Background: Night-shift work (NSW) has been suggested as a possible cause of breast cancer, and its association with mammographic density (MD), one of the strongest risk factors for breast cancer, has been scarcely addressed. This study examined NSW and MD in Spanish women.

Methods: The study covered 2,752 women aged 45–68 years recruited in 2007–2008 in 7 population-based public breast cancer screening centers, which included 243 women who had performed NSW for at least one year. Occupational data and information on potential confounders were collected by personal interview. Two trained radiologist estimated the percentage of MD assisted by a validated semiautomatic computer tool (DM-scan). Multivariable mixed linear regression models with random screening center–specific intercepts were fitted using log-transformed percentage of MD as the dependent variable and adjusting by known confounding variables.

Results: Having ever worked in NSW was not associated with MD |[{e^\beta }$|⁠:0.96; 95% confidence interval (CI), 0.86–1.06]. However, the adjusted geometric mean of the percentage of MD in women with NSW for more than 15 years was 25% higher than that of those without NSW history (MD>15 years:20.7% vs. MDnever:16.5%;|{e^\beta }$|⁠:1.25; 95% CI,1.01–1.54). This association was mainly observed in postmenopausal participants (⁠|{e^\beta }$|⁠:1.28; 95% CI, 1.00–1.64). Among NSW-exposed women, those with ≤2 night-shifts per week had higher MD than those with 5 to 7 nightshifts per week (⁠|{e^\beta }$|⁠:1.42; 95% CI, 1.10–1.84).

Conclusions: Performing NSW was associated with higher MD only in women with more than 15 years of cumulated exposure. These findings warrant replication in futures studies.

Impact: Our findings suggest that MD could play a role in the pathway between long-term NSW and breast cancer. Cancer Epidemiol Biomarkers Prev; 26(6); 905–13. ©2017 AACR.

The study of circadian disruption in the context of cancer development has been receiving increasing attention in cancer epidemiology. Specifically, night-shift workers have been the population of interest in several studies for being one of the main groups exposed to substantial alterations in their circadian rhythm. According to a 2015 European survey, 19% of employees worked at night (1). Similarly, 22% of employees in Spain worked at night at least once a month (1) and 12.5% performed night-shift work (NSW) regularly (2).

In 2007, the International Agency for Research on Cancer (IARC) concluded that shift work involving circadian disruption is probably carcinogenic to humans (Group 2A), mainly due to its association with increased breast cancer risk (3). A recent meta-analysis showed a pooled relative risk of 1.21 for women with a history of NSW exposure (4), and other studies not included in that review have confirmed this positive association (5–8). It is noteworthy to point out that this association is mostly found in women with long term NSW exposure (i.e., 15–20 years; refs. 5, 9). Unfortunately, the biological mechanisms underlying this association are not yet well understood, although the most commonly alleged pathway focuses on the reduced production of melatonin caused by exposure to light at night (10).

Mammographic density (MD) is currently the most interesting marker of risk for breast cancer due to its strong association with this tumor (11). Mammogram images reflect breast tissue composition; light (dense or non-radiolucent) areas on the mammogram represent the fibrous and glandular tissues in the breast, whereas the dark (radiolucent or less dense) areas are primarily fat (12). Breast cancer risk in women with highly dense breasts is four to five times greater compared with those with slightly or not dense breast controlling for age and BMI (11, 13); this association holds regardless of breast cancer subtype (14). Thus, MD has been proposed as an intermediate phenotype for breast cancer (13) and it has been used as outcome variable in prevention trials (15).

MD is highly hereditary (16), but this phenotype has also been independently associated with other breast cancer risk factors such as reproductive and anthropometric characteristics (17, 18); postmenopausal hormone therapy (19), and alcohol or tobacco consumption (20). However, the relationship between MD and NSW has only been explored in a previous study with inconclusive results (21). If confirmed, this association could have practical effects: Being MD the leading potentially modifiable intermediate marker for breast cancer, preventive strategies to reduce MD (22) might moderate the excess of breast cancer risk associated to NSW.

Here, we examine the association between NSW and MD in DDM-Spain/Var-DDM, a research project with over 3,000 women designed to evaluate determinants of MD in Spanish pre/peri- and postmenopausal women attending breast cancer–screening programs.

This is a cross-sectional multicenter study included in DDM-Spain (Determinants of Mammographic Density in Spain)/Var-DDM (Variability of the Mammographic Density in Spanish Women) research project. More detail regarding the design of the study is provided elsewhere (17, 22). Briefly, in Spain, the publicly funded and population-based breast cancer screening network invites all women between 50 and 69 years old (45–69 in some regions) to be screened every 2 years, with attendance rates ranging from 62.9% to 88.0% of its target population, depending on the region (23). For DDM-Spain/Var-DDM women were recruited consecutively at 7 breast cancer screening centers of several Regions (Aragon; Balearic Islands; Castile-Leon; Catalonia; Galicia; Navarre; and Valencia) in 2007 and 2008. Each center had to enroll at least 500 participants; thus, the total number of participants was 3,584 with an average participation rate of 74.5% (range: 64.7%–84.0%). The exclusion criteria were having been diagnosed of breast or ovarian cancer, having had breast surgery or implants and, in a second step, we also excluded 10 women who developed breast cancer within 6 months of the mammogram.

