Environment International 122 (2019) 389–399 Contents lists available at ScienceDirect Environment International journal homepage: www.elsevier.com/locate/envint Alkylphenolic compounds and risk of breast and prostate cancer in the MCC- T Spain study Paula Peremiquel-Trillasa,b, Yolanda Benaventeb,c, Mayte Martín-Bustamanted, Delphine Casabonneb,c, Beatriz Pérez-Gómezc,e,f, Inés Gómez-Aceboc,g, Anna Oliete-Canelad, Marta Diéguez-Rodríguezd, Ignasi Tusquetsh,i, Pilar Amianoc,j, Lourdes Mengualk, Eva Ardanazc,l, Rocío Capelom, Antonio J. Molina de la Torren, Dolores Salas Trejoc,o,p, Guillermo Fernández-Tardónc,q, Virginia Lopec,e, José J. Jimenez-Moleonc,r,s, Rafael Marcos-Gragerac,t,u, Trinidad Dierssen-Sotosc,g, Mikel Azpirij, Montse Muñozv, Marcela Guevarac,l, Tania Fernández-Villan, Ana Molina-Barcelóo, Nuria Aragonésc,w, Marina Pollánc,e, Gemma Castaño-Vinyalsx,y,z, Juan Alguacilm, Manolis Kogevinasx,y,z,aa, Silvia de Sanjoséab, Laura Costasb, ⁎ a Preventive Medicine and Epidemiology Department, Hospital Universitari Vall d'Hebron, Barcelona, Spain b Unit of Molecular Epidemiology and Genetics in Infections and Cancer, IDIBELL, Catalan Institute of Oncology, L'Hospitalet de Llobregat, Barcelona, Spain c CIBER Epidemiología y Salud Pública – CIBERESP, Madrid, Spain d Institut Català de Seguretat i Salut Laboral, Departament de Treball, Afers Socials i Famílies, Generalitat de Catalunya, Barcelona, Spain e Cancer Epidemiology Unit, National Center for Epidemiology, Instituto de Salud Carlos III, Madrid, Spain f Cardiovascular & Metabolic Diseases Unit, National Centre for Epidemiology, Instituto de Salud Carlos III, Madrid, Spain g Universidad de Cantabria – IDIVAL, Santander, Spain h Universitat Autònoma de Barcelona, Parc de Salut Mar Hospital del Mar, Barcelona, Spain i Medical Oncology Department, Parc de Salut Mar Hospital del Mar, Barcelona, Spain j Public Health Division of Gipuzkoa, BioDonostia Research Institute, San Sebastian, Spain k Department and Laboratory of Urology, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Universitat de Barcelona, Barcelona, Spain l Instituto de Salud Pública de Navarra – IdiSNA, Pamplona, Spain m Centro de Investigación en Recursos Naturales, Salud y Medio Ambiente, Universidad de Huelva, Huelva, Spain n Grupo de Investigación en Interacciones Gen-Ambiente y Salud (GIIGAS), Instituto de Biomedicina (IBIOMED), Universidad de León, León, Spain o Cancer and Public Health Area, FISABIO – Public Health, Valencia, Spain p General Directorate Public Health, Valencian Community, Valencia, Spain q Instituto de Oncología de Asturias (IUOPA), Área de Medicina Preventiva, Universidad de Oviedo, Oviedo, Spain r Department of Preventive Medicine and Public Health, Faculty of Medicine, University of Granada, Granada, Spain s Instituto de Investigación Biosanitaria ibs.GRANADA, Hospitales Universitarios de Granada, University of Granada, Granada, Spain t Epidemiology Unit and Girona Cancer Registry (UERCG), Oncology Coordination Plan, Department of Health, Autonomous Government of Catalonia, Catalan Institute of Oncology (ICO), Girona, Spain u Girona Biomedical Research Institute (IDIBGI), University of Girona, Girona, Spain v Translational Genomics and Targeted Therapeutics in Solid Tumors, Hospital Clínic de Barcelona, Barcelona, Spain w Epidemiology Section, Public Health Division, Department of Health of Madrid, Madrid, Spain x ISGlobal, Barcelona, Spain y Hospital del Mar Medical Research Institute – IMIM, Barcelona, Spain z Universitat Pompeu Fabra (UPF), Barcelona, Spain aa National School of Public Health, Athens, Greece ab PATH, Seattle, USA A RT ICLE INFO ABSTRACT Handling Editor: Heather Stapleton Background: Alkylphenolic compounds are chemicals with endocrine disrupting properties that have been Keywords: widely used in industry with important changes in their usage over time. Few epidemiologic studies have Alkylphenols ⁎ Corresponding author at: Unit of Molecular Epidemiology and Genetics in Infections and Cancer, Cancer Epidemiology Research Programme, Institut Català d'Oncologia, Av. Gran Via de l'Hospitalet 199-203, 08908 L'Hospitalet de Llobregat, Spain. E-mail address: lcostas@iconcologia.net (L. Costas). https://doi.org/10.1016/j.envint.2018.12.007 Received 7 August 2018; Received in revised form 30 November 2018; Accepted 4 December 2018 Available online 13 December 2018 0160-4120/ © 2018 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/). P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 Alkylphenolic compounds evaluated the effect of alkylphenolic compounds on human health. Job-exposure matrix Objectives: We investigated whether occupational exposure to alkylphenolic compounds is associated with breast Breast cancer and prostate cancer. Prostate cancer Methods: We carried out a population-based case–control study including 1513 incident cases of breast cancer, Occupational exposure 1095 of prostate cancer, and 3055 controls, frequency matched by sex, age and region. Occupational exposure to alkylphenolic compounds was estimated using a recently developed job-exposure matrix, which considered different scenarios of exposure and different subtypes of alkylphenolic compounds. Results: History of occupational exposure to alkylphenolic compounds was modestly associated with breast cancer (OR = 1.23; 95% CI = 1.01–1.48). Within the different scenarios, the occupational use of domestic tensioactives was positively associated with breast cancer (OR = 1.28; 95% CI = 1.02–1.60), while occupational exposure in other scenarios showed mostly a suggestion of a similar positive associations. Exposure to non- ylphenol ethoxylates was positively associated with breast cancer (OR = 1.21; 95% CI = 1.00–1.47), while exposure to other compounds was uncommon. In general, we did not observe associations between alkylphenolic compounds and prostate cancer, except for a positive association among men occupationally exposed to cos- metic, hair and personal hygiene products. Conclusions: Our findings suggest a modest association between breast cancer risk and occupational exposure to alkylphenolic compounds, and no associations between these compounds and prostate cancer risk. These find- ings warrant further corroboration in other studies. 1. Introduction et al., 2007). The highest estimated exposure would occur with the application of specialty paints (2 mg/kg/day), use pesticides (0.35 μg/ Alkylphenolic compounds are organic chemicals generally produced kg/day), cosmetics (0.1 μg/kg/day) and exposure via food packaging for the manufacture of alkylphenolic ethoxylates (APE), which are materials (0.2 μg/kg/day) and in a textile factory (4.42 mg/kg/day) mainly used as non-ionic surfactants, but also in a wide range of ap- (Bontje et al., 2004). plications. Nonylphenol (NP), octylphenol (OP) and their ethoxylates There are few epidemiologic studies assessing the effect of alkyl- (NPE and OPE, respectively) are the most commonly used alkylphenolic phenolic compounds on human health. Regarding cancer, occupational compounds (Lassen et al., 2013). Exposure to these chemicals can occur exposure to alkylphenolic compounds has been associated with a higher occupationally during their production or with exposure to domestic risk of male and female breast cancer, and lymphoma (Aschengrau and industrial detergents, specialty paints, pesticides, cosmetics and et al., 1998a; Costas et al., 2015; Villeneuve et al., 2010b). Also, higher hair dyes, among others (Lassen et al., 2013), but also as a consequence NP and OP levels in urine among patients with uterine leiomyoma re- of non-occupational exposures, such as diet and water intake, use of lative to controls were observed in two case-control studies in China personal care and household cleaning products. Due to the toxicity and (Shen et al., 2013; Zhou et al., 2013). A study assessed the association bioaccumulation of alkylphenolic compounds in marine organisms of circulating serum levels of NP, OP and other xenoestrogens with (Hansen et al., 2002), in 2003 the European Union limited the com- mammographic breast density, which is a marker of breast cancer risk, mercialization of products containing NP and NPE in concentrations with negative results for alkylphenolic compounds (Sprague et