Abstract
Paraquat (1,1′-dimethyl-4, 4′-bipyridinium dichloride), a nonselective herbicide, was once widely used in North America and is still used in some countries including the U.S.A. It is extremely toxic in animals and humans after acute exposure. Although there is little evidence that paraquat is a carcinogen, exposure has been associated with some types of cancer in humans, including melanoma, leukemia, and cancers of the penis, cervix and lung. We examined the relationship between cancer incidence and lifetime exposure to paraquat among 56,222 licensed pesticide applicators from Iowa and North Carolina enrolled in the Agricultural Health Study. Poisson regression was used to calculate rate ratios (RRs) and 95% confidence intervals (CIs) among paraquat users compared to non-users, while adjusting for potential confounders. There was no risk for cancer overall, nor for any of the cancers suggested by earlier epidemiologic studies. However, risk for non-Hodgkin’s lymphoma (NHL) was significantly elevated (RR=1.51, [95% CI=1.01–2.26]) when we compared those who ever used paraquat to those who never used paraquat. Among the 24,665 applicators (43.9%) who provided more detailed paraquat exposure information, those in the highest tertile of lifetime exposure-days (LE) and intensity-weighted lifetime exposure-days (IWLE) for paraquat had an increased risk of NHL, but the RRs were not significant (RR=1.74 [0.69–4.42] for LE; RR=1.86 [0.68–5.11] for IWLE, respectively) and there was not a significant exposure-response trend. Although we found some evidence for a link between paraquat exposure and NHL in this study, we cannot rule out the possibility that this is a chance finding.
INTRODUCTION
Paraquat (1,1′-dimethyl-4,4′-bipyridylium chloride dichloride; CAS 1910-42-5) is a quick-acting and non-selective herbicide (Bromilow 2004). Its use has been restricted in several countries, including Indonesia (1990), South Korea and Hungary (1991), Germany (1993), the U.S.A (1997), and Chile (2001). Nevertheless, it can still be used in the U.S.A. and some other countries (U.S. Dept. of Agriculture/National Agricultural Statistics Service, 2005; Pesticide Action Network (PAN) International, 2005).
Paraquat is extremely toxic in both animals and humans and has resulted in a high mortality rate after acute exposure (60–80%) (Vale et al., 1987; Lin 1990). In animal studies, paraquat has caused oxidative stress, DNA damage, and inefficient DNA repair (Weidauer et al., 2004; Dusinska et al., 1998; Ross et al., 1979). Unintentional poisoning episodes in humans have resulted in acute and respiratory effects (Hoppin et al., 2002). Paraquat may also induce delayed health effects including Parkinson’s disease and benign lung diseases (Uversky 2004; Wesseling et al., 2001; U.S. Dept. of Agriculture/National Agricultural Statistics Service, 2005). It has been associated with premalignant or malignant skin lesions (Wang et al., 1987), possibly through photosensitivity reaction (Jaworska et al., 1991) and dermal chemical burns (Jee et al., 1995). Excesses of melanoma and leukemia, and cancers of the penis, cervix and lung were observed in a cohort of banana plantation workers, who frequently used paraquat, although only the cervical cancer findings were statistically significant (Wesseling et al., 1996). Another study in Costa Rica assessing cancer incidence in regions with high pesticide use have observed excess overall cancer, skin melanoma, and cancer of the rectum, bladder, ovary, cervix, and lung (Wesseling et al., 1996). The information on potential cancer effects from paraquat exposure is limited and International Association for Research on Cancer (IARC) has not listed paraquat as a potential human carcinogen. Given paraquat’s use, further investigation of the cancer risk is warranted. Enrollment (phase I only) data from the Agricultural Health Study was utilized to evaluate cancer risk associated with paraquat use in the cohort
MATERIALS AND METHODS
Cohort recruitment
The AHS is a prospective cohort study of certified pesticide applicators (n=57,311) and their spouses (n=32,347) who were recruited in Iowaand North Carolina between December, 1993 to December, 1997 (Alavanja et al., 1996). Of the eligible licensed applicators 82.4% (n=57,311) enrolledin the study by completing an enrollment questionnaire. Licensed pesticide applicators include private applicators, who are largely farmers, and commercial applicators, from Iowa only, who are employed by pest control and agricultural crop protection companies. Incident cancers diagnosed from enrollment through December 31, 2004 were identified through the cancer registries in Iowa and North Carolina and coded using the International Classification of Diseases for Oncology (ICD-O-2) (Percy et al., 1990). A total of 24,665 pesticide applicators (43.9% of the enrolled applicators) provided more detailed information on paraquat use in a take-home questionnaire. This study was approved by Institutional Review Boards at the National Institutes of Health, University of Iowa, Westat Inc., and Battelle. Informed consent was obtained from all participants at the time of enrollment into the Agricultural Health Study.
