Abstract
Since its registration in 1994 acetochlor has become a commonly used herbicide in the US, yet no epidemiologic study has evaluated its carcinogenicity in humans. We evaluated use of acetochlor and cancer incidence among licensed pesticide applicators in the Agricultural Health Study. In telephone interviews administered 1999-2005, participants provided information on acetochlor use, use of other pesticides, and additional potential confounders. We used Poisson regression to estimate relative risks (RR) and 95% confidence intervals (95% CI) for cancers that occurred from the time of interview through 2011 in Iowa and 2010 in North Carolina. Among 33,484 men, there were 4,026 applicators who used acetochlor and 3,234 incident cancers, with 304 acetochlor-exposed cases. Increased risk of lung cancer was observed among acetochlor users (RR = 1.74; 95% CI: 1.07-2.84) compared to nonusers, and among individuals who reported using acetochlor/atrazine product mixtures (RR = 2.33; 95% CI: 1.30-4.17), compared to nonusers of acetochlor. Colorectal cancer risk was significantly elevated among the highest category of acetochlor users (RR = 1.75; 95% CI: 1.08-2.83) compared to never users. Additionally, borderline significantly increased risk of melanoma (RR = 1.61; 95% CI: 0.98-2.66) and pancreatic cancer (RR = 2.36; 95% CI: 0.98-5.65) were observed among acetochlor users. The associations between acetochlor use and lung cancer, colorectal cancer, melanoma, and pancreatic cancer are suggestive, however the lack of exposure-response trends, small number of exposed cases, and relatively short time between acetochlor use and cancer development, prohibit definitive conclusions.
Keywords: Acetochlor, Agricultural Health Study, Pesticides, Cancer
INTRODUCTION
Acetochlor is a broad leaf herbicide and a member of the chloracetanilide class1, which also includes alachlor, butachlor, propachlor, and metolachlor.1, 2 In 1994, the US Environmental Protection Agency (EPA) approved acetochlor for use on corn, with continued registration dependent upon a reduction in the use of several other corn herbicides of known health concern, including alachlor, metolachlor, atrazine, 2,4-dichlorophenoxyacetic acid (2,4-D), butylate, and S-Ethyl dipropylthiocarbamate (EPTC).3 Re-registration was not contingent upon reduction of use for any particular herbicide, but rather an overall reduction in the use of these six herbicides.3 EPA re-affirmed acetochlor registration for use on field corn and popcorn in 2006.4 Over 32 million pounds of acetochlor were applied during 1997-2001 on approximately 25% of corn crops in the US; 5 similar amounts continue to be applied in recent years.6
The EPA considers acetochlor to have “suggestive evidence of carcinogenic potential” based on increased incidence of certain tumors in laboratory animals;7 other agencies have not reviewed the carcinogenicity of this chemical. Acetochlor is not considered mutagenic or genotoxic at relevant doses,8 however, evidence suggests potential for acetochlor-induced chromosomal aberration at cytotoxic doses.1 In animal studies, lung tumors and histiocytic sarcomas have been reported in mice, while nasal olfactory tumors and thyroid follicular cell tumors have been reported in Sprague-Dawley rats.1 Liver adenomas and carcinomas, renal adenomas and sarcomas, and benign ovarian tumors have been reported in rats and mice, however, generally only at excessively toxic doses or demonstrated weak associations with acetochlor exposure.1 Bioassays of other chloroacetanilide herbicides have overlapping but not identical tumor profiles.1
Epidemiologic research examining the relationship between acetochlor exposure and non-cancer human health outcomes is limited. Acetochlor levels in umbilical cord blood of full-term infants were associated with risk of low birth weight in a pilot study from a Chinese birth cohort.9 Conversely, acetochlor levels in maternal urine during pregnancy were not associated with any adverse birth outcomes in a French birth cohort.10 A study of various pesticides found an inverse association between acetochlor exposure and semen quality, although this relationship was non-significant after excluding men with other risk factors for reduced semen quality.11 Though there have been no epidemiologic studies of acetochlor and cancer risk to date, studies have evaluated other chloroacetanilide chemicals. Previous findings from the Agricultural Health Study (AHS) noted associations between alachlor exposure and lymphohematopoietic cancers,12 and between metolachor exposure and lung cancer.13, 14
Because of the limited information on cancer risk with acetochlor exposure, we examined occupational exposure to acetochlor and subsequent cancer outcomes among pesticide applicators in the Agricultural Health Study (AHS).
MATERIALS & METHODS
Study Population
The AHS cohort has been described elsewhere in detail.15 Briefly, the AHS is a prospective cohort that includes 57,310 licensed pesticide applicators enrolled during 1993-1997 in Iowa and North Carolina. A follow-up interview was conducted during 1999-2003 for private applicators and 2003-2005 for commercial applicators, with 36,342 total respondents. Incident cancer cases were obtained via linkage with Iowa and North Carolina state cancer registries. Cancer site was classified according to the International Classification of Diseases for Oncology (third revision).16 Due to its recent introduction in 1994, we ascertained acetochlor use only at follow-up interview. Thus, we analyzed first primary cancers diagnosed from the date of follow-up interview through last date of study follow-up (December 31, 2011 for Iowa, December 31, 2010 for North Carolina). The study protocol was approved by the institutional review boards of the National Institutes of Health, University of Iowa, and all other contractors in compliance with applicable requirements of the US.
Exposure Assessment
Computer assisted telephone interviews assessed use of acetochlor and other factors, the texts of which are available at http://aghealth.nih.gov/background/questionnaires.html. Participants were asked to list all pesticides used on each crop during the last year they farmed. Additionally, they were probed to specify if they personally mixed, handled, or applied each pesticide, and how many days per year they applied the pesticide. Where trade names were reported, individual active ingredients were abstracted from the product trade names.
