Original article

Scand J Work Environ Health 2019;45(6):642-650    pdf

doi:10.5271/sjweh.3822 | Issue date:

Postmenopausal breast cancer and occupational exposure to chemicals

by Videnros C, Selander J, Wiebert P, Albin M, Plato N, Borgquist S, Manjer K, Gustavsson P

Objectives The aim of this study was to investigate if exposure to chemicals in the workplace was associated with an increased risk of postmenopausal breast cancer.

Methods The study comprised women born 1923–1950 living in Malmö city, Sweden, 1991–1996, and enrolled for a prospective population cohort study. Occupational exposure to various chemicals was assessed from job-exposure matrices. An extensive set of individual data on hormonal breast cancer risk factors were collected via a baseline questionnaire and used for confounding control. First time diagnoses of invasive breast cancer were identified through the Swedish Cancer Registry until end of follow-up on 31 December 2013.

Results Of 16 084 women, 1011 were diagnosed with breast cancer. Women exposed to chemicals in their occupational environment had a statistically significant increased risk [adjusted hazard ratio (HRadj) 1.26, 95% confidence interval (CI) 1.02–1.54] of breast cancer, and the risk correlated with duration of exposure. Investigation of risk in association with specific chemicals showed a non-significantly elevated risk after exposure to organic solvents. More than ten years of exposure to diesel exhaust was associated with an increased risk (HRadj 1.69, 95% CI 1.01–2.82). Occupational chemical exposures account for 2% of the breast cancer cases in this population.

Conclusions Occupational exposure to chemicals in general was associated with an elevated risk of breast cancer. A slight elevation of risk was seen after exposure to organic solvents. A statistically significant elevation of risk after >10 years of exposure to diesel exhaust was an unexpected finding.

This article refers to the following text of the Journal: 2018;44(3):310-322
The following articles refer to this text: 2021;47(2):154-162; 2022;48(1):81; [online first; 13 June 2022]

Breast cancer is the most common cancer among women worldwide, accounting for 25% of all cancer cases (1). Many established risk factors for breast cancer in women are associated with hormonal factors related to reproduction, such as early menarche, late menopause, hormone replacement therapy (HRT) and oral contraceptive use, while protective factors are early first-time pregnancy and multiparity (2). Alcohol consumption is a strong risk factor for breast cancer (3), while no increased risk was found in association with smoking (2). Family history of breast cancer (4) and high body mass index (BMI) (2) are other risk factors.

Previous studies suggest a link between chemical exposure and breast cancer, acting through three different mechanisms: (i) genotoxic action, (ii) alteration of mammary gland development or hormone responsiveness, or (iii) hormonal tumour promotion (5). Endocrine disrupting chemicals (EDC) can both alter the breast development and increase the susceptibility to breast cancer but also promote tumour growth through oestrogen- or progesterone mediated pathways (5). Even though the interest for studies, especially on EDC and breast cancer, have increased through the years, there is still a need for good quality epidemiological studies investigating the relationship between chemical exposure and the risk of breast cancer.

One of the most studied groups of chemicals in respect to breast cancer is organic solvents. Goldberg et al’s review (6) noted an increased risk of breast cancer in occupations with potential exposure to organic solvents, like dry cleaners, painters and laboratory technicians. This association has further been supported by a few other studies (711). Lifetime cumulative exposure to organic solvents has also been linked to an increased risk of male breast cancer (12). Organic solvents are absorbed systemically from the upper and lower airways but may also be absorbed through the skin and are then distributed throughout the body via the bloodstream (13). It can then, due to their lipophilic properties be stored in the surrounding fat tissue of the breast and migrate into the lobules and be transported to the ductular system (13). The mechanism for organic solvents could possibly be similar to the one for alcohol and breast cancer, since alcohol is an organic solvent (14). It is thought that organic solvents cause breast cancer by acting directly as a genotoxic agent or indirectly through their metabolites (5, 13).

Diesel and gasoline exhaust have shown to increase the risk of cancer in the lung and urinary bladder, however studies on breast cancer are inconsistent (15). Two studies have found an increased risk of breast cancer among women exposed to engine exhaust with a risk ratio of 1.53 [95% confidence interval (CI) 1.00-2.33] (16) and odd ratio of 2.57 (95% CI 1.16-5.69) (17) while three studies have found no increased risk (1820). However many of the studies had methodological problems such as few participants, not enough data on confounders such as reproductive factors or a low validity on exposure data (16, 18, 19). The relationship between diesel and gasoline exhaust and breast cancer needs to be investigated further.

A meta-analysis has found a small but significant increased risk of breast cancer among hairdressers which might be related to the exposure to hair dyes which contain substances that are suspected carcinogenic, like solvents (21).

Although some studies, as mentioned above, indicate an increased risk of breast cancer related to chemical exposure, the results were inconsistent and could potentially be influenced by uncontrolled confounding. Many studies lack detailed information on reproductive factors such as age at first pregnancy and parity, confounding factors that decrease the risk of breast cancer for occupational groups that have a higher parity and are younger at first term pregnancy, which is more common in manual labour where the exposure to chemicals might be higher. Since reproductive factors are crucial risk factors for breast cancer, it is important to have detailed individual reproductive data in order to conduct a good quality study. The aim of this study was to investigate if exposure to chemicals in the workplace were associated with an increased risk of postmenopausal breast cancer, while adjusting for individual data on important potential confounders.