Exposures and other information
At enrollment, we obtained detailed information on 22 pesticides, including years of use, frequency of use in an average year, and decade when use began, and for 28 additional pesticides, including paraquat, we obtained information on ever versus never use. Information regarding application methods and the use of personal protective equipment was collected for pesticides in general and not for specific pesticides. Participants provided information on possible confounding factors, including tobacco use, alcohol consumption, fruit and vegetable consumption, medical conditions, other agricultural activities, and various non-farmoccupational exposures. All participants who completed the enrollment questionnaire were also given a self-administeredtake-home questionnaire that included additional questionsabout exposures, medical conditions, and more detailed questions about the 28 pesticides for which ever versus never use information had been collected in the enrollment questionnaire. All questionnaires are available at www.aghealth.org/questionnaires.
We used two metrics of exposure (lifetime exposure-days and intensity weighted lifetime exposure-days) to characterize paraquat exposure. Lifetime exposure-days (LE) was defined as the number of years a participant personally mixed or applied paraquat times the number of days in an average year that paraquat was used. Intensity-weighted lifetime exposure-days (IWLE) was the product of estimated intensity score of exposure and total LE. The intensity score was calculated as: [(mixing status + application method + equipment repair status) X personal protective equipment used] (Dosemeci et al., 2002). The intensity score heavily weighted the use of protective gloves and thereby dermal exposures and to a lesser extent other protective clothing reflecting our current knowledge of routes of pesticide exposure.
Statistical analysis
Prevalent cancer cases (n = 1,075) and two subjects with missing information on date of birth were excluded, leaving 56,222 cohort members in this analysis. Missing data for specific factors were included in the analyses as categorical variables (7.3% of the 56,222 persons were missing information on ever/never use of paraquat; 2.0% and 2.7% of 24,665 subjects with take-home questionnaire were missing information on paraquat LE and IWLE, respectively). We used Poisson regression to calculate rate ratios (RRs) and 95% confidence intervals (CIs). LE and IWLE were categorized into tertiles based on the distribution for all the cancer cases (0.1–7, 7.1–20, 20.1+ for exposure-days and 0.1–36, 36.1–154, 154.1+ for intensity-weighted exposure-days).
Two reference groups were used for the analyses: pesticide applicators who reported never using paraquat and pesticide applicators whose use of paraquat was in the lowest tertile of exposure. Models were adjusted for age at enrollment (< 40, 40–49, 50–59, 60 years), education ( high school graduate, > high school graduate), cigarette smoking (by pack-years: never, 14, > 14), alcohol consumption during the last 12 months (yes/no), family history of cancer (yes/no), research site (Iowa/North Carolina), and the five pesticides with the highest correlation coefficients with paraquat exposure by IWLE [metribuzin, pendimethalin, Silvex (2,4,5-TP), butylate, and dieldrin]. The correlation coefficients for these five pesticides were small and ranged between 0.30 (pendimethalin) and 0.34 (metribuzin). Linear trends for RR were assessed using the median value for each tertile of exposure in the models, adjusting for covariates. If the sample size in a category for analysis was five or fewer, we combined subjects in the middle and lowest tertiles of LE and IWLE resulting in two categories (0.1–20, 20.1+ exposure-days; 0.1–154, 154.1+ intensity-weighted exposure-days). For colon, lung and NHL analyses, we combined subjects in middle and lowest tertiles of LE and IWLE, since the number of exposed subjects in the middle tertile group was small (n=2).