DDM-Spain/Var-DDM projects were formally approved by the Bioethics and Animal Welfare Committee at the Carlos III Institute of Health. All participants signed an informed consent form and, on the same day of the mammographic exam, they answered a detailed epidemiological questionnaire on obstetric and lifestyle factors administered by trained interviewers. This questionnaire included specific items to explore NSW exposure, namely, whether they had ever worked the night-shift, in which occupation, for how many years, usual number of nights worked per week (None, up to 2, 3–4, 5–7 nights/week), age at beginning of NSW and, if applicable, age when they stopped NSW. Combining this information with the age at mammogram we constructed two additional variables: timing of first childbirth (nulliparous; before onset of NSW; after onset of NSW) and years since last NSW (No NSW; Current or recent NSW (≤1 year); 1–5 years, 5–15 years, >15 years). Postmenopausal status was defined as self-reported absence of menstruation in the last 12 months. Women's height and weight were also measured twice by the interviewer in accordance with pre-established protocols, with a third measurement if the first two were substantially different. Average values were used to compute body mass index (BMI).

We obtained the cranio-caudal mammogram of the left breast for 3309 women (ages 45–68 years; 99.2% Caucasian) in the following formats: (i) Analogical mammograms digitalized with a Totalook MammoAdvantage scanner (maximum optical density 4.2; n = 1,781, 53.8%); (ii) fully digital images using Senographe 2000D Full Field Digital Mammography System; Hologic-Lorad M-IV; or Siemmens MAMMOMAT Novation DR depending on recruitment site (n = 1,376, 41.6%), and (iii) Originally digital mammograms printed on film and digitalized with a Microtek Medi-700scanner (maximum optical density 4.0) in (n = 152, 4.6%).

Two experienced radiologists estimated the percentage of MD in the mammograms in a continuous scale, assisted by DM-Scan, a semiautomated computer tool specifically designed by the Polytechnic University of Valencia for this purpose. This free tool (http://dmscan.iti.upv.es) has shown a high reproducibility and validity, with a substantial discriminative power to predict subsequent breast cancer development (24, 25). One radiologist read the mammograms from Castile-Leon, Balearic Islands and Navarre and the other one read those from Aragon, Catalonia, Galicia and Valencia. To evaluate intra- and inter-rater intraclass correlation (ICC), each radiologist repeated MD estimation in 60 images, and 243 mammograms (∼35 per region) were read by both of them. We obtained high ICCs in both cases (>0.9).

We excluded participants that had never held remunerated jobs (housewives; n = 476) to avoid possible bias due to a healthy worker effect as well as 26 participants with missing information for some of the main confounding variables. In addition, because we decided to set one year as the minimum time of cumulated NSW needed to be considered as ever exposed to NSW, 55 women working NSW for less than a year, which could not be clearly classified, were also excluded.

We adjusted multivariate mixed linear regression models with random screening center-specific intercepts using log-transformed percentage of MD as dependent variable. We also included in the model a number of potential confounder variables as fixed effects, which could be classified into three categories: (i) basic confounders that must be included when studying percent density: BMI (restricted quadratic spline), age at time of mammogram (continuous) and menopausal status (pre/perimenopausal and postmenopausal); (ii) lifestyle and personal variables associated to MD and breast cancer: parity (continuous), HRT(no use, previous use, current use), first-degree relative with breast cancer (yes/no), smoking (never smoker, ex-smoker ≥6 months and current smoker or ex-smoker <6 months), alcohol intake (non-drinker, <10 gram/day, ≥10 gram/day), average level of physical activity (low, moderate and vigorous), daily calorie intake (continuous); and (iii) other possible sources of variability derived of the study design: type of mammogram (analogic, full digital and scanned digital), and radiologist.

We used these log-lineal models to calculate geometric means of MD for nightshift exposure variables adjusted by all the mentioned confounding factors. In addition, the estimated regression coefficients and standard errors of these models were exponentiated to calculate the relative change of the adjusted geometric mean of MD comparing groups of participants across categories of the studied epidemiological variables. Models were also constructed separately for pre/peri- and postmenopausal women and we also explored the possible interactions between menopausal status and the four NSW exposure variables using likelihood ratio test. Interaction terms between cumulative NSW exposure and the demographic, obstetric, and lifestyle characteristics presented in Table 1 were tested as well. In addition, we evaluated the association between MD and NSW among women with current or recent exposure (stopped NSW <1 before mammogram; See Supplementary Table S1).

In a second step, we fitted models restricted only to the ever-exposed women (i.e., the 243 women with a history of NSW >1 year) to study the association of MD with all NSW characteristics simultaneously. These multivariable mixed linear regression models, with random screening center-specific intercepts, were adjusted by the previously mentioned confounding variables and the selected NSW characteristics (time since last NSW, total years of NSW, NSW usual frequency in nights per week, and parity at the beginning of NSW). Again, we calculated adjusted geometric means of MD for nightshift exposure variables.

We also performed additional sensitivity analyses by testing whether a number of factors (education, socioeconomic level, waist to hip ratio, and years since menopause in postmenopausal women) might cause residual confounding and by fitting the models classifying as exposed also those women that had NSW <1 year (n, 55) or including housewives (n, 476) among non-exposed participants as well as excluding (i) all women that had ever had HRT (n, 282) and (ii) those women that were using HRT at the moment of the mammograms (n, 69; See Supplementary Table S2).

Finally, we explored non-linear associations for total years of NSW exposure in all women, using restricted cubic splines with 3 knots located in 5, 10, and 15 years. Departure from linearity was evaluated with non-lineal spline terms. All analyses were carried out with STATA/MP 13.1 software (StataCorp. 2013).