al., over 0.1% (European Union, 2003). Alkylphenols and short-chain al- 2013). However, the conclusions of this study were hampered by the kylphenolic ethoxylates are considered endocrine disruptors, mainly high probability of external contamination, as well as to the suspected because of their effects mediated by estrogen receptors (Isidori et al., short half-lives of these compounds, which thus reflect only recent 2010; Olsen et al., 2005; Sun et al., 2008; White et al., 1994). NP has exposures (Calafat et al., 2013). Current methods to detect alkylphe- shown estrogenic effects in a number of in vitro (yeast, ZR-75 and MCF- nolic compounds in biospecimens commonly used in epidemiologic 7 human breast cancer cell lines) and in vivo assays (among rats and studies, such as blood serum or plasma, are not sensitive enough and mice), although the potency of this activity was moderately lower than they are susceptible to contamination. Although not exempt of limita- that of estradiol (in vitro assays showed an activity between 3 and 6 tions, assessment of occupational exposures may overcome these flaws, orders of lesser magnitude) (Bontje et al., 2004; Rotroff et al., 2014). and it can be used to assess lifetime rather than recent exposures of Sex hormones play critical roles in the development of breast and these compounds, which is relevant when evaluating diseases with long prostate cancers (Acevedo et al., 2005; Gray et al., 2017; Hess-Wilson latency periods, such as cancer (Martín-Bustamante et al., 2017). The and Knudsen, 2006; Risbridger et al., 2010; Villeneuve et al., 2010a) purpose of the present analysis is to explore whether occupational ex- and therefore we hypothesized that alkylphenolic compounds could posure to alkylphenolic compounds influences the risk of breast and have an influence on the development of these neoplasms. Over the last prostate cancer. decades, there is growing interest in this field and a consequent in- creased knowledge of the potential impact of these chemicals on the 2. Methods environment and on human health (Bergman et al., 2013; Casals-Casas and Desvergne, 2011; Damstra et al., 2002). 2.1. Study design Alkylphenolic compounds are ubiquitous, and they have been de- tected in rivers and bottled water (Amiridou and Voutsa, 2011; Bergé Cases were recruited within the MCC-Spain study, a Spanish popu- et al., 2012; Brix et al., 2010; Navarro et al., 2010), as well as in human lation-based multi-case–control study in which population controls fluids or tissues, such as urine (Calafat et al., 2005), blood were enrolled as well as five types of common cancers (breast, prostate, (Gyllenhammar et al., 2012), placenta (Huang et al., 2014), breast milk colorectal, stomach and chronic lymphocytic leukaemia). The recruit- (Ademollo et al., 2008; Chen et al., 2010; Sise and Uguz, 2017), adipose ment started in 2008 and finished in 2013. In this analysis, we included tissue (Ferrara et al., 2011; Geens et al., 2012; Lopez-Espinosa et al., incident patients diagnosed of breast (C50, D05.1, D05.7) or prostate 2009) and hair biospecimens (Nehring et al., 2017). There is con- cancer (C61, D07.5) admitted to 21 hospitals in 11 Spanish regions siderable uncertainty in the estimated human daily intake figures. The (Asturias, Barcelona, Cantabria, Gipuzkoa, Girona, Granada, Huelva, daily intake has been estimated to be between in the range of León, Madrid, Navarra and Valencia). Controls were frequency-mat- 7.5–30 μg/day (0.1–0.4 μg/kg bw/day) from food, and 2.3 × 104 mg/ ched to cases by sex, age, and region of recruitment and were randomly kg bw/day from water (Guart et al., 2014; Guenther et al., 2002; Lu selected from the rosters of General Practitioners at the Primary Health 390 P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 Centers participating in the study, which cover part of the population We applied a job-exposure matrix (JEM), originally developed to living in the corresponding area, and provides a representative sam- apply it to the MCC-Spain study, to assess exposure to alkylphenolic pling frame given the almost universal public coverage of the national compounds which takes into account the changes over time of the al- health system in Spain. We excluded those subjects with communica- kylphenolic compounds used in Spain (Martín-Bustamante et al., 2017). tion difficulties (mental or speaking problems) or physical ability im- The JEM considers the use of these compounds between 1931 and 2014 pairment. Response rates among the eligible subjects were 71% for in Spain, given that the participants in the MCC-Spain study reported breast cancer cases, 72% for prostate cancer cases and 53% for controls. occupations between this range of years. In summary, three hygienists Further details can be found elsewhere (Castaño-Vinyals et al., 2015). coded frequency (minority or majority of workers involved) and in- We excluded non-Caucasian participants (N = 118), subjects with tensity of exposure (including dispersive processes, with shaking or missing information on occupation (N = 62), and homemakers aerosol generation, or otherwise) to alkylphenolic compounds for all (N = 410). A total of 1513 incident breast cancer cases, 1095 incident the job titles by period of time. Intensity and frequency of exposure prostate cancer cases and 3055 controls (1575 females and 1480 males) were combined in a single score as follows: unlikely = 0, occa- were included in these analyses. sional + low intensity = 1, occasional + high intensity = 2, fre- quent + low intensity = 2 and frequent + high intensity = 3. Exposure 2.2. Data collection and exposure assessment assignment was blind to the case-control status and dichotomized into exposed and unexposed. This JEM grouped the different job titles po- Data were collected through direct interviews conducted by trained tentially exposed to alkylphenolic compounds into 6 different scenarios personnel. The questionnaire included basic epidemiologic information of exposure: a) manufacture and use of plastic and rubber products, b) such as age, educational level, family history of cancer, body mass manufacture and use of paints and lubricants, c) use of industrial ten- index (BMI) one year before recruitment, tobacco and alcohol con- sioactives, d) use of domestic tensioactives, e) use of cosmetic and hair sumption one year before recruitment (Castaño-Vinyals et al., 2015; products and personal hygiene products and f) use of pesticides. This Spanish questionnaire available at www.mccspain.org). Lifetime occu- JEM also assessed common types of alkylphenolic compounds: NPE, pational history was assessed for all jobs held for 1 year or more. For OP/NP, APE and others. APE category includes exposure to a mixture of each job reported, information was collected on job title, main activity ethoxylated compounds, and the category ‘others’ includes job titles or task performed, age at start and end, and shift (day, night or ro- which involved a mixture of alkylphenols and alkylphenolic ethox- tating), among others. Each occupation was independently coded by ylates. Occupational exposure to solvents and to pesticides was assigned two industrial hygienists following the national codes of occupation using a JEM designed for Spanish working conditions (MatEmESp) CNO-94, the Spanish adaptation to the International Standard Classi- (García et al., 2013). fication of Occupations (ISCO-88). When discrepancies occurred be- tween the hygienists (20% of job titles), agreement was reached by consensus. Fig. 1. Forest plot of associations on alkylphenolic compounds by scenario and type of compound. *Ajusted for age, region, education level, BMI, smoking, alcohol consumption, occupational shift, exposure to pesticides, exposure to solvents, hormonal contra- ception, postmenopausal hormone therapy, menopausal status and parity. **Ajusted for age, region, education level, BMI, smoking, alcohol consumption, occupational shift, exposure to pesticides and exposure to solvents. 