AMERICAN JOURNAL OF INDUSTRIAL MEDICINE 34:574–580 (1998)

Cosmic Radiation and Magnetic Field

Exposure to Airline Flight Crews

Joyce S. Nicholas, PhD,1,2* Daniel T. Lackland, DrPH,1,2 Gary C. Butler, PhD,3

Lawrence C. Mohr, Jr., MD,2,4 John B. Dunbar, DrPH,1,2 William T. Kaune, PhD,5
Bernd Grosche, PhD,6 and David G. Hoel, PhD1,2

Background Flight crews are exposed to elevated levels of cosmic radiation and to magnetic
fields generated by the aircraft’s electrical system. The purpose of this study was to quantify
these two occupational exposures.
Methods Magnetic fields were measured during 37 flights (23 in the cockpit and 14 in the
cabin) using an Emdex Lite personal dosimeter. All cockpit measurements were taken on the
B737/200. Cabin measurements were taken in several aircraft types, including the B737,
B757, DC9, and L1011. Cosmic radiation was computer estimated for 206 flights using the
Federal Aviation Administration’s program CARI-3C.
Results Magnetic field levels in the cockpit had a mean value of approximately 17 milliGauss
(mG), while cabin measurements were lower (mean values of approximately 3 or less in
economy, 6 in first class, 8 in front serving areas). Cosmic radiation equivalent dose rates to
bone marrow and skeletal tissue ranged from 0.3 to 5.7 microsieverts per hour.
Conclusions Elevated magnetic field levels in front serving areas and the cockpit suggest the
need for further study to evaluate long-term exposure to flight crew members who work in
these areas. Cosmic radiation levels are well below occupational limits for adults, but may
require some pregnant flight crew members to adjust their flying time or routes. Am. J. Ind.
Med. 34:574–580, 1998. r 1998 Wiley-Liss, Inc.

KEY WORDS: cosmic radiation; magnetic fields; exposure assessment; pilots; flight
attendants

INTRODUCTION levels of cosmic radiation and to magnetic fields generated

The airline industry may be an occupational setting with
specific health risks. Flight crews are exposed to elevated
1Department of Biometry and Epidemiology, Medical University of South
by the aircraft’s electrical system. Other factors to be
considered are circadian disruption, and conditions specific
to air travel, such as noise, vibration, mild hypoxia, reduced
atmospheric pressure, low humidity, and air quality. Health
studies have identified certain disease risks. Incidence
studies suggest an increase in hormone-sensitive tumors: for

Carolina, Charleston
2Environmental Hazards Assessment Program, Medical University of South
Carolina, Charleston
3Airline Pilots Association International, Herndon, VA
4Department of Medicine, Medical University of South Carolina, Charleston
5EM Factors, Richland, WA
6BfS, Federal Office for Radiation Protection, Oberschleissheim, Germany
*Correspondence to: Joyce S. Nicholas, Ph.D., Medical University of South
Carolina, Department of Biometry and Epidemiology, 550 MUSC Complex, Suite
1148, Charleston, SC 29425. E-mail nicholjs@musc.edu.
Accepted 1 August 1998
Contract grant sponsor: U.S. Department of Energy; Contract grant number:
DE-FC02-98CH10902.
example, breast cancer in both Finnish [Pukkala et al., 1995]
and Danish [Lynge, 1996] flight attendants, prostate cancer
in Air Canada pilots [Band et al., 1996], and testicular cancer
in United States Air Force pilots [Grayson and Lyons, 1996].
While breast tissue is known to be sensitive to ionizing
radiation, the sensitivity of the prostate appears to be low
[BEIR, 1990]. Suppression of the pineal hormone melatonin
has been implicated in the etiology of both breast and
prostate cancer. Although unproven in humans, it has been
hypothesized that magnetic field exposure may be associated
with reduced melatonin levels [Stevens et al., 1992; Stevens

r 1998 Wiley-Liss, Inc.

