Exposure to Pesticides and Breast Cancer in an Agricultural Region in Brazil

Rural workers are disproportionally exposed to pesticides and might be at an increased risk of developing chronic diseases. Here, we investigated the impact of pesticide exposure on breast cancer (BC) risk and disease profile in rural female workers. This is a case-control study that prospectively included 758 individuals. The study was conducted in the Southwest region of Paraná state in Brazil, a region characterized by family-based agriculture and intensive use of pesticides. We found that this region has a 41% higher BC diagnosis rate and 14% higher BC mortality rate than the mean rates in Brazil, as well as a pesticide trade volume about 6 times higher than the national average. We showed substantial exposure in this population and found that even women who did not work in the fields but performed equipment decontamination and clothes washing of male partners who worked in the fields had urine samples positive for glyphosate, atrazine, and/or 2,4-D. The crude association showed a significantly higher risk of BC among women exposed to pesticides (OR: 1.58, 95% CI 1.18–2.13). Adjusted analyses showed a lower and nonstatistically significant association (OR: 1.30, 95% CI 41 0.87–1.95). Stratification on disease profile showed a significantly higher risk of lymph node metastasis (adjusted OR: 2.19, 95% CI 1.31–3.72) in women exposed to pesticides. Our findings suggest that female populations exposed to pesticides are at a higher risk of developing BC with a more aggressive profile and draw attention to the need to monitor rural populations potentially exposed to pesticides in the field or at home.

Brazil is one of the top global consumers of pesticides. 20Particularly, the Southwest region of the state of Paranáis one of the main pesticide trading regions in Brazil.The region is characterized by widespread family farming and intensive use of agrochemicals, such as glyphosate, atrazine, and 2,4-D. 22In some areas with a predominance of family farming, women play a leading role in the field, while in others, they maintain domestic activities that support family members who work in the field. 17n the Southwest region of ParanáState and other areas in Brazil, handling pesticides in rural work is considered a predominantly male activity. 18Nevertheless, women are not exempt from the risk of direct exposure. 19This is because women could be exposed through their unprotected manipulation of pesticides (e.g., via preparation and dilution of pesticides), as well as through equipment decontamination and handling of contaminated clothing.Indeed, women's participation in rural work through pesticide manipulation and equipment decontamination has been well documented. 21However, these routes of exposure that occur within the household have often been dismissed and neglected.The outcome is that limited attention has been given to the consequences of exposure of female family members not engaged in pesticide application.
−33 Increased risk for BC has been suggested in wives of farmers working with a wide range of pesticides, such as fungicides, 34 organophosphates, 35 and organochlorines. 36However, in such studies, female exposure has often been dismissed.Moreover, the extent of their contamination is not typically documented, with information about the presence of pesticides in women's blood/urine often lacking.All in all, data regarding women's exposure to pesticides at home either via pesticide manipulation or via clothing and equipment decontaminating are scarce.Data concerning disease profiles in women chronically exposed to pesticides and the correlation between exposure and disease outcomes are also similarly limited.
Here, we report the impact of pesticide exposure on women exposed to pesticides via unprotected pesticide manipulation, handling of contaminated clothing, and equipment.We also report the potential implications for pesticide exposure to BC risk.We investigated the relationship between pesticide exposure and BC by documenting women exposure to glyphosate, atrazine, and 2,4-D.We evaluated the odds ratios associated with pesticide exposure and compared exposed and unexposed patients to address the clinicopathological profile of pesticide-exposed BC patients.

