Implications of lifestyle and occupational factors on the risk of breast cancer in shiftwork nurses

healthcare Article Implications of Lifestyle and Occupational Factors on the Risk of Breast Cancer in Shiftwork Nurses Javier Fagundo-Rivera 1,2,3 , Regina Allande-Cussó4, Mónica Ortega-Moreno 5, Juan Jesús García-Iglesias 6,* , Adolfo Romero 7,*, Carlos Ruiz-Frutos 6,8,† and Juan Gómez-Salgado 6,8,†   Citation: Fagundo-Rivera, J.; Allande-Cussó, R.; Ortega-Moreno, M.; García-Iglesias, J.J.; Romero, A.; Ruiz-Frutos, C.; Gómez-Salgado, J. Implications of Lifestyle and Occupational Factors on the Risk of Breast Cancer in Shiftwork Nurses. Healthcare 2021,9, 649. https:// doi.org/10.3390/healthcare9060649 Academic Editor: Alberto Modenese Received: 10 May 2021 Accepted: 28 May 2021 Published: 30 May 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). 1 Health Sciences Doctorate School, University of Huelva, 21071 Huelva, Spain; javier[email protected] 2Centro Universitario de Enfermería Cruz Roja, University of Seville, 41009 Seville, Spain 3Escola Superior de Saúde, Universidade Atlântica, 2730-036 Barcarena, Portugal 4Department of Nursing, University of Sevilla, 41009 Sevilla, Spain; [email protected] 5Department of Economy, Faculty of Labour Sciences, University of Huelva, 21007 Huelva, Spain; [email protected] 6Department of Sociology, Social Work and Public Health, Faculty of Labour Sciences, University of Huelva, 21007 Huelva, Spain; [email protected] (C.R.-F.); [email protected] (J.G.-S.) 7 Nursing and Podiatry Department, Health Sciences School, University of Malaga, Instituto de Investigación Biomédica de Málaga (IBIMA), 29071 Málaga, Spain 8Safety and Health Postgraduate Programme, Espíritu Santo University, Guayaquil 092301, Ecuador *Correspondence: [email protected] (J.J.G.-I.); adolforomer[email protected] (A.R.); Tel.: +34-959219700 (J.J.G.-I.); +34-951952801 (A.R.) † Senior authors, these authors contributed equally to this work and should be regarded as equal. Abstract: Shift work that involves circadian disruption has been highlighted as a likely carcinogenic factor for breast cancer in humans. Also, unhealthy lifestyle habits observed in night work nurses could be causally related to an increase in the incidence of estrogen-positive breast tumours in this population. Assessing baseline risk of breast cancer in nurses is essential. The objective of this study was to analyze the risk of breast cancer that nurses had in relation to their lifestyle and labour factors related to shift work. A cross-sectional descriptive study through a questionnaire about sociodemographic variables, self-perception of health, and working life was designed. The sample consisted of 966 nurses. The relationship between variables was tested. A binary logistic regression and a classification and regression tree were performed. The most significant labour variables in relation to the risk of breast cancer were the number of years worked (more than 16 years; p< 0.01; OR = 8.733, 95% CI = 2.811, 27.134) and the total years performing more than 3 nights per month (10 or more years; p< 0.05; OR = 2.294, 95% CI = 1.008, 5.220). Also, the nights worked throughout life (over 500; OR = 4.190, 95% CI = 2.118, 8.287) were significant in the analysis. Nurses who had or ever had breast cancer valued their self-perceived health more negatively (p< 0.001) and referred a lower quality of sleep (p< 0.001) than the non-cases nurses. The occupational factors derived from night work could have several impacts on nurses’ health and their family-work balance. Promoting healthy lifestyles, informing about shift work risks, and adjusting shift work schedules are critical methods to decrease the possible effects of circadian disruption in nurses. Keywords: breast cancer; night work; shift work; health personnel; occupational disease; working conditions; prevention; carcinogens 1. Introduction Shift work, including night shift work, has been associated with circadian disruption in several epidemiological studies conducted on nurses [ 1 – 5 ] and in several targeted investigations [ 6 – 9 ] in which expression, methylation and polymorphisms of circadian genes that could be associated with breast cancer risk among shift nurses have been studied. In fact, long-term rotating shifts (i.e., 12 h rotating shifts) and night work have been linked to the presence of tumours with oestrogen-positive (ER+) and progesterone (PR+) Healthcare 2021,9, 649. https://doi.org/10.3390/healthcare9060649 https://www.mdpi.com/journal/healthcare Healthcare 2021,9, 649 2 of 18 receptors [ 4 , 10 – 15 ], and to the luminal subtypes of tumour classification (mainly luminal A) in several studies using nursing populations [16,17]. In addition, high levels of oestradiol have been recorded in nurses performing night shifts compared to those performing day shifts [ 10 , 11 , 13 ], and several studies have found significant differences between nurses working on permanent night shifts and those who performed rotating shifts, concluding that long-term, high-intensity shifts for several consecutive years can significantly influence the risk of breast cancer. Under this evidence, shift work and night work were classified as likely carcinogenic factors (Group 2A) by the International Agency for Research on Cancer (IARC), especially incidents in those professions that, such as nursing, must adjust their work throughout 24 h a day [ 18 – 23 ], although in-depth studies confirming these findings are still required. One of the most noted causes has been related to the loss of synchronisation between the circadian rhythm and the sleeping patterns of nurses during continuous