Methods

The Malmö Diet and Cancer Study (MDCS) is a population-based prospective cohort study (22, 23). A total of 74 138 persons living in Malmö, Sweden, between 1991–1996 and 41–73 years old were invited to participate in the study, of which 24 851 did not respond or had unknown address, 16 942 declined to participate and 4247 were excluded due to language problems, leaving 28 098 participants (17 035 women) in MDCS. Exclusion criteria for the present study were women (i) with breast cancer before baseline (N=576), (ii) who had not been employed in an occupation for ≥1 year (N=247), and (iii) who remained premenopausal until the end of follow up (N=128). Premenopausal women were excluded due to the low number of premenopausal breast cancer cases in the cohort and the difficulty to analyse these due to different risk factors in pre- and postmenopausal breast cancer (24). High BMI increases the risk for postmenopausal breast cancer while it seems to decrease the risk in premenopausal women (24). Menopausal status was defined by medical records and questionnaire data. A woman was classified as postmenopausal if: (i) she had undergone bilateral oophorectomy, (ii) the above criteria was absent and she confirmed that her menstruation had ceased two years prior to baseline, or (iii) the above criteria was absent and she was ≥55 years of age. A total of 16 084 women were included in this study. At baseline, each participant filled out an extensive questionnaire containing questions on lifestyle, working history and reproductive factors. Questions on each woman’s three latest occupations and the time period for these were asked as well as detailed questions on the woman’s specific work tasks. Only retrospective occupational data were collected from baseline. Questions on alcohol consumption were asked as regard to the last 30 days using the validated questionnaire AUDIT (25). Health care personnel measured height and weight for each person.

Exposure to chemicals was assessed through the Scandinavian job-exposure matrices (JEM) NOCCA and FINJEM, adapted for Swedish working conditions (26, 27). Chemicals of interest for our study that were available in NOCCA included: 1,1,1-trichloroethane, benzene, benzo(a)pyrene, bitumen fumes, diesel exhaust, gasoline, methylene chloride, perchloroethylene, toluene, and trichloroethylene. Chemicals used from FINJEM included: aliphatic and alicyclic hydrocarbon solvents (ALHC), aromatic hydrocarbon solvents (ARHC), chlorinated hydrocarbon solvents (CHC), other organic solvents (other than defined in ALHC, ARHC, CHC, including alcohols, ketones, esters, glycol ethers), fungicides, herbicides, insecticides, polycyclic aromatic hydrocarbons (PAH), gasoline exhaust, and oil mist. The job-exposure matrices specify two measurements for each chemical and occupational group; intensity level for the exposed individuals and proportion exposed in the specific occupation. The matrices are also divided into time-periods since the exposure to chemicals change over time due to regulations, new methods that influences the usage of certain chemicals or safety precautions affecting the exposure. Being ever exposed to chemicals in our study were defined as having worked in an occupation where ≥5% of the employees were, according to the matrices, exposed to any of these chemicals.

First time diagnoses of invasive breast cancer in 1991−2013 were identified through the Swedish Cancer Registry (28). Breast cancer cases were identified as International Classification of Diseases, 7th Revision (ICD7) code 170. Death and migration status were retrieved from the Swedish National Tax Board.

Statistical analyses

Confounding variables were selected based on a priori knowledge from the literature including age (45–49, 50–54, 55–59, 60–64, 65–69, 70–74 years), parity (0, 1, 2, 3, ≥4), age at first term pregnancy (<20, 20–24, 25–29, 30–34, ≥35 years), months of breastfeeding per child (0, 1–5, 6–12, ≥13), hormone replacement therapy (HRT) (no treatment, oestrogen, progesterone, combined treatment), physical activity at work (quartiles), alcohol consumption (0, 1–14, 15–30, >30 g/day), height (<160, 160–169, ≥170 cm) and BMI (<18.5, 18.5–24.9, 25.0–29.9, ≥30 kg/m2).

BMI categorization was made according to the World Health Organization standard (29). Physical activity was measured with several questions estimating the time of physical activity performed outside of work and multiplied with an intensity factor for each activity (30). Months of breastfeeding were reported for each child and a mean number was calculated. Imputations were made on breastfeeding data for women who had data for at least one child but missing for another, using the mean number as imputation.

Chi-square tests were used to compare the distribution of risk factors for chemically exposed and non-exposed women. Cox proportional hazard model was used to estimate the hazard ratio (HR) for breast cancer in chemically exposed/never chemically exposed women and in duration analysis of 1–10 and >10 years. The underlying time variable was follow-up time from baseline to event or censoring. Minimally adjusted estimates were adjusted for age only while the fully adjusted model included age, parity, age at first term pregnancy, months of breastfeeding/child, HRT, physical activity, alcohol consumption, height, and BMI. We also performed a sensitivity analysis considering the HR for breast cancer by exposure duration, restricted to women who were pre-menopausal at baseline. Cumulative exposure was calculated as intensity level stated in JEM × proportion exposed according to JEM × years worked in that occupation (26, 27). The women were then divided dichotomously at the median and analysed using Cox proportional hazard model. Trend tests in Tables 2 and 3 were calculated using Cox proportional hazard model, creating a variable assigning the unexposed group a value of 0, the low or short-term exposed group a value of 1, and the high or long-term exposed group a value of 2, using the unexposed group as a reference. Pearson correlation analysis was used to investigate correlations between chemical agents and chemical groups (31). Population attributable fraction (PAF) was calculated using the formula AF=proportion of cases exposed to risk factor × (RR-1/RR) (32). All women who were postmenopausal were considered at risk from baseline until a breast cancer diagnosis, death, migration, or end of follow-up at 31 December 2013, whichever occurred first. If premenopausal at baseline, the woman was considered at risk from the time she became postmenopausal according to the earlier specified criteria for postmenopause. Statistical analyses were performed using STATA version 13.0 (StataCorp, College Station, TX, USA) with an α-level for significance tests at 0.05 (33).