 Exposure to Airline Flight Crews 575

TABLE I. Cockpit Magnetic Field (in milliGauss): Statistical Values Based on 44,408 Measurements
(taken at 4-second intervals)*

Standard
Mean Minimum Maximum deviation Median

Broadband resultant 16.8 2.3 160.4 7.7 15.4

X 14.3 0.1 82.1 7.9 13.2
Y 4.3 0.1 39.4 2.4 3.9
Z 6.2 0.2 157.2 4.1 5.6

*Broadband refers to the frequency range 40 to 1,000 Hz. The broadband resultant, Br , is a scalar quantity derived from the magnitude of the three
sensor readings taken from the Emdex Lite. It is defined mathematically for each measurement as Br ⫽ 冑Bx2 ⫹ By2 ⫹ Bz2 where Bx , By , and Bz are the
magnitude of the field in the direction of the x, y, and z sensors, respectively. The above values are the statistical results taken over all 44,408
measurements.

TABLE II. Cockpit Magnetic Field (broadband resultant in milliGauss) METHODS

for Flight Segments in Figure 1: Statistical Values Based on
Measurements Taken at 4-Second Intervals*
Magnetic Fields
Standard
Mean Minimum Maximum deviation Median
Magnetic fields were measured onboard aircraft during
37 flight segments. Measurements were taken in the cockpit
Edmonton-Vancouver 11.3 5.1 19.8 2.9 11.4
during 23 of these flights, and in the passenger cabin during
Vancouver-Calgary 21.0 2.5 37.5 8.6 22.1
the remaining 14 flights. All measurements were taken using
Calgary-Toronto 13.3 3.2 38.7 4.2 13.0
an Emdex Lite personal dosimeter set at a sampling interval
of 4 seconds.
*Broadband refers to the frequency range 40 to 1,000 Hz. The broadband resultant, Br , is a
The Emdex Lite personal dosimeter is capable of
scalar quantity derived from the magnitude of the three sensor readings taken from the measuring magnetic fields in the range 0.1 to 700 mG, with
Emdex Lite. It is defined mathematically for each measurement as Br ⫽ 冑Bx2 ⫹ By2 ⫹ B z2 an optimal resolution of 0.1 mG and a measurement
where Bx , By , and Bz are the magnitude of the field in the direction of the x, y, and z sensors,
respectively. The above values for Edmonton-Vancouver are the statistical results taken over
accuracy of ⫾ 4%. The frequency range is 40 to 1,000 Hz.
1,276 measurements; for Vancouver-Calgary, 1,139 measurements; for Calgary-Toronto, The dimensions of the meter (2.5 x 6.0 x 12.0 cm) permit it
3,691 measurements. to be worn without interfering in the investigator’s activities.
The 23 cockpit segments were flown on a B737/200
during February 1998, with the Emdex Lite carried by the
and Davis, 1996; NRC, 1996], as well as increased radiosen- captain in his shirt pocket. The captain maintained a flight
sitivity [Stevens, 1993; NRC, 1996; Hintenlang, 1993; log, including meter on/off times. The 14 cabin segments
Norimura et al., 1993; Zhang et al., 1995]. Mortality studies were flown on various U.S. commercial aircraft (including
among flight crews in Japan [Kaji et al., 1993], Canada B737, B757, DC9, L1011) during April and May of 1997,
[Band et al., 1996; Salisbury et al., 1991], and England with the Emdex Lite carried in the shirt pocket of a
[Irvine and Davies, 1992] have produced inconsistent re- passenger seated and walking in economy and first class
sults, perhaps due to small sample sizes. A larger mortality cabin compartments. The passenger maintained a log of his
study among United States commercial pilots and navigators movements within the cabin so that measurements could be
