Spiegel - Sephton JNCI 2002 COMMENT
5 Pages
English
Downloading requires you to have access to the YouScribe library
Learn all about the services we offer

Spiegel - Sephton JNCI 2002 COMMENT

Downloading requires you to have access to the YouScribe library
Learn all about the services we offer
5 Pages
English

Description

tors for breast cancer that are associated sleep disruption with cancer risk may bewith sleep disruption. For example, our mediated by a number of pathways, in-laboratory has recently reported, also in cluding, but not limited to, the disrup-tion of melatonin secretion. There arethe Journal, that loss of normal diurnalnow sufficient data to support the notionvariation in cortisol predicts early mor-that the loss of normal diurnal variationtality in metastatic breast cancer (3).in cortisol is another possible mecha-Cortisol normally shows marked diurnalnism that could account for the observedvariation, peaking in early morning andassociation between shift work anddeclining throughout the day. In ourbreast cancer incidence.study, patients with flattened salivarycortisol rhythms or aberrant peaks and DAVID SPIEGELtroughs suffered earlier mortality. The SANDRA SEPHTONprognostic effect of cortisol rhythm onREFERENCESsurvival emerged approximately 1 yearafter cortisol assessment and extended at(1) Schernhammer ES, Laden F, Speizer FE, Wil-least 7 years after the assessment (3). A lett WC, Hunter DJ, Kawachi I, et al. Rotatingsimilar association between 24-hour night shifts and risk of breast cancer in womenparticipating in the nurses’ health study. J Natlrest–activity rhythms and survival hasCancer Inst 2001;93:1563–8.been noted in patients with colorectal(2) Davis S, Mirick DK, Stevens RG. Night shiftcancer (4). Furthermore, patients at highwork, light ...

Subjects

Informations

Published by
Reads 29
Language English

Exrait

tors for breast cancer that are associated sleep disruption with cancer risk may be
with sleep disruption. For example, our mediated by a number of pathways, in-
laboratory has recently reported, also in cluding, but not limited to, the disrup-
tion of melatonin secretion. There arethe Journal, that loss of normal diurnal
now sufficient data to support the notionvariation in cortisol predicts early mor-
that the loss of normal diurnal variationtality in metastatic breast cancer (3).
in cortisol is another possible mecha-Cortisol normally shows marked diurnal
nism that could account for the observedvariation, peaking in early morning and
association between shift work anddeclining throughout the day. In our
breast cancer incidence.study, patients with flattened salivary
cortisol rhythms or aberrant peaks and DAVID SPIEGEL
troughs suffered earlier mortality. The SANDRA SEPHTON
prognostic effect of cortisol rhythm on
REFERENCESsurvival emerged approximately 1 year
after cortisol assessment and extended at
(1) Schernhammer ES, Laden F, Speizer FE, Wil-
least 7 years after the assessment (3). A lett WC, Hunter DJ, Kawachi I, et al. Rotating
similar association between 24-hour night shifts and risk of breast cancer in women
participating in the nurses’ health study. J Natlrest–activity rhythms and survival has
Cancer Inst 2001;93:1563–8.been noted in patients with colorectal
(2) Davis S, Mirick DK, Stevens RG. Night shift
cancer (4). Furthermore, patients at high
work, light at night, and risk of breast cancer.
risk for primary breast cancer show ab-
J Natl Cancer Inst 2001;93:1557–62.