Table 1

Distribution of potential risk factors for breast cancer in the present population of 1492 exposed and 14 592 non-exposed women.

Exposed a (N=1492) % Non-exposed (N=14 592) %
Age b (years)
 45–49 318 21.3 3712 25.4
 50–54 291 19.5 2817 19.3
 55–59 243 16.3 2482 17.0
 60–64 318 21.3 2682 18.4
 65–69 174 11.7 1571 10.8
 70–74 148 9.9 1328 9.1
Parity
 0 172 11.5 1870 12.8
 1 318 21.3 3095 21.2
 2 583 39.1 6061 41.5
 3 274 18.4 2409 16.5
 ≥4 129 8.7 901 6.2
 Missing values 16 1.1 256 1.8
Age at first term pregnancy
 <20 239 16.0 1468 10.1
 20–24 550 36.9 4898 33.6
 25–29 382 25.6 4244 29.1
 30–34 93 6.2 1408 9.7
 ≥35 38 2.6 440 3.0
 No children 172 11.5 1870 13.0
 Missing values 18 1.2 264 1.8
Months of breastfeeding/child
 0 67 4.5 600 4.1
 1–5 770 51.6 7649 52.4
 6–12 364 24.4 3604 24.7
 ≥13 28 1.9 149 1.0
 No children 172 11.5 1870 12.8
 Missing values 91 6.1 720 4.9
Hormone replacement therapy
 No treatment 1 247 83.6 11 808 80.9
 Estrogen 92 6.2 953 6.5
 Progesterone 7 0.5 93 0.6
 Estrogen & progesterone 142 9.5 1696 11.6
 Missing values 4 0.3 42 0.3
Physical activity (percentile)
 0–25 404 27.1 3519 24.1
 25–50 346 23.2 3603 24.7
 50–75 358 24.0 3763 25.8
 75–100 374 25.1 3628 24.9
 Missing values 10 0.7 79 0.5
Alcohol (g/day)
 0 182 12.2 994 6.8
 1–14 1141 76.5 11 107 76.1
 15–30 141 9.5 2129 14.6
 >30 26 1.7 349 2.4
 Missing values 2 0.1 13 0.1
Height (cm)
 <160 483 32.4 3450 23.6
 160–169 836 56.0 8620 59.1
 ≥170 170 11.4 2502 17.2
 Missing values 3 0.2 20 0.1
Body mass index (kg/m2)
 <18.5 underweight 16 1.1 223 1.5
 18.5–24.9 normal weight 661 44.3 7637 52.3
 25–29.9 overweight 538 36.1 4812 33.0
 ≥30 obese 274 18.4 1900 13.0
 Missing values 3 0.2 20 0.1

a Exposed in the work environment to at least one of the following chemicals; aliphatic and alicyclic hydrocarbon solvents, aromatic hydrocarbon solvents, chlorinated hydrocarbon solvents, other organic solvents, fungicides, herbicides, insecticides, polycyclic aromatic hydrocarbons, bitumen fumes, diesel exhaust, gasoline exhaust, oil mist, benzene, trichloroethylene, toluene, perchloroethylene, 1,1,1-trichloroethane, gasoline, benzo(a)pyrene, methylene chloride

b Age at baseline.

Table 2

Hazard ratios (HR) for invasive breast cancer by exposure duration; ever exposed, 1–10 years and >10 years of exposure. [CI=confidence interval.]

Chemical agents N total N cases Ever exposed Minimally adjusted a Ever exposed Fully adjusted b N cases 1-10 year exposure Fully adjusted b N cases >10 year exposure Fully adjusted b Trend test