■ METHODS
This a case-control study that prospectively included a total of 758 individuals.This study collected individual and clinicopathological data from women at a public Oncology Hospital located at the eighth Health Regional of Paranástate (Hospital de Cancer de Francisco Beltraõ, Ceonc) from January 2016 to December 2019.This region comprises an area of 7768 km 2 and a total of 27 municipalities with approximately 330,000 inhabitants.The economy is based on small-scale farms and conventional rural activities, mainly grain crops (soy, corn, and wheat), milk cattle, and chicken livestock, in which pesticide spraying is largely done manually (with spraying machines carried in workers' backs).
The Institutional Ethics Committee on Research of the State University of Western Parana, CAAE 35524814.4.0000.0107,approved this research.All participants signed written informed consents and were informed about the research aims.All women screening for breast cancer were invited to join the study.Based on the analysis of the biopsies by a pathologist, women were categorized according to the presence of breast cancer or benign lesions.Among those diagnosed with BC, the study included women carrying operable tumors (TNM stage II).We included a total of 728 patients for the BC risk study and 30 individuals for the pesticide contamination study.
The study included the following information: Information concerning pesticide exposure was collected through individual interviews based on a validated questionnaire developed for this purpose. 21The study population consists of female rural residents.Women in this region do not typically participate in pesticide spraying.Exposure assessment occurred via a questionnaire that captured both modes of exposure�in the household and through pesticide spraying in the fields�and through assessment of urine samples.For the assessment of urine samples, we chose households where women did not participate in the pesticide spraying.This was a conservative choice because exposure in the household is expected to be lesser than exposure in the fields.It was important for us to assess exposure in the household because this mode of exposure has been neglected and has been suggested to be negligible.Our assessment dispels that notion and shows that even women who work within the household are at risk of significant exposure.
The questions yielded data about continuous contact with pesticides, lifetime working with pesticides, wearing of personal protection equipment (PPE) while spraying or washing/ decontaminating clothes and PPE, and the use of protective gloves when doing these procedures.Based on such questions, women were considered as exposed to pesticides or not exposed to pesticides.The pesticide-exposed group was formed by rural women who positively answered pesticide-contact questions and lived at least 40% of their adult life working with pesticides.In contrast, the unexposed group was composed of women without a history of rural work or pesticide-related occupations in the household.To be most conservative during exposure assessment, we only evaluated women whose husbands or family members worked in the field but who did not themselves work in the fields.The most likely place of exposure for these women is in the household.
To investigate if pesticide exposure in these circumstances was enough to generate women's contamination, urine samples were collected from 30 individuals (10 women + 10 husbands + 10 other family members) at the peak of pulverization of the most commercialized pesticides in Paranástate: glyphosate, atrazine, and 2,4-D.The presence of such residues was analyzed by commercial immunoassay-based kits (Abraxis) from ten distinct farms, randomly selected across the 27 municipalities.Urine samples (30−50 mL) were collected in sterile tubes from all family members who reported directly manipulating pesticides.Samples were collected from one woman and her relatives (applicators #1 and #2) until 6 h after glyphosate, atrazine, or 2,4-D spraying by not-mechanized methods, and the results were expressed as ppb (parts per billion).The applicators Environmental Science & Technology reported using intercostal pumps for spraying pesticides for about six consecutive hours.All exposed women reported having pesticide unprotected contact by diluting the concentrated package of the chemicals, filling the intercostal pump with the diluted pesticides, and washing/decontaminating clothes/PPE used for this end after spraying without wearing gloves or any other protection equipment.For BC risk calculation, data concerning pesticide exposure were collected from women with and without BC diagnosis.
Individual and tumor information was obtained from patient's medical records to evaluate the relationship between pesticide exposure and clinicopathological features.Data from BC patients included age at diagnosis, menopausal status at diagnosis, body mass index values (BMI), BC molecular subtype, presence of lymphatic or distant metastasis, chemoresistance profile, and disease recurrence.Only the age at diagnosis, menopausal status, and body mass index (BMI, kg/ m 2 ) values were collected from patients without a BC diagnosis.
For data analysis, we stratified individual characteristics of participants into exposure groups and calculated the mean and standard deviation for continuous variables and counts with percentages for categorical variables.Means of continuous variables were compared using the Wilcoxon rank sum test and categorical variables were compared using Pearson's χ 2 test.We performed the same calculations for cancer characteristics among participants with breast cancer.We similarly analyzed participants stratified by breast cancer subtype.
An unstratified correlation analysis was conducted across all variables in the study.We also stratified participants by breast cancer diagnosis and exposure to pesticides.All correlations were calculated as Pearson's correlation coefficients and analyzed for significance.
To assess the risk of breast cancer given pesticide exposure, we calculated the univariate odds ratio.We used the complete data set to calculate the odds ratio of developing breast cancer given exposure to pesticides compared to the risk of developing breast cancer among those not exposed to pesticides.Among women who developed breast cancer, we then calculated the odds ratio of specific molecular subtypes, histological grades, and other tumor markers associated with breast cancer given pesticide exposure.To assess molecular subtypes and histological grade� in addition to comparing each subtype to the most benign case (luminal A subtype and grade 1 tumors)�we calculated the odds ratio of various combinations of subtypes and histological grades.
For the clinical variables that had significant univariate odds ratios, we conducted logistic regressions to assess their association with pesticide exposure, adjusted for breast cancer risk factors of age, menopause status, and BMI (body mass index).Raw data for each patient are provided in Supporting Table 1.

