experiences of FAs
Dizziness / fainting
and suffer from a relatively weak study design; such studies have Headache
not included objective measures of cabin air quality and have Earache
tended to make speculations tied to presumed associations such Eye irritation
Skin irritation
as those between cabin air quality and flight- related symptoms.
Note irritation
Although certain FA complaints are consistent with possible Throat irritation
exposure to pollutants in aircraft cabins, such a relationship has Respiratory distress/ complaints (shortness of
not been proven. Moreover, such complaints also are consistent breath and irritation) and pulmonary function
with causes other than poor cabin air quality. In particular, the
Nausea/sickness
collective evidence points to disruption of circadian rhythm, job
duties and lifestyles, and low humidity as likely causative factors.

the aircraft environment than either the cockpit crew or

passengers.
A major limitation of most studies has been the lack of The types of factors that need to be considered for any
objective measures of environmental conditions and outcomes. In investigation of the cabin environment and its potential effects are
cases where such measurements have been included, the summarized in Table III. The environmental factors listed in the
sample size has been quite limited or the focus has been quite table cover the cabin climate and air quality (including odors),
narrow, or both. Also, in some of these studies, beyond providing altitude or pressure, noise, lighting, and vibration or motion. The
descriptive statistics there has been little or no assessment of the outcomes include general systemic symptoms such as fatigue,
extent of correlation between subjective and objective measures, aching/swelling, or nausea/sickness as well as more specific
for either cabin conditions or health symptoms. symptoms such as headache and earache, eye/skin/nose/throat
irritation, and respiratory complaints. Three types of potential
confounding factors are listed—demographics, personal history,
and flight specific; the first two types relate to individual FAs,
FUTURE RESEARCH whereas the third concerns specific aircraft or flights included in
an investigation. Research that does not acknowledge the
To prove or disprove a cause and effect relationship between
potential roles of environmental factors, outcomes, and
the cabin environment and FA health symptoms, it would be
confounding factors, or that does not consider them carefully in the
necessary to monitor a large number of flights for cabin
environmental factors, including air pollutants. In parallel, a health study design, may arrive at misleading results or inferences.
survey of FAs that includes objective outcome measures would be Based-on this review, it is advised that future studies of the
required. The NRC (21) has recommended that cabin crew should airliner cabin environment and FA health or symptoms should
be considered as the primary study group for future surveillance have all or most of the following
and research programs. Indeed, as apparent from this review, FAs
appear to be more sensitive to

108 Aviation, Space, and Environmental Medicine • Vol. 74, No. 2 • February 2003
FLIGHT ATTENDANT HEALTH—NAGDA & KOONTZ
features, if the current state of understanding is to be trip flights over 10 time zones on the circadian variation of
salivary melatonin and cortisol in airline flight attendants.
improved: Ergonomics 1993; 37:1479-89.
Collection of data for all important variables, supplementing 12. Haugli L., Skogstad A., Hellesøy O.H. Health, Sleep, and mood
subjective measures with objective ones where possible and perceptions reported by airline crews flying short, and long hauls.
Aviat Space Environ Med 1994; 65:27-34.
practical;
13. Kraus JF. Epidemiological studies of health effects in commercial
Study design-strengthening features such as control pilots and flight attendants: a review. Los Angeles. CA: School of
groups and multiple measurement points (e.g., preflight in Public Health. University of California, Los Angeles; 1985.
addition to one or more in-flight stage, or before and after an 14. Lee SC, Poon CS, Li XD, Luk F, et al. Questionnaire survey to
evaluate the health and comfort of cabin crew. In: Nagda NL, ed.
intervention); Air quality and comfort in airliner cabins: ASTM STP 1393. West
Adequate sample size; and Conshohocken, PA: American Society for Testing and Materials;
Careful and complete analysis of data, including use of 2000:259-68.
15. Lindgren T, Norbäck D, Andersson K. et al. Cabin environment and
multivariate techniques to assess relationships among perception of cabin air quality among commercial aircrew. Aviat
personal, flight-specific, environmental, and outcome Space Environ Med 2000; 71:774-82.
measures. 16. Marrison C. Muir H. Cabin staff’s perception of the impact of flying
on their physical health. Aviat Med Quarterly 1968; 2:11-7.
Such studies, especially with adequate sample size, will 17. Nagda NL, Fortmann RC, Koontz MD, et al. Airliner cabin
necessarily be quite resource intensive. In the absence of environment contaminant measurements, health risks and mitiga-
such research, the ambiguities surrounding questions related tion options. Washington, DC US Department of Transportation.
Office of the Secretary; 1989 Dec Report No: DOT-P-15- 89-5.
to effects of air quality and other factors on FA health and
18. Nagda NL, Rector HE, Li Z. et al. Aircraft Cabin air quality: a critical
comfort will persist. review of past monitoring studies. In: Nagda NL, ed. Air quality
and comfort in airliner cabins: ASTM STP 1393. West
Conshohocken, PA: American Society for Testing and Materials;
ACKNOWLEDGMENTS 19.
2000:215-39.
Nagda NL, Hodgson M. Low relative humidity and aircraft cabin air
This research was supported by The Boeing Company under quality. Indoor Air 2001; 11:200-14.
contract number KF3308. Richard Johnson and David Space 20. National Research Council The airliner cabin environment air
served as the project managers. Aleksandra Buganski, also of
quality and safety, Washington, DC: National Academy Press;
Boeing, assisted in the identification and initial review of the
1986.
papers. Suneel Nagda, MD., Loyola University Medical Center,
21. National Research Council. The airliner cabin environment and the
provided a medical perspective for this review. Russell Rayman,
health of passengers and crew. Washington, DC National
M.D., Aerospace Medical Association, reviewed a draft al the
Academy Press; 2002.
paper and provided helpful comments.
22. Pelletier F. The perspective of Canadian flight attendants on cabin
air issues. Toronto, Canada: Canadian Union of Public Em-
ployees. Airline Division; 1998 Jun.
23. Pierce MW, Janczewski JN, Roethlisberger B, et al. Air quality on
REFERENCES commercial aircraft. ASHRAE J 1999 Sep; 41(9):26-34.
1. Barnes R.M. Physical energy expenditure in long haul cabin crew. 24. Rayman RB. Passenger safety, health, and comfort: a review. Aviat
Revue de Medecine Aeronautique et Spatiale 1973; 46:335-7. Space Environ Med 1997; 68:432-40
2. Bassett JR. Spillane R. Urinary Cortisol excretion and mood ratings in 25. Reed D, Glaser S., Kaldor J. Oxone toxicity symptoms among flight
aircraft cabin crew during a tour of duty involving a disruption in attendants. Amer J Industrial Medicine 1960; 1:43-54.
circadian rhythm, Pharmacology Biochemistry and Behavior 26. Ross D., Hadi M., Palmer J, et al. Cabin air quality monitoring in
1987; 27:413-20 Boeing 777s. United Kingdom: Centre for Safety, Health, and
3 Cone JE. Cabin air causalities a Hight attendant health survey. San Environment; 2000 Dec Report No. 201543.
Francisco CA: San Francisco General Hospital; 1984. 27. Smolensky MH, Lee E, Mott D, et al. A health profile of american
4. Cottrell JJ, Lebovitz BL, Fennell RG, et al. Inflight arterial satura- flight attendants (FA). J Hum Ergol 1982; 11(Suppl): 103-19.
tion: continuous monitoring by pulse oximetry. Aviat Space 28. Suvanto S., Partinen M., Härmä M., et al. Flight attendants'
Environ Med 1995; 66:126-30. desynchronosis after rapid time sone changes. Aviat Space
5. de Ree H. Bagshaw M, Simons R, et al. Ozone and relative humidity Environ Med 1990; 61543-7.
in airline cabins on polar routes: measurements and physical 29. Suvanto S, Härmä M, Ilmarinen J. et al. Effects of 10 h time zone
symptoms, ln: Nagda NL, ed. Air quality and comfort in airliner changes on female flight attendants' circadian rhythms of body
cabins: ASTM STP 193. West Conshohocken, PA: American temperature, alertness, and visual search. Ergonomics 1993;
Society for Testing and Materials; 2000:243-58. 36:613-25.
6. Enck P, Müller-Sack E., Holtmann G, et al. Gastrointestinal problems 30. Tashkin DP, Coulson AH, Simmons MS, et al. Respiratory symp-
in airline crew members. Z Gastroenterol 1995; 33:513-6. toms of flight attendants during high-altitude flight: possible
7. Eng W.G. Survey on eye comfort in aircraft: I. Flight attendants. relation to cabin ozone exposure. Int Arch Occup Environ Health
Aviat Space Environ Med 1979; 50:401-4. 1963; 52:117-37.
8. Erneling L, Orring R. Joachimsson A. Cabin attendants' working 31. Thibeault C Cabin air quality. Aviat Space Environ Med 1997; 68:80-
environment study. Stockholm: Occupational Health and Safety; 2.
1988 Sept 32. US Environmental Protection Agency. Air quality criteria for ozone
9. Galipault J.B. A study of airline flight attendant sleepiness, fatigue and related photochemical oxidants. EPA/600/P-93/ 004acF.
and stress. Worthington. OH: The Aviation Safety Institute; 1980 Washington. DC: EPA; July 1996.
Sept Report Nor 80-2. 33. Wieslander G, Lindgren T, Norbäck D, et al. Charges in the ocular
10. Gander PH. De Nguyen BE, Rosekind MR. et al. Age, circadian and nasal signs and symptoms of aircrews in relation to the ban
rhythms, and sleep loss in fight crews. Aviat Space Environ Med on smoking on intercontinental flights. Scand J Work Environ
1993; 64:189-95. Health 2000; 26(6)514-22.
11. Härmä M, Laitinen J, Partinen M, et al. The effect of four-day round

109
Aviation, Space, and Environmental Medicine • Vol. 74, No. 2 • February 2003