Overall, 8.8 % of the 2,752 women (n, 243) reported having performed NSW for at least one year in their lifetime. Detailed information of the main characteristics of the women included in this study, stratified by their NSW status, is shown in Table 1.

The prevalence of ever performing NSW was higher among younger women, nulliparous and among those with low or normal weight. Also, NSW was more common in women with a university degree, due to the high proportion of nurses and health care professionals within the NSW group (102 out of 243 women). Other frequent occupational sectors with NSW exposure were industry (35 women), geriatric care (30 women), cleaning services (30 women), and administrative work (25 women). Balearic Islands and Valencia had the highest prevalence of women reporting NSW (11.7% and 11.4%, respectively). Smokers and former smokers were more likely to be engaged in NSW than nonsmokers (Table 1) and, on average, night-shift workers reported slightly higher daily caloric intake, although the difference was not statistically significant (NSWmean: 2112 vs. non-NSWmean: 2061 cals/d; P, 0.112). The median of years in NSW was 8 (Interquartile range: 3–12) and the mean age at onset of NSW was 32 years (Standard Deviation: 11.4). Almost one third of the participants reporting NSW history (76 out of 243) were currently doing NSW at the time of interview.

Table 2 shows the association between MD and NSW overall and stratified by menopausal status. Having ever performed NSW was not associated with MD, neither for the entire sample (MDever:15.9% vs. MDnever:16.5%; |{e^\beta }$|all: 0.96; 95% confidence Intervals (CI), 0.86–1.06), nor by menopausal status (MDever pre:23.3% vs. MDnever pre:25.0%; |{e^\beta }$|pre: 0.93; 95% CI, 0.78–1.11;MDever post:14.3% vs. MDnever post:14.7%; |\ {e^\beta }$|post: 0.98; 95% CI, 0.86–1.11). However, the adjusted geometric mean of the percentage of MD in women with NSW for more than 15 years was 25% higher than that of those without NSW history (MD>15years:20.7% vs. MDnever:16.5%;|{e^\beta }$|⁠:1.25; 95% CI, 1.01–1.54). The association did not change when we also took into account vitamin D consumption (MD>15years:22.1% vs. MDnever:16.4%;|{e^\beta }$|⁠:1.25; 95% CI, 1.01–1.54). We repeated the analysis, including housewives and performed different sensitivity analyses with similar results (See Supplementary Table S1), whereas none of the interaction terms between cumulative NSW exposure and the demographic, obstetric and lifestyle characteristics presented in Table 1 were statistically significant. It is interesting to note that MD was also higher in women who were current/recent night workers or had stopped working night shifts less than 5 years before the mammogram, although these associations were not statistically significant.

Concerning NSW frequency, women working 5 to 7 nights per week had lower MD than those which had never had NSW, both globally (MD5-7 nights:13.6% vs. MDnever:16.5%; |{e^\beta }$|⁠: 0.82; 95% CI, 0.68–0.99), and among postmenopausal women (MD5-7 nights:10.9% vs. MDnever:16.5%; |{e^\beta }$|⁠: 0.75; 95% CI, 0.60–0.92). However, in premenopausal women this association was not observed (Pinteraction:0.029). There were no differences in MD regarding timing of first child (Table 2) or regarding occupational sector of the NSW (Health Services|\ {e^\beta }$|⁠: 1.04; 95% CI, 1.04–0.89; Geriatric Care |{e^\beta }$|⁠: 0.94; 95% CI, 0.71–1.24; Industry |{e^\beta }$|⁠: 0.88; 95% CI, 0.68–1.14; Cleaning Services |{e^\beta }$|⁠: 0.89; 95% CI, 0.67–1.18; Administrative work |{e^\beta }$|⁠: 1.00; 95% CI, 0.74–1.36; Others |{e^\beta }$|⁠: 0.98; 95% CI, 0.64–1.51).

When we restricted the analysis to current workers (See Supplementary Table S2) the percentage of MD was again 20% higher in women who had performed NSW for more than 15 years (eβ:1.20; 95% CI, 0.92–1.59), although the smaller number of exposed women reduced the power to obtain statistical significance. As in the previous analyses, there was a significant interaction between NSW frequency and menopausal status (P, 0.036).

In further analysis, we focused only in the subgroup of 243 women exposed for over a year to be able to evaluate the effect of all the characteristics of NSW within the same model (Table 3). Again, the geometric mean of MD, adjusted by the other NSW exposure variables, was higher in women who had performed NSW for more than 15 years (MD>15years: 23.6% vs. MDnever: 16.0%; |{e^\beta }$|⁠:1.39; 95% CI, 1.05–1.83). In this case, we observed also an inverse association between MD and years since last NSW (eβ5 years decrease: 1.04; 95% CI, 1.00–1.09), which seems more clear in pre/peri-menopausal women (eβ5 years decrease: 1.13; 95% CI, 1.03–1.23). The association between NSW frequency and MD varied by menopausal status (Pinteraction, 0.006). Among postmenopausal women, MD was higher as the number of nightshifts per week reported decreased. Compared with women reporting 5–7 night-shifts per week, risk of having high MD was significantly higher in those working 3-4 nights/week (⁠|{e^\beta }$|⁠: 1.57; 95% CI, 1.11–2.21) or ≤ 2 nights/week (⁠|{e^\beta }$|⁠: 1.83; 95% CI, 1.32–2.52). This trend was not observed in pre/perimenopausal women. There were no differences in MD by timing of first child birth.