The AHS data release version used was “REL0612” (enrollment or phase I data only). All analyses were conducted using SAS version 9.1.
RESULTS
Among the 56,222 private and commercial applicators, 20.0% (n=11,256) reported having ever used paraquat, 72.7% (n=40,861) reported having never used paraquat and 7.3% (n=4,105) did not report on paraquat use (Table 1). The subjects with no information for paraquat exposure were older than the subjects in other categories. Follow-up time and the distribution of gender and race were similar among the non-exposed group and exposed groups. Iowa applicators were less likely to report paraquat use than subjects from North Carolina. The median follow-up time for all subjects was 7.8 years. Commercial applicators were more common among the exposed groups. Of the 50 pesticides, the five most highly correlated with paraquat use were metribuzin, pendimenthalin, silvex (2,4,5-TP), butylate, and dieldrin.
Table 1.
Selected characteristics of pesticide applicators by lifetime exposure-days (LE) to paraquat in the AHS (n=56,222)
| Missing paraquat exposure1 (n=4,105) | Non-exposed1 (n=40,861) | Low-exposed Lowest tertile2 (n=916) | Exposed1 High-exposed High two tertiles2 (n=3,012) |
Exposed but missing LE1 & 2 (n=7,328) | |
|---|---|---|---|---|---|
| No. (%) | No. (%) | No. (%) | No. (%) | No. (%) | |
| Age | |||||
| <40 | 971 (23.7) | 13757 (33.6) | 252 (27.5) | 776 (25.8) | 2466 (33.6) |
| 40–49 | 765 (18.6) | 11315 (27.7) | 249 (27.1) | 878 (29.2) | 2263 (30.9) |
| 50–59 | 956 (23.3) | 8243 (20.2) | 214 (23.4) | 673 (22.3) | 1516 (20.7) |
| >=60 | 1413 (34.4) | 7546 (18.5) | 201 (21.9) | 685 (22.7) | 1083 (14.8) |
| Gender | |||||
| Male | 3918 (95.5) | 39634 (97.0) | 905 (98.8) | 2977 (98.8) | 7247 (98.9) |
| Female | 187 ( 4.5) | 1227 ( 3.0) | 11 ( 1.2) | 35 ( 1.2) | 81 ( 1.1) |
| Race | |||||
| White | 3027 (73.8) | 39449 (96.5) | 879 (96.0) | 2879 (95.6) | 7126 (97.3) |
| Non White | 395 ( 9.6) | 885 (2.2) | 11 ( 1.2) | 30 ( 1.0) | 177 ( 2.4) |
| Missing | 683 (16.6) | 527 (1.3) | 26 ( 2.8) | 103 ( 3.4) | 25 ( 0.3) |
| State | |||||
| Iowa | 1184 (28.8) | 30740 (75.2) | 531 (58.0) | 940 (31.2) | 2644 (36.1) |
| North Carolina | 2921 (71.2) | 10121 (24.8) | 385 (42.0) | 2072 (68.8) | 4684 (63.9) |
| Applicator type | |||||
| Farmer1 | 3990 (97.2) | 37248 (91.2) | 797 (87.0) | 2609 (86.6) | 6711 (91.6) |
| Commercial | 115 (2.8) | 3613 (8.8) | 119 (13.0) | 403 (13.4) | 617 ( 8.4) |
| Education | |||||