and long-duration rotating shift work [ 18 , 19 , 24 – 26 ]. Also, eating at night or referring hunger during hours that are normally dedicated to nocturnal rest may lead to disturbances in the suprachiasmatic nucleus control of hunger and satiety cycles, intrinsically related to metabolic regulation mechanisms and energy activity in peripheral tissues [ 24 ]. In addition, certain levels of exposure to blue or artificial light at night are known to affect the circadian system, altering production times and melatonin levels, among other hormones [27,28]. Circadian and sleep disturbances (i.e., social jet lag [ 29 ] or shift work disorder [ 30 ]) can lead nurses to experience extreme tiredness [ 31 , 32 ], to have a less active life during free time [ 33 ], and to carry out poor dietary control [ 23 ], factors that together increase cardiovascular and diabetes risk [ 13 , 34 – 37 ], and pose an increased risk of breast cancer [ 18 ]. Thus, the duration of the work shift becomes a significant predictor of the quality of nurse care and job safety [ 38 , 39 ], since a longer duration of shifts has been linked to an increase in errors during attendance [ 40 , 41 ]. Therefore, occupational health specialists have the role of providing advice to managers and workers on the best strategies to reduce the negative effects of shift-induced circadian desynchronisation, by evaluating clinical symptoms and behaviours related to sleep-wake patterns, obesity, type 2 diabetes, or dyslipidaemia caused by the shift work disorder [24]. In view of the above, it is necessary to examine the risk profile of breast cancer that nurses working on rotating and night shifts have, as well as their perception of their own health and the factors that can harm or protect it. With this information, conclusions could be drawn that could facilitate health services’ managers’ decision-making when planning more appropriate work shifts to reduce risk factors of occupational breast cancer. In this way, the objective of this study was to analyse the risk of breast cancer and the selfperception of health by nurses in relation to lifestyle and occupational variables associated with shift work (including night shift). 2. Materials and Methods 2.1. Design and Sample Cross-sectional, questionnaire-based descriptive study on the population of Registered Nurses in Spain, both men and women (currently 316,094 subjects) [ 42 ]. The sample selection was made by non-probabilistic snowball sampling, estimating the optimal size at 980 nurses with a 95% confidence level, 3.5% accuracy, and 20% adjustment for losses. The inclusion criteria were (a) Being a nurse and being registered in the General Council of Nurses of Spain; (b) Working in private and/or public centres; (c) Wish to participate, having read and signed the informed consent. On the other hand, the exclusion criteria were (a) Not being Registered Nurses; (b) Being under 18; (c) Working outside Spain. 2.2. Shift Work and Night Work Definition Cambridge Business English Dictionary defines shift work as: “a system in which different groups of workers work somewhere at different times of the day and night”. Healthcare 2021,9, 649 3 of 18 Besides, the IARC defined night work as: “one that requires at least three hours of work between midnight and 5 a.m.” [ 20 , 21 , 43 ]. The main characteristics of shift work are shown in Table S1 [11,20,21,23,44]. 2.3. Instrument To gather the appropriate information, an online questionnaire was used. The main risk variables related to breast cancer and night work which were used in this investigation were extracted and adapted ad hoc based on other questionnaires found in the scientific evidence [ 3 , 11 , 14 , 45 , 46 ]. The final questionnaire was formed by the 8 sections detailed below. The validation of the final instrument was carried out through two rounds of analysis by a panel of experts using a Delphi technique to determine whether the detected variables and the design of the questionnaire were relevant and appropriate in the context of the study. This group was made up of three nurses, two physicians, two psychologists, two members of a healthcare system management board and one methodologist. Subsequently, a piloting was carried out favourably in 10 people from different nursing areas to assess the suitability of the questions, possible grammatical errors or mistakes that were not previously detected. 2.4. Variables The variables considered in this study were: •Sociodemographic data (age, sex and marital relationship). • Self-perception of health. An ad hoc evaluation tool consisting of 5 direct questions was designed to assess nurses’ perception of their own health. The valuation scale was a Likert type with ten response possibilities ranging from 1 “very low” to 10 “very high”. The questions used were: “How do you value your overall health?”, “How do you value the quality of your rest?”, “How do you value the effect shift work has on your health?”, “How do you value your level of work stress?” and “How do you value your satisfaction with your current job?”. • General data on disease and cancer (current disease, oncological disease, number of mammograms, use of oral contraceptives and presence of first-degree familial cancer). • Lifestyle habits (Body Mass Index-BMI-was calculated with the weight and height indicated by the participants. Working exertion, measured by light, moderate, hard or very hard. Free-time physical activity, measured by the time spent in hours). • Family burdens (care for children under 14, and care for dependents or elderly family members at home). •Sleeping aids (“Do you take any medication to sleep?”, “which?”). • Exposure to tobacco (consumption habits (did you ever smoked?), exposure to tobacco in the workplace and at home). • Labour information (type of current working schedule, working experience (throughout life), number of years working regularly 3 nights per month or more, number of worked nights accumulated throughout life, and sick leaves throughout life and in the last year). 