■ RESULTS
High Rates of BC Incidence and Mortality.A comparative analysis concerning mean BC incidence and mortality (Figure 1A−C) shows that the Southwest region of Paranápresents higher rates of BC incidence and mortality relative to the state of Paranáas a whole or Brazil.Specifically, the Southwest of Paranáhas a 31% higher incidence than Paranaś tate and 41% higher incidence than Brazil, as well as 17% higher mortality than Paranástate and 14% higher mortality than Brazil.
During the documented period (2016−2019), Paranaó ccupied the third position in the rank for pesticide trade in the country with 61,712 tons of pesticides traded during the period (Figure 1C).This is only behind the states of Mato Grosso and Saõ Paulo with 106,632 tons and 82,286 tons of pesticides traded, respectively.The mean pesticide trade for Brazil was 590,670 tons in the same period.Due to its agriculture-based economy, the Southwest region of Parana has a per capita consumption of pesticides that is significantly higher than the Brazilian average of about 6.7 kg/hectare (Figure 1D).Importantly, the region is characterized by small farms, in which the extensive use of nonmechanized spraying of pesticides in family-based agriculture is predominant.
Pesticide Contamination in the Female Study Population Occurs during Unprotected Equipment Decontamination and Clothes Washing.Given the region's characteristics, where male family members are primarily engaged in pesticide application, we characterized pesticide exposure in women who are not working in the field.As shown in Figure 2, a total of 10 rural properties located in 9 municipalities from Paranástate were included.We selected farms that had a family member with BC diagnosis.BC patients from the selected properties live in households, where the male partner was primarily occupied in agriculture and had a history of pesticide use over several years.Data from the exposure characterization interview (Figure 2, Box A) showed that most patients were in continuous contact with pesticides at least once a week (57%) and spent more than 40% of their life working with pesticides (53%).
Remarkably, about 95% of exposed patients who did not work in the fields reported that they did not wear protective gloves during clothes/PPE decontamination.We therefore suspect that this might be a significant route of exposure in this population.Indeed, because of their exposure profile, we analyzed urine samples from women after unprotected pesticide manipulation and clothes/PPE decontamination.Results of urine sample analyses obtained from exposed women are shown in Figure 2,  Box B.
Approximately 53% of the samples had some of the three pesticides detected (33.3% of BC patients, 43.3% for applicators #1 and #2).Glyphosate was the most often detected pesticide (at concentrations ranging from 0.25 to 2.12 ppb), followed by atrazine (1.9 to 5.9 ppb) and 2,4-D (29.3 to 80.8 ppb).The Figure 2. Design of the study.A total of 775 women, who attended a public hospital (Francisco Beltraõ Cancer Hospital�Ceonc, Francisco Beltraõ� Parana, Brazil), were screened.After signing consent forms, patients were interviewed to obtain their pesticide exposure profiles.Seventeen women were excluded from the study due to THE lack of follow-up or insufficient information regarding pesticide exposure.Thus, 758 women were included in the study.Urine samples from 30 women were collected in the exposure assessment (contamination) study, aiming to measure glyphosate, atrazine, and 2,4-D residues.For the cancer risk study (n = 728), a total of 373 women were characterized as exposed and 355 as unexposed.Based on the biopsy result, patients were categorized into four groups: Women exposed to pesticides diagnosed with breast cancer (n = 215), women exposed to pesticides without breast cancer (n = 158), women not exposed to pesticides diagnosed with breast cancer (n = 164), and women not exposed to pesticides without breast cancer (n = 191).Data obtention included patient characteristics (age at diagnosis, weight, height, body mass index, menopausal status, disease onset), tumor features molecular subtypes, hormonal receptors status, human epidermal growth factor 2 receptor amplification (HER2), the presence of tumor emboli or lymph nodal metastasis, distant metastasis, chemoresistance, and tumor recurrence.