normal circadian patterns among an ar- (3) Sephton SE, Sapolsky RM, Kraemer HC,
ray of hormones including cortisol (5). Spiegel D. Diurnal cortisol rhythm as a pre-
dictor of breast cancer survival. J Natl CancerSimilar endocrine disturbances have
Inst 2000;92:994–1000.been linked with advancing age and
(4) Mormont MC, Waterhouse J, Bleuzen P, Gi-
with sleep loss in young subjects (6). In
acchetti S, Jami A, Bogdan A, et al. Marked
our study, patient self-reports of sleep
24-h rest/activity rhythms are associated with
disruption were associated with flatten- better quality of life, better response, and
ing of the diurnal cortisol rhythm (3). longer survival in patients with metastatic co-
lorectal cancer and good performance status.There is evidence that the early morning
Clin Cancer Res 2000;6:3038–45.transition from dim to bright light not
(5) Ticher A, Haus E, Ron IG, Sackett-Lundeenonly suppresses melatonin secretion; it
L, Ashkenazi IE. The pattern of hormonal cir-
also induces a marked elevation of cor-
cadian time structure (acrophase) as an asses-
tisol levels (7). Thus, sleep disruption
sor of breast-cancer risk. Int J Cancer 1996;
coupled with exposure to light at night 65:591–3.
may not only suppress the melatonin (6) Van Cauter E, Plat L, Leproult R, Copinschi
G. Alterations of circadian rhythmicity andpeak but may also interfere with the nor-
sleep in aging: endocrine consequences. Hormmal cortisol nadir. Although mere sleepRe: Night Shift Work, Light at
Res 1998;49:147–52.disruption was not associated with el-Night, and Risk of Breast (7) Leproult R, Colecchia EF, L’Hermite-
evated breast cancer risk in the Davis etCancer Baleriaux M, Van Cauter E. Transition from
al. study (2), it is likely that shift work- dim to bright light in the morning induces an
ers have altered diurnal cortisol patterns immediate elevation of cortisol levels. J Clin
Two studies recently published in the engendered through alterations of sleep Endocrinol Metab 2001;86:151–7.
Journal showed that disruption of diur- and nutritional patterns, abnormal photic
NOTESnal sleep–wakefulness rhythms, espe- stimuli, and altered rest–activity cycles.
cially through nighttime shift work, was In addition, the stress incurred from fre- Affiliations of authors: D. Spiegel, Department
associated with higher subsequent breast quent transitions between daytime and of Psychiatry and Behavioral Sciences, Stanford
University School of Medicine, Stanford, CA; S.cancer incidence (1,2). Both of the ar- nighttime shift work should not be over-
Sephton, Department of Psychiatry and Behavioralticles and the accompanying editorial looked. Indeed, chronic sleep debt has
Sciences, University of Louisville School of Medi-cited melatonin suppression as a likely been linked with the disruption of car-
cine, KY.
mediator via reduced suppression of bohydrate metabolism and thyroid hor-
Correspondence to: David Spiegel, M.D., Dept.
ovarian estrogen production due to light mone function, elevation of sympathetic of Psychiatry and Behavioral Sciences, Stanford
exposure during the time when melato- nervous system activity, as well as with University School of Medicine, Stanford, CA
nin would otherwise be highest. It has hypothalamic–pituitary–adrenal axis 94305–5718 (e-mail: dspiegel@stanford.edu; web
address: pstlab.stanford.edu).also been postulated that melatonin dysfunction. Because these effects are
serves as an antioxidant in tumor cells similar to those seen in normal aging, it
and stimulates cytokine release from ac- is reasonable to propose that sleep debt Our interest in the article “Night Shift
tivated T cells. may increase the incidence or severity of Work, Light at Night, and Risk of Breast
The relationship between sleep dis- age-related diseases, including cancer, Cancer” (1) began with the media cov-
ruption and breast cancer risk may alter- by processes similar to those seen with erage of the research in a number of
natively be explained by other risk fac- advancing age. Thus the association of newspapers across the country. Shift
530 CORRESPONDENCE Journal of the National Cancer Institute, Vol. 94, No. 7, April 3, 2002and reduced risk of breast cancer in youngwork is already known to have a nega- over and above that of other workers
women. J Natl Cancer Inst 1994;86:1403–8.must be solidly established. The datative impact on health, work perfor-
mance, and social life (2–5). If this re- displayed in Table 3 show that the study
NOTESyielded 54 night workers in the casesearch (1) is correct, then there is a
group and 37 in the control group. Afurther risk of malignancy, providing the
2 Affiliations of authors: D. Porock, Universitysimple test on these data does notshift work includes sufficient time on
of Missouri Sinclair School of Nursing,reach statistical significance (P>.1).the night shift [no wonder Davis et al.
of Missouri-Columbia; J. Gentry, University ofThere are many possible variables(1) refer to it as the “graveyard shift”].