Whether or not an applicator reported any application of acetochlor (yes/no) was examined in order to broadly classify personal use. Total years of acetochlor use for each applicator was assumed to be number of years between the year the applicator last farmed (follow-up interview reference year) and the enrollment year; we used 1994 (year of acetochlor registration) for applicators enrolled prior to this date. The number of lifetime days of acetochlor use was calculated by multiplying the number of days per year using acetochlor by total years of use. Intensity-weighted lifetime exposure days were computed by multiplying an intensity weighting factor by lifetime exposure days. The intensity weighting factor is a function of pesticide handling practices and personal protective equipment used.17 Exposure days and intensity-weighted exposure days were classified as low (less than the median among all exposed cancer cases) or high (greater than or equal to the median among all exposed cancer cases). Additionally, because acetochlor is frequently applied as a mixture with atrazine (53% of acetochlor applicators reported use of such a mixture), we classified use as 1) never applied acetochlor or atrazine, 2) applied acetochlor but not atrazine, 3) applied atrazine but not acetochlor, or 4) applied both acetochlor and atrazine. Application of both acetochlor and atrazine were further separated into those who reported use of acetochlor/atrazine mixtures (based on trade names of product mixtures including acetochlor and atrazine) and those who reported use of acetochlor and atrazine separately.
Statistical Analysis
Analyses were limited to individuals who responded to the follow-up interview (n = 36,342), and we further restricted to male pesticide applicators due to the small number of female applicators (n = 1006) and few exposed cancer cases (n = 4). We excluded men with cancer diagnoses prior to completion of the follow-up interview (n = 1675) and men with missing or zero person-years of follow-up (n = 177), leaving 33,484 applicators available for analysis. Analyses examining lifetime days and intensity-weighted days of acetochlor use further excluded those missing days of use (n = 456) or intensity (n = 11), leaving 33,028 and 33,017 applicators, respectively.
We report results for all cancer sites with at least 15 exposed cases for days and intensity-weighted days of use, and results for ever use for sites with more than five exposed cases. Colon and rectal tumors were combined as were lymphohematopoietic cancers. Aggressive prostate cancer, as defined by Koutros et al, was examined separately.18 Relative risks (RR) and 95% confidence intervals (CIs) were estimated using Poisson regression in SAS version 9.3 (SAS Institute, Inc., Cary, NC). Subjects contributed person-time from date of follow-up interview through date of first cancer diagnosis, date moved out of state, date of death, or last follow-up, whichever occurred first.
All models were adjusted for potential confounders including age at follow-up interview (continuous), state of residence (Iowa or North Carolina), applicator type (private or commercial), duration of cigarette smoking as reported at enrollment (never, <10 years, 10-19 years, 20-29 years, 30+ years, missing), race (white, other, missing), alcohol use (yes/no in preceding 12 months, missing), educational attainment (less than a high school degree, high school degree, more than a high school degree, missing), use of an enclosed tractor cab (yes, no, missing), and family history of cancer in first degree relatives (yes, no, missing; specific cancer site where available). We also adjusted for ever-use of pesticides most highly associated with acetochlor use (lifetime use: imazethapyr, dicamba, atrazine, EPTC, cyanazine, pendimethalin, trifluralin; follow-up only: flumetsulam, clopyralid, isoxaflutole), pesticides mentioned in the acetochlor EPA registration (alachlor, metolachlor, atrazine, EPTC, 2,4-D, butylate), and pesticides previously found to be associated with specific cancer outcomes in the AHS.19 Tests for trend used the midpoint of each exposure category treated as a continuous variable. Sensitivity analyses were conducted in which the cohort was restricted to cases diagnosed more than two years after completion of follow-up interview. All tests were two-sided with α = 0.05. Data in this analysis are based on the AHS data releases P2REL201209.00 and P1REL201209.00.
RESULTS
Table 1 displays selected demographic, behavioral, health, and farming characteristics of the cohort, stratified by acetochlor use. Acetochlor was used almost exclusively by applicators in Iowa (99%). A larger proportion of commercial applicators used acetochlor compared to private applicators (28.9% and 9.3%, respectively). Applicators who used acetochlor were more likely to be younger, more educated, and white compared to non-users. Applicators in Iowa similarly tend to be younger, more educated, and white compared to applicators in North Carolina, so these differences may be regional and not directly related to acetochlor use. Additionally, acetochlor users were more likely than non-users to be never-smokers and ever-drinkers in the last 12 months, have no family history of cancer, and report that the tractor regularly used to apply pesticides had an enclosed cab.
Table 1.
Selected characteristics of AHS pesticide applicators1 who completed follow-up questionnaire (n=33,017), stratified by lifetime intensity-weighted days of acetochlor use (none, low, high)2
| Acetochlor use |
|||||
|---|---|---|---|---|---|
| Total N=33,017 N (%) |
Never N=29,458 N (%) |
Low Use N= 1,621 N (%) |
High Use N=1,938 N (%) |
p-value3 | |
| State | |||||