Environmental Science & Technology
simultaneous detection of the three pesticides was found in Farms #3 and #4, where the families reported spraying pesticides for more than 6 h/daily.A urine sample from a BC patient living in farm #3 had higher levels of pesticide exposure than those found in the family applicators for all three tested pesticides.In farm #6, only the BC patient had a positive sample (0.91 ppb for glyphosate).Both women from farms #3 and #6 reported decontaminating the family's clothes and PPE after pesticide spraying.Urine samples from BC patients living in farm#10 and #4 were the only negative for pesticides when family members were positive.
Exposure to Pesticides and a Higher Risk of Developing BC and Disease Metastasis.Considering the co-occurrence of high BC rates and elevated pesticide use in the Southwest region of Paranástate, we investigated if there was a relationship between pesticide exposure and BC risk or disease profile.To address the issue, we examined 728 women, who went to a public hospital for breast screening between January 2016 and December 2019.Women were divided into four groups based on their exposure to pesticides as well as biopsy results: exposed with BC (n = 215), exposed without BC (n = 158), unexposed with BC (n = 164), and unexposed without BC (n = 191).Medical records were assessed to obtain data about a patient's characteristics and tumor features.
Table 1 shows the individual characteristics of women included in the study.Women without cancer had a mean age at diagnosis of 40.8 years, and patients having BC showed a mean age at diagnosis of 55.7 years.BMI was 26.6 mg/m 2 in no cancer group and 28.0 kg/m 2 for BC patients.The mean volume of pesticides traded (2011−2016) was 281.6 tons in the municipalities from no cancer women and 276.9 tons in the municipalities from BC patients.
Regarding pesticide exposure (Table 2), in women without BC, the mean age of the negative diagnosis was 38.06 years for unexposed controls and 44.22 years for the exposed ones (p < 0.001).The mean BMI was 26.17 kg/m 2 for the unexposed and 27.01 for the exposed ones (p = 0.072).For the BC groups, both unexposed and exposed women had the same mean age at diagnosis (55.67 years, p > 0.9) and similar BMI (28.19 and 27.85 kg/m 2 , respectively, p = 0.8).BC patients from both unexposed and exposed groups had similar proportions of early disease onset (41 and 38%, respectively, p = 0.6) and menopausal status (71 and 68%, respectively, p = 0.7).
Women exposed to pesticides had a higher risk of developing BC than unexposed controls (OR crude 1.58, CI: 1.18−2.13for all cases, 1.51, CI: 1.05−2.17for complete cases, and OR adjusted 1.30, CI 0.87−1.95,Table 3).The mean volume of pesticide traded in the study period was similar across all groups and ranged from 267.74 tons in the localities of unexposed women without BC to 287.47 tons in the localities of exposed women without breast cancer (p = 0.12).Similar profiles were observed for the tumor molecular subtype in unexposed and exposed BC patients (Table 4).
Exposed BC patients presented more lymphatic metastases than the unexposed group (p = 0.015).Similar profiles were observed when comparing chemoresistance (p = 0.5) and disease recurrence (p = 1).A multivariate analysis of the association between lymph node metastasis with pesticide exposure, adjusting for breast cancer risk factors of age, menopause status, and body mass index (BMI), showed that the risk for being exposed and having lymph nodal metastasis is 2.19 (CI: 1.31−3.72,Table 5).
Spearman correlation analysis among clinicopathological parameters from all BC patients showed a significant positive correlation between pesticide exposure and the presence of lymph nodal metastasis (Figure 4, R = 0.145, p < 0.05) (Figure 3).Other statistically significant correlations were clinically expected (Figure 4).