found significant increases for motor neuron disease and matched to time and cabin locations.
cancer of the kidney and renal pelvis, with suggested Measurement results are given in terms of the broad-
increases for prostate cancer, brain cancer, colon cancer, and band resultant in the unit milliGauss (mG), where 10 mG ⫽
cancer of the lip, buccal cavity, and pharynx [Nicholas, 1 µT (microTesla). Broadband refers to the frequency range
1997]. 40 to 1,000 Hz. The resultant, Br, is a scalar quantity derived
Quantification of occupational exposures is a necessary from the magnitude of the three sensor readings taken from
step in assessing their potential association with disease the Emdex Lite. It is defined mathematically as
findings. The purpose of this study was to quantify flight
crew exposure to cosmic radiation and to magnetic fields. 冑
Br ⫽ B2x ⫹ B2y ⫹ B2z
While cosmic radiation has been addressed by other investi-
gators, little information is available on aircraft magnetic where Bx, By, and Bz are the magnitude of the field in the
field exposures. direction of the x, y, and z sensors, respectively.
576 Nicholas et al.
FIGURE 1. Cockpit magnetic field for three consecutive flights. Magnetic field versus time in the cockpit (B737/200) for three consecutive Canadian flights
(Edmonton-Vancouver, Vancouver-Calgary, Calgary-Toronto). Block time for each flight is marked by dotted lines.
Cosmic Radiation
Cosmic radiation exposure was computer calculated
using the United States Federal Aviation Administration
computer program CARI 3-C. This program calculates the
galactic cosmic radiation equivalent dose to bone marrow
and skeletal tissue received on individual nonstop aircraft
flights between city pairs. Results are presented in terms of
total dose, as well as the contribution to the total by each of
the major radiation components (muons, electrons and
photons, protons, charged pions, and neutrons). Flight-by-
flight doses were used to estimate annual exposure.
Flight profiles for the CARI-3C calculations were
obtained from the FAA for 206 U.S. domestic and interna-
tional flight segments. Flight profile information for each
segment included flight date, origin and destination airports,
number of en route altitudes, minutes climbing to first en
route altitude, time and altitude for each en route altitude,
and minutes descending to destination airport. All flights
were flown during the years 1993–1995 on a variety of US
aircraft types, including B727, B747, B767, DC10, DC980,
MD80, and L1011.
RESULTS
Magnetic Fields
Measured magnetic field strength was found to vary
with stages of flight, location within the aircraft, and type of
aircraft. For the 23 cockpit segments flown on a B737/200,
descriptive statistics were based on block hours. Block hours
for each flight begin when the aircraft leaves the gate before
takeoff and end when it reaches the gate after landing.
Results of an analysis of combined block hours for all 23
flights (44,408 observations taken at 4-second intervals) are
shown in Table I. Table I gives the broadband resultant
(mean value approximately 17 mG), as well as the x, y, and z
magnetic field components. Although the maximum value of
the broadband resultant in Table I is 160.4 mG, 99% of the
values are equal to or below 38.2 mG.

FIGURE 2. Cabin magnetic field: Economy class and front serving area. Magnetic field versus time in the cabin (B737) for a portion of the flight time from New Orleans,
Louisiana, to Washington DC. The passenger was seated in economy class. The region of increased values corresponds to a walk to the front serving area.
As an example of cockpit measurements, Figure 1
shows a ‘‘field vs. time’’ graph for a sample of three
consecutive Canadian flights (Edmonton-Vancouver, Vancou-
ver-Calgary, Calgary-Toronto). Block time for each flight is
marked on the figure by dotted lines. The meter was left
running continuously, including the time between flights.
The spike in the third flight segment occurred in the
washroom directly behind the flight deck. Table II describes
the broadband resultant for each of the flights shown in
Figure 1.
The minimum resultant for the time between flights
(most of which was spent outside of the cockpit) was very
small (less than 1 mG).
For the 14 cabin segments, the magnetic field strength
varied with cabin position and aircraft type (types included
B737, B757, DC9, L1011). In general, the mean resultant
value in economy class was approximately 3 mG or less,
while the mean value in first class was approximately 6 mG.
A walk to the front of a B737, indicated that the mean
Exposure to Airline Flight Crews 577
resultant in the serving area was approximately 8 mG with
spikes as high as 17 mG.
Figures 2 and 3 are given as examples of cabin
measurements. Figure 2 shows the ‘‘field vs. time’’ graph for
a portion of the flight time from New Orleans, Louisiana, to
Washington, DC, onboard a B737. The passenger was seated
in economy class where the mean broadband resultant was
3.1 mG (minimum 1.3, maximum 5.4, standard deviation
0.7). The region of increased values in Figure 2 (maximum
16.9 mG) corresponds to a walk to the front serving area of
the aircraft (no microwave ovens were in operation while the
passenger was in the front serving area). Figure 3 shows
measurement samples in economy class and first class
compartments in two B757s. In the first part of Figure 3, the
passenger was seated in economy class on a B757 flying
from Charleston, South Carolina, to Atlanta, Georgia. The
mean resultant was 1.5 mG (minimum 0.7, maximum 2.3,
standard deviation 0.4). In the second part of Figure 3, the
passenger was seated in first class on a B757 flying from
578 Nicholas et al.
FIGURE 3. Cabin magnetic field: Economy class and first class. Magnetic field versus time in the cabin (B757). In the first part of Figure 3, the passenger was seated in
economy class flying from Charleston, South Carolina, to Atlanta, Georgia. In the second part of Figure 3, the passenger was seated in first class flying from Atlanta,
Georgia, to New Orleams, Louisiana.
Atlanta, Georgia, to New Orleans, Lousiana. The mean
resultant was 6.0 mG (minimum 3.5, maximum 12.3,
standard deviation 1.6).
Cosmic Radiation
Cosmic radiation equivalent dose to bone marrow and
skeletal tissue calculated by CARI-3C varied by altitude,
latitude, phase of solar cycle, and flight duration. Table III
shows altitude, flight time, and CARI-3C dose for a sample
of 22 flights dated March 1995 (near solar minimum).
Within this sample, dose rates varied from 0.3 to 5.7
microsieverts (µSv) per hour. The percent of total dose due
to muons varied from 3 to 20, with the higher percents found
in the shorter, lower altitude flights. In all flights, approxi-
mately one third of total dose was due to neutrons (30 to
37%), while only 1 to 2% was due to charged pions.
Electrons and photons varied from 24 to 41%; protons from
19 to 28%. The distribution of radiation types is important in
that higher energy radiations (neutrons, protons, pions) have
greater potential for biologic damage than lower energy
radiations (muons, electrons, photons).
Based on the total CARI-3C doses given in Table III,
the potential cosmic radiation equivalent dose to bone
marrow and skeletal tissue associated with air travel ranges
from 30 to 570 µSv per 100 flight hours (not including
ground time) depending on altitude, latitude, phase of solar
cycle, and flight duration. Often flight hours for a crew
member are recorded in terms of block hours. If dose rates
are needed in terms of block hours, then additional time
must be added to each flight for time spent on the ground
taxiing or waiting for takeoff. Estimates of ground time were
obtained from the FAA. Adjusting the dose range based on
flight time to account for ground time yields the reduced
range of 20 to 550 µSv per 100 block hours. According to the
1996 to 1997 Occupational Outlook Handbook by the U.S.
Department of Labor [DOT, 1997], flight attendants usually
fly 75 to 85 hours a month. Assuming that these are block
hours and using 960 block hours as a representative work
year, the estimated annual cosmic radiation equivalent dose
to bone marrow and skeletal tissue is 0.2 to 5.3 mSv.
DISCUSSION
The cockpit magnetic fields measured in this study had
a mean resultant value of 16.8 mG. However, it may be more
meaningful to describe the field by a band within which a
certain percentage of the measurements lie. For example,
 Exposure to Airline Flight Crews 579