| <=High school | 3287 (80.0) | 23913 (58.5) | 475 (51.9) | 1525 (50.6) | 3970 (54.2) |
| > High school | 808 (19.7) | 16855 (41.3) | 439 (47.9) | 1482 (49.2) | 3341 (45.6) |
| Missing | 10 ( 0.3) | 93 (0.2) | 2 (0.2) | 5 (0.2) | 17 (0.2) |
| Smoking | |||||
| Never | 1335 (32.5) | 22207 (54.3) | 428 (46.7) | 1348 (44.7) | 3445 (47.0) |
| Ex | 1181 (28.8) | 11717 (28.7) | 319 (34.8) | 1038 (34.5) | 2261 (30.8) |
| Current | 710 (17.3) | 6296 (15.4) | 143 (15.6) | 512 (17.0) | 1587 (21.7) |
| Missing | 879(21.4) | 641 ( 1.6) | 26 ( 2.8) | 114 (3.8) | 35 ( 0.5) |
| Alcohol use | |||||
| No | 1004 (24.5) | 11863 (29.0) | 292 (31.9) | 1157 (38.4) | 2701 (36.9) |
| Yes | 916 (22.3) | 27888 (68.3) | 586 (64.0) | 1691 (56.1) | 4529 (61.8) |
| Missing | 2185 (53.2) | 1110 ( 2.7) | 38 ( 4.1) | 164 ( 5.5) | 98 ( 1.3) |
| Family history of cancer | |||||
| No | 922 (22.4) | 22959 (56.2) | 454 (49.6) | 1517 (50.4) | 4004 (54.6) |
| Yes | 431 (10.5) | 15043 (36.8) | 377 (41.2) | 1184 (39.3) | 2783 (38.0) |
| Missing | 2752 (67.1) | 2859 ( 7.0) | 85 ( 9.2) | 311 (10.3) | 541 ( 7.4) |
| Corn production | |||||
| No | 2470 (60.2) | 10482 (25.6) | 312 (34.1) | 1242 (41.2) | 2527 (34.5) |
| Yes | 1635 (39.8) | 30379 (74.4) | 604 (65.9) | 1770 (58.8) | 4801 (65.5) |
| Take-homequestionnaire | |||||
| No | 3846 (93.7) | 20623 (50.5) | 0 (0.0) | 0 (0.0) | 7088 (96.7) |
| Yes | 259 ( 6.3) | 20238 (49.5) | 916 (100.0) | 3012 (100.0) | 240 ( 3.3) |
| Other pesticide use Metribuzin | |||||
| No | 12 ( 0.3) | 12952 (45.8) | 489 (53.4) | 1601 (53.3) | 76 ( 1.0) |
| Yes | 5 ( 0.1) | 6682 ( 2.2) | 401 (43.7) | 1289 (42.8) | 35 ( 0.5) |
| Missing | 4088(99.7) | 21227 (52.0) | 26 ( 2.9) | 118 ( 3.9) | 7217 (98.5) |
| Pendimethalin | |||||
| No | 36 ( 0.9) | 13035 (31.9) | 439 (47.9) | 1299 (43.1) | 74 ( 1.0) |
| Yes | 28 ( 0.7) | 6471 (15.8) | 453 (49.5) | 1611 (53.5) | 56 ( 0.8) |
| Missing | 4041 (98.4) | 21355 (52.3) | 24 ( 2.6) | 102 ( 3.4) | 7198 (98.2) |
| Silvex (2,4,5-TP) | |||||
| No | 33 ( 0.8) | 18997 (46.5) | 835 (91.2) | 2564 (85.1) | 121 ( 1.7) |
| Yes | 1 ( 0.0) | 684 ( 1.7) | 59 ( 6.4) | 354 (11.8) | 8 ( 0.1) |
| Missing | 4071 (99.2) | 21180 (51.8) | 22 ( 2.4) | 94 ( 3.1) | 7199 (98.2) |
| Butylate | |||||
| No | 26 ( 0.6) | 15105 (37.0) | 590 (64.4) | 1036 (62.0) | 97 ( 1.3) |
| Yes | 6 ( 0.2) | 4538 (11.1) | 306 (33.4) | 1868 (34.4) | 33 ( 0.5) |
| Missing | 4073 (99.2) | 21218 (51.9) | 20 ( 2.2) | 108 ( 3.6) | 7198 (98.2) |
| Dieldrin | |||||
| No | 54 ( 1.3) | 19069 (46.7) | 846 (92.4) | 2766 (91.8) | 142 ( 1.9) |