HR 95% CI HR 95% CI HR 95% CI HR 95% CI P-value
Any of the included chemicals c 1492 109 1.21 0.99–1.47 1.26 1.02–1.54 43 1.05 0.77–1.45 66 1.43 1.10–1.85 0.01
Organic solvents d 957 67 1.16 0.90–1.48 1.15 0.89–1.50 34 1.14 0.80–1.63 33 1.18 0.81–1.71 0.28
Aliphatic and alicyclic hydrocarbon solvents 284 13 0.74 0.42–1.28 0.85 0.49–1.47 5 0.55 0.23–1.32 8 1.28 0.64–2.58 0.93
Aromatic hydrocarbon solvents 524 35 1.08 0.77–1.51 1.10 0.77–1.55 19 0.99 0.62–1.58 16 1.26 0.75–2.09 0.47
Benzene 567 37 1.04 0.75–1.44 1.06 0.76–1.49 23 1.24 0.82–1.87 14 0.84 0.47–1.48 0.97
Toluene 259 15 0.94 0.56–1.56 1.00 0.60–1.67 8 0.90 0.45–1.81 7 1.15 0.55–2.42 0.87
Chlorinated hydrocarbon solvents 628 42 1.08 0.79–1.47 1.12 0.82–1.55 21 1.09 0.70–1.70 21 1.17 0.74–1.84 0.45
Methylene chloride 208 17 1.44 0.90–2.34 1.39 0.83–2.32 6 1.03 0.46–2.31 11 1.80 0.93–3.49 0.11
1,1,1-trichloroethane 198 11 0.93 0.51–1.68 0.97 0.53–1.76 6 0.94 0.42–2.10 5 1.01 0.42–2.44 0.96
Other organic solvents e 292 24 1.37 0.92–2.06 1.34 0.87–2.06 13 1.31 0.74–2.31 11 1.38 0.71–2.67 0.20
Pesticides f 150 10 1.08 0.58–2.02 1.17 0.62–2.18 3 0.89 0.29–2.77 7 1.34 0.64–2.83 0.52
Fumes g 466 35 1.21 0.86–1.69 1.25 0.89–1.76 14 0.94 0.54–1.63 21 1.57 1.01–2.42 0.09
Polycyclic aromatic hydrocarbons 175 11 1.02 0.56–1.84 1.07 0.59–1.94 6 1.00 0.45–2.24 5 1.16 0.48–2.79 0.78
Diesel exhaust 303 24 1.27 0.85–1.91 1.33 0.88–2.01 9 0.95 0.47–1.92 15 1.69 1.01–2.82 0.08
Oil mist 395 19 0.76 0.48–1.20 0.89 0.56–1.40 9 0.71 0.37–1.37 10 1.14 0.61–2.13 0.87

a Adjusted for age.

b Adjusted for age, parity, age at first term pregnancy, months of breastfeeding per child, hormonal replacement therapy, physical activity, alcohol consumption, height and BMI.

c Exposed in the work environment to at least one of the following chemicals; aliphatic and alicyclic hydrocarbon solvents, aromatic hydrocarbon solvents, chlorinated hydrocarbon solvents, other organic solvents, fungicides, herbicides, insecticides, polycyclic aromatic hydrocarbons, bitumen fumes, diesel exhaust, gasoline exhaust, oil mist, benzene, trichloroethylene, toluene, perchloroethylene, 1,1,1-trichloroethane, gasoline, benzo(a)pyrene, methylene chloride.

d Aliphatic and alicyclic hydrocarbon solvents, aromatic hydrocarbon solvents, chlorinated hydrocarbon solvents, other organic solvents

e Includes alcohols, ketones, esters, glycol ethers etc.

f Fungicides, herbicides, insecticides.

g Polycyclic aromatic hydrocarbons, bitumen fumes, diesel exhaust, gasoline exhaust.

Table 3

Hazard ratios (HR) for invasive breast cancer by high/low cumulative exposure to chemicals. [CI=confidence interval.]

Chemical agents Mean cumulative exposure in class N total N cases Fully adjusted a Trend test

HR 95% CI P-value
Organic solvents b (ppm-years)
 Unexposed 0 15 127 944 1 0.42
 >0–17.74 5.97 474 35 1.26 0.89–1.79
 17.78–2802.58 140.38 483 32 1.05 0.72–1.54
Aliphatic and alicyclic hydrocarbon solvents (ppm–years)
 Unexposed 0 15 800 998 1 0.70
 >0–20.00 9.20 140 5 0.70 0.29–1.68
 20.34–487.92 106.53 144 8 0.98 0.49–1.97
Aromatic hydrocarbon solvents (ppm–years)
 Unexposed 0 15 560 976 1 0.78
 >0–3.13 0.68 259 21 1.21 0.77–1.91
 3.34–1253.36 95.83 265 14 0.97 0.57–1.65
Benzene (ppm–years)
 Unexposed 0 15 517 974 1 0.55
 >0–1.43 0.57 280 16 0.92 0.55–1.53
 1.45–16.85 4.08 287 21 1.20 0.77–1.88
Toluene (ppm–years)
 Unexposed 0 15 825 996 1 0.64
 >0–11.70 4.24 124 10 1.42 0.76–2.66
 12.00–770.00 85.53 135 5 0.63 0.26–1.52
Chlorinated hydrocarbon solvents (ppm–years)
 Unexposed 0 15 456 969 1 0.59
 >0–8.11 3.63 311 21 1.20 0.77–1.87
 8.26–285.00 25.83 317 21 1.05 0.67–1.66
Methylene chloride (ppm–years)
 Unexposed 0 15 876 994 1 0.33
 >0–8.40 4.26 99 8 1.62 0.81–3.27
 8.50–84.00 18.94 109 9 1.19 0.56–2.51
1,1,1–trichloroethane (ppm–years)
 Unexposed 0 15 886 1000 1 0.95
 >0–3.90 1.83 98 5 0.84 0.35–2.03
 4.00–108.75 11.95 100 6 1.12 0.50–2.50
Other organic solvents c (ppm–years)
 Unexposed 0 15 792 987 1 0.42
 >0–23.21 10.69 144 14 1.70 0.98–2.95
 23.86–1053.39 118.78 148 10 0.99 0.49–1.99
Pesticides d (mg–years/m3)
 Unexposed 0 15 934 1001 1 0.61
 >0–0.74 0.16 74 5 1.11 0.46–2.68
 0.77–7.28 2.36 76 5 1.23 0.51–2.96
Fumes e (mg–years/m3)
 Unexposed 0 15 618 976 1 0.18
 >0–1.13 0.43 232 17 1.16 0.71–1.91
 1.15–397.11 44.94 234 18 1.34 0.84–2.14
Polycyclic aromatic hydrocarbons (µg–years/m3)
 Unexposed 0 15 909 1000 1 0.51
 >0–1.17 0.39 84 3 0.60 0.19–1.87
 1.23–397.11 24.87 91 8 1.50 0.75–3.02
Diesel exhaust (mg–years/m3)
 Unexposed 0 15 781 987 1 0.11
 >0–0.42 0.23 151 10 1.09 0.58–2.03
 0.43–3.57 0.89 152 14 1.61 0.93–2.79
Oil mist (mg–years/m3)
 Unexposed 0 15 689 992 1 0.67
 >0–0.70 0.34 200 9 0.84 0.44–1.63
 0.72–78.00 7.19 195 10 0.93 0.50–1.74