TABLE III. Cosmic Radiation Equivalent Dose to Bone Marrow and Skeletal Tissue*

Total
Altitude % % Total Total dose
max:average % electrons % charged % dose air time rate in
Flight in 1,000s of fta muons ⴙ photons protons pions neutrons in ␮Sv in hrs ␮Sv/hr

Houston, TX, to Austin, TX 20:20 20 25 19 2 34 0.1 0.4 0.3

Seattle, WA, to Portland, OR 21:21 18 24 20 2 36 0.1 0.4 0.3
Miami, FL, to Tampa, FL 24:24 13 28 22 2 35 0.3 0.6 0.5
St Louis, MO, to Tulsa, OK 35:35 6 31 25 2 36 1.3 0.9 1.4
San Juan, Puerto Rico, to Miami, FL 35:34 5 36 24 2 33 4.3 2.2 2.0
Los Angeles, CA, to Honolulu, Hawaii 35:35 4 39 24 2 31 11.8 5.2 2.3
Honolulu, Hawaii, to Los Angeles, CA 40:38 3 41 24 2 30 13.7 5.1 2.7
New York, NY, to San Juan, Puerto Rico 37:36 4 35 25 2 34 8.4 3.0 2.8
Tokyo, Japan, to Los Angeles, CA 37:35 4 36 25 2 33 27.0 8.8 3.1
Washington, DC, to Los Angeles, CA 35:35 4 33 26 2 35 15.6 4.7 3.3
Chicago, IL, to New York, NY 37:37 4 31 27 2 36 5.5 1.6 3.4
London, UK, to Dallas Fort Worth, TX 39:33 4 30 27 2 37 34.0 9.7 3.5
Lisbon, Portugal, to New York, NY 39:36 4 32 27 2 35 24.8 6.5 3.8
Seattle, WA, to Anchorage, AK 35:35 4 30 27 2 37 13.5 3.4 4.0
New York, NY, to Chicago, IL 39:39 3 31 28 2 36 7.6 1.8 4.2
Chicago, IL, to San Francisco, CA 39:38 3 33 27 2 35 16.1 3.8 4.2
Seattle, WA, to Washington, DC 37:37 3 31 28 2 36 18.5 4.1 4.5
Tokyo, Japan, to New York, NY 41:37 3 31 28 2 36 56.4 12.2 4.6
New York, NY, to Tokyo, Japan 43:38 3 32 27 2 36 61.1 13.0 4.7
New York, NY, to Seattle, WA 39:37 3 31 28 2 36 23.0 4.9 4.7
Chicago, IL, to London, UK 37:37 3 30 28 2 37 34.8 7.3 4.8
Athens, Greece, to New York, NY 41:41 3 33 28 1 35 53.3 9.4 5.7

*Flight profiles provided by Wallace Friedberg, PhD, Federal Aviation Administration. All profiles dates 3-95, heliocentric potential 561 MV.
aTime averaged en route altitude.