2.5. Data Collection Procedure The study development took place from December 2019 to November 2020. Google Forms © (Google, Mountain View, CA, USA) was used to create the online questionnaire. Participants could not access the questionnaire until they had previously done the following: (a) Having read and understood an introductory letter to the study and its objectives; (b) Having confirmed voluntary and anonymous participation in the study; (c) Declaring working as a nurse in Spain and being currently registered. The data obtained from the questionnaires were entered in Excel (Microsoft © , Redmond, WA, USA), R Commander 4.0.0 [47], and SPSS version 26.0 (IBM©, Armonk, NY, USA) for the statistical analysis. The online questionnaire link was provided via email to the nurses who were registered in the General Nursing Council of Spain and previously had accepted to receive Healthcare 2021,9, 649 4 of 18 emails, news and information bulletins from this institution. The General Council of Official Nursing Colleges of Spain is the only regulatory body and competent authority of the nursing profession in Spain, therefore the registration in this government-authorized licensing body is obligatory to obtain a nursing license and be legally authorized to work, thus being considered a Registered Nurse (RN). Given this fact, a high number of nurses were consulted via email thanks to this collaboration, although we were not able to control this intervention due to the Data Protection policy of this institution. The social networks of official entities and professional groups of renowned prestige in the area of nursing in Spain (i.e., University of Huelva or Spanish Nurses Syndicate) also collaborated with the dissemination of the questionnaire. 2.6. Statistical Analysis Absolute frequencies, percentages, and measures of position and dispersion, depending on the type of variable, described the variables of interest. Student’s T-test for independent samples and the Chi-squared test were used to contrast differences and relationship (OR) between the variables and breast cancer risk. The Mann–Whitney U nonparametric test for independent samples was used to analyse heterogeneity in the self-perception of health category. Binary logistic regression allowed building a model to study the presence of breast cancer and identify those risk variables that played a relevant role. The Hosmer-Lemeshow test was used, and Odds Ratios (OR) were estimated with their confidence intervals. Finally, the CART (Classification and Regression Trees) data mining method [ 48 , 49 ] was used to design a binary algorithm to predict which variables of the self-perception of health category played a significant role in breast cancer. The CART methodology refers to a model where the target variable and the algorithm itself are used to predict values based on several categorical or continuous input variables. It is shown as a Decision Tree Classifier (Figure 1), where each round is known as Node. Each node will have a question or if-clause according to which the subjects of study will be routed to a specific internal or terminal/leaf-node that will tell the final prediction. Each node shows the predicted class, the predicted probability of cases within the node, and the percentage of cases of the node over the total sample. In summary, there are three types of nodes: • Root Node: represents a single input variable and a split point on that variable. It does not have any parent node and gives two children nodes based on the question. •Internal Node: it will have a parent node and gives two children nodes. • Terminal or Leaf Node: contains an output variable which is used to make a prediction. This node will have a parent node but will not have any children nodes. Healthcare 2021, 9, x 5 of 18 Figure 1. Decision Tree Classifier (for learning purpose). 3. Results 3.1. Two-Dimensional Analysis for Healthy Participants and Those Who Have or Ever Had Breast Cancer The questionnaire was answered by a total of 966 nursing professionals, of whom 502 (51.97%) were healthy, 99 (10.25%) had or ever had some form of cancer, and 365 (37.78%) had another type of disease. Of those who had or ever had cancer, for 56.57%, it was breast cancer (56 subjects). Common diseases found in both men and women were hypertension, diabetes, obesity, asthma, or seasonal allergy. In women, isolated cases of thyroid cancer, cervical cancer, melanoma, leukaemia, and lymphoma were also found. For the bivariate analysis, healthy individuals (502) and breast cancer patients (56) were compared with the variables of interest. The median age of the sample was 41 years (p = 0.367) and 89.6% of the sample were women (p = 0.705). Subsequent bivariate analysis categorized by sex revealed only age (χ2 = 9.669; p 0.002; OR = 2.466; 95%CI= 1.376, 4.419) as significant in relation to breast cancer (see Table S2). According to the current workplace, a 19.2% of the sample worked on permanent shifts, mostly morning shifts, and a 74.5% worked on rotating shifts. Only a 3.8% worked on 24h-shifts. Approximately a 72.4% of the sample was working on night shifts at the time of the survey (Table 1). Table 1. Description of the sample’s work organization. Cases Percentage Permanent Shift 107 19.2% Only morning 73 68.2% Only afternoon/evening 5 4.7% Only night 29 27.1% Rotative 3 Shifts/24 h Cycles (M/AE/N) 212 37.9% Rotative 2 Shifts/24 h Cycles 204 36.6% Only morning + eventual extra-duty (+17 h) 33 16.2% 12-h shifts 71 34.8% Rotative M and N 31 15.2% Rotative M and AE 62 30.4% Rotative AE and N 7 3.4% 24-h Shifts 21 3.8% Irregular 14 2.5% Total general 558 100% M: morning (7-h shift); AE: afternoon/evening (7-h shift); N: night (10-h shift). Figure 1. Decision Tree Classifier (for learning purpose). Healthcare 2021,9, 649 5 of 18 2.7. Ethical Considerations For this study, the Declaration of Helsinki (Fortaleza, 2013) was taken into consideration and explicit written permission was obtained from participants through their informed consent. Data obtained in this study were to be duly guarded by the research team, ensuring the confidential use and processing in accordance with the Organic Law on Protection of Personal Data and the Guarantee of Digital Rights. Approval was obtained from the Research Ethics Committee of the province of Huelva, belonging to the Regional Government of Andalusia, with code TD-CMTE-2020. Likewise, approval was obtained from the Research Ethics Committee of the General Council of Official Nursing Colleges of Spain with code PI 2109/02/CE. 3. Results 3.1. Two-Dimensional Analysis for Healthy Participants and Those Who Have or Ever Had Breast Cancer The questionnaire was answered by a total of 966 nursing professionals, of whom 502 (51.97%) were healthy, 99 (10.25%) had or ever had some form of cancer, and 365 (37.78%) had another type of disease. Of those who had or ever had cancer, for 56.57%, it was breast cancer (56 subjects). Common diseases found in both men and women were hypertension, diabetes, obesity, asthma, or seasonal allergy. In women, isolated cases of thyroid cancer, cervical cancer, melanoma, leukaemia, and lymphoma were also found. For the bivariate analysis, healthy individuals (502) and breast cancer patients (56) were compared with the variables of interest. The median age of the sample was 41 years (p= 0.367) and 89.6% of the sample were women (p= 0.705). Subsequent bivariate analysis categorized by sex revealed only age ( χ2 = 9.669; p0.002; OR = 2.466; 95%CI= 1.376, 4.419) as significant in relation to breast cancer (see Table S2). According to the current workplace, a 19.2% of the sample worked on permanent shifts, mostly morning shifts, and a 74.5% worked on rotating shifts. Only a 3.8% worked on 24h-shifts. Approximately a 72.4% of the sample was working on night shifts at the time of the survey (Table 1). Table 1. Description of the sample’s work organization. Cases Percentage Permanent Shift 107 19.2% Only morning 73 68.2% Only afternoon/evening 5 4.7% Only night 29 27.1% Rotative 3 Shifts/24 h Cycles (M/AE/N) 212 37.9% Rotative 2 Shifts/24 h Cycles 204 36.6% Only morning + eventual extra-duty (+17 h) 33 16.2% 12-h shifts 71 34.8% Rotative M and N 31 15.2% Rotative M and AE 62 30.4% Rotative AE and N 7 3.4% 24-h Shifts 21 3.8% Irregular 14 2.5% Total general 558 100% M: morning (7-h shift); AE: afternoon/evening (7-h shift); N: night (10-h shift). No statistically significant differences were found for parity, either in the general (p= 0.684) and sex-categorized analysis (p= 0.648). On the other hand, significant associations with breast cancer were found in participants with a partner (p= 0.041) and those who cared for dependents at home outside the working hours (p< 0.001). Indeed, being single (OR = 0.541; 95% CI = 0.298, 0.982) and without family caring responsibilities (OR = 0.288 ; 95% CI = 0.146, 0.569) were related with breast cancer risk reduction (Table 2). Healthcare 2021,9, 649 6 of 18 Table 2. Sociodemographic and lifestyle characteristics of the sample. N (%) Healthy Cases (%) (N = 502) Breast Cancer Cases (%) (N = 56) χ2pOdds Ratio (CI = 95%) Sex Male 58 (10.4) 53 (10.6) 5 (8.9) 0.144 0.705 1.203 (0.460, 3.145) Female 500 (89.6) 449 (89.4) 51 (91.1) Age 41 or younger 281 (50.4) 256 (51.0) 25 (44.6) 0.813 0.367 1.290 (0.741, 2.247) Older than 41 277 (49.6) 246 (49.0) 31 (55.4) Marital relationship With partner 317 (56.8) 278 (55.4) 39 (69.6) 4.178 0.041 0.541 (0.298, 0.982) Single 241 (43.2) 224 (44.6) 17 (30.4) Children under 14 years Yes 225 (40.3) 201 (40.0) 24 (43.0) 0.166 0.684 0.890 (0.509, 1.558) No 333 (59.7) 301 (60.0) 32 (57.1) Care for dependents at home Yes 58 (10.4) 44 (8.8) 14 (25.0) 14.257 <0.001 0.288 (0.146, 0.569) No 500 (89.6) 458 (91.2) 42 (75.0) Mammography (N = 497) Yes 211 (42.5) 156 (35.3) 55 (100) * <0.001 1.353 (1.248, 1.466) Never 286 (57.5) 286 (64.7) 0 (0) Family history of cancer (N = 551) Yes 72 (13.1) 58 (11.7) 14 (25.0) 7.814 0.005 0.398 (0.205, 0.773) No 479 (86.9) 437 (88.3) 42 (75.0) BMI Underweight 10 (1.8) 8 (1.6) 2 (3.6) 8.074 0.045 - Normal 376 (67.4) 347 (69.1) 29 (51.8) Overweight 128 (22.9) 111 (22.1) 17 (30.4) Obese 44 (7.9) 36 (7.2) 8 (14.3) Physical activity at work Light 124 (22.2) 114 (22.7) 10 (17.9) 30.175 <0.001 - Moderate 313 (56.1) 283 (56.4) 30 (53.6) Hard 113 (20.3) 103 (20.5) 10 (17.9) Very hard 8 (1.4) 2 (0.4) 6 (10.7) Physical activity during leisure time Two hours or less 286 (51.25) 259 (51.6) 27 (48.2) 0.230 0.631 1.144 (0.659, 1.887) More than 2 h 272 (28.75) 243 (48.4) 29 (51.8) Tobacco consumption Yes 301 (53.9) 271 (54.0) 30 (53.6) 0.003 0.953 1.016 (0.584, 1.770) No 257 (46.1) 231 (46.0) 26 (46.4) Compliance with the smoking ban at work Totally 124 (22.2) 104 (20.7) 20 (35.7) 11.377 0.010 - Almost always 239 (42.8) 213 (42.4) 26 (46.4) Hardly ever 141 (25.3) 132 (26.3) 9 (16.1) Never 54 (9.7) 53 (10.6) 1 (1.8) Exposure to tobacco smoke at home More than 5 h a day 22 (3.9) 15 (3.0) 7 (12.5) 15.967 