Wisconsin-Madison Comprehensive Cancer Care
that have a relationship to the develop-The study by Davis et al. (1) is based Center.
ment of breast cancer that were not re-on 1) the link between exposure to es- Correspondence to: Davina Porock, Ph.D., Uni-
ported, for example, ethnicity or socio- versity of Missouri/Columbia, MU Sinclair Schooltrogen and the risk of breast cancer and
economic status (6) and physical of Nursing, S428 Nursing Bldg., Columbia, MO2) the fact that exposure to light at night
65211–2222 (e-mail: porockD@missouri.edu).exercise (7). In fact, economic status inincreases exposure to estrogen through
particular may affect the need to workmechanisms associated with the circa-
on the graveyard shift. A statisticallydian rhythm. This is not an illogical train A very thoughtful and thorough edi-
significant difference between the caseof thought. Four studies were cited [ref- torial (1) and two most interesting
patients and control subjects was founderences within (1)], reporting an in- and valuable publications (2,3) focus
for those having worked the graveyardcreased risk of breast cancer in women attention on the role of light and mela-
shift for 5.7 years or more. This may
who work during the night. Of the four tonin production and the risk of breast
indeed show a need to obtain more data
studies, three were conducted in Scandi- cancer.
on this interesting potential risk factor.
navia, a geographic location known for In 1990, our group put forward a
Questions regarding the risks to
its long nights in winter, and two fo- comprehensive hypothesis that one of
women for breast cancer are important,
cused on airline crews, who have greater the most important etiologic factors in
and the need for research that investi-
exposure to cosmic radiation. Further- the increasing rate of cancers is the
gates all possible factors, particularly
more, in the “Discussion” section, Davis change in light exposure that has taken
those that are potentially under a per-
et al. (1) make reference to research that place during the last 100 years (4). The
son’s control, are essential. Neverthe-
found women with bilateral blindness introduction of electric light generally
less, the need for research with robust
had a statistically significantly lower increased the average daily light expo-
data is imperative if the scientific com-
risk of breast cancer, thus supporting the sure by 4–6 hours. The increase in light
munity is to ensure that women have
notion that there is an association be- exposure decreases the amount of time
access to the quality information that
tween light at night and breast cancer that is available for melatonin produc-
they deserve.
risk. tion, which reduces the nonspecific on-
However, in Davis et al. (1), the data costatic effect of the pineal gland. ThisDAVINA POROCK
collection is based on recall of occupa- adverse effect of light exposure, accord-JOHN GENTRY
tional history, sleeping habits, bedroom ing to our hypothesis, may increase the
REFERENCESenvironment over a decade, and other risk of other cancers. Beral et al. (5) re-
known risk factors. To minimize the ported that the exposure to fluorescent
(1) Davis S, Mirick DK, Stevens RG. Night shift limitations of recalled data, the light in the workplace was associated
work, light at night, and risk of breast cancer.
manuscript would have benefited from a with a twofold increase in the risk ofJ Natl Cancer Inst 2001;93:1557–62.
discussion of the reliability of the re- melanoma compared with a matched(2) Fitzpatrick JM, While AE, Roberts JD. Shift
called data, the development of the work and its impact upon nurse performance: control group. The risk increased with
current knowledge and research issues. J Advquestionnaire, any pilot work, and the increasing duration of exposure to fluo-
Nurs 1999;29:18–27.training of the interviewers. rescent light, with a relative excess of
(3) Karlsson B, Knutsson A, Lindahl B. Is thereTable 3 in Davis et al. (1) reported lesions on the trunk (5). By contrast,
an association between shift work and having
the findings on occupational history— blind women who are not ocularly re-
a metabolic syndrome? Results from a popu-
the one variable that is likely to have ceptive to light are not influenced by thelation based study of 27,485 people. Occup
some reliability because, understand- extended light exposure and have a re-Environ Med 2001;58:747–52.
ably, women who have worked the (4) Munakata M, Ichi S, Nunokawa T, Saito Y, Ito duced risk of breast cancer (6). Women
N, Fukudo S, et al. Influence of night shift“graveyard shift” would remember the with a milder degree of visual impair-
work on psychologic state and cardiovascularyears of that employment. The major ment did not have a similar reduced
and neuroendocrine responses in healthythrust of the authors’ work is predicated risk (6).
nurses. Hypertens Res 2001;24:25–31.
on a small difference between the num- Our group surveyed workers who,(5) van Amelsvoort LG, Schouten EG, Maan AC,
ber of women among the case patients (n every day, worked for 6–7 hours inSwenne CA, Kok FJ. Changes in frequency of
54 of 767 [7%]) and those among the premature complexes and heart rate variability darkness, producing films at the Cana-
related to shift work. Occup Environ Medcontrol group (n 37 of 743 [5%]) who dian Kodak factory. We found that the
2001;58:678–81.had ever worked a night shift. Although film production workers had a reduced
(6) Baquet CR, Horm JW, Gibbs T, Greenwald P.the difference may be statistically sig- risk of breast cancer and of malignant
Socioeconomic factors and cancer incidence
nificant in this analysis, the question is melanoma compared with workers in
among blacks and whites. J Natl Cancer Inst
whether a 2% difference is clinically administration or other manufacturing1991;83:551–7.
significant. A in the rates of areas (Kerenyi N: unpublished find-(7) Bernstein L, Henderson BE, Hanisch R, Sul-
breast cancer for night shift workers livan-Halley J, Ross RK. Physical exercise ings).
Journal of the National Cancer Institute, Vol. 94, No. 7, April 3, 2002 CORRESPONDENCE 531The conclusion of the editorial by factors for breast cancer, such as stress specifically, the control of the hormone
Dr. J. Hansen (1) emphasizes that there and related cortisol levels. For example, melatonin (7), a phenomenon that has
is an urgent need for further exploration studies have shown that cortisol levels been poorly understood until very re-
regarding the effects of light exposure are low during the night shift but that cently. It was further determined that
on cancers. I strongly agree with this married nurses had higher cortisol levels wavelengths of light in the blue region
conclusion. In 1973, el-Domeiri and Das than single nurses (4). Such observa- of the visible spectrum (e.g., fluorescent
Gupta (7) found that the accelerated tions support the notion that domestic light, commonly used in environments
growth of transplanted melanoma in stress, possibly mediated through role with artificial light) have the highest po-
pinealectomized hamsters could be re- conflicts among married nurses, alters tency in causing changes in melatonin
versed by the addition of exogenous cortisol levels in shift workers. Dr. Spie- levels.
melatonin. Considering all these facts, gel’s and recent other work have impli- Light at night is one of the most com-
raising the question of whether melato- mon occupational exposures in our in-cated that altered cortisol levels predict
nin replacement could be beneficial for survival time among breast cancer pa- dustrialized societies. Therefore, further
the light-exposed night shift workers is research to improve our current under-tients. Yet, their appealing theory that a
justifiable. A controlled clinical study standing of the effects of light at nighthealthy woman’s breast cancer risk may
may answer this question. on human health is prudent.also be elevated through altered cortisol
levels (caused by sleep deprivation andNORBERT KERENYI EVA SCHERNHAMMER
anticipated higher stress levels among
GRAHAM COLDITZREFERENCES shift workers) still requires confirma-
tion.(1) Hansen J. Light at night, shiftwork, and breast REFERENCES
Alternatively, there is evidence re-cancer risk. J Natl Cancer Inst 2001;93:
1513–5. garding the responsiveness of the pineal
(1) Hansen J. Increased breast cancer risk among(2) Davis S, Mirick DK, Stevens RG. Night shift gland to changes in emotional state,
women who work predominantly at night. Epi-work, light at night, and risk of breast cancer. thereby inducing alterations in melato-
demiology 2001;12:74–7.J Natl Cancer Inst 2001;93:1557–62.
nin levels. Rodent studies suggest that (2) Davis S, Mirick DK, Stevens RG. Night shift(3) Schernhammer ES, Laden F, Speizer FE, Wil-
the suppression of melatonin by light work, light at night, and risk of breast cancer.lett WC, Hunter DJ, Kawachi I, et al. Rotating
J Natl Cancer Inst 2001;93:1557–62.could be modified by changes in emo-night shifts and risk of breast cancer in women
(3) Schernhammer ES, Laden F, Speizer FE, Wil-participating in the nurses’ health study. J Natl tional state, produced by adversive con-
lett WC, Hunter DJ, Kawachi I, et al. RotatingCancer Inst 2001;93:1563–8. ditioning. On the contrary, in humans,
night shifts and risk of breast cancer in women(4) Kerenyi NA, Pandula E, Feuer G. Why the activation induced by physical stress in
participating in the nurses’ health study. J Natlincidence of cancer is increasing: the role of the middle of the dark phase did not alter Cancer Inst 2001;93:1563–8.
‘light pollution’. Med Hypotheses 1990;33:
melatonin levels (5). (4) Goldstein IB, Shapiro D, Chicz-DeMet A,
75–8.
Guthrie D. Ambulatory blood pressure, heartThe strongest biologic evidence to(5) Beral V, Evans S, Shaw H, Milton G. Malig-
rate, and neuroendocrine responses in womendate supports cancer risk emerging di-nant melanoma and exposure to fluorescent
nurses during work and off work days. Psy-
lighting at work. Lancet 1982;2:290–3. rectly from the exposure to light at night
chosom Med 1999;61:387–96.
(6) Kliukiene J, Tynes T, Andersen A. The risk of through the melatonin pathway. Artifi-
(5) Monteleone P, Maj M, Fuschino A, Kemali D.
breast cancer among Norwegian women with cial light was repeatedly shown to pro- Physical stress in the middle of the dark phase
visual impairment. Br J Cancer 2001;84: foundly suppress humans’ melatonin does not affect light-depressed plasma mela-
397–9.
levels and, in particular, those of women tonin levels in humans. Neuroendocrinology
(7) el-Domeiri AA, Das Gupta TK. Reversal
1992;55:367–71.(6). On average, melatonin concentra-by melatonin of the effect of pinealectomy
(6) Graham C, Cook MR, Gerkovich MM, Sastreon tumor growth. Cancer Res 1973;33: tions decrease by approximately 35%
A. Examination of the melatonin hypothesis in2830–3. after 2 weeks of intermittent nightly
women exposed at night to EMF or bright
light exposures. Moreover, the widely
light. Environ Health Perspect 2001;109:NOTE established antiproliferative effect of 501–7.
melatonin through its antioxidative ac-Correspondence to: Norbert Kerenyi, M.D., (7) Brainard GC, Hanifin JP, Greeson JM, Byrne
FRCP(C), FRC Path., Dept. of Laboratory tivity and potential other immunomodu- B, Glickman G, Gerner E, et al. Action spec-
Medicine and Pathobiology, St. Michael’s Hospi- trum for melatonin regulation in humans: evi-lating mechanisms is not only limited to
tal, 30 Bond St., Toronto, Ontario M5B 1W8 dence for a novel circadian photoreceptor.modulations on estrogen receptors,
Canada. J Neurosci 2001;21:6405–12.
thereby affecting breast cancer risk, but
has already been extended to other can-RESPONSES
NOTEScers.
We thank Dr. Spiegel et al. for their Based on these novel findings, we
interesting contribution. propose that the exposure to light at Affiliations of authors: E. Schernhammer,
The elevated breast cancer risk ob- night and subsequent melatonin sup- Channing Laboratory, Department of Medicine,
Brigham and Women’s Hospital and Harvardserved among women who work rotat- pression may also affect other cancers in
Medical School, Boston, MA; G. Colditz, Depart-ing night shifts (1–3) stimulates novel humans. We believe that our hypothesis
ment of Epidemiology, Harvard School of Publichypotheses for underlying biologic is given further importance by recent
Health, Channing Laboratory, Department of
causes of potential health risks associ- evidence for a unique photopigment in
Medicine, Brigham and Women’s Hospital and
ated with exposure to artificial light at the human eye that mediates circadian Harvard Medical School, Harvard Center for Can-
night. One avenue is to view work at photoreception. The new photoreceptor cer Prevention, and Epidemiology Program, Dana-
night as a proxy for other potential risk influences the biologic effects of light, Farber/Harvard Cancer Center, Boston.
532 CORRESPONDENCE Journal of the National Cancer Institute, Vol. 94, No. 7, April 3, 2002Correspondence to: Eva Schernhammer, M.D., may be important in the development of that such approaches may lead to a bet-
M.P.H., Channing Laboratory, 181 Longwood Ave., breast cancer and express concern that ter understanding of the health impacts
Boston, MA 02115 (e-mail: eva.schernhammer@
our results are derived from self- of exposure to LAN and the resulting
channing.harvard.edu).
disruption of normal circadian biology.reported data obtained through an in-
person interview. The limitations of this
Drs. Spiegel and Sephton rightly SCOTT DAVIS
approach are well recognized, and we
point out that the association we re- DANA MIRICK
used several techniques (e.g., structured
ported (1) between being employed in RICHARD STEVENS
interview, pilot tests, extensive inter-graveyard shift work and an increased
viewer training and monitoring, and a REFERENCESrisk of breast cancer might be accounted
random sample of abbreviated re-for by mechanisms other than or in ad-
(1) Davis S, Mirick DK, Stevens RG. Night shiftinterviews) to help ensure that the datadition to suppression of the normal noc-
work, light at night, and risk of breast cancer.were collected in an unbiased mannerturnal rise in melatonin, and they sug-
J Natl Cancer Inst 2001;93:1557–62.and were as accurate as possible. Thegest that loss of diurnal variation in (2) Czeisler CA, Klerman EB. Circadian and
potential impacts of imperfect or biasedcortisol levels might be one such alter- sleep-dependent regulation of hormone release
recall were considered in our originalnative. Women who engage in night in humans. Recent Prog Horm Res 1999;54:
paper (1). Porock and Gentry are also 97–130; discussion 130–2.shift work are subject to the influence of
(3) Stevens RG, Rea MS. Light in the built envi-concerned that our primary findings areboth sleep disruption and exposure to
ronment: potential role of circadian disruptionbased on a small difference in graveyardlight-at-night (LAN). Sleep deprivation
in endocrine disruption and breast cancer.
shift work between cases and controlscan have a profound effect on endocrine
Cancer Causes Control 2001;12:279–87.
and that a simple statistical test of thefunction and hormones such as melato- (4) Leproult R, Colecchia EF, L’Hermite-
difference is not significant. Although anin and cortisol (2), and it is well estab- Baleriaux M, Van Cauter E. Transition from
relatively small number of study partici- dim to bright light in the morning induces anlished that exposure to LAN can affect
immediate elevation of cortisol levels. J Clinpineal function and the production of pants worked the graveyard shift, the
Endocrinol Metab 2001;86:151–7.more relevant comparison is that a con-melatonin (3). Spiegel and Sephton cite
(5) Ticher A, Haus E, Ron IG, Sackett-Lundeen
recent evidence (4) of a similar effect on siderably higher proportion of those
L, Ashkenazi IE. The pattern of hormonal cir-
cortisol. In addition to the prognostic who did work the graveyard shift were cadian time structure (acrophase) as an asses-
significance of altered cortisol patterns cases (60%) than were controls (40%). sor of breast cancer risk. Int J Cancer 1996;
that they describe, there is beginning to These and similar statistically signifi- 65:591–3.
(6) Zheng W, Jin F, Dunning LA, Shu XO, Dai Q,emerge some limited evidence (5,6) that cant results summarized in Table 3 of
Wen WQ, et al. Epidemiological study of uri-our original paper (1) are more informa-cortisol may be associated with the risk
nary 6beta-hydroxycortisol to cortisol ratiosof developing breast cancer as well. tive and also allow for the independent
and breast cancer risk. Cancer Epidemiol Bio-
If breast cancer risk is affected by the effects of other factors related to breast
markers Prev 2001;10:237–42.
suppression of the normal nocturnal rise cancer risk when evaluating the effect of (7) Funk D, Amir S. Conditioned fear attenuates
in melatonin, one would predict that 1) graveyard shift work. We concur that light-induced suppression of melatonin release
in rats. Physiol Behav 1999;67:623–6.persons who do not perceive light on the the potential importance of our findings
(8) Stevens RG. Electric power use and breastretina (e.g., those who are blind) would dictates that future research be thought-
cancer: a hypothesis. Am J Epidemiol 1987;be at a decreased risk of breast cancer fully designed and carefully executed,
125:556–61.
and 2) those who work in occupations and we hope that our results help direct
that are characterized by circadian dis- the design of the next generation of epi-
NOTES
ruption (e.g., flight attendants) and those demiologic and laboratory studies.
who work the graveyard shift in what- We appreciate the thoughtful com- Affiliations of authors: S. Davis, Ph.D., Pro-
ever occupation would be at an in- ments of Dr. Kerenyi and encourage him gram in Epidemiology, Division of Public Health
creased risk of developing the disease. to more fully describe and publish his Sciences, Fred Hutchinson Cancer Research Cen-
ter and Department of Epidemiology, School ofOur results, as well as those from a num- findings regarding workers who spend
Public Health and Community Medicine, Univer-ber of other studies, are consistent with much of their time in darkness. We have
sity of Washington, Seattle; D. K. Mirick, Pro-these predictions, but they could also be also had a long-standing interest in the
gram in Epidemiology, Fred Hutchinson Cancer
consistent with other mechanisms. In- etiologic role of exposure to LAN in the
Research Center; R. G. Stevens, Department of
deed, Funk and Amir (7) have raised the development of cancer (8) and agree that Community Medicine, University of Connecticut
intriguing possibility that factors such as accumulating evidence warrants further Health Center, Farmington.
stress and fear may modify the effect of study of the potential effects of such ex- Correspondence to: Scott Davis, Ph.D., Fred
Hutchinson Cancer Research Center, 1100 Fair-light on hormone production. Epidemio- posure. We currently have a proposal
view Ave. North MP-474, P.O. Box 19024, Se-logic studies cannot distinguish with under review to initiate a study to mea-
attle, WA 98109–1024 (e-mail: sdavis@fhcrc.much certainty which underlying bio- sure melatonin and reproductive hor-
org).
logic mechanisms account for or can mone levels in women who work at
best explain associations observed on a night and in women who work during
population basis. However, they can in- the day. The results of such a study may I thank Dr. Kerenyi for his comments
form the design of laboratory-based in- be useful in designing a trial of the po- regarding the recent editorial on light at
vestigations that are capable of identify- tential beneficial effect of melatonin re- night and breast cancer risk (1), and ap-
ing specific mechanisms. placement by providing information on preciate the opportunity to further com-
Dr. Porock and Mr. Gentry also point hormone patterns associated with differ- ment on this issue of light as a potential
out that a number of different factors ent work and sleep schedules. We agree carcinogen.