a Adjusted for age, parity, age at first child, months of breastfeeding per child, hormonal replacement therapy, physical activity, alcohol consumption, height and BMI. The crude model gave essentially the same results as the fully adjusted model and was therefore not included in the table.

b Aliphatic and alicyclic hydrocarbon solvents, aromatic hydrocarbon solvents, chlorinated hydrocarbon solvents, other organic solvents.

c Includes alcohols, ketones, esters, glycol ethers etc.

d Fungicides, herbicides, insecticides.

e Polycyclic aromatic hydrocarbons, bitumen fumes, diesel exhaust, gasoline exhaust.

Results

The cohort comprised 16 084 women, of which 1011 women were diagnosed with breast cancer during the study period, and 1492 women were exposed to any of the chemicals we included in the analysis.

Distribution of risk factors for breast cancer among women exposed to chemicals and non-exposed women are presented in table 1. Women who were ever exposed to chemicals were older and had higher BMI (positive confounding) compared to unexposed women. Women ever exposed also had more children, were younger at first term pregnancy, and drank less alcohol (negative confounding) compared to unexposed women. These differences were all statistically significant.

Table 2 shows that women being exposed to any of the included chemicals had a higher risk of breast cancer compared to unexposed women (HR 1.26, 95% CI 1.02–1.54). Analysis by duration of exposure indicated that the risk increased with duration in the occupation, with a HR of 1.43 (95% CI 1.10–1.85) for women exposed >10 years compared to non-exposed women. The P-for-trend test showed a significant result with a P-value of 0.01. Calculation of PAF indicated that 2% of the breast cancer cases in this specific population could be attributed to occupational exposure to chemicals.

The analysis by duration of exposure, also in Table 2, showed statistically significantly increased risk of breast cancer for women being exposed >10 years to fumes [polycyclic aromatic hydrocarbons (PAH), bitumen fumes, diesel exhaust and gasoline exhaust] with a HR of 1.57 (95% CI 1.01–2.42), while no increased risk was found for women exposed 1–10 years. The P-for-trend test was non-significant. Among the components in fumes, specifically exposure to diesel exhaust showed a statistically increased risk for women being exposed >10 years (HR 1.69, 95% CI 1.01–2.82). The sensitivity analysis restricted to pre-menopausal women at baseline showed similar associations as in the full model (not shown).

The risk of breast cancer in relation to the cumulative exposure to each chemical agent is presented in Table 3. The trend of increased risk of breast cancer with cumulative exposure to chemicals showed no clear association, with non-significant HR both in the minimally adjusted (for age) and the fully adjusted analysis. Since the results were essentially the same in the minimally and fully adjusted model, only the latter is presented in Table 3.

Discussion

This study showed that women exposed to chemicals in their occupation were at an increased risk of breast cancer compared to women who never had been exposed. Women working longer time in an occupation with chemical exposure had higher risk of breast cancer compared to women working shorter time. Specifically women exposed to diesel exhaust for >10 years seemed to have an increased risk of breast cancer; 2% of the breast cancer cases in this population could be attributed to occupational exposure to chemicals.

The main analysis showed an increased risk of breast cancer for women who had been working in an occupation with exposure to any of the included chemicals. This result included increased HR in organic solvents, methylene chloride, other organic solvents, pesticides, fumes, and diesel exhaust, although all non-significant when evaluated individually.

The previous literature indicates that organic solvents could increase the risk of breast cancer (611). Even though our results showed a small increased risk, the HR were not statistically significant.

Women exposed to diesel exhaust for >10 years had a significantly increased risk of breast cancer (HR 1.69 95% CI 1.01–2.82). Breast cancer also seemed to increase with cumulative dose of diesel exposure, however those results were statistically non-significant. Not surprisingly, a quite strong correlation was found between diesel and gasoline exhaust (r=0.64) indicating that the results should be interpreted cautiously. Our findings have some support in previous studies; a cohort study in the US found increased risk of breast cancer (RR 1.53, 95% CI 1.00–2.33) associated with self-reported occupational exposure to engine exhaust (16). Another study found increased risk of breast cancer (OR 2.57, 95% CI 1.16–5.69) for women exposed to motor vehicle exhaust (17). Nevertheless, most previous studies on diesel exhaust found no increased risk of breast cancer (1820). Animal studies have shown possible mechanisms for gasoline exposure and breast cancer by either mimicking oestrogen to promote tumour growth, directly damaging DNA or promoting cancer cell proliferation. The International Agency for Research on Cancer (IARC) have classified diesel exhaust as carcinogenic to humans based on sufficient evidence linking it to an increased risk of lung cancer and urinary bladder cancer, but lacking enough quality studies on breast cancer (15). The findings of diesel exhaust and increased risk of breast cancer is a clear finding in our study but needs to be investigated further.