Finally, the analysis of the dose–response relationship for cumulative years of NSW is presented in Fig. 1. Starting from 15 to 20 years of cumulative NSW, the percentage of MD was higher than in women without NSW history, suggesting the existence of an exposure threshold, although departure from linearity was not statistically significant (P, 0.209).

Up to date this is the only study examining NSW and MD carried out in general population. Our results show that having ever worked in NSW is not related to MD; however, performing NSW for more than 15 years is associated to higher MD, with no evidence of a lineal dose-response relationship. Our findings suggest that MD may play an intermediate role in the not fully understood association between long term NSW and breast cancer. In addition, our data indicate that the frequency of weekly night-shifts may affect MD, although the effect may be substantially different in pre/perimenopausal and postmenopausal women. Shorter total years of NSW or other characteristics of this exposure do not seem to be associated with MD.

As far as we are aware of, the relationship between MD and NSW has only been explored in a cross-sectional study with 640 middle aged nurses/midwives. In that study, there was no association with cumulated NSW, whereas women currently working in rotating night-shifts had non-significant higher MD and women with higher number of night-shifts per month had slightly lower MD than women with lower night-shift frequency. As 90% of the participants had performed NSW for over 5 years, the very low proportion of unexposed women could explain the lack of concordance with our findings (21).

Circadian disruptive NSW has been classified as probably carcinogenic to humans by the IARC (3). Even though some studies, particularly in Asian populations, have reported no association (26), most evidence shows an increased risk of breast cancer after NSW exposure (4–8, 27–31). Up to six meta-analyses published in recent years have reported risk ratios for breast cancer ranging from 1.06 to 1.21 (4, 28–32) after NSW exposure. Two of the most recent meta-analyses observed a dose-response effect, with an increase in breast cancer risk per 5-year rise in NSW exposure (28, 30). However, often the association was only observed when long exposure to NSW was considered (5, 8, 9, 33, 34). MD is a dynamic trait, but our results seem to suggest that prolonged exposures might induce effects in MD that could remain after the end of NSW, and supports the inclusion of MD in the biological pathways that connect long term NSW to breast cancer. In addition, we observed an inverse relationship between years since last NSW and MD among those that had ever had NSW, which could reflect a decreasing effect of the exposure with time, but this trend is only clear in premenopausal women.

There are several biological mechanisms that could contribute to the observed higher MD of long-term night-shift workers (see Supplementary Fig. S1). Recent research suggests that long-term exposure to NSW-related light at night (LAN) may induce epigenetic dysregulation of microRNA that are involved in apoptotic and proliferative gene networks in breast cancer cells (35). There is also the hypothesis that phase shift (alteration of the peripheral functions due to the desynchronization with sleep-wake cycle) may confuse the body's master clock and produce variations in the circadian clock genes expression, promoting cellular proliferation and breast cancer (36, 37). However, the association of MD with cellular proliferation markers such as Ki67 is not clear (38, 39).

Other possible mechanism is LAN-related melatonin suppression. Melatonin, which is regulated by the light/dark cycle, presents antioxidant, antimitotic, antiangiogenic and antiestrogenic activity (10). Night-shift workers have substantially reduced 6-sulfatoxy melatonin levels during night work and daytime sleep, remaining low even when sleeping at night (40, 41). However, recent studies have not found any association between this biomarker and breast cancer risk (42, 43), or with MD (44).

Growth hormone (GH) and insulin-like growth factor 1 (IGF -1) may also play a role in the NSW and MD association. GH, which stimulates IGF-1 production, is secreted in a pulsatile fashion, with sleep and exercise being their strongest physiologic stimuli (45). Sleep deprivation due to NSW can de‐synchronize GH release timing. Individuals adapted to NSW experience lower GH nocturnal release, more frequent daily pulses and sporadic random releases (46). GH and IGF-1 promote breast cell proliferation and both have been associated with higher MD among women with high levels of estrogen (47).

Vitamin D could also mediate the association between MD and NSW as shift workers may have lower vitamin D levels due to reduced sun exposure (36, 37). Vitamin D is known to inhibit cell proliferation and induce apoptosis and differentiation (48). Inconclusive evidence links high circulating vitamin D levels with lower breast cancer risk (49) and it has also been hypothesized that it could reduce MD (50). It should be noted that our participants had low intake of vitamin D, as reported in a prior paper (51), but these levels did not differ by NSW history and the inclusion of vitamin D consumption in the model did not change our results

Lifestyle factors may also mediate the association between NSW and MD. There is some evidence that night-shift workers are more likely than others to smoke, to be less physically active, and to report poor diet choices (52), three known breast cancer risk factors (53–55). Indeed, in our sample night-shift workers were more likely to smoke and report a slightly, although non-significant, higher caloric intake than other workers. However, those in NSW had lower BMI, and alcohol consumption and physical activity did not differ between the two groups. In any case, our results are adjusted for these lifestyle factors, so they are not likely to explain the association observed.

Interestingly, number of night-shifts per week had different effects on pre/peri- and postmenopausal women. Postmenopausal women working 5 to 7 night-shifts per week had even lower MD than those unexposed to NSW. Among women exposed to NSW, two or less night-shifts per week were associated with higher MD. Five to seven shifts per week may correspond to regular night-shift workers who may not be at higher breast cancer risk according to a recent meta-analysis of cohort studies. This meta-analysis concluded that women in rotating NSW showed an increased risk of breast cancer whereas women with regular night work showed a marginally statistically significant decrease in breast cancer risk (28); a differential effect explained by variations in adaptation to NSW (56). In contrast, our results for pre/perimenopausal participants present a different pattern: MD was slightly higher in women working 5 to 7 night shifts per week compared with those unexposed to NSW.

This study has several strengths. The population-based nature of the study sample and the relatively high participation rate of women to breast cancer screening network in the Spanish regions analyzed (23) support the external validity of our results. In fact, according to the data supplied by the Spanish National Health Survey (57) our participants were very similar to the national population in the same age range in terms of lifestyle factors such as smoking habits, alcohol consumption, use of HRT and prevalence of night-shift workers (2). With regard to MD estimations, we have used a validated software to obtain a continuous measure of MD which allows to explore associations more accurately, as recommended by previous authors (58). We would like to emphasize that DDM-Spain logistic organization made it possible to use mammograms performed as part of regular clinical practice, that is, no additional mammograms were performed for this study.

Our study also has several limitations. First, we have a reduced number of NSW women in the subgroup of long-term exposure. Second, our sample reflects the Spanish target population of a breast cancer screening program (women between the ages of 45 and 68 years), so we had relatively small numbers of premenopausal women. This limited the power of our analyses to assess possible effect modification due to menopausal status. Third, as a cross-sectional study, changes in MD patterns across time and through changes in NSW could not be investigated. Fourth, NSW exposure information was self-reported and collected retrospectively and, thus, susceptible to recall bias. Nevertheless, since density assessment was blind and anonymous, any recall bias would not be differential. In turn, it may result in an underestimation of the effects studied. In addition, the possibility that the associations found could be due to other confounders not included in the analysis (e.g., sleep duration) can never be ruled out, although we made sensitivity analyses to evaluate the more plausible candidates. Fifth, we combined different images (digital and digitized digital or analogical images) as well as the estimations of two radiologists. We have adjusted for these possible sources of error by including these variables (type of mammogram and radiologist) into the model. Finally, we lack sufficient information to distinguish between rotating or permanent shifts, and information on chronotypes and sleep characteristics was not registered.

Our results show that long-term NSW is associated with MD, which could play an intermediate role in the still not fully understood relationship between NSW and breast cancer. Future studies with more detailed information on NSW and with larger sample size for night shift workers, especially pre/perimenopausal ones, are needed to confirm our results and to help identify the biological pathways involved in this association.

No potential conflicts of interest were disclosed.

Conception and design: A.M. Pedraza-Flechas, M. Pollán, B. Pérez-Gómez

Development of methodology: A.M. Pedraza-Flechas, C. Vidal, B. Pérez-Gómez

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C. Sánchez-Contador, C. Santamariña, C. Pedraz-Pingarrón, P. Moreo, M. Ederra, J. Miranda-García, D. Salas-Trejo

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.M. Pedraza-Flechas, V. Lope, C. Vidal, R. Llobet, N. Aragonés, M. Pollán, B. Pérez-Gómez

Writing, review, and/or revision of the manuscript: A.M. Pedraza-Flechas, V. Lope, C. Santamariña, C. Pedraz-Pingarrón, P. Moreo, M. Ederra, J. Miranda-García, C. Vidal, N. Aragonés, M. Pollán, B. Pérez-Gómez

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Pedraz-Pingarrón, P. Moreo, M. Pollán

Study supervision: D. Salas-Trejo, M. Pollán, B. Pérez-Gómez

We would like to thank the participants in the DDM-Spain/Var-DDM-Spain study for their contribution to breast cancer research. Other members of DDM-Spain/Var-DDM: Gonzalo. López-Abente, Roberto Pastor-Barriuso, Pablo Fernández-Navarro, Anna Cabanes, Soledad Laso, Manuela Alcaraz, María Casals, Inmaculada Martínez, Juan Carlos Pérez Cortés, Joaquín Antón, Nieves Ascunce, Isabel González-Román, Ana Belén Fernández, Montserrat Corujo, Soledad Abad, and Jesús Vioque.

A.M. Pedraza-Flechas FI14CIII/00013 PFIS; B. Perez-Gomez FIS PS09/0790; M. Pollán FIS PI060386, EPY1306/06 collaboration agreement between Astra-Zeneca and ISCIII, and FECMA 485 EPY 1170–10; R. LLobet Gent per Gent Fund (EDEMAC Project); All authors: European Regional Development Fund (FEDER).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Eurofound
.
Sixth European Working Conditions Survey (EWCS) [cited 2016 February 20]
.
Available from:
http://www.eurofound.europa.eu/publications/resume/2015/working-conditions/first-findings-sixth-european-working-conditions-survey-resume
2.
Instituto Nacional de Estadística
.
Encuesta de Población Activa (EPA). Variables de submuestra. Año 2013
.
[cited 2015 November 23]. Available from:
http://www.ine.es/prensa/np841.pdf
3.
IARC Working Group on the Evaluation of Carcinogenic Risks to Humans
. 
Painting, firefighting, and shiftwork
.
IARC Monogr Eval Carcinog Risks Hum
2010
;
98
:
9
764
.
4.
Kamdar
BB
,
Tergas
AI
,
Mateen
FJ
,
Bhayani
NH
,
Oh
J
. 
Night-shift work and risk of breast cancer: a systematic review and meta-analysis
.
Breast Cancer Res Treat
2013
;
138
:
291
301
.
5.
Hansen
J
,
Lassen
CF
. 
Nested case-control study of night shift work and breast cancer risk among women in the Danish military
.
Occup Environ Med
2012
;
69
:
551
6
.
6.
Hansen
J
,
Stevens
RG
. 
Case–control study of shift-work and breast cancer risk in Danish nurses: impact of shift systems
.
Eur J Cancer
2012
;
48
:
1722
9
.
7.
Knutsson
A
,
Alfredsson
L
,
Karlsson
B
,
Akerstedt
T
,
Fransson
EI
,
Westerholm
P
, et al
Breast cancer among shift workers: results of the WOLF longitudinal cohort study
.
Scand J Work Environ Health
2013
;
39
:
170
7
.
8.
Menegaux
F
,
Truong
T
,
Anger
A
,
Cordina-Duverger
E
,
Lamkarkach
F
,
Arveux
P
, et al
Night work and breast cancer: a population-based case-control study in France (the CECILE study)
.
Int J Cancer
2013
;
132
:
924
31
.
9.
Grundy
A
,
Richardson
H
,
Burstyn
I
,
Lohrisch
C
,
Sengupta
SK
,
Lai
AS
, et al
Increased risk of breast cancer associated with long-term shift work in Canada
.
Occup Environ Med
2013
;
70
:
831
8
.
10.
Di Bella
G
,
Mascia
F
,
Gualano
L
,
Di Bella
L
. 
Melatonin anticancer effects: review
.
Int J Mol Sci
2013
;
14
:
2410
30
.
11.
McCormack
VA
,
dos Santos Silva
I
. 
Breast density and parenchymal patterns as markers of breast cancer risk: a meta-analysis
.
Cancer Epidemiol Biomarkers Prev
2006
;
15
:
1159
69
.
12.
Boyd
NF
,
Martin
LJ
,
Bronskill
M
,
Yaffe
MJ
,
Duric
N
,
Minkin
S
. 
Breast tissue composition and susceptibility to breast cancer
.
J Natl Cancer Inst
2010
;
102
:
1224
37
.
13.
Boyd
NF
,
Rommens
JM
,
Vogt
K
,
Lee
V
,
Hopper
JL
,
Yaffe
MJ
, et al
Mammographic breast density as an intermediate phenotype for breast cancer
.
Lancet Oncol
2005
;
6
:
798
808
.
14.
Antoni
S
,
Sasco
AJ
,
dos Santos Silva
I
,
McCormack
V
. 
Is mammographic density differentially associated with breast cancer according to receptor status? A meta-analysis
.
Breast Cancer Res Treat
2013
;
137
:
337
47
.
15.
Cuzick
J
,
Warwick
J
,
Pinney
E
,
Duffy
SW
,
Cawthorn
S
,
Howell
A
, et al
Tamoxifen-induced reduction in mammographic density and breast cancer risk reduction: a nested case-control study
.
J Natl Cancer Inst
2011
;
103
:
744
52
.
16.
Martin
LJ
,
Melnichouk
O
,
Guo
H
,
Chiarelli
AM
,
Hislop
TG
,
Yaffe
MJ
, et al
Family history, mammographic density, and risk of breast cancer
.
Cancer Epidemiol Biomarkers Prev
2010
;
19
:
456
63
.
17.
Lope
V
,
Pérez-Gómez
B
,
Sánchez-Contador
C
,
Santamariña
MC
,
Moreo
P
,
Vidal
C
, et al
Obstetric history and mammographic density: a population-based cross-sectional study in Spain (DDM-Spain)
.
Breast Cancer Res Treat
2012
;
132
:
1137
46
.
18.
Pollán
M
,
Lope
V
,
Miranda-García
J
,
García
M
,
Casanova
F
,
Sánchez-Contador
C
, et al
Adult weight gain, fat distribution and mammographic density in Spanish pre- and post-menopausal women (DDM-Spain)
.
Breast Cancer Res Treat
2012
;
134
:
823
38
.
19.
Couto
E
,
Qureshi
SA
,
Hofvind
S
,
Hilsen
M
,
Aase
H
,
Skaane
P
, et al
Hormone therapy use and mammographic density in postmenopausal Norwegian women
.
Breast Cancer Res Treat
2012
;
132
:
297
305
.
20.
Cabanes
A
,
Pastor-Barriuso
R
,
García-López
M
,
Pedraz-Pingarrón
C
,
Sánchez-Contador
C
,
Vázquez Carrete
JA
, et al
Alcohol, tobacco, and mammographic density: a population-based study
.
Breast Cancer Res Treat
2011
;
129
:
135
47
.
21.
Peplonska
B
,
Bukowska
A
,
Sobala
W
,
Reszka
E
,
Gromadzinska
J
,
Wasowicz
W
, et al
Rotating night shift work and mammographic density
.
Cancer Epidemiol Biomarkers Prev
2012
;
21
:
1028
37
.
22.
Castelló
A
,
Prieto
L
,
Ederra
M
,
Salas-Trejo
D
,
Vidal
C
,
Sánchez-Contador
C
, et al
Association between the Adherence to the International Guidelines for Cancer Prevention and Mammographic Density
.
PLoS ONE
2015
;
10
:
e0132684
.
23.
Castells
X
,
Sala
M
,
Ascunce
N
,
Salas
D
,
Zubizarreta
R
,
Casamitjana
M
. 
Descripción del cribado del cáncer en España. Proyecto DESCRIC
.
Madrid
:
Plan de Calidad para el Sistema Nacional de Salud. Ministerio de Sanidad y Consumo. Agència d'Avaluació de Tecnologia i Recerca Mèdiques de Cataluña
; 
2007
.
Informes de Evaluación de Tecnologías Sanitarias, AATRM núm. 2006/01
.
24.
Llobet
R
,
Pollán
M
,
Antón
J
,
Miranda-García
J
,
Casals
M
,
Martínez
I
, et al
Semi-automated and fully automated mammographic density measurement and breast cancer risk prediction
.
Comput Methods Programs Biomed
2014
;
116
:
105
15
.
25.
Pollán
M
,
Llobet
R
,
Miranda-García
J
,
Antón
J
,
Casals
M
,
Martínez
I
, et al
Validation of DM-Scan, a computer-assisted tool to assess mammographic density in full-field digital mammograms
.
Springerplus
2013
;
2
:
242
.
26.
Li
W
,
Ray
RM
,
Thomas
DB
,
Davis
S
,
Yost
M
,
Breslow
N
, et al
Shift work and breast cancer among women textile workers in Shanghai, China
.
Cancer Causes Control
2015
;
26
:
143
50
.
27.
Papantoniou
K
,
Castaño-Vinyals
G
,
Espinosa
A
,
Aragonés
N
,
Pérez-Gómez
B
,
Ardanaz
E
, et al
Breast cancer risk and night shift work in a case–control study in a Spanish population
.
Eur J Epidemiol
2015
;
9
:
867
78
.
28.
Lin
X
,
Chen
W
,
Wei
F
,
Ying
M
,
Wei
W
,
Xie
X
. 
Night-shift work increases morbidity of breast cancer and all-cause mortality: a meta-analysis of 16 prospective cohort studies
.
Sleep Med
2015
;
16
:
1381
7
.
29.
Jia
Y
,
Lu
Y
,
Wu
K
,
Lin
Q
,
Shen
W
,
Zhu
M
, et al
Does night work increase the risk of breast cancer? A systematic review and meta-analysis of epidemiological studies
.
Cancer Epidemiol
2013
;
37
:
197
206
.
30.
Wang
F
,
Yeung
KL
,
Chan
WC
,
Kwok
CCH
,
Leung
SL
,
Wu
C
, et al
A meta-analysis on dose-response relationship between night shift work and the risk of breast cancer
.
Ann Oncol
2013
;
24
:
2724
32
.
31.
He
C
,
Anand
ST
,
Ebell
MH
,
Vena
JE
,
Robb
SW
. 
Circadian disrupting exposures and breast cancer risk: a meta-analysis
.
Int Arch Occup Environ Health
2015
;
88
:
533
47
.
32.
Ijaz
S
,
Verbeek
J
,
Seidler
A
,
Lindbohm
M-L
,
Ojajärvi
A
,
Orsini
N
, et al
Night-shift work and breast cancer—a systematic review and meta-analysis
.
Scand J Work Environ Health
2013
;
39
:
431
47
.
33.
Åkerstedt
T
,
Knutsson
A
,
Narusyte
J
,
Svedberg
P
,
Kecklund
G
,
Alexanderson
K
. 
Night work and breast cancer in women: a Swedish cohort study
.
BMJ Open
2015
;
5
:
e008127
.
34.
Schernhammer
ES
,
Kroenke
CH
,
Laden
F
,
Hankinson
SE
. 
Night work and risk of breast cancer
.
Epidemiology
2006
;
17
:
108
11
.
35.
Liu
R
,
Jacobs
DI
,
Hansen
J
,
Fu
A
,
Stevens
RG
,
Zhu
Y
. 
Aberrant methylation of miR-34b is associated with long-term shiftwork: a potential mechanism for increased breast cancer susceptibility
.
Cancer Causes Control
2015
;
26
:
171
8
.
36.
Dickerman
B
,
Liu
J
. 
Does current scientific evidence support a link between light at night and breast cancer among female night-shift nurses? Review of evidence and implications for occupational and environmental health nurses
.
Workplace Health Saf
2012
;
60
:
273
281
.
37.
Fritschi
L
,
Glass
DC
,
Heyworth
JS
,
Aronson
K
,
Girschik
J
,
Boyle
T
, et al
Hypotheses for mechanisms linking shiftwork and cancer
.
Med Hypotheses
2011
;
77
:
430
6
.
38.
Huo
CW
,
Chew
G
,
Hill
P
,
Huang
D
,
Ingman
W
,
Hodson
L
, et al
High mammographic density is associated with an increase in stromal collagen and immune cells within the mammary epithelium
.
Breast Cancer Res
2015
;
17
:
79
.
39.
Harvey
JA
,
Santen
RJ
,
Petroni
GR
,
Bovbjerg
VE
,
Smolkin
ME
,
Sheriff
FS
, et al
Histologic changes in the breast with menopausal hormone therapy use: correlation with breast density, estrogen receptor, progesterone receptor, and proliferation indices
.
Menopause
2008
;
15
:
67
73
.
40.
Marie Hansen
A
,
Helene Garde
A
,
Hansen
J
. 
Diurnal urinary 6-sulfatoxymelatonin levels among healthy Danish nurses during work and leisure time
.
Chronobiol Int
2006
;
23
:
1203
15
.
41.
Davis
S
,
Mirick
DK
,
Chen
C
,
Stanczyk
FZ
. 
Night shift work and hormone levels in women
.
Cancer Epidemiol Biomarkers Prev
2012
;
21
:
609
18
.
42.
Sturgeon
SR
,
Doherty
A
,
Reeves
KW
,
Bigelow
C
,
Stanczyk
FZ
,
Ockene
JK
, et al
Urinary levels of melatonin and risk of postmenopausal breast cancer: women's health initiative observational cohort
.
Cancer Epidemiol Biomarkers Prev
2014
;
23
:
629
37
.
43.
Brown
SB
,
Hankinson
SE
,
Eliassen
AH
,
Reeves
KW
,
Qian
J
,
Arcaro
KF
, et al
Urinary melatonin concentration and the risk of breast cancer in Nurses' Health Study II
.
Am J Epidemiol
2015
;
181
:
155
62
.
44.
Nagata
C
,
Matsubara
T
,
Hara
T
,
Fujita
H
,
Nagao
Y
,
Yamamoto
S
, et al
Urinary 6-sulfatoxymelatonin and mammographic density in Japanese women
.
Cancer Epidemiol Biomarkers Prev
2007
;
16
:
2509
12
.
45.
Ritsche
K
,
Nindl
BC
,
Wideman
L
. 
Exercise-Induced growth hormone during acute sleep deprivation
.
Physiol Rep
2014
;
2
.
46.
Brandenberger
G
,
Gronfier
C
,
Chapotot
F
,
Simon
C
,
Piquard
F
. 
Effect of sleep deprivation on overall 24 h growth-hormone secretion
.
Lancet
2000
;
356
:
1408
.
47.
Frydenberg
H
,
Flote
VG
,
Iversen
A
,
Finstad
SE
,
Furberg
AS
,
Torjesen
PA
, et al
Insulin-like growth factor-1, growth hormone, and daily cycling estrogen are associated with mammographic density in premenopausal women.
Cancer Causes Control
2014
;
25
:
891
903
.
48.
Deeb
KK
,
Trump
DL
,
Johnson
CS
. 
Vitamin D signalling pathways in cancer: potential for anticancer therapeutics
.
Nat Rev Cancer
2007
;
7
:
684
700
.
49.
Chen
P
,
Hu
P
,
Xie
D
,
Qin
Y
,
Wang
F
,
Wang
H
. 
Meta-analysis of vitamin D, calcium and the prevention of breast cancer
.
Breast Cancer Res Treat
2010
;
121
:
469
77
.
50.
Ellingjord-Dale
M
,
dos-Santos-Silva
I
,
Grotmol
T
,
Sakhi
AK
,
Hofvind
S
,
Qureshi
S
, et al
Vitamin D intake, month the mammogram was taken and mammographic density in Norwegian women aged 50–69
.
PLoS ONE
2015
;
10
:
e0123754
.
51.
García-Arenzana
N
,
Navarrete-Muñoz
EM
,
Vázquez-Carrete
JA
,
Moreno
MP
,
Vidal
C
,
Salas
D
, et al
Compliance with current dietary recommendations and geographical variability of diet in women participating in 7 screening programs for breast cancer in Spain
.
Nutr Hosp
2011
;
26
:
863
73
.
52.
Puttonen
S
,
Härmä
M
,
Hublin
C
. 
Shift work and cardiovascular disease - pathways from circadian stress to morbidity
.
Scand J Work Environ Health
2010
;
36
:
96
108
.
53.
Gaudet
MM
,
Gapstur
SM
,
Sun
J
,
Diver
WR
,
Hannan
LM
,
Thun
MJ
. 
Active smoking and breast cancer risk: original cohort data and meta-analysis
.
J Natl Cancer Inst
2013
;
105
:
515
25
.
54.
Pizot
C
,
Boniol
M
,
Mullie
P
,
Koechlin
A
,
Boniol
M
,
Boyle
P
, et al
Physical activity, hormone replacement therapy and breast cancer risk: A meta-analysis of prospective studies
.
Eur J Cancer
2016
;
52
:
138
54
.
55.
Castelló
A
,
Martín
M
,
Ruiz
A
,
Casas
AM
,
Baena-Cañada
JM
,
Lope
V
, et al
Lower breast cancer risk among women following the world cancer research fund and American Institute for cancer research lifestyle recommendations: EpiGEICAM case-control study
.
PLoS ONE
2015
;
10
:
e0126096
.
56.
Stevens
RG
,
Hansen
J
,
Costa
G
,
Haus
E
,
Kauppinen
T
,
Aronson
KJ
, et al
Considerations of circadian impact for defining “shift work” in cancer studies: IARC Working Group Report
.
Occup Environ Med
2011
;
68
:
154
62
.
57.
Instituto Nacional de Estadística
.
Encuesta Nacional de Salud 2011 - 2012. [cited 2016 March 2]
.
Available from:
http://www.ine.es/prensa/np770.pdf.
58.
Santen
RJ
,
Boyd
NF
,
Chlebowski
RT
,
Cummings
S
,
Cuzick
J
,
Dowsett
M
, et al
Critical assessment of new risk factors for breast cancer: considerations for development of an improved risk prediction model
.
Endocr Relat Cancer
2007
;
14
:
169
87
.