■ DISCUSSION
The role of women in family-based agriculture is particularly underappreciated in regions where fieldwork is typically  conducted by male family members. 37Because of this, studies focusing on women's health issues due to pesticide exposure are rare in these settings.In the present study, we detailed how pesticide exposure occurred in women not engaged in pesticide application and investigated their increased risk for BC development and metastasis.We reported that women exposed to pesticides have higher breast cancer risk than the unexposed ones.Also, urine analysis for glyphosate, atrazine, and 2,4-D residues showed that the unprotected manipulation of pesticide-containing items results in women's contamination.Furthermore, pesticide exposure significantly correlated to disease metastasis.
The main route of pesticide exposure was unprotected PPE decontamination and washing of contaminated clothes after pulverization by family members.Exposure to pesticides occurs mainly due to their misperceptions about the hazardous risk of such chemicals and the wrong notion that PPE decontamination and handling of clothing is not a significant source of exposure.Further, the lack of adequate spaces to manage contaminated equipment and low adherence to good pesticide management practices are also reported as primary sources of widespread female contamination in these small-scale family owned properties. 19tudies of female exposure to pesticides have mainly focused on reproductive and hormonal health and include outcomes, such as changes in placenta, 38 female infertility, 39 and spontaneous abortion occurrence. 40Previous research in Brazil has considered that women were minimally exposed, by virtue of their living on farms or having limited contact with pesticides through their husbands who engaged in the spraying of pesticides; however, no research to date has evaluated women pesticide exposure in the household and the implications for disease risk profile.Our analysis indicates that women exposure to pesticides is substantial even when they are not directly engaged in pesticide spraying and that the implications of exposure to cancer risk are not negligible.This is because exposure levels through contact with contaminated equipment, clothes, and pesticide dilution at home can be substantial.Further, little attention has been given to other chronic, lifethreatening pathologies such as BC.The gap is particularly relevant because BC is the leading cause of women's death worldwide, a cancer for which most cases result from environmental factors and life habits, with a minority of cases attributable to heritable high-risk genes. 41Evidence addressing BC risk and pesticide exposure for potentially exposed women is not well documented.Indeed, data concerning specific pesticide exposures or pesticide measurement in biological samples from female rural workers are typically unavailable.We show that BC risk for exposed women is 32% higher compared to unexposed ones in the crude analysis.This risk was lower and not significant     when adjusting for age, BMI, and menopause at diagnosis (adjusted OR: 1.30, IC 95 : 0.87−1.95).
Pesticides may pose a substantial risk for BC development.−46 Risk assessment studies about occasional pesticide exposure and BC development have also shown a positive association in multiple populations, with hazard risks reaching more than 4 times in some cases but not others. 7,25,34,43,45,47,48This variability may result from various factors, such as epidemiological studies that analyze the overall population with individuals who are occasionally/accidentally exposed to pesticides versus target populations with high exposure, such as rural workers.The pesticide residues found in the present study were very similar to others identified in surveys of urine samples in populations worldwide.−55 Detected levels are reported in the magnitude of ppb, though they could be conservatively viewed as underestimates of exposure levels due to the short half-life of such substances in the body.These findings reinforce that all sources (food, drinking water, occupational) can lead to pesticide contamination, even though it needs to be clarified how to define the impact of such concentrations on people's health.
Glyphosate is the most used herbicide and is classified by International Agency for Research on Cancer (IARC) as a probable human carcinogen.The presence of the glyphosate metabolite aminomethylphosphonic acid (AMPA) in women's urine samples has been associated with BC development. 56lausible carcinogenic mechanisms for this substance include changes in DNA methylation 57 and alterations in the estrogen pathway 58 that can occur under environmentally relevant concentrations. 592,4-D is classified by IARC as a possible human carcinogen.Its carcinogenic-linked mechanisms include the generation of oxidative stress 61 and endocrine disruption. 62ne study in California showed an increased risk for BC development in 2,4-D-exposed female workers. 60Furthermore, it has been suggested that 2,4-D has potentially genotoxic effects when coexposed with glyphosate. 63Finally, atrazine is classified by IARC as lacking a carcinogenic risk designation for humans.−69 No data regarding female exposure to atrazine was found in the literature.
Investigating environmental factors that enhance metastasis occurrence is of great value to preventive medicine and public health.A major finding of our study is that pesticide-exposed women diagnosed with BC have a 54% higher risk of metastasis than unexposed women with BC.It is relevant because metastases are responsible for most cancer-related deaths. 70he metastatic process is characterized by a series of events triggered by the detachment of cancer cells from the primary tumor, which join the lymphatic and circulatory systems to invade distant organs. 71Interestingly, we also observed that exposed BC patients exhibit more lymph nodal metastasis than unexposed BC patients, and pesticide exposure correlates positively to lymph nodal invasion.These findings reinforce the plausible role of chronic pesticide exposure on disease aggravation.It is essential to highlight that lymph nodal-positive women are submitted to cytotoxic chemotherapy protocols that, despite mitigating BC cells, have deleterious effects on noncancerous cells.It means that BC women undergoing chemotherapy are at risk for developing drug-related toxicities that can have poor outcomes and even death.
All in all, we suggest that BC metastases may result from a series of events triggered by chronic pesticide exposure.While the direct effect of glyphosate or 2,4-D on metastasis is not reported, it has been suggested that atrazine has the potential to promote metastasis in ovarian and liver cells. 72,73However, no data is available for BC.Nevertheless, chronic exposure to pesticide mixtures may collectively affect BC behavior.For example, hormones of the circadian cycle are suggested as keys to promoting the dissemination of circulating tumor cells in BC 74 and are also reported to be dysregulated during pesticide exposure. 75Furthermore, the relationship between pesticide exposure and metastasis might affect specific groups of BC patients and influence their risk for disease recurrence and death.Additional mechanisms linking pesticide exposure to tumor metastasis include depletion of antitumor proteins, such as interleukin 12, and augmented expression of protumoral molecules, such as CLTA-4 and TGF-β1. 76These mechanisms are not reported in unexposed BC women and are suggested as a worse prognosis signature linked to pesticide exposure.
This study has some limitations.A potential selection bias for the study is that the control group in the risk analysis came from the hospital-attended population and not from the general population.Also, the age at diagnosis and the menopausal status could be considered as risk factors for breast cancer.Because of this, they were considered as confounding factors in the multivariate analysis.
For instance, we have not measured all potential pesticides used in the region across all women included in the study.Other pesticides reported as frequent contaminants in Paranástate include mancozeb, diuron, and legacy pollutants, such as DDT and lindane. 33In addition, we cannot access their correlation to clinicopathological features.Furthermore, a longer-term followup and monitoring would be relevant to re-evaluate the positive and negative observations.Eventually, we envision performing a 10-year follow-up in our study population to evaluate the impact of exposure on long-term disease survival and disease risk.Accordingly, no significant differences were observed regarding disease survival and chemoresistance in exposed and unexposed BC women, possibly due to the short-term follow-up of patients (5 years on average).Further, it is difficult to point out which pesticide has a major contribution to BC development and aggressiveness since patients are exposed to mixtures of such substances.
Despite those limitations, our findings caution against dismissing the pesticides studied here as plausible sources of increased BC cancer in women regularly exposed to them.The data suggest that pesticide exposure increases the risk for BC diagnosis and metastasis development in women continuously exposed.The study also highlights the educational challenge of promoting best practices for pesticide handling and equipment/ clothing decontamination in rural populations with family-based agriculture.
In conclusion, this study points to glyphosate, atrazine, and 2,4-D contamination in women exposed to pesticides at home.The relationship between increased BC and metastases risk in the exposed population needs attention and would benefit from long-term monitoring of potentially exposed women populations.Our findings reinforce the concerns regarding the increase in pesticide use over the past few years and the potential risks Environmental Science & Technology they pose for human health, especially when considering that our population study is exposed to a cocktail of pesticides that mainly include glyphosate, atrazine, and 2,4-D.The safe use of some pesticides is a matter of concern, including the widespread contamination of family members who are not working in the field.Long-term monitoring of these populations is urgently needed.Safety assessments for substances such as glyphosate and other widely used pesticides should be ongoing to mitigate reliance on safety profiles, and practices established decades ago that are yet to be revised.

Figure 1 .
Figure 1.Breast cancer epidemiology and pesticide trade in Brazil and Paranástate.Mean breast cancer (BC) cases among the 27 Brazilian States from 2009 to 2018.The black bar highlights Paranástate as the fourth state with higher BC incidence.The dotted line represents the mean of BC cases in Brazil (per million people).The comparative mean percentage incidence (B) and mortality (C) are presented for Brazil, Paranástate, and the ParanaŚ outhwest region from 2016 to 2019 (the period of data collection for this study).(D) Map of Brazil showing the mean pesticide trade for each state during the 2016−2019 interval; Paranástate is the third biggest consumer of pesticides.(E) the 27 municipalities from ParanáSouthwest regions were distributed according to their pesticide trade for 2016−2019; the darker the color, the larger the amount of pesticides used by the municipality.Acre� AC; Alagoas�AL; Amapa�AP; Amazonas�AM; Bahia�BA; Ceara�CE; Distrito Federal�DF; Espi ́rito Santo�ES; Goiaś�GO; Maranhaõ� MA; Mato Grosso�MT; Mato Grosso do Sul�MS; Minas Gerais�MG; Para�PA; Parai ́ba�PB; Parana�PR; Pernambuco�PE; Piaui ́�PI; Roraima�RR; Rondonia�RO; Rio de Janeiro�RJ; Rio Grande do Norte�RN; Rio Grande do Sul�RS; Santa Catarina�SC; Saõ Paulo�SP; Sergipe�SE; Tocantins�TO.