Exposure to methylene chloride showed a non-significant increased risk of breast cancer, and with no indication of dose–response. Most previous epidemiological studies have found little or no association between breast cancer and methylene chloride (34, 35) although an US study found an increased risk of breast cancer mortality for women exposed to methylene chloride (36). Animal studies have found increased mammary tumour in mice exposed to methylene chloride (37). The IARC recently classified methylene chloride as 2A carcinogenic to humans (cancer of the biliary tract and non-Hodgkin lymphoma) (38).

The analysis by duration of exposure (Table 2) showed a clearer trend than cumulative analysis using the full FINJEM/NOCCA JEM (Table 3). This is an unexpected result, but could be due to the misclassification of exposure that is inherent to using JEM. Since duration of exposure is measured at an individual level (self-reported) but intensity and proportion exposed are measured and estimated on group level, this adds inaccuracy to the estimate. A non-differential misclassification of exposure is therefore likely, resulting in an attenuation of the HR.

It has been suggested that exposure to chemicals can affect the risk of adverse outcomes differentially depending on at what age the women are exposed (39). During pregnancy there is a shift in the breast tissue of more stem cells to a stage where they become less sensitive (39). Our data contained women exposed both before and after menopause, however the majority of exposure took place before menopause, in reproductive age. Sensitivity analysis showed essentially the same results for women exposed only before menopause and those exposed before and after. The loss in power made the CI wider, but the point estimates remained close to the original. However, there was a significant increase in risk for women exposed to pesticides; from a non-significant HR for the full cohort to a significant increased risk for women only exposed before menopause (HR 2.90 95% CI 1.43–5.86). Pesticides have been researched increasingly over the last years and there is an indication that exposure to pesticides in early life might increase the risk of breast cancer (40), however the research is far from consistent (37, 41). The sensitivity analysis indicates a possible increased risk if exposed in reproductive years and the relationship between pesticides and breast cancer should be investigated further.

Aside from the correlation between diesel exhaust and gasoline exhaust, low correlations were found between individual chemical agents (r=0.01–0.47) and between main chemical groups (r=0.01–0.23).

This study has some other weaknesses and strengths that need to be considered. Occupational data were only available and measured until the day of enrolment in the study. However, the participants were quite old at recruitment and many were close to retirement, leaving an almost complete occupational exposure history for each participant. Another weakness is the power of this study. As we find predominantly positive but non-significant associations, several of these might have been significant with a better power. Since the exposure is quite rare, especially regarding specific chemical agents, we would have needed a larger sample in order to observe an effect for many of the chemical agents. This is especially true when analysing duration and cumulative exposure. This study lacks information on the estrogen-receptor status of breast cancer, which may play a role in the development of breast cancer. Strength of this study is the extensive individual data on hormonal and reproductive factors allowing for a good confounder control. However, there is always a possibility of residual confounding that need to be considered. Night-shift work is a possible confounder that is strongly associated with occupation and seems to be associated with increased risk of breast cancer (42). It might explain the increased risk seen in occupations where exposure to chemicals and night-shift work exist, eg, among bus drivers or nurses. However, most of the results in this study are probably attenuated towards HR 1.00 due to the misclassification of exposure described earlier.

A strength of this study is that the design used a prospective cohort with a follow-up of >20 years. Outcome data on breast cancer cases are close to 100% coverage in the Swedish Cancer Registry.

Occupational chemical exposure is a risk factor that can be regulated and eliminated if found to be harmful, and is therefore important to investigate thoroughly. This paper indicates that chemical exposure is associated with an increased risk of breast cancer and that 2% of the breast cancer cases in this specific population could be attributed to occupational chemical exposure. This is equivalent to approximately 20 of the 1011 cases. However, this figure should be interpreted cautiously due to the risk of residual confounding discussed above.

In conclusion, women exposed to chemicals in their occupational environment had a statistically significant increased risk of breast cancer compared to women who were not exposed. The risk tended to be higher after >10 years of exposure than after 1–10 years of exposure for many of the studied exposures. Specifically women exposed to diesel exhaust for >10 years had a statistically significant increased risk of breast cancer. Further research is needed to identify associations between exposure to individual chemicals and breast cancer.

Acknowledgements

The authors thank Håkan Andersson, IT coordinator at Skåne University Hospital, for data management.

Competing interest

Authors declare no conflicts of interest

Funding

The Swedish Research Council for Health, Working Life and Welfare (FORTE) has funded this study (grant no 2013-04-02). The sponsor had no role in the design or conduct of the study, the collection, management, analysis, or interpretation of data, the preparation, review, or approval of the manuscript, or the decision to submit the manuscript for publication.

Ethical approval

Lund ethical review board approved the MDCS study (LU 51-90), and the Stockholm ethical review board approved the present study (Dnr 2014/233-31/4). All participants gave informed consent before taking part in this study.

References

1 

Bray, F, Ferlay, J, Soerjomataram, I, Siegel, RL, Torre, LA, & Jemal, A. (2018, Nov). Global cancer statistics. 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin, 68(6), 394-424, https://doi.org/10.3322/caac.21492.

2 

Schottenfeld, D, & Fraumeni, JF. (2006). Cancer epidemiology and prevention. New York: Oxford university press.