| Iowa | 21810 (66.06) | 18297 (62.11) | 1601 (98.77) | 1912 (98.66) | p < 0.0001 |
| North Carolina | 11207 (33.94) | 11161 (37.89) | 20 (1.23) | 26 (1.34) | |
| Type of applicator | |||||
| Commercial | 2527 (7.65) | 1797 (6.10) | 141 (8.70) | 589 (30.39) | p < 0.0001 |
| Private | 30490 (92.35) | 27661 (93.9) | 1480 (91.30) | 1349 (69.61) | |
| Age at follow-up | |||||
| Quartile 1: <43.4 | 8319 (25.20) | 7058 (23.96) | 543 (33.50) | 718 (37.05) | p < 0.0001 |
| Quartile 2: 43.4-51.2 | 8260 (25.02) | 7234 (24.56) | 430 (26.53) | 596 (30.75) | |
| Quartile 3: 51.3-61.4 | 8227 (24.92) | 7431 (25.23) | 362 (22.33) | 434 (22.39) | |
| Quartile 4: ≥61.5 | 8211 (24.87) | 7735 (26.26) | 286 (17.64) | 190 (9.80) | |
| Educational Attainment 4 | |||||
| < High School | 2624 (7.95) | 2532 (8.60) | 46 (2.84) | 46 (2.37) | p < 0.0001 |
| High School Graduate | 14774 (44.75) | 13168 (44.70) | 734 (45.28) | 872 (44.99) | |
| > High School | 14383 (43.56) | 12596 (42.76) | 813 (50.15) | 974 (50.26) | |
| Other/Missing | 1236 (3.74) | 1162 (3.94) | 28 (1.73) | 46 (2.37) | |
| Race | |||||
| White | 31685 (95.97) | 28166 (95.61) | 1602 (98.83) | 1917 (98.92) | p < 0.0001 |
| Non-white | 728 (2.20) | 718 (2.44) | 7 (0.43) | 3 (0.15) | |
| Other/Missing | 604 (1.83) | 574 (1.95) | 12 (0.74) | 18 (0.93) | |
| Tobacco use 4 | |||||
| Never Smoker | 17302 (52.40) | 15069 (51.15) | 1015 (62.62) | 1218 (62.85) | p < 0.0001 |
| Former Smoker, ≤10 yrs | 4896 (14.83) | 4380 (14.87) | 255 (15.73) | 261 (13.47) | |
| Former Smoker, >10 yrs | 4548 (13.77) | 4225 (14.34) | 159 (9.81) | 164 (8.46) | |
| Current Smoker, ≤10 yrs | 917 (2.78) | 813 (2.76) | 41 (2.53) | 63 (3.25) | |
| Current Smoker, >10 yrs | 3885 (11.77) | 3574 (12.13) | 114 (7.03) | 197 (10.17) | |
| Missing | 1469 (4.45) | 1397 (4.74) | 37 (2.28) | 35 (1.81) | |
| Alcohol consumption | |||||
| No | 17702 (53.61) | 16291 (55.30) | 711 (43.86) | 700 (36.12) | p < 0.0001 |
| Yes | 15025 (45.51) | 12886 (43.74) | 905 (55.83) | 1234 (63.67) | |
| Missing | 290 (0.88) | 281 (0.95) | 5 (0.31) | 4 (0.21) | |
| Family History of Cancer 4 | |||||
| No | 17390 (52.67) | 15347 (52.10) | 876 (54.04) | 1167 (60.22) | p < 0.0001 |
| Yes | 12971 (39.29) | 11579 (39.31) | 678 (41.83) | 714 (36.84) | |
| Missing | 2656 (8.04) | 2532 (8.60) | 67 (4.13) | 57 (2.94) | |
| Use of Enclosed Cab | |||||
| No | 20063 (60.77) | 19048 (64.66) | 434 (26.77) | 581 (29.98) | p < 0.0001 |
| Yes | 12954 (39.23) | 10410 (35.34) | 1187 (73.23) | 1357 (70.02) | |
| Atrazine use | |||||
| No | 22579 (68.39) | 21932 (74.45) | 329 (20.30) | 318 (16.41) | p < 0.0001 |
| Yes | 10438 (31.61) | 7526 (25.55) | 1292 (79.7) | 1620 (83.59) | |
Reported on follow-up questionnaire
Low: 72-879 intensity-weighted days, High: 880- 47520 intensity-weighted days
Chi-square test for homogeneity
Reported on enrollment questionnaire
We evaluated the use of acetochlor in association with a variety of cancer sites controlling the covariates described above (Table 2). We observed an association with ever-use of acetochlor and increased risk of lung cancer (RR = 1.74; 95% CI: 1.07-2.84), melanoma (RR = 1.61; 95% CI: 0.98-2.66), and pancreatic cancer (RR = 2.36; 95% CI: 0.98-5.65), although the associations for melanoma and pancreatic cancer were of borderline statistical significance.
Table 2.
Relative risks (RR) and 95% confidence intervals (95% CI)1 for ever-use of acetochlor, relative to never-use, and risk for various cancers among AHS applicators (n=33,484).
| Nunexposed | Nexposed | RR (95% CI) | |
|---|---|---|---|
| All Sites | 2930 | 304 | 1.06 (0.94-1.21) |
| Bladder | 167 | 15 | 1.08 (0.61-1.91) |
| Lymphohematopoietic | 293 | 23 | 0.77 (0.49-1.20) |
| Colon | 206 | 21 | 1.07 (0.66-1.73) |
| Colorectal | 300 | 31 | 1.03 (0.69-1.53) |
| Esophagus | 42 | 6 | 1.28 (0.47-3.44) |
| Kidney | 98 | 13 | 1.14 (0.61-2.11) |
| Leukemia | 93 | 9 | 0.89 (0.43-1.84) |
| Lung | 251 | 23 | 1.74 (1.07-2.84)** |
| Melanoma | 124 | 23 | 1.61 (0.98-2.66)* |
| NHL | 249 | 18 | 0.70 (0.43-1.16) |
| Pancreas | 55 | 7 | 2.36 (0.98-5.65)* |
| Prostate | 1232 | 130 | 0.99 (0.82-1.20) |
| Prostate, Aggressive | 644 | 69 | 0.98 (0.75-1.28) |
| Rectum | 94 | 10 | 0.97 (0.48-1.95) |
Significant at p<0.1;
Significant at p<0.05
Adjusted for age, race, state, applicator type, smoking, family history of cancer, alcohol consumption, BMI, use of an enclosed cab, education, and correlated /associated pesticide use
Compared to never use of acetochlor, low use of acetochlor, as measured by lifetime days used, was associated with a decreased risk of colorectal cancer (RR = 0.31; 95% CI: 0.11-0.83) while high use was associated with an increased risk of colorectal cancer (RR = 1.75; 95% CI: 1.08-2.83), with a borderline significant exposure-response trend (p-trend = 0.07, Table 3). Results were similar for colorectal cancer risk and lifetime intensity-weighted days of acetochlor use. Low use of acetochlor, as measured by lifetime days used, was associated with an increased risk of lung cancer (RR = 2.64; 95% CI: 1.47-4.74) compared to never use; no association was seen for high use and no exposure-response trend was observed (p-trend = 0.36). Similarly, low use of acetochlor, as measured by lifetime intensity-weighted days used, was associated with an increased risk of lung cancer (RR = 2.26; 95% CI: 1.24-4.14, p-trend = 0.16). High use of acetochlor, as measured by lifetime days used, was associated with borderline increased risk of melanoma (RR = 1.78; 95% CI: 0.90-3.52, p-trend = 0.08) compared to never use. However, when measured using lifetime intensity-weighted days used, low acetochlor use was associated with borderline increased risk of melanoma (RR = 1.79; 95% CI: 0.93-3.45, p-trend = 0.29).
Table 3.
Relative risks (RR) and 95% confidence intervals (95% CI)1 by days (n= 33,028) and intensity-weighted days (n= 33,017) of acetochlor use, compared to never-use, for various cancers.
| Acetochlor days use2 |
Acetochlor intensity-weighted days use3 |
||||
|---|---|---|---|---|---|
| N | RR (95% CI) | N | RR (95% CI) | ||
| All Sites | Never | 2930 | 1.00 (REF) | 2930 | 1.00 (REF) |
| Low | 119 | 0.99 (0.82-1.20) | 127 | 1.00 (0.83-1.20) | |
| High | 139 | 1.19 (1.00-1.43)* | 131 | 1.20 (1.00-1.45)* | |
| p-trend | 0.07* | 0.06* | |||
| Bladder | Never | 167 | 1.00 (REF) | 167 | 1.00 (REF) |
| Low | 7 | 1.11 (0.51-2.44) | 6 | 0.91 (0.39-2.10) | |
| High | 4 | 0.74 (0.26-2.09) | 5 | 1.01 (0.40-2.57) | |
| p-trend | 0.63 | 0.98 | |||
| Lymphohematopoietic | Never | 293 | 1.00 (REF) | 293 | 1.00 (REF) |
| Low | 12 | 0.93 (0.51-1.68) | 10 | 0.76 (0.40-1.44) | |
| High | 9 | 0.80 (0.40-1.59) | 11 | 1.02 (0.55-1.91) | |
| p-trend | 0.51 | 0.91 | |||
| Colon | Never | 206 | 1.00 (REF) | 206 | 1.00 (REF) |
| Low | 4 | 0.46 (0.17-1.27) | 4 | 0.44 (0.16-1.20) | |
| High | 12 | 1.56 (0.83-2.91) | 12 | 1.67 (0.90-3.11) | |
| p-trend | 0.28 | 0.20 | |||
| Colorectal | Never | 300 | 1.00 (REF) | 300 | 1.00 (REF) |
| Low | 4 | 0.31 (0.11-0.83)** | 7 | 0.51 (0.24-1.09)* | |
| High | 21 | 1.75 (1.08-2.83)** | 18 | 1.59 (0.95-2.64)* | |
| p-trend | 0.07* | 0.16 | |||
| Kidney | Never | 98 | 1.00 (REF) | 98 | 1.00 (REF) |
| Low | 4 | 0.81 (0.29-2.25) | 3 | 0.57 (0.18-1.86) | |
| High | 7 | 1.53 (0.67-3.47) | 8 | 1.87 (0.86-4.03) | |
| p-trend | 0.36 | 0.16 | |||
| Lung | Never | 251 | 1.00 (REF) | 251 | 1.00 (REF) |
| Low | 14 | 2.64 (1.47-4.74)** | 13 | 2.26 (1.24-4.14)** | |
| High | 7 | 1.14 (0.51-2.57) | 8 | 1.47 (0.68-3.16) | |
| p-trend | 0.36 | 0.16 | |||
| Melanoma | Never | 124 | 1.00 (REF) | 124 | 1.00 (REF) |
| Low | 8 | 1.41 (0.67-2.97) | 11 | 1.79 (0.93-3.45)* | |
| High | 11 | 1.78 (0.90-3.52)* | 8 | 1.38 (0.64-2.99) | |
| p-trend | 0.08* | 0.29 | |||
| NHL | Never | 249 | 1.00 (REF) | 249 | 1.00 (REF) |
| Low | 11 | 1.01 (0.54-1.87) | 10 | 0.89 (0.47-1.71) | |
| High | 5 | 0.50 (0.20-1.25) | 6 | 0.63 (0.28-1.46) | |
| p-trend | 0.16 | 0.27 | |||
| Prostate | Never | 1232 | 1.00 (REF) | 1232 | 1.00 (REF) |
| Low | 52 | 0.95 (0.72-1.27) | 57 | 0.99 (0.75-1.30) | |
| High | 59 | 1.10 (0.84-1.46) | 54 | 1.08 (0.81-1.44) | |
| p-trend | 0.54 | 0.63 | |||
| Prostate, Aggressive | Never Low |
644 27 |
1.00 (REF) 0.92 (0.62-1.36) |
644 30 |
1.00 (REF) 0.97 (0.67-1.42) |
| High | 33 | 1.20 (0.83-1.74) | 30 | 1.16 (0.78-1.70) | |
| p-trend | 0.40 | 0.50 | |||
Significant at p<0.1
Significant at p<0.05
Adjusted for age, race, state, applicator type, smoking, family history of cancer, alcohol consumption, BMI, use of an enclosed cab, education, and correlated /associated pesticide use
Low: 2-19 days, High: 20-1080 days
Low: 72-879 intensity-weighted days, High: 880- 47,520 intensity-weighted days
Over 80% of acetochlor users also reported use of atrazine at follow-up interview (Table 1); 53% of applicators reporting ever use of acetochlor report using it in a product also containing atrazine. We found that applicators using both atrazine and acetochlor were at increased risk of lung cancer (RR = 2.01; 95% CI: 1.17-3.46) compared to those with no use of either (Table 4). This association was similarly strong for persons using acetochlor/atrazine product mixtures (RR = 2.33; 95% CI: 1.30-4.17), compared to use of neither atrazine nor acetochlor. Among those exposed to acetochlor as a mixture only, we saw elevated risks among both low (RR = 2.71; 95% CI: 1.14-5.78) and high (RR = 1.96; 95% CI: 0.75-5.11) categories of intensity-weighted days of exposure (results not shown). We also observed a borderline significant excess risk of pancreatic cancer among users of atrazine/acetochlor product mixtures (RR = 2.62; 95% CI: 0.95-7.20).
Table 4.
Relative risk (RR) and 95% confidence intervals (95% CI)1 for ever applying acetochlor and atrazine alone or as a mixture, compared to never use, for selected cancers.
| N | RR (95% CI) | ||
|---|---|---|---|
| All Sites | Neither acetochlor or atrazine | 2282 | 1.00 (REF) |
| Acetochlor | 59 | 0.99 (0.76-1.29) | |
| Atrazine | 648 | 1.07 (0.97-1.17) | |
| Both acetochlor and atrazine | 245 | 1.12 (0.97-1.29) | |
| Applied separately | 83 | 1.20 (0.95-1.50) | |
| Applied as a mixture | 162 | 1.08 (0.91-1.28) | |
| Colorectal | Neither acetochlor or atrazine | 240 | 1.00 (REF) |
| Acetochlor | 6 | 0.87 (0.38-1.99) | |
| Atrazine | 60 | 0.91 (0.67-1.23) | |
| Both acetochlor and atrazine | 25 | 1.03 (0.66-1.61) | |
| Applied separately | 8 | 1.01 (0.49-2.11) | |
| Applied as a mixture | 17 | 1.04 (0.62-1.75) | |
| Lung | Neither acetochlor or atrazine | 204 | 1.00 (REF) |
| Acetochlor | 4 | 1.96 (0.70-5.45) | |
| Atrazine | 47 | 1.32 (0.93-1.88) | |
| Both acetochlor and atrazine | 19 | 2.01 (1.17-3.46)** | |
| Applied separately | 4 | 1.29 (0.45-3.68) | |
| Applied as a mixture | 15 | 2.33 (1.30-4.17)** | |
| Melanoma | Neither acetochlor or atrazine | 89 | 1.00 (REF) |
| Acetochlor | 6 | 2.24 (0.94-5.33)* | |
| Atrazine | 35 | 1.39 (0.91-2.14) | |
| Both acetochlor and atrazine | 17 | 1.75 (0.97-3.14)* | |
| Applied separately | 6 | 1.87 (0.77-4.51) | |
| Applied as a mixture | 11 | 1.69 (0.86-3.34) | |
| Pancreas | Neither acetochlor or atrazine | 49 | 1.00 (REF) |
| Acetochlor | 2 | 2.62 (0.6-11.44) | |
| Atrazine | 6 | 0.63 (0.26-1.55) | |
| Both acetochlor and atrazine | 5 | 1.84 (0.67-5.08) | |
| Applied separately | 0 | -- | |
| Applied as a mixture | 5 | 2.62 (0.95-7.20)* |
Significant at p<0.1;
Significant at p<0.05
Adjusted for age, race, state, applicator type, smoking, family history of cancer, alcohol consumption, BMI, use of an enclosed cab, education, and correlated /associated pesticide use
The findings for lung cancer, colorectal cancer, melanoma, and pancreatic cancer were similar in magnitude and direction after restricting the study population to those diagnosed with cancer at least two years after completion of the follow-up interview, though the risk estimates for ever use of lung and pancreatic cancer no longer reached statistical significance. We limited lung and pancreatic cancer analyses to never smokers to examine the results independent of smoking status. The results were in a similar direction and magnitude for low acetochlor exposure (days of use) and risk of lung cancer (RR = 3.30; 95% CI: 0.89-12.27), though the association was only borderline statistically significant. It is possible that the small number of exposed cases (n = 3) caused the estimates to be unstable. We saw a stronger association for ever use of acetochlor and risk of pancreatic cancer when restricting to never smokers (RR = 5.61; 95% CI: 1.58-19.97). We also restricted our analyses to Iowa applicators, as the majority of acetochlor was applied there (98.8%); the results were overall very similar. Approximately half of our study population had complete information regarding potential confounders for colorectal cancer (physical activity, fruit and vegetable consumption) and melanoma (sun sensitivity and sun protection). When we limited our analyses to these men, the relationship for high days of acetochlor use and colorectal cancer was similar in magnitude and direction (RR = 1.71; 95% CI: 0.88-3.33) after controlling for physical activity and dietary factors, and the relationship between ever use of acetochlor and melanoma risk was strengthened (RR = 2.55; 95% CI: 1.45-4.48) after controlling for sun sensitivity and exposure factors. We also examined days and intensity-weighted days of acetochlor use per year, classified as high or low based on the median among cancer cases, in order to determine if number of years between enrollment and follow-up was skewing our results. The results were consistent with findings for days and intensity-weighted days, and we additionally observed a significant exposure-response trend for intensity-weighted days per year of acetochlor use and lung cancer risk (p-trend = 0.05).
DISCUSSION
In this analysis of reported use of acetochlor among AHS applicators, we found elevated risk for several cancer sites. To our knowledge, this is the first epidemiologic study attempting to evaluate cancer risk from occupational use of acetochlor. Specifically, we observed an increased risk of colorectal cancer among pesticide applicators with high lifetime use of acetochlor, and increased risk of lung cancer among ever users and low lifetime users of acetochlor. Additionally, ever use and increasing days of acetochlor use were associated with borderline significantly increased risk of melanoma. Though the number of exposed cases was small, there was a borderline significant increased risk of pancreatic cancer among ever users of acetochlor.
The finding that use of acetochlor is associated with increased risk of lung cancer is complicated because it was observed among those with below median use, but not among those with higher exposure. However, this finding was robust in sensitivity analyses restricting the cohort to persons diagnosed two or five years after interview. This suggests that reporting of acetochlor use was not likely to be influenced by early disease. A previous AHS analysis found that high use of metolachlor, another chloracetanilide herbicide, was associated with an increased risk of lung cancer.14 The observed association in our study did not appear to be related to previous metolachlor use, as there was very little correlation between metolachlor and acetochlor use (ρ = 0.11). Concerned about the possibility of residual confounding, we adjusted for smoking using several metrics, including duration (years smoked), total cigarette exposure (pack-years), and current/former/never smoking status; all produced similar results and duration was chosen based on model fit. We saw results of a similar direction and magnitude for low acetochlor exposure and risk of lung cancer when we restricted our sample to never smokers. Because we did not see an excess of lung cancer among the more highly exposed individuals and there was no significant exposure-response trend, the association may be spurious. Though we were underpowered to do so, future studies should attempt to examine lung cancer risk by histologic subtype; we noted slight differences in lung cancer histologic subtype by acetochlor exposure in descriptive analyses.
Colorectal cancer is not often associated with occupational exposures;20 however, we observed an increased risk between reported acetochlor use and risk of colorectal cancer, with some indication of an exposure-response trend. In the AHS, several pesticides have been associated with colon and rectal cancers; aldicarb, dicamba, EPTC, imazethapyr, and trifluralin were associated with colon cancer, and chlordane, chlorpyrifos, pendimethalin, and toxaphene were associated with rectal cancer.19 A small study of manufacturing workers exposed to alachlor, a chemical structurally related to acetochlor, found elevated risks of colorectal cancer.21 A limitation in our study is that while we could control for body mass index, an established risk factor for colorectal cancer,22 we were not able to control for other known risk factors such as dietary habits or physical activity,20 which were only collected for a portion of the cohort. When restricted to individuals who provided information on leisure time physical activity and average fruit and vegetable consumption, our results remained relatively unchanged.
We observed an association between melanoma and ever use of acetochlor, though the exposure-response relationship was not consistent for days and intensity-weighted days of use. The relationship between pesticides and melanoma is complicated by other occupational exposures, particularly extended exposure to natural ultraviolet radiation,23 however several pesticides have been associated with melanoma in previous AHS analyses.24, 25 Sun sensitivity and exposure characteristics were only available for about half of our cohort; however when we limited our analyses to men with valid response and controlled for these factors, the relationship between ever use of acetochlor and melanoma risk was strengthened.
Occupationally, pancreatic cancer has been associated with chlorinated hydrocarbon solvents,26 though findings for chlorinated pesticides and pancreatic cancer have been less clear.27-30 In the AHS, EPTC and pendimethalin were associated with increased risk of pancreatic cancer, possibly via the formation of N-nitroso-compounds.31 Smoking, an established risk factor for pancreatic cancer, was controlled for in our analyses and we observed a significantly increased risk of pancreatic cancer among never smokers. While we did not present the results due to small number of exposed cases, we did see a significant exposure-response trend for intensity-weighted days of acetochlor use and pancreatic cancer (p-trend = 0.01). Future studies should evaluate these findings.
We evaluated atrazine and acetochlor product mixtures because these active ingredients are often applied as a mixture in our cohort. Previous studies in the AHS have not been able to explicitly examine product mixtures because of how pesticide use was assessed at enrollment.14, 32, 33 At follow-up interview, participants self-reported specific pesticide product names, and based on this information we determined if acetochlor was being applied alone or as a mixture with atrazine. These analyses did not account for applicators who mixed individual ingredients or products on their own (as opposed to using pre-mixed formulations); this information was not collected. Little is known about the toxicology and potential carcinogenicity of pesticide mixtures. Epidemiologic studies usually examine pesticides either independently, or more often by chemical class. Findings here suggest that future epidemiologic studies should consider the effects of pesticide mixtures, particularly acetochlor and atrazine, and future toxicological studies should attempt to understand whether exposure as a mixture influences genotoxicity and mutagenicity.
Acetochlor, like most pesticides licensed for use today, does not appear to be mutagenic or genotoxic in laboratory studies.1 Therefore, other mechanisms of carcinogenesis should be considered. Occupational use of other chloracetanilide herbicides alachlor and metolachlor was significantly associated with shortened telomere length in a recent AHS study,34 acetochlor was not examined. Telomere length has been implicated in cancer etiology.35 Another proposed mechanism of carcinogenicity for chloracetanilide herbicides involves bioactivation through several steps to DNA-reactive metabolites (2-methyl-6-ethylbenzoquinone imine for acetochlor and metolachlor and 2,6-diethylbenzoquinone imine for alachlor).7 These chloracetanilide metabolites, 2-methyl-6-ethylbenzoquinone imine and 2,6-diethylbenzoquinone imine, may lead to formation of DNA adducts or single-strand DNA breaks.7, 36, 37 Studies suggest that triazine herbicides, such as atrazine, stimulate cytochrome P450 (CYP450) system activity.38 Acetochlor is metabolized and potentially bioactivated by members of the CYP450 family of enzymes, CYP2B6 and CYP34A, in the liver.7 Thus, when atrazine and acetochlor are applied together, it is possible that there is increased acetochlor bioactivation as a result of atrazine-induced CYP450 activity.
Our analysis is the first epidemiologic study to examine the effect of acetochlor on cancer outcomes in humans. Strengths include longitudinal study design with regular follow-up of participants for cancer and mortality outcomes, as well as detailed and validated self-reported pesticide exposure and intensity information.39 An important limitation for our analysis was sample size; although AHS is a large cohort, only about 12% (n = 4026) of our sample reported any acetochlor use. Additionally because acetochlor was introduced after AHS enrollment and therefore assessed only in follow-up interviews, the follow-up period is shorter than analyses of other chemicals, and thus fewer cancer cases were accrued. Because acetochlor was registered for use relatively recently, the observed associations for acetochlor and solid tumors may be due to earlier exposures and reflect long latency periods for these cancers. However, we restricted our cohort to those diagnosed at least two and five years after the start date of the study, and did not see marked differences in the magnitude and direction of the effects for lung cancer, colorectal cancer, pancreatic cancer, and melanoma. Applicators in our cohort were typically applying other pesticides prior to acetochlor registration and concurrently at follow-up interview. We controlled for use of pesticides that were highly correlated with acetochlor, as well as herbicides that were specified in the acetochlor EPA registration, in order to minimize these possible sources of confounding. Due to limited numbers of cases, we were unable to evaluate rarer cancer sites, as well as histologic subtypes of more common cancers. Due to few exposed women, we were unable to evaluate cancer sites more common in females such as breast or thyroid cancer.
In conclusion, we observed associations between acetochlor use and lung cancer, colorectal cancer, pancreatic cancer, and melanoma. However, due to minimal evidence for exposure-response relationships, the observed associations could reflect chance findings. Findings regarding the carcinogenicity of acetochlor need further evaluation in populations with sufficient numbers of exposed subjects. Additionally, future studies should attempt to evaluate potential mechanisms of carcinogenesis for acetochlor and other chloroacetanilide herbicides by examining genetic markers of susceptibility, as well as markers of cellular and DNA damage.
Novelty and Impact.
This study is the first to examine the relationship between occupational exposure to acetochlor, a commonly used herbicide registered for use on corn, and human cancer risk in a large prospective cohort. Acetochlor use was associated with increased risk of lung cancer, colorectal cancer, pancreatic cancer, and melanoma. Use of mixtures of acetochlor and atrazine, another widely used herbicide often applied concurrently with acetochlor, was associated with increased risk of lung cancer.
ACKNOWLEDGEMENTS
This work was supported by the intramural research program of the National Institutes of Health, the National Cancer Institute (Z01-CP010119), and the National Institute of Environmental Health Sciences (Z01-ES049030).
Abbreviations
- EPA
Environmental Protection Agency
- 2,4-D
2,4-dichlorophenoxyacetic acid
- EPTC
S-Ethyl dipropylthiocarbamate
- AHS
Agricultural Health Study
- DNA
Deoxyribonucleic acid
- CYP450
Cytochrome P450
- NHL
Non-Hodgkin Lymphoma
- BMI
Body Mass Index
Footnotes
DISCLAIMER
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health.
REFERENCES
- 1.Evaluation of the Carcinogenic Potential of Acetochlor (Fifth Assessment) (121601) Environmental Protection Agency, Cancer Assessment Review Committee, Health Effects Division, Office of Pesticide Programs; 2007. pp. 1–66. [Google Scholar]
- 2.Green T, Lee R, Moore RB, Ashby J, Willis GA, Lund VJ, Clapp MJ. Acetochlor-induced rat nasal tumors: further studies on the mode of action and relevance to humans. Regul Toxicol Pharmacol. 2000;32:127–33. doi: 10.1006/rtph.2000.1413. [DOI] [PubMed] [Google Scholar]
- 3.Acetochlor (EPA-HQ-OPP-2007-0725-0002) Environmental Protection Agency; 2007. [Google Scholar]
- 4.Acetochlor; Tolerance Reassessment Decision (EPA-HQ-OPP-2005-0227) Vol. 71. Environmental Protection Agency; 2006. pp. 15730–1. [Google Scholar]
- 5.Usage Report in Support of Registration for the Herbicide Acetochlor (121601) Environmental Protection Agency, Office of Prevention, Pesticides, and Toxic Substances; Washington, DC: 2004. [Google Scholar]
- 6.Pesticide Industry Sales and Usage: 2006 and 2007. Market Estimates Office of Pesticide Programs, US Environmental Protection Agency; Washington, DC: 2011. [Google Scholar]
- 7.Coleman S, Linderman R, Hodgson E, Rose RL. Comparative metabolism of chloroacetamide herbicides and selected metabolites in human and rat liver microsomes. Environ Health Perspect. 2000;108:1151–7. doi: 10.1289/ehp.001081151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ashby J, Kier L, Wilson AG, Green T, Lefevre PA, Tinwell H, Willis GA, Heydens WF, Clapp MJ. Evaluation of the potential carcinogenicity and genetic toxicity to humans of the herbicide acetochlor. Hum Exp Toxicol. 1996;15:702–35. doi: 10.1177/096032719601500902. [DOI] [PubMed] [Google Scholar]
- 9.Wickerham EL, Lozoff B, Shao J, Kaciroti N, Xia Y, Meeker JD. Reduced birth weight in relation to pesticide mixtures detected in cord blood of full-term infants. Environ Int. 2012;47:80–5. doi: 10.1016/j.envint.2012.06.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Chevrier C, Limon G, Monfort C, Rouget F, Garlantezec R, Petit C, Durand G, Cordier S. Urinary biomarkers of prenatal atrazine exposure and adverse birth outcomes in the PELAGIE birth cohort. Environ Health Perspect. 2011;119:1034–41. doi: 10.1289/ehp.1002775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Swan SH, Kruse RL, Liu F, Barr DB, Drobnis EZ, Redmon JB, Wang C, Brazil C, Overstreet JW. Semen quality in relation to biomarkers of pesticide exposure. Environ Health Perspect. 2003;111:1478–84. doi: 10.1289/ehp.6417. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lee WJ, Hoppin JA, Blair A, Lubin JH, Dosemeci M, Sandler DP, Alavanja MC. Cancer incidence among pesticide applicators exposed to alachlor in the Agricultural Health Study. Am J Epidemiol. 2004;159:373–80. doi: 10.1093/aje/kwh040. [DOI] [PubMed] [Google Scholar]
- 13.Rusiecki JA, Hou L, Lee WJ, Blair A, Dosemeci M, Lubin JH, Bonner M, Samanic C, Hoppin JA, Sandler DP, Alavanja MC. Cancer incidence among pesticide applicators exposed to metolachlor in the Agricultural Health Study. Int J Cancer. 2006;118:3118–23. doi: 10.1002/ijc.21758. [DOI] [PubMed] [Google Scholar]
- 14.Alavanja MC, Dosemeci M, Samanic C, Lubin J, Lynch CF, Knott C, Barker J, Hoppin JA, Sandler DP, Coble J, Thomas K, Blair A. Pesticides and lung cancer risk in the agricultural health study cohort. Am J Epidemiol. 2004;160:876–85. doi: 10.1093/aje/kwh290. [DOI] [PubMed] [Google Scholar]
- 15.Alavanja MC, Sandler DP, McMaster SB, Zahm SH, McDonnell CJ, Lynch CF, Pennybacker M, Rothman N, Dosemeci M, Bond AE, Blair A. The Agricultural Health Study. Environ Health Perspect. 1996;104:362–9. doi: 10.1289/ehp.96104362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Fritz AG. International Classification of Diseases for Oncology: ICD-O. World Health Organization; 2000. [Google Scholar]
- 17.Coble J, Thomas KW, Hines CJ, Hoppin JA, Dosemeci M, Curwin B, Lubin JH, Beane Freeman LE, Blair A, Sandler DP, Alavanja MC. An updated algorithm for estimation of pesticide exposure intensity in the agricultural health study. Int J Environ Res Public Health. 2011;8:4608–22. doi: 10.3390/ijerph8124608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Koutros S, Beane Freeman LE, Lubin JH, Heltshe SL, Andreotti G, Barry KH, DellaValle CT, Hoppin JA, Sandler DP, Lynch CF, Blair A, Alavanja MC. Risk of total and aggressive prostate cancer and pesticide use in the Agricultural Health Study. Am J Epidemiol. 2013;177:59–74. doi: 10.1093/aje/kws225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Weichenthal S, Moase C, Chan P. A review of pesticide exposure and cancer incidence in the Agricultural Health Study cohort. Environ Health Perspect. 2010;118:1117–25. doi: 10.1289/ehp.0901731. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Schottenfeld D, Fraumeni JF. Cancer Epidemiology and Prevention. Oxford University Press; USA: 2006. [Google Scholar]
- 21.Acquavella JF, Delzell E, Cheng H, Lynch CF, Johnson G. Mortality and cancer incidence among alachlor manufacturing workers 1968-99. Occup Environ Med. 2004;61:680–5. doi: 10.1136/oem.2003.010934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA: A Cancer Journal for Clinicians. 2014;64:9–29. doi: 10.3322/caac.21208. [DOI] [PubMed] [Google Scholar]
- 23.MacKie RM, Hauschild A, Eggermont AM. Epidemiology of invasive cutaneous melanoma. Annals of oncology: official journal of the European Society for Medical Oncology / ESMO. 2009;20(Suppl 6):vi1–7. doi: 10.1093/annonc/mdp252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Dennis LK, Lynch CF, Sandler DP, Alavanja MC. Pesticide use and cutaneous melanoma in pesticide applicators in the agricultural heath study. Environ Health Perspect. 2010;118:812–7. doi: 10.1289/ehp.0901518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Purdue MP, Hoppin JA, Blair A, Dosemeci M, Alavanja MC. Occupational exposure to organochlorine insecticides and cancer incidence in the Agricultural Health Study. Int J Cancer. 2007;120:642–9. doi: 10.1002/ijc.22258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ojajarvi A, Partanen T, Ahlbom A, Boffetta P, Hakulinen T, Jourenkova N, Kauppinen T, Kogevinas M, Vainio H, Weiderpass E, Wesseling C. Risk of pancreatic cancer in workers exposed to chlorinated hydrocarbon solvents and related compounds: a meta-analysis. Am J Epidemiol. 2001;153:841–50. doi: 10.1093/aje/153.9.841. [DOI] [PubMed] [Google Scholar]
- 27.Ji BT, Silverman DT, Stewart PA, Blair A, Swanson GM, Baris D, Greenberg RS, Hayes RB, Brown LM, Lillemoe KD, Schoenberg JB, Pottern LM, et al. Occupational exposure to pesticides and pancreatic cancer. American journal of industrial medicine. 2001;39:92–9. doi: 10.1002/1097-0274(200101)39:1<92::aid-ajim9>3.0.co;2-p. [DOI] [PubMed] [Google Scholar]
- 28.Fryzek JP, Garabrant DH, Harlow SD, Severson RK, Gillespie BW, Schenk M, Schottenfeld D. A case-control study of self-reported exposures to pesticides and pancreas cancer in southeastern Michigan. Int J Cancer. 1997;72:62–7. doi: 10.1002/(sici)1097-0215(19970703)72:1<62::aid-ijc9>3.0.co;2-2. [DOI] [PubMed] [Google Scholar]
- 29.Garabrant DH, Held J, Langholz B, Peters JM, Mack TM. DDT and related compounds and risk of pancreatic cancer. Journal of the National Cancer Institute. 1992;84:764–71. doi: 10.1093/jnci/84.10.764. [DOI] [PubMed] [Google Scholar]
- 30.Hoppin JA, Tolbert PE, Holly EA, Brock JW, Korrick SA, Altshul LM, Zhang RH, Bracci PM, Burse VW, Needham LL. Pancreatic cancer and serum organochlorine levels. Cancer Epidemiol Biomarkers Prev. 2000;9:199–205. [PubMed] [Google Scholar]
- 31.Andreotti G, Freeman LE, Hou L, Coble J, Rusiecki J, Hoppin JA, Silverman DT, Alavanja MC. Agricultural pesticide use and pancreatic cancer risk in the Agricultural Health Study Cohort. Int J Cancer. 2009;124:2495–500. doi: 10.1002/ijc.24185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Freeman LE, Rusiecki JA, Hoppin JA, Lubin JH, Koutros S, Andreotti G, Zahm SH, Hines CJ, Coble JB, Barone-Adesi F, Sloan J, Sandler DP, et al. Atrazine and cancer incidence among pesticide applicators in the agricultural health study (1994-2007) Environ Health Perspect. 2011;119:1253–9. doi: 10.1289/ehp.1103561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Rusiecki JA, De Roos A, Lee WJ, Dosemeci M, Lubin JH, Hoppin JA, Blair A, Alavanja MC. Cancer incidence among pesticide applicators exposed to atrazine in the Agricultural Health Study. Journal of the National Cancer Institute. 2004;96:1375–82. doi: 10.1093/jnci/djh264. [DOI] [PubMed] [Google Scholar]
- 34.Hou L, Andreotti G, Baccarelli AA, Savage S, Hoppin JA, Sandler DP, Barker J, Zhu ZZ, Hoxha M, Dioni L, Zhang X, Koutros S, et al. Lifetime Pesticide Use and Telomere Shortening among Male Pesticide Applicators in the Agricultural Health Study. Environ Health Perspect. 2013;121:919–24. doi: 10.1289/ehp.1206432. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ma H, Zhou Z, Wei S, Liu Z, Pooley KA, Dunning AM, Svenson U, Roos G, Hosgood HD, 3rd, Shen M, Wei Q. Shortened telomere length is associated with increased risk of cancer: a meta-analysis. PLoS One. 2011;6:e20466. doi: 10.1371/journal.pone.0020466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Bonfanti M, Taverna P, Chiappetta L, Villa P, D’Incalci M, Bagnati R, Fanelli R. DNA damage induced by alachlor after in vitro activation by rat hepatocytes. Toxicology. 1992;72:207–19. doi: 10.1016/0300-483x(92)90113-s. [DOI] [PubMed] [Google Scholar]
- 37.Brown MA, Kimmel EC, Casida JE. DNA adduct formation by alachlor metabolites. Life sciences. 1988;43:2087–94. doi: 10.1016/0024-3205(88)90358-x. [DOI] [PubMed] [Google Scholar]
- 38.Hernandez AF, Parron T, Tsatsakis AM, Requena M, Alarcon R, Lopez-Guarnido O. Toxic effects of pesticide mixtures at a molecular level: their relevance to human health. Toxicology. 2013;307:136–45. doi: 10.1016/j.tox.2012.06.009. [DOI] [PubMed] [Google Scholar]
- 39.Thomas KW, Dosemeci M, Coble JB, Hoppin JA, Sheldon LS, Chapa G, Croghan CW, Jones PA, Knott CE, Lynch CF, Sandler DP, Blair AE, et al. Assessment of a pesticide exposure intensity algorithm in the agricultural health study. J Expo Sci Environ Epidemiol. 2010;20:559–69. doi: 10.1038/jes.2009.54. [DOI] [PMC free article] [PubMed] [Google Scholar]