391 P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 2.3. Statistical analyses family history of breast cancer, and have menarche at a younger age. Male cases were more likely to have a lower educational level, to have The distribution of potential risk factors between cases and controls been occupational exposed to pesticides and to report a family history was compared using the Pearson's chi-squared test. Duration, age at of prostate cancer, than male controls (Table 1). first exposure, time since first exposure and time since last exposure to Female controls exposed to alkylphenols were more likely to be alkylphenolic compounds were calculated based on the years at start older than the unexposed, to live in the Barcelona region, to have a and stop reported for each job and the date of interview. These vari- lower educational level, to have a higher BMI, to be non-smoker and ables were categorized in three groups using tertiles based on the dis- never drinkers, to have been occupationally exposed to solvents or tribution among exposed controls. Multivariate unconditional logistic pesticides, to be parous, to have never used hormonal contraceptives, to regression models were used to estimate odds ratios (OR) and 95% be younger at first delivery and to be postmenopausal at the moment of confidence intervals (95% CI) for the association between occupational the interview. Compared to the unexposed, exposed male controls were exposures and cancer. The variables considered for inclusion in the more likely to be younger, to have a higher educational level, to live in multivariable models are shown in the Directed Acyclic Graphs the Barcelona region, to report occupational exposure to solvents and/ (Supplemental Fig. 1). Basic adjusted models included age at interview or to pesticides and to have worked in permanent night shifts (Table 2). (< 60, 60–69, ≥70), region of recruitment and educational level (primary or less, secondary, higher). Final models, in addition to the 3.2. Associations between breast and prostate cancer and alkylphenolic basic adjustment, included BMI (< 18.5, 18.5–24.9, 25–29.9 or ≥30), compounds smoking status (never, current, former), alcohol consumption (never, current, past), occupational shift (day, permanent night, rotating night, Occupational exposure to alkylphenolic compounds was observed other rotating shifts), ever previous occupational exposure to solvents among 412 breast cancer cases (27.2%) and 342 female controls and/or to pesticides, and family history of cancer (breast or prostate, (21.7%). Patterns of estimates were in general similar between basic accordingly). For breast cancer, multivariate models were further ad- and fully adjusted models; ORs for fully-adjusted models are reported in justed for hormonal contraceptives use (ever, never), parity (nulli- the text unless otherwise specified. Subjects occupationally exposed to parous, 1 child, 2 children, ≥3 children), menopause status (pre- alkylphenolic compounds had a modestly higher risk of breast cancer menopause, postmenopause) and postmenopausal hormone therapy use (OR = 1.23; 95% CI = 1.01–1.48), compared to the unexposed. (ever, never), given that these variables are clearly related with breast Associations were significant among those who ever worked frequently cancer risk (Risbridger et al., 2010). For all variables, missing data was exposed and with high intensity of exposure (OR = 1.25; 95% ≤10% of subjects, except for alcohol consumption which was 15%. CI = 1.01–1.55). A short duration (< 6 years) was positively associated Missing values were introduced in models as independent categories. with breast cancer (OR tertile 1 vs. never = 1.34; 95% CI = 1.01–1.77), To test for linear trend, ordinal variables were treated as continuous although no linear trend was observed with duration of exposure (p- using midpoints in the categories as category values. Multiplicative trend = 0.30). Those starting exposure at an older age (≥29) and those interactions were tested by means of likelihood ratio tests comparing who started < 30 years since the interview showed positive associa- models with and without interactions. Several sensitivity analyses were tions with breast cancer (OR tertile 3 vs. never = 1.35; 95% CI = 1.04–1.84 performed. Given that exposure to solvents has been previously asso- and OR tertile 3 vs. never = 1.50; 95% CI = 1.14–1.98, respectively). ciated to these cancers and solvents and alkylphenolic compounds may Estimates were higher among those whose last exposure occurred later be correlated, we excluded participants who reported occupational in the calendar (OR1984–2000 vs. never = 1.32; 95% CI = 0.99–1.76; OR exposure to solvent compounds to ensure that occupational exposure to ≥2001 = 1.30; 95% CI = 0.98–1.72; Table 3, p-trend = 0.035). Among these chemicals was not driving associations between exposure to al- those exposed < 30 years ago for the first time, associations were ob- kylphenols and cancer risk. Also, we classified those who reported oc- served regardless of the duration of exposure (OR duration < 6 = 1.53, casional and low occupational exposure to alkylphenolic compounds as 95% CI = 1.04–2.27, and OR duration≥6 = 1.43, 95% CI = 1.04–2.27; non-exposed, and thus we considered only frequent and high exposures, data not shown). Similarly, associations were stronger among those as a measure of high probability of exposure. We performed analyses on who were exposed for the first time during the last 30 years and their sub-phenotypes of breast and prostate cancer (hormone receptors status last use was on 2001 or onwards (OR = 1.56, 95% CI = 1.12–2.17, and Gleason score, respectively). Finally, we further explored adjusting compared with never exposed; data not shown). our models for other measures of socioeconomic status (SES) than A total of 331 prostate cancer cases (30.2%) and 377 male controls educational level, such as self-reported maternal and paternal SES. All (25.5%) were occupationally exposed to alkylphenolic compounds, but analyses were conducted using Stata version 14.0 and the forest plot no significant association was observed (OR = 1.15; 95% (Fig. 1) was graphed with R 3.4.1. CI = 0.96–1.39). Associations were not significant in relation to the frequency and intensity of the exposure (OR = 1.18; 95% 2.4. Ethics CI = 0.88–1.59 among those frequently exposed and with high in- tensity of exposure compared with never exposed). Duration of ex- The MCC-Spain Study followed the national and international di- posure, age at first exposure, time since first exposure were not clearly rectives on ethics and data protection (declaration of Helsinki and associated with prostate cancer. A shorter time since last occupational Spanish law on confidentiality of data, Organic Law on Data Protection exposure (< 13 years) to alkylphenolic compounds was associated with 15/1999-LOPD). All eligible subjects signed an informed consent form prostate cancer (OR tertile 1 vs. never = 1.36; 95% CI = 1.02–1.81), but no for participation. Study protocol was approved by the Ethical and linear trend was observed (p-trend = 0.709, Table 3). Similarly, asso- Research Committees of each participating center. ciations were significant for those whom last exposure occurred on 1997 or onwards (OR ≥1997 vs. never = 1.33; 95% CI = 1.01–1.77; p- 3. Results trend = 0.396). 