3 

Longnecker, MP. (1994, Jan). Alcoholic beverage consumption in relation to risk of breast cancer:meta-analysis and review. Cancer Causes Control, 5(1), 73-82, https://doi.org/10.1007/BF01830729.

4 

Morrow, M, & Gradishar, W. (2002, Feb). Breast cancer. BMJ, 324(7334), 410-4, https://doi.org/10.1136/bmj.324.7334.410.

5 

Rodgers, KM, Udesky, JO, Rudel, RA, & Brody, JG. (2018, Jan). Environmental chemicals and breast cancer:an updated review of epidemiological literature informed by biological mechanisms. Environ Res, 160, 152-82, https://doi.org/10.1016/j.envres.2017.08.045.

6 

Goldberg, MS, & Labrèche, F. (1996, Mar). Occupational risk factors for female breast cancer:a review. Occup Environ Med, 53(3), 145-56, https://doi.org/10.1136/oem.53.3.145.

7 

Hansen, J. (1999, Jul). Breast cancer risk among relatively young women employed in solvent-using industries. Am J Ind Med, 36(1), 43-7, https://doi.org/10.1002/(SICI)1097-0274(199907)36:1<43::AID-AJIM6>3.0.CO;2-A.

8 

Peplonska, B, Stewart, P, Szeszenia-Dabrowska, N, Lissowska, J, Brinton, LA, Gromiec, JP, et al. (2010, Nov). Occupational exposure to organic solvents and breast cancer in women. Occup Environ Med, 67(11), 722-9, https://doi.org/10.1136/oem.2009.046557.

9 

Gustavsson, P, Andersson, T, Gustavsson, A, & Reuterwall, C. (2017, Nov). Cancer incidence in female laboratory employees:extended follow-up of a Swedish cohort study. Occup Environ Med, 74(11), 823-6, https://doi.org/10.1136/oemed-2016-104184.

10 

Ekenga, CC, Parks, CG, D'Aloisio, AA, DeRoo, LA, & Sandler, DP. (2014, Jun). Breast cancer risk after occupational solvent exposure:the influence of timing and setting. Cancer Res, 74(11), 3076-83, https://doi.org/10.1158/0008-5472.CAN-13-2430.

11 

Labrèche, F, Goldberg, MS, Valois, MF, & Nadon, L. (2010, Apr). Postmenopausal breast cancer and occupational exposures. Occup Environ Med, 67(4), 263-9, https://doi.org/10.1136/oem.2009.049817.

12 

Laouali, N, Pilorget, C, Cyr, D, Neri, M, Kaerlev, L, Sabroe, S, et al. (2018, May). Occupational exposure to organic solvents and risk of male breast cancer:a European multicenter case-control study. Scand J Work Environ Health, 44(3), 310-22, https://doi.org/10.5271/sjweh.3717.

13 

Labrèche, FP, & Goldberg, MS. (1997, Jul). Exposure to organic solvents and breast cancer in women:a hypothesis. Am J Ind Med, 32(1), 1-14, https://doi.org/10.1002/(SICI)1097-0274(199707)32:1<1::AID-AJIM1>3.0.CO;2-3.

14 

Dumitrescu, RG, & Shields, PG. (2005, Apr). The etiology of alcohol-induced breast cancer. Alcohol, 35(3), 213-25, https://doi.org/10.1016/j.alcohol.2005.04.005.

15 

IARC. Monographs Diesel and Gasoline Engine Exhausts and Some Nitroarenes, vol 105. (2013). IARC. Lyon: .

16 

Van Den Eeden, SK, & Friedman, GD. (1993, Mar). Exposure to engine exhaust and risk of subsequent cancer. J Occup Med, 35(3), 307-11.

17 

Nie, J, Beyea, J, Bonner, MR, Han, D, Vena, JE, Rogerson, P, et al. (2007, Nov). Exposure to traffic emissions throughout life and risk of breast cancer:the Western New York Exposures and Breast Cancer (WEB) study. Cancer Causes Control, 18(9), 947-55, https://doi.org/10.1007/s10552-007-9036-2.

18 

Rai, R, Glass, DC, Heyworth, JS, Saunders, C, & Fritschi, L. (2016, Jun). Occupational exposures to engine exhausts and other PAHs and breast cancer risk:A population-based case-control study. Am J Ind Med, 59(6), 437-44, https://doi.org/10.1002/ajim.22592.

19 

Raaschou-Nielsen, O, Andersen, ZJ, Hvidberg, M, Jensen, SS, Ketzel, M, Sørensen, M, et al. (2011, Jul). Air pollution from traffic and cancer incidence:a Danish cohort study. Environ Health, 10, 67, https://doi.org/10.1186/1476-069X-10-67.

20 

Weiderpass, E, Pukkala, E, Kauppinen, T, Mutanen, P, Paakkulainen, H, Vasama-Neuvonen, K, et al. (1999, Jul). Breast cancer and occupational exposures in women in Finland. Am J Ind Med, 36(1), 48-53, https://doi.org/10.1002/(SICI)y1097-0274(199907)36:1<48::AID-AJIM7>3.0.CO;2-2.

21 

Takkouche, B, Regueira-Méndez, C, & Montes-Martínez, A. (2009, Dec). Risk of cancer among hairdressers and related workers:a meta-analysis. Int J Epidemiol, 38(6), 1512-31, https://doi.org/10.1093/ije/dyp283.