0.001 - Between 1 and 5 h a day 36 (6.5) 36 (7.2) 0 (0) Less than 1 h a day 42 (7.5) 39 (7.8) 3 (5.4) Never or hardly ever 458 (82.1) 412 (82.1) 46 (82.1) Healthcare 2021,9, 649 7 of 18 Table 2. Cont. N (%) Healthy Cases (%) (N = 502) Breast Cancer Cases (%) (N = 56) χ2pOdds Ratio (CI = 95%) Use of medication to sleep Yes 116 (20.8) 83 (16.5) 33 (58.9) 54.988 <0.001 0.138 (0.077, 0.247) No 442 (79.2) 419 (83.5) 23 (41.1) Hormone-based oral contraceptives (women only) (N = 504) Yes 334 (66.3) 295 (65.7) 39 (70.9) 0.594 0.441 0.786 (0.425, 1.451) Never 170 (33.7) 154 (34.3) 16 (29.1) *: Fisher. BMI: Body Mass Index. Under 18.5: Underweight; (18.5, 25) Normal; (25, 29.9) Overweight; equal or higher to 30: Obese. Having ever had a mammogram and the number of performed mammograms also resulted significative for breast cancer risk (mean = 2.27; SD = 4.43; p< 0.001; OR = 1.353 , 95% CI = 1.248 , 1.466). Similarly, the presence of familial cancer was significant (p= 0.005) , resulting in risk reduction when there was no family cancer history (OR = 0.398; 95% CI = 0.205, 0.773) (Table 2). In terms of lifestyle habits, BMI showed significant differences (p= 0.045). The higher number of breast cancer cases had normal BMI. Similarly, statistically significant differences (p< 0.001) were detected depending on the intensity of physical activity at work. Most breast cancer cases classified their exertion at work as moderate. However, no significant differences were found for the physical activity during free time (p= 0.631). Statistically significant differences were found regarding the frequency of exposure to tobacco smoke at home (p= 0.001) and the compliance with the smoking ban in the workplace (p= 0.010), but not with having ever smoked (p= 0.953). Lastly, the 79.2% of the nurses claimed not to take any sleep medication, although this variable was relevant in breast cancer cases (p< 0.001). In fact, not taking sleep medications resulted in risk reduction (OR = 0.138; 95% CI = 0.077, 0.247) (Table 2). The most common medication was oral melatonin (60 subjects, of which 11 had breast cancer). Among other remedies, infusions (valerian, melissa), hypnotics (doxylamine, zolpidem), anxiolytics (alprazolam, bromazepam), antidepressants (trazodone), and other benzodiazepines (lorazepam, lormetazepam, diazepam) were used. Oral contraceptives were only used by women in this study, not being significant for breast cancer (p= 0.441). With respect to labour data (Table 3), being not currently working at night (p< 0.001; OR = 2.708, 95% CI = 1.548, 4.735) and being not currently shift working (p< 0.001; OR = 3.148 , 95% CI = 1.765, 5.615) showed a statistical association with the risk of breast cancer. Considering the working history of the participating subjects, having exceeded 16 years of work was presented as one of the most significant variables in this study (p< 0.001 ; OR = 12.346, 95% CI = 4.854, 31.250). Nurses who had or ever had breast cancer had worked a mean of 26.1 years (SD = 8.1), while healthy nurses had worked a mean of 15.0 years (SD = 9.2). On the other hand, the percentage of cases with breast cancer was also higher in professionals with 500 or more nights worked (OR = 4.190, 95% CI = 2.118 , 8.287) and when more than 3 nights per month had been worked for more than 10 years (OR = 4.132 ; 95% CI = 2.227, 7.634), finding a statistically significant difference in both situations (p< 0.001). The mean number of nights worked was 627.9 (SD = 639.4) in the case of healthy subjects and 1017.4 nights in those who had or ever had breast cancer (SD = 837.9) . 2.3% of respondents never worked night shifts. Finally, the cumulative number of sick leaves, both in the last year (mean 0.35; SD = 0.542) and throughout the professional life (mean 2.20; SD = 2.118), have shown a statistically significant association with breast cancer (p< 0.001 in all cases) (Table 3). Healthcare 2021,9, 649 8 of 18 Table 3. Labour variables and risk of breast cancer. N (%) Healthy Cases (%) (N = 502) Breast Cancer Cases (%) (N = 56) χ2pOdds Ratio (CI = 95%) Shift work at this moment Yes 444 (79.6) 411 (81.9) 33 (58.9) 16.315 <0.001 3.148 (1.765, 5.615) No 114 (20.4) 91 (18.1) 23 (41.1) Night work at this moment Yes 378 (67.7) 352 (70.1) 26 (46.4) 12.940 <0.001 2.708 (1.548, 4.735) No 180 (32.3) 150 (29.9) 30 (53.6) Working experience 16 years or less 280 (50.2) 275 (54.8) 5 (8.9) 42.369 <0.001 12.346 (4.854, 31.250) More than 16 years 278 (49.8) 227 (45.2) 51 (91.1) Total years performing more than 3 nights a month 10 years or less 317 (56.8) 302 (60.2) 15 (26.8) 22.870 <0.001 4.132 (2.227, 7.634) More than 10 years 241 (43.2) 200 (39.8) 41 (73.2) Total worked nights Less than 500 nights 265 (47.5) 254 (50.6) 11 (19.6) 19.358 <0.001 4.190 (2.118, 8.287) 500 nights or more 293 (52.5) 248 (49.4) 45 (80.4) Total sick leaves over lifespan (N = 550) 2 or less 342 (62.2) 329 (66.3) 13 (24.1) 36.977 <0.001 6.211 (3.236, 11.905) More than 2 208 (37.8) 167 (33.7) 41 (75.9) Sick leaves in the last year (N = 554) Without sick leave 385 (69.5) 368 (73.6) 17 (31.4) 40.782 <0.001 6.061 (3.300, 11.111) With sick leave 169 (30.5) 132 (26.4) 37 (68.5) 3.2. Self-Perception of Health Descriptive Analysis Regarding the self-perception of health, the overall health was rated 7.94 (SD = 1.26) among the sample nurses, being lower in breast cancer cases (6.45) and higher in the healthy cases (8.11). The lowest value of this category was found in the quality of sleep and rest, with a mean of 6.28 (SD = 1.96) and decreasing to 5.29 in breast cancer cases. The highest value was identified when considering whether shifts affect the health of people, with a mean of 9.08 (SD = 1.37). In relation to stress at work, the mean value was 7.57 (SD = 1.86). The stress perception was higher among breast cancer cases (8.23) as compared to healthy cases (7.49). Finally, the satisfaction with the working conditions was valued with a mean of 7.28 (SD = 1.87). Among all the health self-perception variables, statistical differences were found in terms of overall health (p< 0.001), sleep and rest quality (p< 0.001) , and stress at work (p= 0.002) (Table 4). Table 4. Sample profile according to the health self-perception variables. From 1 to 10 . . . M (SD) (N= 558) Breast Cancer Cases (N= 56) Non Cases (N= 502) Mann Whitney-U p How do you value your overall health? 