3.1. Demographic features of participants 3.3. Associations by scenarios and type of compound Baseline characteristics of participants are described in Table 1. Occupational use of domestic tensioactives, the most common sce- Compared to female controls, female cases were more likely to be nario of alkylphenolic compounds use among women, was positively younger, to have a higher BMI, to be non-smokers and never drinkers, associated with breast cancer (OR = 1.28; 95% CI = 1.02–1.60). to have been occupationally exposed to solvents, as well as to have a Women exposed to NPE, the most frequently used compound, had 21% 392 P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 Table 1 Descriptive characteristics of controls and cases. Breast cancer Prostate cancer b Controls Cases p-Value Controls Cases p-Valueb No. (%)a No. (%)a No. (%)a No. (%)a Overallc 1575 (51.0) 1513 (49.0) 1480 (57.5) 1095 (42.5) Age < 0.001 0.003 < 60 844 (53.6) 974 (64.4) 273 (18.4) 187 (17.1) 60–69 375 (23.8) 344 (22.7) 636 (43.0) 544 (49.7) ≥70 356 (22.6) 195 (12.9) 571 (38.6) 364 (33.2) Region 0.016 0.004 Barcelona 342 (21.7) 272 (18.0) 591 (39.9) 399 (36.4) Madrid 325 (20.6) 301 (19.9) 332 (22.4) 308 (28.1) Others 908 (57.7) 940 (62.1) 557 (37.6) 388 (35.4) Educational level 0.085 < 0.001 Primary 686 (43.6) 672 (44.4) 767 (51.8) 695 (63.5) Secondary 516 (32.8) 531 (35.1) 401 (27.1) 234 (21.4) Higher 373 (23.7) 310 (20.5) 312 (21.1) 166 (15.2) BMId 0.029 0.575 Underweight < 18.5 34 (2.2) 25 (1.7) 6 (0.4) 2 (0.2) Normal 18.5–24.99 748 (47.5) 669 (44.2) 359 (24.3) 267 (24.4) Overweight 25–29.99 451 (28.6) 492 (32.5) 759 (51.3) 584 (53.3) Obesity > 30 230 (14.6) 258 (17.1) 322 (21.8) 224 (20.5) Diabetes 0.524 < 0.001 No 1447 (91.9) 1398 (92.4) 1158 (78.2) 921 (84.1) Yes 122 (7.7) 108 (7.1) 320 (21.6) 170 (15.5) Alcohol consumption 0.011 0.445 Never 371 (23.6) 321 (21.2) 75 (5.1) 50 (4.6) Ever - current 884 (56.1) 824 (54.5) 1077 (72.8) 793 (72.4) Ever - past 97 (6.2) 130 (8.6) 102 (6.9) 89 (8.1) Smoking 0.002 0.166 Never smoker 878 (55.7) 781 (51.6) 390 (26.4) 323 (29.5) Current smoker 328 (20.8) 298 (19.7) 296 (20.0) 199 (18.2) Former smoker 367 (23.3) 430 (28.4) 793 (53.6) 571 (52.1) Occupational exposure to solvents 0.011 0.358 Never 1419 (90.1) 1319 (87.2) 988 (66.8) 712 (65.0) Ever 156 (9.9) 194 (12.8) 492 (33.2) 383 (35.0) Occupational exposure to pesticides 0.055 < 0.001 Never 1426 (90.5) 1337 (88.4) 1171 (79.1) 791 (72.2) Ever 148 (9.4) 174 (11.5) 308 (20.8) 304 (27.8) Occupational shifts 0.440 0.171 Day shifts 1183 (75.1) 1175 (77.7) 933 (63.0) 704 (64.3) Ever permanent night shifts 84 (5.3) 96 (6.3) 161 (10.9) 150 (13.7) Ever rotating night shiftse 115 (7.3) 122 (8.1) 188 (12.7) 171 (15.6) Other rotating schedulese 128 (8.1) 108 (7.1) 95 (6.4) 69 (6.3) Cancer family historyf < 0.001 < 0.001 No 1239 (78.7) 988 (65.3) 1257 (84.9) 828 (75.6) Yes 274 (17.4) 481 (31.8) 112 (7.6) 226 (20.6) Age at menarche 0.041 ≤13 1002 (63.6) 1038 (68.6) Not applicable > 13 513 (32.6) 453 (29.9) Hormonal contraceptives 0.176 Never 744 (47.2) 750 (49.6) Not applicable Ever 829 (52.6) 758 (50.1) Parity (number of children) 0.113 Nulliparous 274 (17.4) 291 (19.2) Not applicable 1 239 (15.2) 283 (18.7) 2 571 (36.3) 590 (39.0) ≥3 350 (22.2) 313 (20.7) Age at first deliveryg 0.694 ≤26 584 (50.3) 586 (49.4) Not applicable > 26 573 (49.4) 594 (50.1) Menopause status 0.519 Postmenopause 1119 (71.0) 1075 (71.1) Not applicable Premenopause 433 (27.5) 438 (28.9) Postmenopausal hormone therapyh 0.426 Never 948 (84.7) 942 (87.6) Not applicable Ever 114 (10.2) 101 (9.4) No. = number, % = percentage. a Percentages do not sum to the total due to missing values. b Chi squared, calculated without missing values. c Row percentage, the rest of percentages in the table are column percentages. d Expressed as weight (kg)/height2 (m2). e Never permanent night shifts. f Family history of breast/prostate cancer. g Among parous women. h Among postmenopausal women. 393 P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 Table 2 Descriptive characteristics among controls by occupational exposure to alkylphenolic compounds. Breast cancer Prostate cancer b Never exposed Ever exposed p-Value Never exposed Ever exposed p-Valueb No. (%)a No. (%)a No. (%)a No. (%)a Overallc 1233 (78.3) 342 (21.7) 1103 (74.5) 377 (25.5) 0.008d Age 0.001 0.031 < 60 692 (56.1) 152 (44.4) 220 (19.9) 53 (14.1) 60–69 277 (22.5) 98 (28.7) 460 (41.7) 176 (46.7) ≥70 264 (21.4) 92 (26.9) 423 (38.3) 148 (39.3) Region 0.069 0.001 Barcelona 254 (20.6) 88 (25.7) 409 (37.1) 182 (48.3) Madrid 265 (21.5) 60 (17.5) 263 (23.8) 69 (18.3) Others 714 (57.9) 194 (56.7) 431 (39.1) 126 (33.4) Educational level < 0.001 < 0.001 Primary 427 (34.6) 259 (75.7) 515 (46.7) 252 (66.8) Secondary 449 (36.4) 67 (19.6) 296 (26.8) 105 (27.9) Higher 357 (29.0) 16 (4.7) 292 (26.5) 20 (5.3) BMIe < 0.001 0.954 Underweight < 18.5 30 (2.4) 4 (1.2) 5 (0.5) 1 (0.3) Normal 18.5–24.99 624 (50.6) 124 (36.3) 268 (24.3) 91 (24.1) Overweight 25–29.99 348 (28.2) 103 (30.1) 566 (51.3) 193 (51.2) Obesity > 30 155 (12.6) 75 (21.9) 243 (22.0) 79 (21.0) Diabetes 0.133 0.817 No 1140 (92.5) 307 (89.8) 865 (78.4) 293 (77.7) Yes 89 (7.2) 33 (9.6) 237 (21.5) 83 (22.0) Alcohol consumption 0.019 0.145 Never 278 (22.5) 93 (27.2) 60 (5.4) 15 (4.0) Ever - current 719 (58.3) 165 (48.2) 808 (73.3) 269 (71.4) Ever - past 72 (5.8) 25 (7.3) 69 (6.3) 33 (8.8) Smoking 0.002 0.234 Never smoker 664 (53.9) 214 (62.6) 292 (26.5) 98 (26.0) Current smoker 256 (20.8) 72 (21.1) 231 (20.9) 65 (17.2) Former smoker 311 (25.2) 56 (16.4) 579 (52.5) 214 (56.8) Occupational exposure to solvents 0.002 < 0.001 Never 1126 (91.3) 293 (85.7) 858 (77.8) 130 (34.5) Ever 107 (8.7) 49 (14.3) 245 (22.2) 247 (65.5) Occupational exposure to pesticides < 0.001 < 0.001 Never 1161 (94.2) 265 (77.5) 897 (81.3) 274 (72.7) Ever 71 (5.8) 77 (22.5) 206 (18.7) 102 (27.1) Occupational shifts 0.082 < 0.001 Day shifts 933 (75.7) 250 (73.1) 728 (66.0) 205 (54.4) Ever permanent night shifts 57 (4.6) 27 (7.9) 95 (8.6) 66 (17.5) Ever rotating night shiftsf 93 (7.5) 22 (6.4) 127 (11.5) 61 (16.2) Other rotating schedulesf 104 (8.4) 24 (7.0) 74 (7.2) 21 (5.6) Cancer family historyg 0.912 0.392 No 973 (78.9) 266 (77.8) 930 (84.3) 327 (86.7) Yes 216 (17.5) 58 (17.0) 87 (7.9) 25 (6.6) Age at menarche 0.275 ≤13 794 (64.4) 208 (60.8) Not applicable > 13 394 (32.0) 119 (34.8) Hormonal contraceptives 0.022 Never 564 (45.7) 180 (52.6) Not applicable Ever 668 (54.2) 161 (47.1) Parity (number of children)h < 0.001 Nulliparous 238 (19.3) 36 (10.5) Not applicable 1 189 (15.3) 50 (14.6) 2 447 (36.3) 124 (36.3) ≥3 255 (20.7) 95 (27.8) Age at first deliveryg < 0.001 ≤26 418 (46.9) 166 (61.7) Not applicable > 26 470 (52.7) 103 (38.3) Menopause status < 0.001 Postmenopause 849 (68.9) 270 (78.9) Not applicable Premenopause 370 (30.0) 63 (18.4) Postmenopausal hormone therapyi 0.204 Never 714 (84.1) 234 (86.7) Not applicable Ever 92 (10.8) 22 (8.1) No. = number, % = percentage. a Percentages do not sum to the total due to missing values. b Chi squared, calculated without missing values. c Row percentage, the rest of percentages in the table are column percentages. d Chi-squared test of difference of exposure by sex. e Expressed as weight (kg)/height2 (m2). f Never permanent night shifts. g Family history of breast/prostate cancer. h Among parous women. i Among postmenopausal women. 394 P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 increased probability of cancer that the unexposed (OR = 1.21; 95% could suggest that the hormone-dependent responses of alkylphenolic CI = 1.00–1.47). Occupational exposure in other scenarios or to other compounds might produce the potential effects in a relatively short compounds showed mostly similar non-statistically significant asso- period of time. However, we observed increased risks among women ciations (Fig. 1). with triple negative cancer rather than with hormonal receptors posi- In general, we did not observe associations between the different tive cancer, which is not in line with our a priori hypothesis of alkyl- scenarios and specific types of alkylphenolic compounds and prostate phenols promoting mammary cell proliferation by binding to estrogen cancer (Fig. 1). An exception was the occupational use of cosmetic, hair receptors. Our findings are in accordance with the scarce published and personal hygiene products, which was associated with nearly three- literature on the subject (Aschengrau et al., 1998b; Gray et al., 2017; fold risk compared to the unexposed (OR = 2.97; 95% CI = 1.15–7.69), Villeneuve et al., 2010a). Occupational exposure to 4-octylphenol and a although sample size was small (8 controls and 15 cases). higher risk of breast cancer in females has been previously reported in one study performed in Massachusetts using data from the NIOSH Na- 3.4. Sensitivity analyses tional Occupational Exposure Survey. However, these findings need to be interpreted cautiously as the exposed number of subjects was very Among those unexposed to solvents, estimates were usually above small (6 cases and 5 controls) (Aschengrau et al., 1998b). A European the unity, and the ORs for the association between ever exposure to case-control study reported an increased risk between occupational alkylphenolic compounds and breast cancer (OR = 1.19; 95% exposure to alkylphenolic compounds and breast cancer among males, CI = 0.97–1.47) was similar to the overall ORs for whole population although with certain