| Yes | 2 ( 0.1) | 572 ( 1.4) | 43 ( 4.7) | 151 ( 5.0) | 7 ( 0.1) |
| Missing | 4049 (98.6) | 21220 (51.9) | 27 ( 2.9) | 95 ( 3.2) | 7179 (98.0) |
based on enrollment questionnaire
based on the take-home questionnaire only
The rate ratios (RRs) of selected cancers among AHS applicators are shown in Table 2. Ever use of paraquat was not associated with cancer incidence overall, or most individual cancers. It was associated with non-Hodgkin’s lymphoma (NHL) (RR [95% CI] = 1.51 [1.01–2.26]). There was a non-significant excess risk of leukemia and pancreatic cancer among paraquat users (RR= 1.36 [0.72–2.58]; RR= 1.66 [0.79–3.49], respectively). Paraquat use was not significantly associated with lung, kidney, bladder, and female breast cancers, or skin melanoma.
Table 2.
Rate ratios for selected cancers, by exposure status to paraquat among AHS applicators
| Total subjects (n=56,222) | Subjects who returned a take homequestionnaire (n=24,665) | |||
|---|---|---|---|---|
| N of cases | RR (95% CI)1 | N of cases | RR (95% CI)1 | |
| All Cancers | ||||
| Non-exposed | 2,441 | 1.00 (reference) | 1,459 | 1.00 (reference) |
| Exposed | 667 | 0.96 (0.88–1.05) | 306 | 0.99 (0.86–1.13) |
| Prostate cancer | ||||
| Non-exposed | 1,002 | 1.00 (reference) | 625 | 1.00 (reference) |
| Exposed | 252 | 0.96 (0.83–1.18) | 122 | 0.95 (0.77–1.18) |
| Lung cancer | ||||
| Non-exposed | 221 | 1.00 (reference) | 137 | 1.00 (reference) |
| Exposed | 85 | 1.09 (0.83–1.43) | 33 | 0.95 (0.63–1.43) |
| Colon cancer | ||||
| Non-exposed | 193 | 1.00 (reference) | 112 | 1.00 (reference) |
| Exposed | 46 | 0.89 (0.63–1.25) | 25 | 1.32 (0.82–2.13) |
| Pancreas cancer | ||||
| Non-exposed | 38 | 1.00 (reference) | 21 | 1.00 (reference) |
| Exposed | 14 | 1.66 (0.79–3.49) | 7 | 1.52 (0.6–3.76)2 |
| Kidney cancer | ||||
| Non-exposed | 67 | 1.00 (reference) | 37 | 1.00 (reference) |
| Exposed | 20 | 0.99 (0.58–1.71) | 9 | 1.08 (0.48–2.42) |
| Bladder cancer | ||||
| Non-exposed | 105 | 1.00 (reference) | 64 | 1.00 (reference) |
| Exposed | 25 | 0.92 (0.57–1.49) | 12 | 0.95 (0.48–1.90) |
| Non-Hodgkin's lymphoma | ||||
| Non-exposed | 95 | 1.00 (reference) | 47 | 1.00 (reference) |
| Exposed | 41 | 1.51 (1.01–2.26)** | 18 | 1.72 (0.94–3.18)* |
| Leukemia | ||||
| Non-exposed | 77 | 1.00 (reference) | 45 | 1.00 (reference) |
| Exposed | 19 | 1.36 (0.72–2.58) | 11 | 1.41 (0.70–2.84)2 |
| Skin melanoma | ||||
| Non-exposed | 88 | 1.00 (reference) | 56 | 1.00 (reference) |
| