Advertisement

Circadian Disruption and Fatal Ovarian Cancer

      Background

      The International Agency for Research on Cancer determination that shift work is a “probable” human carcinogen was based primarily on studies of breast cancer but it was also noted that additional aspects of circadian disruption and other cancer sites deserved further research.

      Purpose

      To examine possible associations of three measures of circadian disruption: nontypical work schedules, nightly sleep duration, and monthly frequency of insomnia with risk of fatal ovarian cancer in a sample of American women.

      Methods

      Several measures of circadian disruption and other information were assessed in 1982 from 161,004 employed women in the American Cancer Society’s Cancer Prevention Study–II, a cohort that has been followed for mortality through 2010. In 2013, Cox proportional hazards regression was used to model the relative risks (RRs) and 95% CIs of death from ovarian cancer for categories of each indicator of circadian disruption.

      Results

      Over 28 years of follow-up, 1289 deaths from ovarian cancer occurred in the at-risk cohort. Compared to fixed daytime work, a rotating schedule was associated with an elevated risk of fatal ovarian cancer (RR=1.27, 95% CI=1.03, 1.56). No significant associations were observed for sleep duration (p trend=0.24) or insomnia (p trend=0.44).

      Conclusions

      In this large prospective study, there was a higher risk of fatal ovarian cancer in women who reported a rotating work schedule. These findings and the high prevalence of rotating shift schedules underscore the need for further research examining the role of work schedule and risk of ovarian cancer.

      Introduction

      In 2004, the Bureau of Labor Statistics (BLS) reported that 83.9% of employed women worked typical daytime schedules, defined by the BLS as between the hours of 6:00am and 6:00pm. Fewer women, 11.5%, worked schedules that took them into the evening (between 2:00pm and 12mn); night (between 9:00pm and 8:00am); or that rotated periodically.
      • McMenamin T.M.
      A time to work: recent trends in shift work and flexible schedules.
      Accumulated evidence suggests that working during the night has important physiological consequences and in 2007 the International Agency for Research on Cancer (IARC) concluded that shift work involving circadian disruption was “probably carcinogenic.”
      IARC
      IARC based this report on animal studies and limited evidence from occupational cohorts. The association is most convincing for breast cancer
      • Hansen J.
      Increased breast cancer risk among women who work predominantly at night.
      • Lie J.A.
      • Roessink J.
      • Kjaerheim K.
      Breast cancer and night work among Norwegian nurses.
      • Megdal S.P.
      • Kroenke C.H.
      • Laden F.
      • Pukkala E.
      • Schernhammer E.S.
      Night work and breast cancer risk: a systematic review and meta-analysis.
      • Pronk A.
      • Ji B.T.
      • Shu X.O.
      • et al.
      Night-shift work and breast cancer risk in a cohort of Chinese women.
      • Schernhammer E.S.
      • Kroenke C.H.
      • Laden F.
      • Hankinson S.E.
      Night work and risk of breast cancer.
      • Schernhammer E.S.
      • Laden F.
      • Speizer F.E.
      • et al.
      Rotating night shifts and risk of breast cancer in women participating in the Nurses’ Health Study.
      but also was observed for prostate,
      • Kubo T.
      • Ozasa K.
      • Mikami K.
      • et al.
      Prospective cohort study of the risk of prostate cancer among rotating-shift workers: findings from the Japan collaborative cohort study.
      colorectal,
      • Schernhammer E.S.
      • Laden F.
      • Speizer F.E.
      • et al.
      Night-shift work and risk of colorectal cancer in the Nursesʼ Health Study.
      and endometrial
      • Viswanathan A.N.
      • Hankinson S.E.
      • Schernhammer E.S.
      Night shift work and the risk of endometrial cancer.
      cancers. The working group discussed the need to explore these associations with additional cancer sites as well as other aspects of circadian disruption such as other nontraditional work schedules or sleep duration and quality.
      • Stevens R.G.
      • Hansen J.
      • Costa G.
      • et al.
      Considerations of circadian impact for defining “shift work” in cancer studies: IARC Working Group Report.
      Retinal exposure to light stimulates the pineal gland to modulate levels of melatonin,
      • Gachon F.
      • Nagoshi E.
      • Brown S.A.
      • Ripperger J.
      • Schibler U.
      The mammalian circadian timing system: from gene expression to physiology.
      and the response to the 24-hour day–night cycle is a powerful physiological synchronizer.
      • Pandi-Perumal S.R.
      • Srinivasan V.
      • Spence D.W.
      • Cardinali D.P.
      Role of the melatonin system in the control of sleep: therapeutic implications.
      • Zeitzer J.M.
      • Dijk D.J.
      • Kronauer R.
      • Brown E.
      • Czeisler C.
      Sensitivity of the human circadian pacemaker to nocturnal light: melatonin phase resetting and suppression.
      In the late 1970s, Cohen et al.
      • Cohen M.
      • Lippman M.
      • Chabner B.
      Role of pineal gland in aetiology and treatment of breast cancer.
      hypothesized that pineal gland dysfunction was linked to breast cancer after they observed that melatonin inhibits pituitary gonadotropins. Later studies found that women working night shifts were more likely to report irregular menstrual cycles, fertility problems, and negative birth outcomes due to disrupted sex hormone levels
      • Baker F.C.
      • Driver H.S.
      Circadian rhythms, sleep, and the menstrual cycle.
      • Voordouw B.C.
      • Euser R.
      • Verdonk R.E.
      • et al.
      Melatonin and melatonin–progestin combinations alter pituitary–ovarian function in women and can inhibit ovulation.
      ; these hormones are important in the etiology of reproductive cancers.
      • Martin A.M.
      • Weber B.L.
      Genetic and hormonal risk factors in breast cancer.
      Moreover, animal experiments have shown that nocturnal physiological levels of melatonin have direct anti-proliferative and anti-metastatic effects on cancer cells.
      • Blask D.E.
      • Brainard G.C.
      • Dauchy R.T.
      • et al.
      Melatonin-depleted blood from premenopausal women exposed to light at night stimulates growth of human breast cancer xenografts in nude rats.
      • Shah P.N.
      • Mhatre M.C.
      • Kothari L.S.
      Effect of melatonin on mammary carcinogenesis in intact and pinealectomized rats in varying photoperiods.
      There is limited evidence demonstrating an association between circadian rhythm disruption and risk of ovarian cancer. A case–control study of 1101 borderline and invasive ovarian cancer cases found significantly increased odds of ever working night shifts in women with cancer compared to the controls
      • Bhatti P.
      • Cushing-Haugen K.L.
      • Wicklund K.G.
      • Doherty J.A.
      • Rossing M.A.
      Nightshift work and risk of ovarian cancer.
      ; however, there was no dose–response with longer duration. In contrast, an earlier analysis of Nurses’ Health Studies (NHS)–I and –II found no association between duration of rotating shift schedules and incident ovarian cancer.
      • Poole E.M.
      • Schernhammer E.S.
      • Tworoger S.S.
      Rotating night shift work and risk of ovarian cancer.
      Only one study has examined sleep duration as a measure of circadian disruption. A small Japanese cohort reported an inverse association between sleep duration and incident ovarian cancer
      • Weiderpass E.
      • Sandin S.
      • Inoue M.
      • et al.
      Risk factors for epithelial ovarian cancer in Japan—results from the Japan public health center-based prospective study cohort.
      ; however, that analysis included only 86 total cases and did not examine work schedule. None of the studies of circadian disruption–related factors and ovarian cancer have investigated sleep and work schedule concurrently.
      The Cancer Prevention Study–II (CPS–II) is a large nationwide prospective mortality cohort of men and women followed for 28 years. The primary aims of this analysis were to examine whether three indicators of circadian disruption (i.e., work schedule, average sleep duration, and frequency of insomnia) were associated with risk of dying from ovarian cancer in this cohort. Secondary aims were to evaluate whether the association between sleep duration and ovarian cancer mortality are modified by BMI or work schedule, and whether results differed by follow-up period.

      Methods

      The Cancer Prevention Study–II Cohort

      In 1982, the American Cancer Society initiated the CPS–II mortality cohort to identify risk factors for and opportunities to prevent cancer. Volunteers enrolled 1.2 million men and women in all 50 states, the District of Columbia, and Puerto Rico. Each returned a mailed four-page questionnaire that provided information on demographic factors, reproductive health, diet and nutrition, height and weight, current and past use of tobacco products, as well as personal and family medical histories. More detailed discussions of the Cancer Prevention Studies are available elsewhere.
      • Garfinkel L.
      Selection, follow-up, and analysis in the American Cancer Society prospective studies.
      • Calle E.E.
      • Rodriguez C.
      • Jacobs E.J.
      • et al.
      The American Cancer Society Cancer Prevention Study II Nutrition Cohort: rationale, study design, and baseline characteristics.
      Every aspect of CPS–II is approved and monitored by the Emory University School of Medicine IRB.
      Women were excluded from this analysis if they reported any of the following at baseline: prevalent cancer other than nonmelanoma skin cancer (n=57,094); history of ovarian surgeries or hysterectomies (n=172,892); women who did not indicate their menopausal status or reported it as “artificially” induced (n=45,541); did not respond to questions of current rotating work schedule (n=171,366); fixed-schedule workers who did not disclose the time of day they began working (n=22,508); or who were not currently employed (n=45,755). The final analytic cohort included 161,004 employed women with a mean age of 50.3 years.

      Assessment of Circadian Disruption

      Three indicators of circadian disruption were considered in this analysis: work schedule, average sleep duration, and frequency of insomnia. The CPS–II questionnaire asked participants to answer Do you work rotating shifts? and What time of day do you start working? Their responses were combined to create a single variable for work schedule based on BLS definitions
      • McMenamin T.M.
      A time to work: recent trends in shift work and flexible schedules.
      and assuming an 8-hour workday. Rotating schedule workers were those who self-identified as so; all others were considered to work fixed schedules. Fixed daytime workers started work between 6:00am and 10:00am; fixed afternoon/evening workers began 2:00pm to 4:00pm; fixed night workers began 9:00pm to 12mn. Very few women began their workday outside of these hours and were not well defined within BLS descriptions; they were classified as “other.”
      To evaluate sleep quality and duration, participants were asked: On the average, how many hours do you sleep each night? and responses were categorized as 3–5, 6, 7, 8, and 9–12 hours; outliers or missing responses were combined as a separate category. Statistical models compared each group to participants who sleep 7 hours, similar to previous studies of sleep in this cohort.
      • Kripke D.F.
      • Garfinkel L.
      • Wingard D.L.
      • Klauber M.R.
      • Marler M.R.
      Mortality associated with sleep duration and insomnia.
      • Kripke D.F.
      • Marler M.R.
      • Calle E.E.
      Epidemiological health impact.
      Insomnia was assessed by asking: On the average, how many times per month do you have insomnia? and categorized as ≤1, 2, 3–9, and ≥10 nights per month; each was compared to participants reporting no insomnia.

      Mortality Follow-Up

      Prior to 1988, American Cancer Society volunteers made personal inquiries to determine the vital status of each CPS–II participant. Death certificates were obtained and the underlying cause of death was recorded. Subsequently, automatic linkages with the National Death Index have been completed biennially and mortality follow-up is current through December 31, 2010. Underlying cause of death is established for 99.4% of all known deaths. Ovarian cancer was coded according to ICD-9

      Medicode. Physician ICD-9-CM: 1999–2000 International Classification of Diseases clinical modification. 6th ed. Ingenix Inc, 1999–2000.

      codes 183.0–183.9, and ICD-10

      WHO. International statistical classification of diseases and related health problems, 10th revision. 2007. apps.who.int/classifications/apps/icd/icd10online/.

      codes C56, and C57.0–C57.4. There were a total of 1289 ovarian cancer deaths identified in the at-risk cohort.

      Statistical Analysis

      All statistical analyses were conducted in SAS v9.3. Follow-up time in years was computed for each individual as the time since enrollment in 1982 until the date of death or December 31, 2010, whichever was earliest. Age- and multivariable-adjusted Cox proportional hazards regression modeling was used to evaluate possible associations between each of the indicators of circadian disruption (i.e., work schedule, duration of sleep, and frequency of insomnia) and death from ovarian cancer. Trend variables for sleep duration and frequency of insomnia were derived by using the median value from each category and then modeling this as a continuous variable. All models stratified on single year of age at enrollment. Multivariable models adjusted for covariates found to be significant risk factors for ovarian cancer in this cohort
      • Rodriguez C.
      • Tatham L.M.
      • Calle E.E.
      • Thun M.J.
      • Jacobs E.J.
      • Heath Jr, C.W.
      Infertility and risk of fatal ovarian cancer in a prospective cohort of U.S. women.
      ; these included categorical variables for race, family history of breast or ovarian cancer, age at menarche, menopausal status, age at menopause, age at first birth, parity, duration of oral contraceptive use, postmenopausal estrogen use, and previous tubal ligation. Continuous measures were used for BMI and height. Education, alcohol use, smoking status, and use of sleeping pills were not significantly associated with the primary exposure variables or fatal ovarian cancer; consequently, they were not included in the final model.
      Two sensitivity analyses were conducted. Effect modification of the association between sleep and fatal ovarian cancer by BMI or work schedule was assessed by creating cross-product interaction terms for categories of sleep and WHO BMI categories and sleep and work schedule; the results were compared to the base model using likelihood ratio tests. To evaluate the degree to which misclassification would bias the results as the cohort aged and left the workforce, follow-up time was divided in half, from 1982 to 1995 and 1996 to 2010.

      Results

      The distribution of baseline characteristics according to work schedule for the analytic cohort is shown in Table 1. The mean age in 1982 was 50.3 years. The majority of the women were white (92%); educated beyond high school (64.4%); not menopausal at baseline (53.4%); reported never taking oral contraceptives (57.6%); got moderate exercise (61.7%); and had a BMI <25.0 (65.2%). Only 6.6% of the cohort reported working a rotating schedule in 1982 and much smaller proportions worked fixed afternoon/evening or night shifts. Nearly all (90.9%) women reported that they slept between 6 and 8 hours a night and more than half (56.1%) did not report insomnia. Between 1982 and 2010, there were 1289 deaths from ovarian cancer over 4,146,706 person-years of follow-up.
      Table 1Baseline characteristics, CPS–II 1982 mortality cohort, % unless otherwise noted
      N/n (%)Work schedule
      Fixed day (n=141,637)Rotating shifts (n=10,552)Fixed afternoon/evening (n=2288)Fixed night (n=1754)Other shifts (n=4773)
      Age (M [SD])161,00450.3 (8.6)49.7 (9.4)50.6 (6.5)49.9 (9.1)51.1 (9.5)
      BMI (M [SD])24.4 (4.5)24.8 (4.7)25.2 (5.0)26.3 (5.7)25.0 (4.9)
      Height (cm; M [SD])163.9 (6.6)163.7 (6.7)163.8 (6.9)163.8 (6.9)163.5 (6.6)
      Hours of sleep (M [SD])7.2 (0.9)7.2 (1)7.2 (1)6.8 (1.2)7.2 (1.1)
      Monthly frequency of insomnia (M [SD])1.5 (3.1)1.7 (3.5)1.6 (3.3)1.4 (3.0)1.6 (3.3)
      Race
       White148,436 (92.2)88.26.41.41.12.9
       Black8,085 (5.0)84.58.61.71.73.6
       Other4,483 (2.8)86.18.11.50.93.4
      Family history of breast/ovarian cancers
       No149,274 (92.7)87.96.61.41.12.9
       Yes11,730 (7.3)88.46.31.51.02.9
      Age at menarche
       ≤1271,990 (44.7)88.36.31.41.22.9
       1346,113 (28.6)88.26.41.41.02.9
       ≥1340,017 (24.9)87.27.21.41.03.1
      Age at menopause
       Pre-/peri-menopausal85,903 (53.4)88.06.81.31.12.8
       <50 years30,471 (18.9)87.96.31.51.13.1
       ≥50 years39,232 (24.4)88.25.91.61.13.2
      Parity/age of first birth
       Nulliparous21,937 (13.6)90.45.71.20.82.0
       1–2 births, <25 years old30,011 (18.6)89.15.91.30.92.8
       1–2 births, ≥25 years old29,465 (18.3)90.15.31.10.92.6
       ≥3 births, <25 years old52,710 (32.7)85.87.61.81.33.5
       ≥3 births, ≥25 years old20,683 (12.8)87.26.51.61.53.2
      Duration of oral contraceptive use
       Never92,771 (57.6)87.96.31.51.23.1
       <5 years32,763 (20.3)87.87.01.41.12.8
       ≥5 years29,926 (18.6)89.16.41.20.92.5
      Postmenopausal ERT use
       Pre-/peri-menopausal85,903 (53.4)88.06.81.31.12.8
       Never47,294 (29.4)88.45.91.51.13.1
       Ever21,183 (13.2)88.26.21.51.03.0
      Tubal ligation
       No143,937 (89.4)88.16.51.41.13.0
       Yes17,067 (10.6)87.17.01.61.42.9
      Education
       <High school degree9,691 (6.0)79.110.62.11.56.6
       High school graduate46,423 (28.8)87.76.51.30.83.8
       Vocational/some college51,224 (31.8)87.37.01.71.42.7
       College graduate52,418 (32.6)90.85.31.20.91.8
      Exercise
       None/slight50,285 (31.2)91.84.60.80.72.1
       Moderate99,288 (61.7)86.97.11.61.23.2
       Heavy9,383 (5.8)79.711.22.61.45.2
      ERT, estrogen replacement therapy
      Women working each schedule did not differ remarkably in age, BMI, height, or frequency of insomnia. A greater proportion of black women, those who did not complete high school, and heavy exercisers reported rotating schedules than did other women. The same groups were more likely to indicate either extreme of sleep duration (Appendix A, available online at www.ajpmonline.org). White women and ever postmenopausal estrogen users reported more frequent insomnia; otherwise, there were no differences (Appendix B, available online at www.ajpmonline.org).
      A significant elevated risk of fatal ovarian cancer was observed with women reporting rotating work schedules in 1982 compared to fixed day workers after adjusting for reproductive, anthropometric, and other risk factors (RR=1.27, 95% CI=1.03, 1.56; Table 2), an association that was robust throughout the entire follow-up (Figure 1). Fixed afternoon/evening or night shifts were not associated with fatal ovarian cancer. There were no associations observed with sleep duration (p trend=0.2416) or insomnia (p trend=0.4438).
      Table 2Results of Cox proportional hazards analyses for indicators of circadian disruption and fatal ovarian cancer
      Deaths
      Number of deaths do not equal total because “missing” categories are excluded.
      Person-yearsRR
      Age-adjusted
      95% CIRR
      Multivariable results are adjusted for oral contraceptive use, age at menarche and menopause, tubal ligation, parity, postmenopausal estrogen use, race, family history of breast/ovarian cancers, exercise, BMI, and height.
      95% CI
      Work schedule
       Fixed day1,1263,655,9861.001.00
       Rotating shifts101269,7421.27(1.04, 1.56)1.27(1.03, 1.56)
       Fixed afternoon/evening1157,5060.62(0.34, 1.11)0.62(0.34, 1.12)
       Fixed night1544,3941.14(0.68, 1.89)1.12(0.67, 1.87)
      Sleep duration (hours)
       3–541135,4990.99(0.72, 1.37)1.01(0.73, 1.40)
       6230681,7931.13(0.96, 1.32)1.13(0.97, 1.33)
       74531,566,8801.001.00
       84931,526,6601.16(1.02, 1.31)1.16(1.02, 1.31)
       9–1258203,8721.08(0.82, 1.42)1.08(0.82, 1.42)
      p trend=0.2689p trend=0.2416
      Insomnia (nights/month)
       Never7002,329,2851.001.00
       ≤1115378,7780.98(0.81, 1.20)0.98(0.8, 1.20)
       2138432,3780.99(0.83, 1.19)1.01(0.84, 1.21)
       3–9228662,4911.02(0.88, 1.19)1.04(0.89, 1.21)
       ≥1030129,3900.69(0.48, 0.99)0.71(0.49, 1.03)
      p trend=0.3213p trend=0.4438
      RR, relative risk
      a Number of deaths do not equal total because “missing” categories are excluded.
      b Age-adjusted
      c Multivariable results are adjusted for oral contraceptive use, age at menarche and menopause, tubal ligation, parity, postmenopausal estrogen use, race, family history of breast/ovarian cancers, exercise, BMI, and height.
      Figure thumbnail gr1
      Figure 1Cumulative hazards of fatal ovarian cancer associated with rotating shift schedules and fixed day schedules
      In sensitivity analyses, the association between sleep duration and ovarian cancer was not modified by BMI categories (p interaction=0.513), or by rotating work schedule (p interaction=0.483). When the analyses were stratified by follow-up time, results of sleep duration and insomnia remained unremarkable. The general pattern observed with rotating shift schedules persisted in both time periods: it did not reach significance in the first half, although it did in the second: RR=1.14 (95% CI=0.79, 1.63) and RR=1.34 (95% CI=1.05, 1.72).

      Discussion

      In this large prospective study of employed women, working rotating shifts was significantly associated with a moderately elevated risk of fatal ovarian cancer compared to a fixed day schedule beginning between 6:00am and 10:00am. No associations were observed for sleep duration or insomnia. Although the RR associated with working a rotating schedule is modest (1.27), it is of similar magnitude to other ovarian cancer risk factors observed in this cohort.
      • Rodriguez C.
      • Tatham L.M.
      • Calle E.E.
      • Thun M.J.
      • Jacobs E.J.
      • Heath Jr, C.W.
      Infertility and risk of fatal ovarian cancer in a prospective cohort of U.S. women.
      • Rodriguez C.
      • Calle E.E.
      • Coates R.J.
      • Miracle-McMahill H.L.
      • Thun M.J.
      • Heath Jr, C.W.
      Estrogen replacement therapy and fatal ovarian cancer.
      • Miracle-McMahill H.L.
      • Calle E.E.
      • Kosinski A.S.
      • et al.
      Tubal ligation and fatal ovarian cancer in a large prospective cohort study.
      • Rodriguez C.
      • Henley S.J.
      • Calle E.E.
      • Thun M.J.
      Paracetamol and risk of ovarian cancer mortality in a prospective study of women in the USA.
      • Cottreau C.M.
      • Ness R.B.
      • Kriska A.M.
      Physical activity and reduced risk of ovarian cancer.
      • Rodriguez C.
      • Patel A.V.
      • Calle E.E.
      • Jacob E.J.
      • Thun M.J.
      Estrogen replacement therapy and ovarian cancer mortality in a large prospective study of U.S. women.
      • Rodriguez C.
      • Calle E.E.
      • Fakhrabadi-Shokoohi D.
      • Jacobs E.J.
      • Thun M.J.
      Body mass index, height, and the risk of ovarian cancer mortality in a prospective cohort of postmenopausal women.
      These findings and those from other studies do not provide clear evidence of the relationship between circadian disruption and ovarian cancer. Differences in study design, exposure definitions, and outcomes might explain, in part, inconsistencies across studies. In this analysis of CPS–II, there was no association of current night shift work with fatal ovarian cancer and a positive association of rotating shifts with fatal ovarian cancer. However, a large case–control study found a significant positive association for night work but did not specifically collect information on rotating schedules.
      • Bhatti P.
      • Cushing-Haugen K.L.
      • Wicklund K.G.
      • Doherty J.A.
      • Rossing M.A.
      Nightshift work and risk of ovarian cancer.
      Importantly, very few women in CPS–II reported fixed night work and therefore power was limited for the exposure. In CPS–II and in the NHS,
      • Poole E.M.
      • Schernhammer E.S.
      • Tworoger S.S.
      Rotating night shift work and risk of ovarian cancer.
      self-reported work schedule information was collected prospectively; however, in the NHS, detailed information on work schedule was collected in repeated follow-ups whereas in CPS–II, information was collected only at baseline. Therefore, the definitions of rotating shift work differed between the two studies. More specifically, in the NHS, rotating shift was defined as at least 3 nights per month working at night in addition to day or evening shifts; in contrast, CPS–II participants were asked to simply report if they were currently working rotating shift schedules.
      Given the repeated measures over time in the NHS, there is likely to be less misclassification over time, whereas in CPS–II this misclassification is likely to attenuate any associations. This nondifferential misclassification in CPS–II over time does not explain the observed association between rotating shift work and risk of ovarian cancer. Further, the case–control and NHS studies evaluated association between shift work and ovarian cancer incidence, whereas the outcome for the CPS–II was mortality. Although most risk factors are observed consistently across most subtypes of ovarian cancer,
      • Kurian A.W.
      • Balise R.R.
      • McGuire V.
      • Whittemore A.S.
      Histologic types of epithelial ovarian cancer: have they different risk factors?.
      it is possible that rotating shift work might be differentially associated with highly fatal disease compared to less fatal ovarian cancer. Importantly, despite the differences in design among studies, the null results from NHS are compelling, because in that cohort, significant positive associations were reported of shift work with risk of other cancers including breast,
      • Schernhammer E.S.
      • Kroenke C.H.
      • Laden F.
      • Hankinson S.E.
      Night work and risk of breast cancer.
      • Schernhammer E.S.
      • Laden F.
      • Speizer F.E.
      • et al.
      Rotating night shifts and risk of breast cancer in women participating in the Nurses’ Health Study.
      colorectal,
      • Schernhammer E.S.
      • Laden F.
      • Speizer F.E.
      • et al.
      Night-shift work and risk of colorectal cancer in the Nursesʼ Health Study.
      and endometrial
      • Viswanathan A.N.
      • Hankinson S.E.
      • Schernhammer E.S.
      Night shift work and the risk of endometrial cancer.
      cancers. Regardless, taken together, findings from these three studies cannot rule out a possible association between work schedule and risk of ovarian cancer, and further research investigating a comprehensive history of lifetime work is warranted.
      For sleep duration, the inverse association with incident ovarian cancer reported in a cohort of Japanese women,
      • Weiderpass E.
      • Sandin S.
      • Inoue M.
      • et al.
      Risk factors for epithelial ovarian cancer in Japan—results from the Japan public health center-based prospective study cohort.
      was not observed in CPS–II. These two studies differed most notably in the total number of cases available for analysis. In CPS–II, there were 1289 ovarian cancer deaths over 28 years of follow-up, whereas in the Japanese study there were only 86 cases over 16 years. Therefore it is unclear whether their findings are due to chance, or due to underlying cultural and genetic differences between these populations.
      • Herrinton L.J.
      • Stanford J.L.
      • Schwartz S.M.
      • Weiss N.S.
      Ovarian cancer incidence among Asian migrants to the United States and their descendants.
      Unfortunately, CPS–II does not have sufficient numbers of Asian Americans to explore possible racial differences. However, this issue deserves further research in other cohorts.
      The biologic mechanisms underlying these associations are likely complex. There is considerable evidence that circadian disruption, through exposure to light at night, has profound effects on human health and risk of cancer. Even small periods of nighttime light exposure can significantly affect plasma melatonin levels
      • Zeitzer J.M.
      • Dijk D.J.
      • Kronauer R.
      • Brown E.
      • Czeisler C.
      Sensitivity of the human circadian pacemaker to nocturnal light: melatonin phase resetting and suppression.
      and disrupt the body’s natural pacemaker.
      • Pandi-Perumal S.R.
      • Srinivasan V.
      • Spence D.W.
      • Cardinali D.P.
      Role of the melatonin system in the control of sleep: therapeutic implications.
      Human studies show that hormone surges follow predictable circadian patterns, and that disruptions in these rhythms are strongly associated with reproductive dysfunction in women
      • Mahoney M.M.
      Shift work, jet lag, and female reproduction.
      ; these hormones are particularly important in the development of reproductive cancers.
      • Moorman P.G.
      • Calingaert B.
      • Palmieri R.T.
      • et al.
      Hormonal risk factors for ovarian cancer in premenopausal and postmenopausal women.
      More directly, exposure to light at night can have nonhormonal effects on cancer risk. Blask et al.
      • Blask D.E.
      • Brainard G.C.
      • Dauchy R.T.
      • et al.
      Melatonin-depleted blood from premenopausal women exposed to light at night stimulates growth of human breast cancer xenografts in nude rats.
      exposed human breast cancer xenografts in nude rats to melatonin enriched or deficient blood and found that melatonin directly suppressed proliferative activity but that even short-term light exposure would eliminate this benefit. The circadian system is modulated through at least nine “clock” genes,
      • Stevens R.G.
      Light-at-night, circadian disruption and breast cancer: assessment of existing evidence.
      many of which are important in cell cycle regulation. There is active research linking dysfunction in these genes to cancer risk, and polymorphisms in the CLOCK, PER, and CRY genes are important in the development of some breast
      • Dai H.
      • Zhang L.
      • Cao M.
      • et al.
      The role of polymorphisms in circadian pathway genes in breast tumorigenesis.
      and ovarian
      • Tokunaga H.
      • Takebayashi Y.
      • Utsunomiya H.
      • et al.
      Clinicopathological significance of circadian rhythm-related gene expression levels in patients with epithelial ovarian cancer.
      tumors.
      The strengths of this study are its prospective design and long-term follow-up; it is remarkable for its large size and inclusion of multiple measures of circadian disruption examined in relation to fatal ovarian cancer. One limitation is that all of these exposures were self-reported and assessed only at baseline. As people age, sleeping patterns change and CPS–II participants would be expected to have long sleep latency and shorter duration over follow-up.
      • Ohayon M.M.
      • Carskadon M.A.
      • Guilleminault C.
      • Vitiello M.V.
      Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan.
      Additionally, as women left the workforce, rotating shift work would become increasingly misclassified. Despite this, analyses stratified by follow-up period did not suggest that misclassification was a major problem in this study. A second limitation was that follow-up included only fatal ovarian cancers. Data suggest that ovarian cancer is an aggregate of several distinct diseases
      • Ohayon M.M.
      • Carskadon M.A.
      • Guilleminault C.
      • Vitiello M.V.
      Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan.
      • Soslow R.A.
      Histologic subtypes of ovarian carcinoma: an overview.
      ; analyses by subtype are advised where possible.
      In this study, there was a 27% higher risk of fatal ovarian cancer in women working rotating shifts compared to fixed daytime schedules, whereas sleep duration and frequency of insomnia were not associated. Further research is needed to identify subgroups of women in which circadian disruption may be more clinically meaningful, for instance, women with a family history of cancer or women who might not physiologically adapt well to frequent rotating work shifts.

      Acknowledgments

      Publication of this article was supported by the American Cancer Society.
      The publication of this supplement was made possible through the CDC and the Association for Prevention Teaching and Research (APTR) Cooperative Agreement No. 1 U360E000005-01. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the CDC or the APTR.
      The authors of this paper express their sincere appreciation to all CPS–II study participants and to each member of the study management group. The American Cancer Society (ACS) funds the creation, maintenance, and updating of the Cancer Prevention Study II cohort. All of the authors were employed by the ACS during the course of this analysis.
      No financial disclosures were reported by the authors of this paper.

      Appendix. Supplementary data

      References

        • McMenamin T.M.
        A time to work: recent trends in shift work and flexible schedules.
        Monthly Labor Statistics. December 2007;
        • IARC
        Painting, firefighting, and shiftwork. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Volume 98. WHO, Geneva2010
        • Hansen J.
        Increased breast cancer risk among women who work predominantly at night.
        Epidemiology. 2001; 12: 74-77
        • Lie J.A.
        • Roessink J.
        • Kjaerheim K.
        Breast cancer and night work among Norwegian nurses.
        Cancer Causes Control. 2006; 17: 39-44
        • Megdal S.P.
        • Kroenke C.H.
        • Laden F.
        • Pukkala E.
        • Schernhammer E.S.
        Night work and breast cancer risk: a systematic review and meta-analysis.
        Eur J Cancer. 2005; 41: 2023-2032
        • Pronk A.
        • Ji B.T.
        • Shu X.O.
        • et al.
        Night-shift work and breast cancer risk in a cohort of Chinese women.
        Am J Epidemiol. 2010; 171: 953-959
        • Schernhammer E.S.
        • Kroenke C.H.
        • Laden F.
        • Hankinson S.E.
        Night work and risk of breast cancer.
        Epidemiology. 2006; 17: 108-111
        • Schernhammer E.S.
        • Laden F.
        • Speizer F.E.
        • et al.
        Rotating night shifts and risk of breast cancer in women participating in the Nurses’ Health Study.
        J Natl Cancer Inst. 2001; 93: 1563-1568
        • Kubo T.
        • Ozasa K.
        • Mikami K.
        • et al.
        Prospective cohort study of the risk of prostate cancer among rotating-shift workers: findings from the Japan collaborative cohort study.
        Am J Epidemiol. 2006; 164: 549-555
        • Schernhammer E.S.
        • Laden F.
        • Speizer F.E.
        • et al.
        Night-shift work and risk of colorectal cancer in the Nursesʼ Health Study.
        J Natl Cancer Inst. 2003; 95: 825-828
        • Viswanathan A.N.
        • Hankinson S.E.
        • Schernhammer E.S.
        Night shift work and the risk of endometrial cancer.
        Cancer Res. 2007; 67: 10618-10622
        • Stevens R.G.
        • Hansen J.
        • Costa G.
        • et al.
        Considerations of circadian impact for defining “shift work” in cancer studies: IARC Working Group Report.
        Occup Environ Med. 2011; 68: 154-162
        • Gachon F.
        • Nagoshi E.
        • Brown S.A.
        • Ripperger J.
        • Schibler U.
        The mammalian circadian timing system: from gene expression to physiology.
        Chromosoma. 2004; 113: 103-112
        • Pandi-Perumal S.R.
        • Srinivasan V.
        • Spence D.W.
        • Cardinali D.P.
        Role of the melatonin system in the control of sleep: therapeutic implications.
        CNS Drugs. 2007; 21: 995-1018
        • Zeitzer J.M.
        • Dijk D.J.
        • Kronauer R.
        • Brown E.
        • Czeisler C.
        Sensitivity of the human circadian pacemaker to nocturnal light: melatonin phase resetting and suppression.
        J Physiol. 2000; 526: 695-702
        • Cohen M.
        • Lippman M.
        • Chabner B.
        Role of pineal gland in aetiology and treatment of breast cancer.
        Lancet. 1978; 2: 814-816
        • Baker F.C.
        • Driver H.S.
        Circadian rhythms, sleep, and the menstrual cycle.
        Sleep Med. 2007; 8: 613-622
        • Voordouw B.C.
        • Euser R.
        • Verdonk R.E.
        • et al.
        Melatonin and melatonin–progestin combinations alter pituitary–ovarian function in women and can inhibit ovulation.
        J Clin Endocrinol Metab. 1992; 74: 108-117
        • Martin A.M.
        • Weber B.L.
        Genetic and hormonal risk factors in breast cancer.
        J Natl Cancer Inst. 2000; 92: 1126-1135
        • Blask D.E.
        • Brainard G.C.
        • Dauchy R.T.
        • et al.
        Melatonin-depleted blood from premenopausal women exposed to light at night stimulates growth of human breast cancer xenografts in nude rats.
        Cancer Res. 2005; 65: 11174-11184
        • Shah P.N.
        • Mhatre M.C.
        • Kothari L.S.
        Effect of melatonin on mammary carcinogenesis in intact and pinealectomized rats in varying photoperiods.
        Cancer Res. 1984; 44: 3403-3407
        • Bhatti P.
        • Cushing-Haugen K.L.
        • Wicklund K.G.
        • Doherty J.A.
        • Rossing M.A.
        Nightshift work and risk of ovarian cancer.
        Occup Environ Med. 2013; 70: 231-237
        • Poole E.M.
        • Schernhammer E.S.
        • Tworoger S.S.
        Rotating night shift work and risk of ovarian cancer.
        Cancer Epidemiol Biomarkers Prev. 2011; 20: 934-938
        • Weiderpass E.
        • Sandin S.
        • Inoue M.
        • et al.
        Risk factors for epithelial ovarian cancer in Japan—results from the Japan public health center-based prospective study cohort.
        Int J Oncol. 2012; 40: 21-30
        • Garfinkel L.
        Selection, follow-up, and analysis in the American Cancer Society prospective studies.
        Natl Cancer Inst Monogr. 1985; 67: 49-52
        • Calle E.E.
        • Rodriguez C.
        • Jacobs E.J.
        • et al.
        The American Cancer Society Cancer Prevention Study II Nutrition Cohort: rationale, study design, and baseline characteristics.
        Cancer. 2002; 94: 2490-2501
        • Kripke D.F.
        • Garfinkel L.
        • Wingard D.L.
        • Klauber M.R.
        • Marler M.R.
        Mortality associated with sleep duration and insomnia.
        Arch Gen Psychiatry. 2002; 59: 131-136
        • Kripke D.F.
        • Marler M.R.
        • Calle E.E.
        Epidemiological health impact.
        in: Kushida C.A. Sleep deprivation: clinical issues, pharmacology, and sleep loss effects. Lung biology in health and disease. Marcel Dekker, New York2005: 195-208 (193)
      1. Medicode. Physician ICD-9-CM: 1999–2000 International Classification of Diseases clinical modification. 6th ed. Ingenix Inc, 1999–2000.

      2. WHO. International statistical classification of diseases and related health problems, 10th revision. 2007. apps.who.int/classifications/apps/icd/icd10online/.

        • Rodriguez C.
        • Tatham L.M.
        • Calle E.E.
        • Thun M.J.
        • Jacobs E.J.
        • Heath Jr, C.W.
        Infertility and risk of fatal ovarian cancer in a prospective cohort of U.S. women.
        Cancer Causes Control. 1998; 9: 645-651
      3. WHO. Obesity and overweight, 2013. www.who.int/mediacentre/factsheets/fs311/en/.

        • Rodriguez C.
        • Calle E.E.
        • Coates R.J.
        • Miracle-McMahill H.L.
        • Thun M.J.
        • Heath Jr, C.W.
        Estrogen replacement therapy and fatal ovarian cancer.
        Am J Epidemiol. 1995; 141: 828-835
        • Miracle-McMahill H.L.
        • Calle E.E.
        • Kosinski A.S.
        • et al.
        Tubal ligation and fatal ovarian cancer in a large prospective cohort study.
        Am J Epidemiol. 1997; 145: 349-357
        • Rodriguez C.
        • Henley S.J.
        • Calle E.E.
        • Thun M.J.
        Paracetamol and risk of ovarian cancer mortality in a prospective study of women in the USA.
        Lancet. 1998; 352: 1354-1355
        • Cottreau C.M.
        • Ness R.B.
        • Kriska A.M.
        Physical activity and reduced risk of ovarian cancer.
        Obstet Gynecol. 2000; 96: 609-614
        • Rodriguez C.
        • Patel A.V.
        • Calle E.E.
        • Jacob E.J.
        • Thun M.J.
        Estrogen replacement therapy and ovarian cancer mortality in a large prospective study of U.S. women.
        JAMA. 2001; 285: 1460-1465
        • Rodriguez C.
        • Calle E.E.
        • Fakhrabadi-Shokoohi D.
        • Jacobs E.J.
        • Thun M.J.
        Body mass index, height, and the risk of ovarian cancer mortality in a prospective cohort of postmenopausal women.
        Cancer Epidemiol Biomarkers Prev. 2002; 11: 822-828
        • Kurian A.W.
        • Balise R.R.
        • McGuire V.
        • Whittemore A.S.
        Histologic types of epithelial ovarian cancer: have they different risk factors?.
        Gynecol Oncol. 2005; 96: 520-530
        • Herrinton L.J.
        • Stanford J.L.
        • Schwartz S.M.
        • Weiss N.S.
        Ovarian cancer incidence among Asian migrants to the United States and their descendants.
        J Natl Cancer Inst. 1994; 86: 1336-1339
        • Mahoney M.M.
        Shift work, jet lag, and female reproduction.
        Int J Endocrinol. 2010; (2010:813764)
        • Moorman P.G.
        • Calingaert B.
        • Palmieri R.T.
        • et al.
        Hormonal risk factors for ovarian cancer in premenopausal and postmenopausal women.
        Am J Epidemiol. 2008; 167: 1059-1069
        • Stevens R.G.
        Light-at-night, circadian disruption and breast cancer: assessment of existing evidence.
        Int J Epidemiol. 2009; 38: 963-970
        • Dai H.
        • Zhang L.
        • Cao M.
        • et al.
        The role of polymorphisms in circadian pathway genes in breast tumorigenesis.
        Breast Cancer Res Treat. 2011; 127: 531-540
        • Tokunaga H.
        • Takebayashi Y.
        • Utsunomiya H.
        • et al.
        Clinicopathological significance of circadian rhythm-related gene expression levels in patients with epithelial ovarian cancer.
        Acta Obstet Gynecol Scand. 2008; 87: 1060-1070
        • Ohayon M.M.
        • Carskadon M.A.
        • Guilleminault C.
        • Vitiello M.V.
        Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan.
        Sleep. 2004; 27: 1255-1273
        • Soslow R.A.
        Histologic subtypes of ovarian carcinoma: an overview.
        Int J Gynecol Pathol. 2008; 27: 161-174