approximately 90% of the cockpit measurements were
between 6 and 30 mG, with approximately 45% of these
above the mean. Regardless of how the cockpit field is
described, it is substantially above the field typically found
in the home or office (1 and 3 mG, respectively) [Kaune,
1993]. Further study is needed to evaluate long-term expo-
sure to these elevated fields.
Additional measurements are needed to clarify variation
in magnetic field levels in both the cabin and the cockpit. For
example, variation in the mean cockpit resultant value for
the three flights shown in Figure 1 is unexplained. In each
case, measurements were taken during night-time hours on
Canadian routes in the same cockpit position on the same
aircraft type. Mean resultant values (calculated separately)
for all 23 cockpit flights ranged from approximately 10 to
25 mG.
The computer-calculated cosmic radiation equivalent
dose to bone marrow and skeletal tissue associated with the
U.S. domestic and international flights modeled in this study
ranged from 30 to 570 µSv per 100 flight hours (not
including ground time). This range depends on altitude,
latitude, phase of solar cycle, and flight duration. Annual
doses estimated from these values range from 0.2 to 5.3
mSv. For comparison, the ICRP [ICRP, 1991] recommends
as exposure limits an effective dose of 1 mSv per year for
members of the general public and 20 mSv per year for an
occupationally exposed adult. Additional restrictions apply
to the occupational exposure of pregnant women. The
Commission’s policy is that the methods of protection at
work for women who may be pregnant should provide a
standard of protection for any conceptus that is broadly
comparable with that provided for members of the general
public. Therefore, once pregnancy has been declared, the
conceptus should be protected by applying a supplementary
equivalent dose limit to the surface of the woman’s abdomen
of 2 mSv for the remainder of the pregnancy [ICRP, 1991].
As an example of potential ionizing radiation exposure to a
pregnant crew member, consider a woman who declares
pregnancy after one month and flies to the end of the sixth
month. Assuming 80 block hours per month for each of the 5
remaining months, this woman would reach the 2 mSv limit
only if she received an average of 5 µSv per block hour.
580 Nicholas et al.
Using the dose rates in Table III as surrogates for dose rates
to the surface of the abdomen, an average of 5 µSv per block
hour would require a flight schedule consisting entirely of
long, high altitude flights (examples: Athens–New York,
Chicago–London).
In conclusion, additional information is needed to
assess the impact of occupational exposures (and their
potential synergy of effects) on flight crew members.
Extensive magnetic field measurements are currently being
made in the cockpits of several aircraft types to investigate
both the magnitude and frequency of the exposure fields.
Flight histories and health information are being collected
for comparison of career cosmic radiation dose to disease
outcomes on an individual basis. Any suggested associations
must then be supported with further study on biologic
mechanisms.
ACKNOWLEDGMENTS
The authors acknowledge Wallace Friedberg, PhD,
Federal Aviation Administration, for his consultation and for
providing the CARI-3C program and flight profiles used in
this study.
This study was supported in part by funds from the U.S.
Department of Energy cooperative agreement DE-FC02-
98CH10902.
REFERENCES
Band PR, Nhu DL, Fang R, Deschamps M, Coldman A J, Gallagher, RP,
Moody J (1996): Cohort study of Air Canada pilots: mortality, cancer
incidence, and leukemia risk. Am J Epidemiol 143:137–143.
Committee on the Biological Effects of Ionizing Radiations (BEIR) (1990):
‘‘Health Effects of Exposure to Low Levels of Ionizing Radiation.’’
Washington, DC: BEIR V, National Academy Press.
Grayson JK, Lyons TJ (1996): Cancer incidence in United States Air Force
Aircrew, 1975–89. Aviat Space Environ Med 67:101–104.
Hintenlang DE (1993): Synergistic effects of ionizing radiation and 60 Hz
magnetic fields. Bioelectromagnetics 14:545–551.
International Commission on Radiological Protection (1991): ‘‘1990 Rec-
ommendations of the ICRP, ICRP Publicaton 60.’’ Annals of the ICRP
21(1-3), Oxford: Pergamon Press.
Irvine D, Davies M (1992): The mortality of British Airways pilots,
1966–1989: A proportional mortality study. Aviat Space Environ Med
63:276–279.
Kaji M, Tango T, Asukata I, Tajima N, Yamamoto K, Yamamoto Y, Hokari
M (1993): Mortality experience of cockpit crewmembers from Japan
Airlines. Aviat Space Environ Med 64:748–750.
Kaune W T (1993): Introduction to power-frequency electric and magnetic
fields. Environ Health Perspect 101(Suppl 4):73–81.
Lynge E (1996): Risk of breast cancer is also increased among Danish
female airline cabin attendants. BMJ 312(7025):253.
National Research Council (NRC) (1996): ‘‘Possible Health Effects of
Exposure to Residential Electric and Magnetic Fields.’’ Washington, DC:
National Academy Press.
Nicholas JS (1997): ‘‘Cosmic Radiation and Magnetic Fields: Exposure
Assessment and Health Outcomes Among Airline Flight Crews.’’ Doctoral
dissertation, Medical University of South Carolina. Ann Harbor: UMI.
Norimura T, Imada H, Kunugita N, Yoshida N, Nikaido M (1993): Effects
of strong magnetic fields on cell growth and radiation response of human
T-lymphocytes in culture. Sangyo Ika Daigaku Zasshi 15:103–112.
Pukkala E, Auvinen A, Wahlberg G (1995): Incidence of cancer among
Finnish airline cabin attendants, 1967–92. BMJ 311:649–652.
Salisbury DA, Band PR, Threlfall WJ, Gallagher RP (1991): Mortality
among British Columbia pilots. Aviat Space Environ Med 62:351–352.
Stevens RG (1993): Biologically based epidemiological studies of electric
power and cancer. Environ Health Perspect 101(Suppl 4):93–100.
Stevens RG, Davis S (1996): The melatonin hypothesis: Electric power and
breast cancer. Environ Health Perspect 104(Suppl 1):135–140.
Stevens RG, Davis S, Thomas DB, Anderson LE, Wilson BW (1992):
Electric power, pineal function, and the risk of breast cancer. FASEB J
6:853–860.
U.S. Department of Labor (DOT) (1997): ‘‘1996–97 Occupational Outlook
Handbook.’’ Washington, DC 352.367-010.
Zhang XR, Kobayashi H, Hayakawa A, Ishigake T (1995): An evaluation of
the biological effects of three different modes of magnetic fields on cultured
mammalian cells. Nagoya J Med Sci 58:157–164.