3 .
Pesticide contamination study.A total of 10 farms were chosen to conduct the exposure assessment (contamination) study, as shown on the map.This step was carried out during the peak of pulverization of the three most traded pesticides in the region (glyphosate, atrazine, and 2,4-D).An interview was conducted to obtain their occupational exposure profile to pesticides, and the results are shown in Box A. Urine samples were collected to evaluate pesticide concentration after its pulverization, and the results are shown in Box B. ND = not detected, ppb = parts per billion.
assessment was aimed at understanding if this mode of contact was sufficient to generate contamination by glyphosate, atrazine, and 2,4-D; (iii) A cancer risk analysis, which included 728 women diagnosed with breast cancer or not diagnosed with breast cancer, and who were exposed to pesticides or not exposed to pesticides; and(iv) A clinicopathological characterization of women with breast cancer.

Table 1 .
Individual Characteristics of Breast Cancer Cases and Controls a,b a Proportions are presented for categorical variables with percentages in parentheses.b Means are presented for numerical variables with standard deviations in parentheses.

Table 2 .
Individual Characteristics of Women Included in the Study and Municipatlity Information Concerning Pesticide Trade a aProportions are presented for categorical variables with percentages in parentheses.Means are presented for numerical variables with standard deviations in parentheses.bMunicipalityinformation is averaged and weighted by the number of patients from each municipality.cWilcoxonrank sum test; Pearson's χ 2 test.dWeuse the Yates continuity correction when estimating p-values comparing the proportion of exposed vs unexposed to avoid overestimating the significance of these small proportions.(1).e p < 0.05.

Table 3 .
Counts and Odds Ratios of Breast Cancer by Exposure Cases with no missing responses for age, BMI, and menopause at diagnosis.b Adjusted for age, BMI, and menopause at diagnosis. a

Table 4 .
Counts and Odds Ratios of Breast Cancer Molecular Subtypes by Exposure aORs are calculated as given molecular subtypes vs all other subtypes.b Adjusted for age, BMI, and menopause at diagnosis.

Table 5 .
Clinicopathological Characterization of Patients with Breast Cancer Occupationally Exposed or Not to Pesticides a We use the Yates continuity correction when estimating p-values comparing the proportion of exposed vs unexposed to avoid overestimating the significance of these small proportions (1).c Pearson's Chi-squared test.3. Multivariate analysis of the association between breast cancer and lymph node metastasis with pesticide exposure, adjusting for breast cancer risk factors of age, menopause status, and body mass index (BMI).d Adjusting for breast cancer risk factors of age, menopause status, and body mass index (BMI).e Adjusted for age, BMI, and menopause at diagnosis.f p < 0.05.
a NA = data not available.b