22 

Manjer, J, Carlsson, S, Elmståhl, S, Gullberg, B, Janzon, L, Lindström, M, et al. (2001, Dec). The MalmöDiet and Cancer Study:representativity, cancer incidence and mortality in participants and non-participants. Eur J Cancer Prev, 10(6), 489-99, https://doi.org/10.1097/00008469-200112000-00003.

23 

Kullberg, C, Selander, J, Albin, M, Borgquist, S, Manjer, J, & Gustavsson, P. (2017, Sep). Female white-collar workers remain at higher risk of breast cancer after adjustments for individual risk factors related to reproduction and lifestyle. Occup Environ Med, 74(9), 652-8, https://doi.org/10.1136/oemed-2016-104043.

24 

Chen, Y, Liu, L, Zhou, Q, Imam, MU, Cai, J, Wang, Y, et al. (2017, Dec). Body mass index had different effects on premenopausal and postmenopausal breast cancer risks:a dose-response meta-analysis with 3,318,796 subjects from 31 cohort studies. BMC Public Health, 17(1), 936, https://doi.org/10.1186/s12889-017-4953-9.

25 

Piccinelli, M, Tessari, E, Bortolomasi, M, Piasere, O, Semenzin, M, Garzotto, N, et al. (1997, Feb). Efficacy of the alcohol use disorders identification test as a screening tool for hazardous alcohol intake and related disorders in primary care:a validity study. BMJ, 314(7078), 420-4, https://doi.org/10.1136/bmj.314.7078.420.

26 

Kauppinen, T, Heikkilä, P, Plato, N, Woldbaek, T, Lenvik, K, Hansen, J, et al. (2009). Construction of job-exposure matrices for the Nordic Occupational Cancer Study (NOCCA). Acta Oncol, 48(5), 791-800, https://doi.org/10.1080/02841↡0271∫.

27 

Kauppinen, T, Uuksulainen, S, Saalo, A, Mäkinen, I, & Pukkala, E. (2014, Apr). Use of the Finnish Information System on Occupational Exposure (FINJEM) in epidemiologic, surveillance, and other applications. Ann Occup Hyg, 58(3), 380-96.

28 

Barlow, L, Westergren, K, Holmberg, L, & Talbäck, M. (2009). The completeness of the Swedish Cancer Register:a sample survey for year 1998. Acta Oncol, 48(1), 27-33, https://doi.org/10.1080/02841860802247664.

29 

Physical status: the use and interpretation of anthropometry. Report of a WHO Expert Committee. (1995). World Health Organ Tech Rep Ser, 854, 1-452.

30 

Questionnaire for Malmo Diet and Cancer study. Questions regarding physical activity, Available from: https://www.med. lu.se/malmoe_kost_cancer_och_malmoe_foerebyggande_ medicin/malmoe_kost_cancer/mkc_fraageformulaer .

31 

Woodward, M. (2005). Epidemiology: study design and data analysis (2nd ed). Chapman & Hall.

32 

Rockhill, B, Newman, B, & Weinberg, C. (1998, Jan). Use and misuse of population attributable fractions. Am J Public Health, 88(1), 15-9, https://doi.org/10.2105/AJPH.88.1.15.

33 

StataCorp. Stata Statistical Software (Release 13 ed). (2013). College Station, TX: StataCorp LP.

34 

Liu, T, Xu, QE, Zhang, CH, & Zhang, P. (2013, Dec). Occupational exposure to methylene chloride and risk of cancer: a meta-analysis. Cancer Causes Control, 24(12), 2037-49, https://doi.org/10.1007/s10552-013-0283-0.

35 

Cooper, GS, Scott, CS, & Bale, AS. (2011, Aug). Insights from epidemiology into dichloromethane and cancer risk. Int J Environ Res Public Health, 8(8), 3380-98, https://doi.org/10.3390/ijerph8083380.

36 

Cantor, KP, Stewart, PA, Brinton, LA, & Dosemeci, M. (1995, Mar). Occupational exposures and female breast cancer mortality in the United States. J Occup Environ Med, 37(3), 336-48, https://doi.org/10.1097/00043764-199503000-00011.

37 

Brody, JG, & Rudel, RA. (2003, Jun). Environmental pollutants and breast cancer. Environ Health Perspect, 111(8), 1007-19, https://doi.org/10.1289/ehp.6310.

38 

IARC. (2017). Some Chemicals Used as Solvents and in Polymer Manufacture, 110.

39 

Russo, J, Moral, R, Balogh, GA, Mailo, D, & Russo, IH. (2005). The protective role of pregnancy in breast cancer. Breast Cancer Res, 7(3), 131-42, https://doi.org/10.1186/bcr1029.

40 

Fenga, C. (2016, Mar). Occupational exposure and risk of breast cancer. Biomed Rep, 4(3), 282-92, https://doi.org/10.3892/br.2016.575.

41 

Khanjani, N, Hoving, JL, Forbes, AB, & Sim, MR. (2007, Jan-Mar). Systematic review and meta-analysis of cyclodiene insecticides and breast cancer. J Environ Sci Health C Environ Carcinog Ecotoxicol Rev, 25(1), 23-52, https://doi.org/10.1080/10590500701201711.

42 

Kamdar, BB, Tergas, AI, Mateen, FJ, Bhayani, NH, & Oh, J. (2013, Feb). Night-shift work and risk of breast cancer:a systematic review and meta-analysis. Breast Cancer Res Treat, 138(1), 291-301, https://doi.org/10.1007/s10549-013-2433-1.