inconsistencies, suggesting that alkylphenolic (Table 3) although it was no longer statistically significant (Supple- compounds could also play a role in the development of this cancer in mental Table 1). When we considered only those exposures that were men (Villeneuve et al., 2010a). frequent and/or of high intensity, conclusions were mostly similar In accordance with previous literature, breast cancer was associated (Supplemental Table 2). Adjusting models for maternal or paternal SES in our sample with BMI, tobacco, alcohol consumption, age at me- yielded similar conclusions. No significant interactions were detected narche, and family history of breast cancer (Sun et al., 2017; Winters between BMI, menopausal status or education and ever exposure to et al., 2017). Other exposures related to estrogens trough lifetime alkylphenolic compounds (data not shown). Analyses excluding 138 in (parity, age at menopause, use of hormonal contraceptives or post- situ breast tumors yielded similar results (ORever vs. never = 1.25; 95% menopausal hormonal therapy) are also relevant predictors of breast CI = 1.03–1.51). Analyses by tumor receptors revealed significant as- cancer (Sun et al., 2017; Winters et al., 2017). Although these variables sociations with ever exposure for triple negative cancers (OR = 1.83; were not statistically significantly associated with breast cancer in our 95% CI = 1.17–2.87), but not for HER-2 positive or hormonal receptors bivariate analyses, they were introduced in the fully adjusted model, positive cancers (OR = 0.91; 95% CI = 0.62–1.35; and OR = 1.18; 95% similarly to educational level, in order to potentially control for con- CI = 0.95–1.47, respectively; Supplemental Table 3). Analyses by founding and selection bias for these factors. Given that participation prostate cancer aggressiveness revealed no significant associations rates were lower in controls than cases, selection bias could have oc- when comparing Gleason scores ≤7 and > 7 versus controls curred as controls have a higher SES than cases. We therefore controlled (OR = 1.17; 95% CI = 0.95–1.45; and OR = 0.75; 95% all analyses for educational level, and explored other different measures CI = 0.50–1.13, respectively; Supplemental Table 4). of SES, obtaining similar results, which suggests that our results are not strongly influenced by this potential bias. Sensitivity analyses re- 4. Discussion stricting analyses to the unexposed to solvents, as well as reclassifying those with occasional and low intensity as unexposed, yielded similar Using data from a large epidemiologic case-control study, we ob- conclusions. Similarly, associations were not restricted to a particular served a moderate positive association between occupational exposure scenario, such as the occupational use of pesticides. This suggests that to alkylphenolic compounds and breast cancer, and in general, no as- our results are not driven by confounding for pesticides or organic sociations with prostate cancer. In support of the hypothesis of an as- solvents, which were previously associated with breast cancer, although sociation between breast cancer and these chemicals, estimates were with inconsistences (Engel, 2005; Peplonska et al., 2010). None of the stronger among those who ever worked frequently exposed and with previous studies examining associations between alkylphenolic com- high intensity of exposure, compared with those exposed to alkylphe- pounds and cancer took into account relevant exposure changes over nolic compounds but never with a frequent and high intensity. Within time. Given the short half-live of these chemicals (Müller et al., 1998) the different scenarios of exposure, the occupational use of domestic and the long latency of diseases such as cancer, it is important to use a tensioactives was positively associated with breast cancer, while other method that allows estimating lifetime exposure and that considers the scenarios showed similar, but not statistically significant associations. relevant changes in their use over the years. The use of these com- The type of compound most frequently used among exposed women pounds has experienced relevant changes throughout last decades was nonylphenol ethoxylates, which were positively associated with worldwide. For instance, in the United Kingdom, the first use of these breast cancer. Overall, no associations were observed for prostate compounds was detected in 1944, while in Spain they were introduced cancer and occupational use of alkylphenolic compounds. However, in 1959 with the opening to international markets. In the 70s, the first exposures occurred on 1997 or onwards revealed significant associa- concerns about the toxicity of these compounds and their potential to tions, as well as occupational use of cosmetic, hair and personal hygiene bioaccumulate in marine organisms and in the environment appeared, products, although based on a small sample size. and voluntary agreements were undertaken to reduce or eliminate the Despite the growing interest in the potential impact of these che- use and manufacture of these compounds. In 2002–2003 the European micals on the environment and on health, data among humans are still Union established regulations to limit the use of NP and NPE (Bontje scarce, especially for specific EDC (Bergman et al., 2013; Casals-Casas et al., 2004; European Union, 2003; Martín-Bustamante et al., 2017). and Desvergne, 2011; Damstra et al., 2002). Alkylphenolic compounds This is relevant because 92% of the exposed women and 88% of ex- are considered EDC because of their estrogenic and weak anti-andro- posed males in our study had an exposure between 1965 and 2002, i.e. genic activities (Bontje et al., 2004). Recent in vitro and in vivo data before regulations took place to reduce these exposures. We found suggested that low-dose alkylphenol exposure promotes mammary al- significant positive associations between breast cancer and both NPE terations, and stated that exposure to alkylphenols could be considered exposure and use of domestic tensioactives. This is probably due to the as a tumor promoting environment, while the effect as tumor initiators correlation between these two variables, as NPE is the most frequent was not evident (Chamard-Jovenin et al., 2017). Although with certain compound within the domestic tensioactives scenario (Martín- inconsistencies, risks were increased with recent exposures, which Bustamante et al., 2017). 395 Table 3 Associations between breast and prostate cancer and occupational exposure to alkylphenolic compounds. P. Peremiquel-Trillas et al. Breast cancer Prostate cancer a b Controls (n = 1575) Cases (n = 1513) OR (95% CI) OR (95% CI) Controls (n = 1480) Cases (n = 1095) ORa (95% CI) ORc (95% CI) Never exposed to alkylphenolic compounds 1233 1101 Ref Ref 1103 764 Ref Ref Ever exposed to alkylphenolic compounds 342 412 1.28 (1.07–1.54) 1.23 (1.01–1.48) 377 331 1.09 (0.89–1.33) 1.15 (0.96–1.39) Intensity and frequency of exposure Occasional and/or low intensity (scores 1 and 2d) 93 115 1.21 (0.90–1.63) 1.17 (0.86–1.60) 261 209 1.08 (0.87–1.34) 1.04 (0.83–1.31) Frequent and high intensity (score 3) 249 297 1.31 (1.07–1.61) 1.25 (1.01–1.55) 116 122 1.32 (0.99–1.75) 1.18 (0.88–1.59) 0.007 0.035 0.061 0.284 Duration (years)e Tertile 1 109 148 1.45 (1.11–1.90) 1.34 (1.01–1.77) 113 104 1.24 (0.92–1.66) 1.07 (0.79–1.44) Tertile 2 115 135 1.17 (0.89–1.54) 1.18 (0.88–1.58) 130 95 0.98 (0.73–1.31) 0.91 (0.67–1.25) Tertile 3 118 129 1.23 (0.93–1.62) 1.15 (0.86–1.55) 134 132 1.25 (0.96–1.64) 1.28 (0.96–1.71) Ptrend 0.134 0.296 0.168 0.141 Age at first exposuref Tertile 1 111 150 1.28 (0.97–1.68) 1.16 (0.87–1.54) 102 84 1.02 (0.74–1.41) 0.94 (0.67–1.32) Tertile 2 113 119 1.17 (0.88–1.55) 1.14 (0.85–1.54) 143 135 1.27 (0.98–1.66) 1.18 (0.90–1.56) Tertile 3 118 143 1.40 (1.07–1.83) 1.35 (1.04–1.84) 131 112 1.13 (0.85–1.50) 1.10 (0.82–1.49) Ptrend 0.007 0.020 0.131 0.337 Time since first exposure (years)g Tertile 1 113 182 1.47 (1.13–1.91) 1.50 (1.14–1.98) 100 95 1.26 (0.93–1.72) 1.22 (0.88–1.68) Tertile 2 111 136 1.20 (0.91–1.59) 1.09 (0.81–1.47) 139 113 1.02 (0.78–1.35) 1.00 (0.75–1.35) 396 Tertile 3 118 94 1.13 (0.83–1.54) 1.05 (0.76–1.44) 138 123 1.21 (0.92–1.60) 1.08 (0.81–1.44) Ptrend 0.061 0.265 0.157 0.502 