Journal of the National Cancer Institute, Vol. 94, No. 7, April 3, 2002 CORRESPONDENCE 533lett WC, Hunter DJ, Kawachi I, et al. RotatingRecent studies on light and breast year of employment and a lag period of
night shifts and risk of breast cancer in womenat least 10 years) was statistically sig-cancer risk have primarily focused on
participating in the nurses’ health study. J Natl
people who work at night (2–4). Dr. nificantly decreased (OR 0.4; 95%
Cancer Inst 2001;93:1563–8.
confidence interval [CI] 0.2 to 0.9;Kerenyi refers to an interesting unpub- (3) Davis S, Mirick DK, Stevens RG. Night shift
N 11) among the photo laboratorylished study regarding the opposite situ- work, light at night, and risk of breast cancer.
workers compared with other employ-ation, i.e., of people who worked in J Natl Cancer Inst 2001;93:1557–62.
(4) Hansen J. Increased breast cancer risk amongees. The corresponding OR values fordarkness during the day in a film-
women who work predominantly at night. Epi-melanomas among women and men, re-producing factory in Canada. Compared
demiology 2001;12:74–7.spectively, were 1.4 (95% CI 0.5 towith other workers with normal light ex-
(5) Stevens RG. Electric power use and breast
3.7; N 6) and 0.4 (95% CI 0.1 toposure within the same factory, the cancer: a hypothesis. Am J Epidemiol 1987;
1.8; N 2). Considering 1) the limita-workers with low light exposure had a 125:556–61.
tions of a small number of includedreduced risk of breast cancer and mela- (6) Hahn RA. Profound bilateral blindness and the
cases and 2) that no further information incidence of breast cancer. Epidemiologynomas, thus supporting the “melatonin-
1991;2:208–10.is available on exposure to light for thishypothesis”, which, in this opposite situ-
(7) Kerenyi NA, Pandula E, Feuer G. Why theoccupational group, we found that theation with low exposure to light, may
incidence of cancer is increasing: the role ofresults for breast cancer are in line withargue that excess darkness, including
‘light pollution’. Med Hypotheses 1990;33:
the findings by Dr. Kerenyi. This mayblindness, may decrease the risk of cer- 75–8.
put further impetus on the need for stud-tain cancers, including breast cancer and (8) Stevens RG, Rea MS. Light in the built envi-
ies with detailed information on light ronment: potential role of circadian disruptionmelanomas (5–8).
exposure and cancer risk among work- in endocrine disruption and breast cancer.To explore the results of Dr. Kerenyi,
Cancer Causes Control 2001;12:279–87.ers with extreme exposure to light eitherwe estimated the relative risk of breast
(9) Olsen JH, Jensen OM. Occupation and risk ofas darkness during the day or as artificialcancer and malignant melanomas
cancer in Denmark. An analysis of 93,810light during the night.among workers employed in photo- cases, 1970–1979. Scand J Work
graphic laboratories in Denmark, on the Environ Health 1987;13 Suppl 1:1–91.JOHNNI HANSEN
basis of our comprehensive nationwide
REFERENCES NOTEScase–control data linkage on occupa-
tional history (1964–1989) and cancer
(1) Hansen J. Light at night, shift work, and breast Correspondence to: Johnni Hansen, Ph.D.,
(1970–1989) (9). The odds ratio (OR), cancer risk. J Natl Cancer Inst 2001;93: Danish Cancer Society, Institute of Cancer Epide-
adjusted for sex and birth year, for fe- 1513–5. miology, Strandboulevarden 49, DK-2100 Copen-
male breast cancer (including at least 1 (2) Schernhammer ES, Laden F, Speizer FE, Wil- hagen Ø, Denmark (e-mail: johnni@cancer.dk).
534 CORRESPONDENCE Journal of the National Cancer Institute, Vol. 94, No. 7, April 3, 2002