7.94 (1.26) 6.45 (1.61) 8.11 (1.09) 5920.500 <0.001 How do you value your sleepingquality? 6.28 (1.96) 5.29 (2.06) 6.39 (1.91) 9741.500 <0.001 How do you value the effect shift work has on your health? 9.08 (1.37) 9.16 (1.60) 9.07 (1.35) 15,223.500 0.262 How do you value your level of work stress? 7.57 (1.86) 8.23 (1.67) 7.49 (1.87) 17,571.000 0.002 How do you value your satisfaction with your current job? 7.28 (1.87) 7.02 (2.09) 7.31 (1.85) 12,903.500 0.305 Healthcare 2021,9, 649 9 of 18 3.3. Breast Cancer Prediction The following two regression methods were performed in this study to identify those variables that play a relevant role in breast cancer risk. 3.3.1. Considering Labour Variables and Sleep Medication Binary logistic regression analysis predicts breast cancer among nurses through the following variables: total years performing more than 3 nights per month, sleep medication, sick leaves, years worked, and actual exposition to night work. This model was validated with the Hosmer-Lemeshov test (p= 0.811), correctly classifying 91.3% of cases. In addition, all variables included in this model had significant values lower than 0.05 and OR values greater than the unit (Table 5). Table 5. Logistic regression analysis for breast cancer. Coefficient OR CI = 95% for OR Inferior Superior Number of working years 12.167 ** 8.733 2.811 27.134 Medication to sleep 1.765 ** 5.841 2.848 11.978 Night work at this moment 1.701 ** 5.479 2.520 11.915 Sick leave last year 1.684 ** 5.387 2.527 11.484 Total years performing more than 3 nights per month 20.830 * 2.294 1.008 5.220 Constant −1.814 ** 0.163 Sensitivity/Specificity 52.8%/95.9% Correctly classified percentage 91.3% R2 Cox and Snell/R2 Nagelkerke 0.228/0.461 Hosmer-Lemeshov Test 0.811 Omnibus test < 0.001 OR: Odds ratio. * p< 0.05; ** p< 0.01. 1: (more than 16 years); 2: (10 or more years). 3.3.2. Self-Perception of Health Variable With regard to the self-perception of health among nurses, the CART showed the 558 cases at a root node, of which 10.03% (0.10; 56 subjects) have breast cancer. A second node differed according to the valuation of the overall health (higher or equal to 5.5; yes or no), resulting in a breast cancer percentage of 68% (0.68) in the 25 subjects (4% of the total cases) who valued their overall health below 5.5, and 7.3% of breast cancer cases in the subsequent 533 subjects (96% of the sample) whose scores were equal to or higher than 5.5. For the 4% of cases with worse health perception (overall health below 5.5), an internal node differed according to the satisfaction with the current working conditions. The percentage of breast cancer cases reached 82.4% (0.82) when the level of job satisfaction was higher or equal than 5.5, and otherwise the percentage of cases was 38% (0.38). Returning to those cases with overall health higher or equal to 5.5 (96%; 533 subjects), the following internal node differed again according to the valuation of the overall health (higher or equal to 7; yes or no). 502 nurses (90%) whose self-perception of health was ≥ 7 showed a 5.8% (0.06) of breast cancer cases. Otherwise, the 6% (33) of nurses whose self-perception of health was between 5.5 and 7 points comprised 29.4% (0.29) of cancer cases. Finally, an internal node for sleep quality (<6; yes or no) is shown for the cases with overall health between 5.5 and 7. 62% (0.62) of breast cancer cases occurred in 1% (6) of nurses who perceived their sleep quality under 6. On the other hand, 19% of breast cancer cases occurred in the 5% (28) of nurses who perceived their sleep quality over 6 (Figure 2). Healthcare 2021,9, 649 16 of 18 43. Stevens, R.G.; Hansen, J.; Costa, G.; Haus, E.; Kauppinen, T.; Aronson, K.J.; Castaño-Vinyals, G.; Davis, S.; Frings-Dresen, M.H.W.; Fritschi, L.; et al. Considerations of circadian impact for defining ’shift work’ in cancer studies: IARC Working Group Report. Occup. Environ. Med. 2010,68, 154–162. [CrossRef] 44. Directiva 2003/88/CE del Parlamento Europeo y del Consejo, de 4 de Noviembre de 2003, Relativa a Determinados Aspectos de la Ordenación del Tiempo de Trabajo. Available online: https://www.boe.es/doue/2003/299/L00009-00019.pdf (accessed on 31 March 2021). 45. Nebot, M.; Manzanares, S.; López, M.J.; Ariza, C.; Galán, I.; Moncada, A.; Montes, A.; Pérez-Ríos, M.; Schiaffino, A.; Fernández, E. Estimación de la exposición al humo ambiental de tabaco: Revisión de cuestionarios utilizados en España. Gac. Sanit. 2011 ,25, 322–328. [CrossRef] [PubMed] 46. Fagundo-Rivera, J.; Gómez-Salgado, J.; García-Iglesias, J.; Gómez-Salgado, C.; Camacho-Martín, S.; Ruiz-Frutos, C. Relationship between Night Shifts and Risk of Breast Cancer among Nurses: A Systematic Review. Medicina 2020 ,56, 680. [CrossRef] [PubMed] 47. Fox, J. Using the R Commander: A Point-and-Click Interface for R; CRC Press: Boca Raton, FL, USA, 2017. 48. Breiman, L.; Friedman, J.H.; Olshen, R.A.; Stone, C.J. Classification and Regression Trees; Routledge: Boca Raton, FL, USA, 2017. 49. Krzywinski, M.; Altman, N. Classification and regression trees. Nat. Methods 2017,14, 757–758. [CrossRef] 50. Szkiela, M.; Kusideł, E.; Makowiec-D ˛abrowska, T.; Kaleta, D. Night Shift Work—A Risk Factor for Breast Cancer. Int. J. Environ. Res. Public Health 2020,17, 659. [CrossRef] 51. Bustamante-Montes, L.P.; Flores-Meza, B.; Hernández-Valero, M.A.; Cárdenas-López, A.; Dolores-Velázquez, R.; Borja- Bustamante, P.; Borja-Aburto, V.H. Night Shift Work and Risk of Breast Cancer in Women. Arch. Med. Res. 2019 ,50, 393–399. [CrossRef] 52. López-Abente, G.; Aragonés, N.; Pérez-Gómez, B.; Pollán, M.; García-Pérez, J.; Ramis, R.; Fernández-Navarro, P. Time trends in municipal distribution patterns of cancer mortality in Spain. BMC Cancer 2014,14, 1–15. [CrossRef] 53. Schernhammer, E.S.; Hankinson, S.E.; Rosner, B.; Kroenke, C.H.; Willett, W.C.; Colditz, G.A.; Kawachi, I. Job Stress and Breast Cancer Risk: The Nurses’ Health Study. Am. J. Epidemiol. 2004,160, 1079–1086. [CrossRef] [PubMed] 54. Kossek, E.E.; Lee, K.-H. Work-Family Conflict and Work-Life Conflict. Oxf. Res. Encycl. Bus. Manag. 2017, 1–23. [CrossRef] 55. Labrague, L.; Ballad, C.; Fronda, D. Predictors and outcomes of work–family conflict among nurses. Int. Nurs. Rev. 2020 . [CrossRef] 56. Masuda, A.D.; Sortheix, F.; Beham, B.; Naidoo, L.J. Cultural value orientations and work–family conflict: The mediating role of work and family demands. J. Vocat. Behav. 2019,112, 294–310. [CrossRef] 57. Arlinghaus, A.; Bohle, P.; Iskra-Golec, I.; Jansen, N.; Jay, S.; Rotenberg, L. Working Time Society consensus statements: Evidencebased effects of shift work and non-standard working hours on workers, family and community. Ind. Health 2019 ,57, 184–200. [CrossRef] 58. Pinto, K.A.; Menezes, G.M.D.S.; Griep, R.H.; Lima, K.T.R.D.S.; Almeida, M.D.C.C.D.; Aquino, E.M.L. Work-family conflict and time use: Psychometric assessment of an instrument in ELSA-Brazil. Rev. Saúde Públ. 2016,50, 39. [CrossRef] 59. Svedberg, P.; Mather, L.; Bergström, G.; Lindfors, P.; Blom, V. Time pressure and sleep problems due to thoughts about work as risk factors for future sickness absence. Int. Arch. Occup. Environ. Health 2018,91, 1051–1059. [CrossRef] 60. Savard, J.; Morin, C.M. Insomnia in the Context of Cancer: A Review of a Neglected Problem. J. Clin. Oncol. 2001 ,19, 895–908. [CrossRef] [PubMed] 61. Costa, G.; Haus, E.; Stevens, R. Shift work and cancer—Considerations on rationale, mechanisms, and epidemiology. Scand. J. Work. Environ. Health 2010,36, 163–179. [CrossRef] 62. Khan, W.A.A.; Jackson, M.L.; Kennedy, G.A.; Conduit, R. A field investigation of the relationship between rotating shifts, sleep, mental health and physical activity of Australian paramedics. Sci. Rep. 2021,11, 1–11. [CrossRef] 63. Pietroiusti, A.; Neri, A.; Somma, G.; Coppeta, L.; Iavicoli, I.; Bergamaschi, A.; Magrini, A. Incidence of metabolic syndrome among night-shift healthcare workers. Occup. Environ. Med. 2009,67, 54–57. [CrossRef] 64. Eldevik, M.F.; Flo, E.; Moen, B.E.; Pallesen, S.; Bjorvatn, B. Insomnia, Excessive Sleepiness, Excessive Fatigue, Anxiety, Depression and Shift Work Disorder in Nurses Having Less than 11 Hours In-Between Shifts. PLoS ONE 2013,8, e70882. [CrossRef] 65. Costa, G.; Anelli, M.M.; Castellini, G.; Fustinoni, S.; Neri, L. Stress and sleep in nurses employed in “3 × 8” and “2 × 12” fast rotating shift schedules. Chrono Int. 2014,31, 1169–1178. [CrossRef] [PubMed] 66. Tahghighi, M.; Brown, J.A.; Breen, L.J.; Kane, R.; Hegney, D.; Rees, C.S. A comparison of nurse shift workers’ and non-shift workers’ psychological functioning and resilience. J. Adv. Nurs. 2019,75, 2570–2578. [CrossRef] [PubMed] 67. Ríos-Risquez, M.I.; Godoy-Fernández, C. Association between occupational satisfaction and perceived general health in emergency nurses. Enferm. Clin. 2008,18, 134–141. [CrossRef] 68. McHugh, M.D.; Kutney-Lee, A.; Cimiotti, J.P.; Sloane, D.M.; Aiken, L.H. Nurses’ Widespread Job Dissatisfaction, Burnout, and Frustration with Health Benefits Signal Problems for Patient Care. Health Aff. 2011,30, 202–210. [CrossRef] [PubMed] 69. Ferri, P.; Guadi, M.; Marcheselli, L.; Balduzzi, S.; Magnani, D.; Di Lorenzo, R. The impact of shift work on the psychological and physical health of nurses in a general hospital: A comparison between rotating night shifts and day shifts. Health Policy 2016 ,9, 203–211. [CrossRef] [PubMed] 70. Van Bogaert, P.; Kowalski, C.; Weeks, S.M.; Van Heusden, D.; Clarke, S.P. The relationship between nurse practice environment, nurse work characteristics, burnout and job outcome and quality of nursing care: A cross-sectional survey. Int. J. Nurs. Stud. 2013,50, 1667–1677. [CrossRef] [PubMed] Healthcare 2021,9, 649 17 of 18 71. Stimpfel, A.W.; Sloane, D.M.; Aiken, L.H. The Longer the Shifts for Hospital Nurses, the Higher the Levels of Burnout and Patient Dissatisfaction. Health Aff. 2012,31, 2501–2509. [CrossRef] [PubMed] 72. Stimpfel, A.W.; Lake, E.T.; Barton, S.; Gorman, K.C.; Aiken, L.H. How Differing Shift Lengths Relate to Quality Outcomes in Pediatrics. JONA 2013,43, 95–100. [CrossRef] 73. Ramoo, V.; Abdullah, K.L.; Piaw, C.Y. The relationship between job satisfaction and intention to leave current employment among registered nurses in a teaching hospital. J. Clin. Nurs. 2013,22, 3141–3152. [CrossRef] [PubMed] 74. Aiken, L.H.; Sermeus, W.; Heede, K.V.D.; Sloane, D.M.; Busse, R.; McKee, M.; Bruyneel, L.; Rafferty, A.M.; Griffiths, P.; Moreno- Casbas, M.T.; et al. Patient safety, satisfaction, and quality of hospital care: Cross sectional surveys of nurses and patients in 12 countries in Europe and the United States. BMJ 2012,344, e1717. [CrossRef] [PubMed] 75. Aiken, L.H.; Sloane, D.M.; Bruyneel, L.; Heede, K.V.D.; Sermeus, W. Nurses’ reports of working conditions and hospital quality of care in 12 countries in Europe. Int. J. Nurs. Stud. 2013,50, 143–153. [CrossRef] [PubMed] 76. Chan, D.S.M.; Abar, L.; Cariolou, M.; Nanu, N.; Greenwood, D.C.; Bandera, E.V.; McTiernan, A.; Norat, T. World Cancer Research Fund International: Continuous Update Project—Systematic literature review and meta-analysis of observational cohort studies on physical activity, sedentary behavior, adiposity, and weight change and breast cancer risk. Cancer Causes Control 2019 ,30, 1183–1200. [CrossRef] 77. Brum, M.C.B.; Filho, F.F.D.; Schnorr, C.C.; Bottega, G.B.; Rodrigues, T.C. Shift work and its association with metabolic disorders. Diabetol. Metab. Syndr. 2015,7, 1–7. [CrossRef] [PubMed] 78. Peplonska, B.; Bukowska, A.; Sobala, W. Association of Rotating Night Shift Work with BMI and Abdominal Obesity among Nurses and Midwives. PLoS ONE 2015,10, e0133761. [CrossRef] 79. Di Sibio, A.; Abriata, G.; Buffa, R.; Viniegra, M.; Forman, D.; Sierra, M.S. Etiology of breast cancer (C50) in Central and South America. In Cancer in Central and South America; International Agency for Research on Cancer: Lyon, France, 2016. 80. Dossus, L.; Boutron-Ruault, M.-C.; Kaaks, R.; Gram, I.T.; Vilier, A.; Fervers, B.; Manjer, J.; Tjonneland, A.; Olsen, A.; Overvad, K.; et al. Active and passive cigarette smoking and breast cancer risk: Results from the EPIC cohort. Int. J. Cancer 2013 ,134, 1871–1888. [CrossRef] 81. Luo, J.; Margolis, K.L.; Wactawski-Wende, J.; Horn, K.; Messina, C.; Stefanick, M.L.; Tindle, H.; Tong, E.; Rohan, T. Association of active and passive smoking with risk of breast cancer among postmenopausal women: A prospective cohort study. BMJ 2011 , 342, d1016. [CrossRef] [PubMed] 82. López-Guillén, A.G.; Pardo, J.M.V. Return to Work after Breast Cancer. Med. Segur. Trap. 2017,63, 51–67. 83. Guinan, E.M.; Connolly, E.M.; Kennedy, M.J.; Hussey, J. The presentation of metabolic dysfunction and the relationship with energy output in breast cancer survivors: A cross-sectional study. Nutr. J. 2013,12, 99. [CrossRef] 84. World Cancer Research Fund/American Institute for Cancer Research. Diet, Nutrition, Physical Activity and Cancer: A Global Perspective; World Cancer Research Fund International: London, UK, 2018. 85. Bruno, E.; Gargano, G.; Villarini, A.; Traina, A.; Johansson, H.; Mano, M.P.; De Magistris, M.S.; Simeoni, M.; Consolaro, E.; Mercandino, A.; et al. Adherence to WCRF/AICR cancer prevention recommendations and metabolic syndrome in breast cancer patients. Int. J. Cancer 2015,138, 237–244. [CrossRef] 86. Toribio, M.J.; Lope, V.; Castelló, A.; Salas, D.; Vidal, C.; Ascunce, N.; Santamariña, C.; Moreo, P.; Pedraz-Pingarrón, C.; Sánchez- Contador, C.; et al. Prevalence of healthy lifestyles against cancer in Spanish women. Sci. Rep. 2019,9, 10638. [CrossRef] 87. Barrios-Rodríguez, R.; Toledo, E.; Martinez-Gonzalez, M.A.; Aguilera-Buenosvinos, I.; Romanos-Nanclares, A.; Jiménez-Moleón, J.J. Adherence to the 2018 World Cancer Research Fund/American Institute for Cancer Research Recommendations and Breast Cancer in the SUN Project. Nutrients 2020,12, 2076. [CrossRef] [PubMed] 88. Albala, L.; Bober, T.; Hale, G.; Warfield, B.; Collins, M.L.; Merritt, Z.; Steimetz, E.; Nadler, S.; Lev, Y.; Hanifin, J. Effect on nurse and patient experience: Overnight use of blue-depleted illumination. BMJ Open Qual. 2019,8, e000692. [CrossRef] [PubMed] 89. Borg, G. Borg’s Perceived Exertion and Pan Scales; Human Kinetics: Champaign, IL, USA, 1998. 90. Sala, E.; Bonfiglioli, R.; Fostinellil, J.; Tomasi, C.; Graziosi, F.; Violante, F.S.; Apostoli, P. Risk assessment comparison of biomechanical overloading of the musculoskeletal system: 10 years’ applied experience. Ital. Med. Lav. Ergon. 2014 ,36, 260–266. 91. Sala, E.; Lopomo, N.F.; Tomasi, C.; Romagnoli, F.; Morotti, A.; Apostoli, P.; De Palma, G. Importance of Work-Related Psychosocial Factors in Exertion Perception Using the Borg Scale Among Workers Subjected to Heavy Physical Work. Front. Public Health 2021 , 9. [CrossRef] 92. Ware, J.E.; Kosinski, M.; Keller, S.D. A 12-Item Short-Form Health Survey. Med. Care 1996,34, 220–233. [CrossRef] 93. Cella, D.F.; Tulsky, D.S.; Gray, G.; Sarafian, B.; Linn, E.; Bonomi, A.; Silberman, M.; Yellen, S.B.; Winicour, P.; Brannon, J. The Functional Assessment of Cancer Therapy scale: Development and validation of the general measure. J. Clin. Oncol. 1993 ,11, 570–579. [CrossRef] [PubMed] 94. Sprangers, M.A.G.; Cull, A.; Bjordal, K.; Groenvold, M.; Aaronson, N.K. The European Organization for Research and treatment of cancer approach to quality of life assessment: Guidelines for developing questionnaire modules. Qual. Life Res. 1993 ,2, 287–295. [CrossRef] 95. Sprangers, M.; Groenvold, M.; I Arraras, J.; Franklin, J.; Velde, A.T.; Muller, M.; Franzini, L.; Williams, A.; De Haes, H.C.; Hopwood, P.; et al. The European Organization for Research and Treatment of Cancer breast cancer-specific quality-of-life questionnaire module: First results from a three-country field study. J. Clin. Oncol. 1996,14, 2756–2768. [CrossRef] [PubMed] Healthcare 2021,9, 649 18 of 18 96. Buysse, D.J.; Reynolds, C.F.; Monk, T.H.; Berman, S.R.; Kupfer, D.J. The Pittsburgh sleep quality index: A new instrument for psychiatric practice and research. Psychiatry Res. 1989,28, 193–213. [CrossRef] 97. Randler, C.; Díaz-Morales, J.F.; Rahafar, A.; Vollmer, C. Morningness–eveningness and amplitude—Development and validation of an improved composite scale to measure circadian preference and stability (MESSi). Chrono Int. 2016 ,33, 832–848. [CrossRef]