Time since last exposure (years)h Tertile 1 114 161 1.33 (1.01–1.73) 1.25 (0.95–1.66) 120 132 1.38 (1.05–1.82) 1.36 (1.02–1.81) Tertile 2 114 149 1.39 (1.07–1.83) 1.39 (1.05–1.85) 130 89 0.97 (0.72–1.31) 0.93 (0.68–1.28) Tertile 3 114 102 1.10 (0.82–1.48) 1.02 (0.75–1.39) 127 110 1.11 (0.84–1.48) 0.98 (0.73–1.32) Ptrend 0.219 0.493 0.583 0.709 Year of last exposurei Tertile 1 113 102 1.12 (0.83–1.51) 1.05 (0.77–1.42) 122 108 1.16 (0.87–1.54) 1.02 (0.76–1.38) Tertile 2 112 142 1.35 (1.03–1.78) 1.32 (0.99–1.76) 126 82 0.91 (0.67–1.24) 0.88 (0.64–1.22) Tertile 3 117 168 1.35 (1.04–1.76) 1.30 (0.98–1.72) 129 141 1.38 (1.05–1.80) 1.33 (1.01–1.77) Ptrend 0.007 0.035 0.125 0.396 OR = odds ratio, CI = confidence interval. Numbers do not add up to the totals due to missing values. a Adjusted for region, age and educational level. b Adjusted for age, region, educational level, BMI, smoking, alcohol consumption, occupational shift, exposure to pesticides, exposure to solvents, hormonal contraception, postmenopausal hormone therapy, menopausal status and parity. c Adjusted for age, region, educational level, BMI, smoking, alcohol consumption, occupational shift, exposure to pesticides and exposure to solvents. d Never score 3. e Tertiles for breast cancer: < 6, 6–14.99, ≥15, for prostate cancer: < 6, 6–24.99, ≥25. f Tertiles for breast cancer: < 19, 19–28.99, ≥29, for prostate cancer: < 18, 18–25.99, ≥26. g Tertiles for breast cancer: < 30, 30–43.99, ≥44, for prostate cancer: < 40, 40–48.99, ≥49. h Tertiles for breast cancer: < 9.5, 9.5–26.49, ≥26.5, for prostate cancer: < 13, 13–34.99, ≥35. i Tertiles for breast cancer: < 1984, 1984–2000, ≥2001, for prostate cancer: < 1976, 1976–1996, ≥1997. Environment International 122 (2019) 389–399 P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 In general, we found no associations between prostate cancer and relationship between occupational exposure to alkylphenolic com- exposure to alkylphenolic compounds. Previous studies suggested a pounds and breast and prostate cancer. We examined these associations possible relationship between exposure to endocrine disrupting com- using a method that considered relevant changes over time in the use of pounds with estrogenic activity, other than alkylphenols, and an in- these compounds, and we adjusted for multiple potential confounding creased prostate cancer risk (Hess-Wilson and Knudsen, 2006). How- variables. Our findings suggest a modest association between breast ever, no epidemiologic study has ever evaluated the relationship cancer risk and occupational exposure to alkylphenolic compounds, between alkylphenolic compounds and prostate cancer. Besides their while no association with prostate cancer. The elucidation of the role of estrogenic activity, alkylphenolic compounds may also have weak anti- alkylphenolic compounds on breast and prostate cancers should shed a androgenic effects, as they can act as partial agonists of androgenic light on the etiology of those tumors and help informing future public receptors thus inhibiting the effect of full agonists (Weiss et al., 2009). health decisions. Androgens are necessary for cancer prostate development, and an- drogen receptor signaling has been well characterized in the develop- Funding ment of prostate cancer metastasis (Zhou et al., 2014). Although they have been less studied in cancer prostate risk compared with breast This work was partially funded by the public grants from the cancer, there is evidence that high doses of estrogens also have direct Catalan Government (2014SGR756, 2017SGR1085, 2017SGR733, effects on the development of the prostate gland and also indirect ef- SLT006/17/76), and European Regional Development Fund-ERDF, by fects through the suppression of the hypothalamic pituitary gonadal the “Accion Transversal del Cancer”, approved on the Spanish Ministry axis (Jarred et al., 2000). Metabolism of these sex hormones is complex Council on the 11th October 2007, by the Instituto de Salud Carlos III- and correlated and conversion of androgens to estrogens is mediated by FEDER (PI08/1770, PI08/0533, PI08/1359, PS09/00773-Cantabria, an enzyme encoded by CYP19A1 (Risbridger et al., 2010; Rothenberger PS09/01286-León, PS09/01903-Valencia, PS09/02078-Huelva, PS09/ et al., 2018). Given the complexity of the hormonal metabolism, and 01662-Granada, PI11/01403, PI11/01889-FEDER, PI11/00226, PI11/ the estrogenic and anti-androgenic effects of alkylphenolic compounds, 01810, PI11/02213, PI12/00488, PI12/00265, PI12/01270, PI12/ further research is needed to disentangle whether and how these 00715, PI12/00150, PI14/01219, PI14/0613, PI15/00069, PI15/ compounds could influence prostate cancer risk. 00914, PI15/01032, PI11/01810, PI14/01219, PI11/02213, PIE16/ JEMs for occupational exposure assessment are useful and cost-ef- 00049, PI17/01179), by the Fundación Marqués de Valdecilla (API 10/ ficient tool in large scale studies (Kauppinen, 1996; Kim et al., 2011; Le 09), by the Red Temática de Investigación del Cáncer (RTICC) del ISCIII Moual et al., 2000; Martín-Bustamante et al., 2017), but they can lead (RD12/0036/0036), by the Junta de Castilla y León (LE22A10-2), by to substantial misclassification of exposures (Rothman et al., 2007). In the Consejería de Salud of the Junta de Andalucía (PI-0571-2009, PI- the present analysis, this misclassification would be non-differential as 0306-2011, salud201200057018tra), by the Conselleria de Sanitat of the exposure assessment was blind to the case-control status. In case of the Generalitat Valenciana (AP_061/10), by the Recercaixa (2010ACUP truly positive associations, it would result in the attenuation of the 00310), by the Regional Government of the Basque Country, by the estimates for binary and continuous exposures (Pearce et al., 2007; Consejería de Sanidad de la Región de Murcia, by the European Steenland et al., 2000). The JEM used in the present analyses was de- Commission grants FOOD-CT-2006-036224-HIWATE, by the Spanish veloped undertaking several strategies to minimize potential sources of Association Against Cancer (AECC) Scientific Foundation, by the misclassification. Among others, it considered relevant changes over Catalan Government-Agency for Management of University and time in use and manufacture of alkylphenolic compounds in the same Research Grants (AGAUR) grants 2014SGR647 and 2014SGR850. population that it is applied, which probably results in better estimates ISGlobal is a member of the CERCA Programme, Generalitat de of the exposure. However, we could not validate the JEM using biologic Catalunya. measurements given the short life of these compounds, and the changes in use over time. Evaluating levels among workers may help better Conflicts of interest understanding current exposures, but the scores in our JEM reflect lifetime exposures which have greatly varied over time. We had the The authors have no conflict of interests to declare. opportunity to adjust for many potential confounders, including well- established and other potential risk factors for breast and prostate Acknowledgements cancer. We did not observe clear evidence of confounding by any of them, although residual confounding cannot be completely discarded in We appreciate the collaboration of all those who participated pro- explaining some of our results. Non-occupational sources of exposure, viding questionnaire data. We also thank Laia Alemany, Claudia Robles, such as diet, use of personal care and household cleaning products, Emilia Molinero-Ruiz, Nereida Gras, Santos Hernández, and Ferran could be relevant sources of alkylphenolic compounds. The JEM ex- Calduch for their helpful contributions. We thank all MCC-Spain study posure scores reflect the probability that occupational exposure con- group members and collaborators. tributes significantly to an individual's body burden in comparison to other sources of exposure. Although we do not expect big differences in Appendix A. Supplementary data background levels, we were not able to assess non-occupational sources of these compounds and therefore certain misclassification of the ex- Supplementary data to this article can be found online at https:// posure cannot be ruled out. The relatively large sample size of the doi.org/10.1016/j.envint.2018.12.007. study, allowed us to perform several subgroup analysis. Nevertheless, some of the analyses by scenarios and compound subtypes were based References on small sample sizes. Some of the job titles involving alkylphenols actually involved a mixture of alkylphenols and alkylphenolic ethox- Acevedo, R., Parnell, P.G., Villanueva, H., Chapman, L.M., Gimenez, T., Gray, S.L., ylates, and therefore the JEM that we used did not assign a more spe- Baldwin, W.S., 2005. The contribution of hepatic steroid metabolism to serum es- tradiol and estriol concentrations in nonylphenol treated MMTVneu mice and its cific category of compound. Some of the reported associations could be potential effects on breast cancer incidence and latency. J. Appl. Toxicol. 25, spurious as multiple comparisons were performed in our analysis. 339–353. https://doi.org/10.1002/jat.1078. Ademollo, N., Ferrara, F., Delise, M., Fabietti, F., Funari, E., 2008. Nonylphenol and octylphenol in human breast milk. Environ. Int. 34, 984–987. https://doi.org/10. 5. Conclusion 1016/j.envint.2008.03.001. Amiridou, D., Voutsa, D., 2011. Alkylphenols and phthalates in bottled waters. J. Hazard. To our knowledge, this is the largest study examining the Mater. 185, 281–286. https://doi.org/10.1016/j.jhazmat.2010.09.031. 397 P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 Aschengrau, A., Coogan, P.F., Quinn, M., Cashins, L.J., 1998a. Occupational exposure to Nonylphenol and bisphenol A in Swedish food and exposure in Swedish nursing estrogenic chemicals and the occurrence of breast cancer: an exploratory analysis. women. Environ. Int. 43, 21–28. https://doi.org/10.1016/j.envint.2012.02.010. Am. J. Ind. Med. 34, 6–14. Hansen, Munn, De Bruijn, Pakalin, Luotarno, Berthault, Vegro, Heidorn, Pellegrini, Aschengrau, A., Coogan, P.F., Quinn, M., Cashins, L.J., 1998b. Occupational exposure to Vormann, Allanou, Scheer, 2002. European Union risk assessment report: 4-non- estrogenic chemicals and the occurrence of breast cancer: an exploratory analysis. ylphenol (branched) and nonylphenol. In: Inst. Health Consum. Prot. Eur. Chem. Am. J. Ind. Med. 34, 6–14. Bur. 10. Bergé, A., Cladière, M., Gasperi, J., Coursimault, A., Tassin, B., Moilleron, R., 2012. Meta- Hess-Wilson, J.K., Knudsen, K.E., 2006. Endocrine disrupting compounds and prostate analysis of environmental contamination by alkylphenols. Environ. Sci. Pollut. Res. cancer. Cancer Lett. 241, 1–12. https://doi.org/10.1016/j.canlet.2005.10.006. 19, 3798–3819. https://doi.org/10.1007/s11356-012-1094-7. Huang, Y.-F., Wang, P.-W., Huang, L.-W., Yang, W., Yu, C.-J., Yang, S.-H., Chiu, H.-H., Bergman, Å., Heindel, J.J., Kasten, T., Kidd, K.A., Jobling, S., Neira, M., Zoeller, R.T., Chen, M.-L., 2014. Nonylphenol in pregnant women and their matching fetuses: Becher, G., Bjerregaard, P., Bornman, R., Brandt, I., Kortenkamp, A., Muir, D., Drisse, placental transfer and potential risks of infants. Environ. Res. 134, 143–148. https:// M.-N.B., Ochieng, R., Skakkebaek, N.E., Byléhn, A.S., Iguchi, T., Toppari, J., doi.org/10.1016/j.envres.2014.07.004. Woodruff, T.J., 2013. The impact of endocrine disruption: a consensus statement on Isidori, M., Cangiano, M., Palermo, F.A., Parrella, A., 2010. E-screen and vitellogenin the state of the science. Environ. Health Perspect. 121, a104–a106. https://doi.org/ assay for the detection of the estrogenic activity of alkylphenols and trace elements. 10.1289/ehp.1205448. Comp. Biochem. Physiol. Toxicol. Pharmacol. 152, 51–56. https://doi.org/10.1016/j. Bontje, D., Hermens, J., Vermeire, T., Damstra, T., 2004. United Nations Environment cbpc.2010.02.011. Programme. International Labour Organisation. International Programme on Jarred, R.A., Cancilla, B., Prins, G.S., Thayer, K.A., Cunha, G.R., Risbridger, G.P., 2000. Chemical Safety. Integrated Risk Assessment: Nonylphenol Case Study. WHO/IPCS/ Evidence that estrogens directly alter androgen-regulated prostate development. IRA/12/04. Endocrinology 141, 3471–3477. https://doi.org/10.1210/endo.141.9.7648. Brix, R., Postigo, C., González, S., Villagrasa, M., Navarro, A., Kuster, M., de Alda, M.J.L., Kauppinen, T., 1996. Exposure assessment - a challenge for occupational epidemiology. Barceló, D., 2010. Analysis and occurrence of alkylphenolic compounds and estro- Scand. J. Work Environ. Health 22, 401–403. https://doi.org/10.5271/sjweh.160. gens in a European river basin and an evaluation of their importance as priority Kim, H.-M., Richardson, D., Loomis, D., Van Tongeren, M., Burstyn, I., 2011. Bias in the pollutants. Anal. Bioanal. Chem. 396, 1301–1309. https://doi.org/10.1007/s00216- estimation of exposure effects with individual- or group-based exposure assessment. 009-3358-8. J. Expo. Sci. Environ. Epidemiol. 21, 212–221. https://doi.org/10.1038/jes.2009.74. Calafat, A.M., Kuklenyik, Z., Reidy, J.A., Caudill, S.P., Ekong, J., Needham, L.L., 2005. Lassen, C., Jensen, A.A., Maag, J., Christensen, F., Kjolholt, J., Jeppesen, C.N., Mikkelsen, Urinary concentrations of bisphenol A and 4-nonylphenol in a human reference po- S.H., Innanen, S., 2013. Survey of Alkylphenols and Alkylphenol Ethoxylates. Part of pulation. Environ. Health Perspect. 113, 391–395. the LOUS-review. The Danish Environmental Protection Agency, Copenhagen, Calafat, A.M., Koch, H.M., Swan, S.H., Hauser, R., Goldman, L.R., Lanphear, B.P., Denmark. Longnecker, M.P., Rudel, R.A., Teitelbaum, S.L., Whyatt, R.M., Wolff, M.S., 2013. Le Moual, N., Bakke, P., Orlowski, E., Heederik, D., Kromhout, H., Kennedy, S.M., Misuse of blood serum to assess exposure to bisphenol A and phthalates. Breast Rijcken, B., Kauffmann, F., 2000. Performance of population specific job exposure Cancer Res. 15, 403. https://doi.org/10.1186/bcr3494. matrices (JEMs): European collaborative analyses on occupational risk factors for Casals-Casas, C., Desvergne, B., 2011. Endocrine disruptors: from endocrine to metabolic chronic obstructive pulmonary disease with job exposure matrices (ECOJEM). Occup. disruption. Annu. Rev. Physiol. 73, 135–162. https://doi.org/10.1146/annurev- Environ. Med. 57, 126–132. physiol-012110-142200. Lopez-Espinosa, M.J., Freire, C., Arrebola, J.P., Navea, N., Taoufiki, J., Fernandez, M.F., Castaño-Vinyals, G., Aragonés, N., Pérez-Gómez, B., Martín, V., Llorca, J., Moreno, V., Ballesteros, O., Prada, R., Olea, N., 2009. Nonylphenol and octylphenol in adipose Altzibar, J.M., Ardanaz, E., de Sanjosé, S., Jiménez-Moleón, J.J., Tardón, A., Alguacil, tissue of women in Southern Spain. Chemosphere 76, 847–852. https://doi.org/10. J., Peiró, R., Marcos-Gragera, R., Navarro, C., Pollán, M., Kogevinas, M., MCC-Spain 1016/j.chemosphere.2009.03.063. Study Group, 2015. Population-based multicase-control study in common tumors in Lu, Y.-Y., Chen, M.-L., Sung, F.-C., Wang, P.S.-G., Mao, I.-F., 2007. Daily intake of 4- Spain (MCC-Spain): rationale and study design. Gac. Sanit. 29, 308–315. https://doi. nonylphenol in Taiwanese. Environ. Int. 33, 903–910. https://doi.org/10.1016/j. org/10.1016/j.gaceta.2014.12.003. (SESPAS). envint.2007.04.008. Chamard-Jovenin, C., Thiebaut, C., Chesnel, A., Bresso, E., Morel, C., Smail-Tabbone, M., Martín-Bustamante, M., Oliete-Canela, A., Diéguez-Rodríguez, M., Benavente, Y., Devignes, M.-D., Boukhobza, T., Dumond, H., 2017. Low-dose alkylphenol exposure Casabonne, D., Alguacil, J., Kogevinas, M., de Sanjosé, S., Costas, L., 2017. Job-ex- promotes mammary epithelium alterations and transgenerational developmental posure matrix for the assessment of alkylphenolic compounds. Occup. Environ. Med. defects, but does not enhance tumorigenic behavior of breast cancer cells. Front. 74, 52–58. https://doi.org/10.1136/oemed-2016-103614. Endocrinol. 8, 272. https://doi.org/10.3389/fendo.2017.00272. Müller, S., Schmid, P., Schlatter, C., 1998. Pharmacokinetic behavior of 4-nonylphenol in Chen, G.-W., Ding, W.-H., Ku, H.-Y., Chao, H.-R., Chen, H.-Y., Huang, M.-C., Wang, S.-L., humans. Environ. Toxicol. Pharmacol. 5, 257–265. https://doi.org/10.1016/S1382- 2010. Alkylphenols in human milk and their relations to dietary habits in central 6689(98)00009-X. Taiwan. Food Chem. Toxicol. 48, 1939–1944. https://doi.org/10.1016/j.fct.2010.04. Navarro, A., Tauler, R., Lacorte, S., Barceló, D., 2010. Occurrence and transport of pes- 038. ticides and alkylphenols in water samples along the Ebro River Basin. J. Hydrol. 383, Costas, L., Infante-Rivard, C., Zock, J.-P., Van Tongeren, M., Boffetta, P., Cusson, A., 18–29. https://doi.org/10.1016/j.jhydrol.2009.06.039. Robles, C., Casabonne, D., Benavente, Y., Becker, N., Brennan, P., Foretova, L., Nehring, I., Staniszewska, M., Falkowska, L., 2017. Human hair, Baltic grey seal Maynadié, M., Staines, A., Nieters, A., Cocco, P., de Sanjosé, S., 2015. Occupational (Halichoerus grypus) fur and herring gull (Larus argentatus) feathers as accumulators exposure to endocrine disruptors and lymphoma risk in a multi-centric European of bisphenol A and alkylphenols. Arch. Environ. Contam. Toxicol. 72, 552–561. study. Br. J. Cancer (112 Suppl), 1251–1256. https://doi.org/10.1038/bjc.2015.83. https://doi.org/10.1007/s00244-017-0402-0. Damstra, T., Barlow, S., Bergman, A., Kavloc, R., Van Der Kraak, G., 2002. Global Olsen, C.M., Meussen-Elholm, E.T.M., Hongslo, J.K., Stenersen, J., Tollefsen, K.-E., 2005. Assessment of the State-of-the-Science of Endocrine Disruptors: An Assessment Estrogenic effects of environmental chemicals: an interspecies comparison. Comp. Prepared by an Expert Group on Behalf of the World Health Organization, the Biochem. Physiol. Toxicol. Pharmacol. 141, 267–274. https://doi.org/10.1016/j.cca. International Labour Organisation, and the United Nations Environment Programme. 2005.07.002. World Health Organization. Pearce, N., Checkoway, H., Kriebel, D., 2007. Bias in occupational epidemiology studies. Engel, L.S., 2005. Pesticide use and breast cancer risk among farmers' wives in the Occup. Environ. Med. 64, 562–568. https://doi.org/10.1136/oem.2006.026690. agricultural health study. Am. J. Epidemiol. 161, 121–135. https://doi.org/10.1093/ Peplonska, B., Stewart, P., Szeszenia-Dąbrowska, N., Lissowska, J., Brinton, L.A., aje/kwi022. Gromiec, J.P., Brzeznicki, S., Yang, R., Sherman, M., García-Closas, M., Blair, A., European Union, 2003. Directive 2003/53/EC of the European Parliament and of the 2010. Occupational exposure to organic solvents and breast cancer in women. Occup. Council of 18 June 2003 Amending for the 26th Time Council Directive 76/769/EEC Environ. Med. 67, 722–729. https://doi.org/10.1136/oem.2009.046557. Relating to Restrictions on the Marketing and Use of Certain Dangerous Substances Risbridger, G.P., Davis, I.D., Birrell, S.N., Tilley, W.D., 2010. Breast and prostate cancer: and Preparations (Nonylphenol, Nonylphenol Ethoxylate and Cement). more similar than different. Nat. Rev. Cancer 10, 205–212. https://doi.org/10.1038/ Ferrara, F., Ademollo, N., Orrù, M.A., Silvestroni, L., Funari, E., 2011. Alkylphenols in nrc2795. adipose tissues of Italian population. Chemosphere 82, 1044–1049. https://doi.org/ Rothenberger, N.J., Somasundaram, A., Stabile, L.P., 2018. The role of the estrogen 10.1016/j.chemosphere.2010.10.064. pathway in the tumor microenvironment. Int. J. Mol. Sci. 19, 611. https://doi.org/ García, A.M., González-Galarzo, M.C., Kauppinen, T., Delclos, G.L., Benavides, F.G., 2013. 10.3390/ijms19020611. A job-exposure matrix for research and surveillance of occupational health and safety Rothman, K., Greenland, S., Lash, T.L., 2007. Modern Epidemiology, 3rd edition. in Spanish workers: MatEmESp. Am. J. Ind. Med. 56, 1226–1238. https://doi.org/10. Lippincott Williams Wilkins, Phila. PA. 1002/ajim.22213. Rotroff, D.M., Martin, M.T., Dix, D.J., Filer, D.L., Houck, K.A., Knudsen, T.B., Sipes, N.S., Geens, T., Neels, H., Covaci, A., 2012. Distribution of bisphenol-A, triclosan and n-non- Reif, D.M., Xia, M., Huang, R., Judson, R.S., 2014. Predictive endocrine testing in the ylphenol in human adipose tissue, liver and brain. Chemosphere 87, 796–802. 21st century using in vitro assays of estrogen receptor signaling responses. Environ. https://doi.org/10.1016/j.chemosphere.2012.01.002. Sci. Technol. 48, 8706–8716. https://doi.org/10.1021/es502676e. Gray, J.M., Rasanayagam, S., Engel, C., Rizzo, J., 2017. State of the evidence 2017: an Shen, Y., Xu, Q., Ren, M., Feng, X., Cai, Y., Gao, Y., 2013. Measurement of phenolic update on the connection between breast cancer and the environment. Environ. environmental estrogens in women with uterine leiomyoma. PLoS ONE 8, e79838. Health 16. https://doi.org/10.1186/s12940-017-0287-4. https://doi.org/10.1371/journal.pone.0079838. Guart, A., Bono-Blay, F., Borrell, A., Lacorte, S., 2014. Effect of bottling and storage on the Sise, S., Uguz, C., 2017. Nonylphenol in human breast milk in relation to socio- migration of plastic constituents in Spanish bottled waters. Food Chem. 156, 73–80. demographic variables, diet, obstetrics histories and lifestyle habits in a Turkish https://doi.org/10.1016/j.foodchem.2014.01.075. population. Iran. J. Public Health 46, 491–499. Guenther, K., Heinke, V., Thiele, B., Kleist, E., Prast, H., Raecker, T., 2002. Endocrine Sprague, B.L., Trentham-Dietz, A., Hedman, C.J., Wang, J., Hemming, J.D., Hampton, disrupting nonylphenols are ubiquitous in food. Environ. Sci. Technol. 36, J.M., Buist, D.S., Aiello Bowles, E.J., Sisney, G.S., Burnside, E.S., 2013. Circulating 1676–1680. serum xenoestrogens and mammographic breast density. Breast Cancer Res. 15, R45. Gyllenhammar, I., Glynn, A., Darnerud, P.O., Lignell, S., van Delft, R., Aune, M., 2012. 4- https://doi.org/10.1186/bcr3432. 398 P. Peremiquel-Trillas et al. Environment International 122 (2019) 389–399 Steenland, K., Deddens, J.A., Zhao, S., 2000. Biases in estimating the effect of cumulative Occup. Environ. Med. 67, 837–844. https://doi.org/10.1136/oem.2009.052175. exposure in log-linear models when estimated exposure levels are assigned. Scand. J. Weiss, J.M., Hamers, T., Thomas, K.V., van der Linden, S., Leonards, P.E.G., Lamoree, Work Environ. Health 26, 37–43. https://doi.org/10.5271/sjweh.508. M.H., 2009. Masking effect of anti-androgens on androgenic activity in European Sun, H., Xu, X.-L., Qu, J.-H., Hong, X., Wang, Y.-B., Xu, L.-C., Wang, X.-R., 2008. 4- river sediment unveiled by effect-directed analysis. Anal. Bioanal. Chem. 394, Alkylphenols and related chemicals show similar effect on the function of human and 1385–1397. https://doi.org/10.1007/s00216-009-2807-8. rat estrogen receptor alpha in reporter gene assay. Chemosphere 71, 582–588. White, R., Jobling, S., Hoare, S.A., Sumpter, J.P., Parker, M.G., 1994. Environmentally https://doi.org/10.1016/j.chemosphere.2007.09.031. persistent alkylphenolic compounds are estrogenic. Endocrinology 135, 175–182. Sun, Y.-S., Zhao, Z., Yang, Z.-N., Xu, F., Lu, H.-J., Zhu, Z.-Y., Shi, W., Jiang, J., Yao, P.-P., https://doi.org/10.1210/endo.135.1.8013351. Zhu, H.-P., 2017. Risk factors and preventions of breast cancer. Int. J. Biol. Sci. 13, Winters, S., Martin, C., Murphy, D., Shokar, N.K., 2017. Breast cancer epidemiology, 1387–1397. https://doi.org/10.7150/ijbs.21635. prevention, and screening. Prog. Mol. Biol. Transl. Sci. 151, 1–32. https://doi.org/10. Villeneuve, S., Cyr, D., Lynge, E., Orsi, L., Sabroe, S., Merletti, F., Gorini, G., Morales- 1016/bs.pmbts.2017.07.002. Suarez-Varela, M., Ahrens, W., Baumgardt-Elms, C., Kaerlev, L., Eriksson, M., Zhou, F., Zhang, L., Liu, A., Shen, Y., Yuan, J., Yu, X., Feng, X., Xu, Q., Cheng, C., 2013. Hardell, L., Fevotte, J., Guenel, P., 2010a. Occupation and occupational exposure to Measurement of phenolic environmental estrogens in human urine samples by HPLC- endocrine disrupting chemicals in male breast cancer: a case-control study in Europe. MS/MS and primary discussion the possible linkage with uterine leiomyoma. J. Occup. Environ. Med. 67, 837–844. https://doi.org/10.1136/oem.2009.052175. Chromatogr. B Anal. Technol. Biomed. Life Sci. 938, 80–85. https://doi.org/10. Villeneuve, S., Cyr, D., Lynge, E., Orsi, L., Sabroe, S., Merletti, F., Gorini, G., Morales- 1016/j.jchromb.2013.08.032. Suarez-Varela, M., Ahrens, W., Baumgardt-Elms, C., Kaerlev, L., Eriksson, M., Zhou, Y., Bolton, E.C., Jones, J.O., 2014. Androgens and androgen receptor signaling in Hardell, L., Févotte, J., Guénel, P., 2010b. Occupation and occupational exposure to prostate tumorigenesis. J. Mol. Endocrinol. 54, R15–R29. https://doi.org/10.1530/ endocrine disrupting chemicals in male breast cancer: a case-control study in Europe. JME-14-0203. 399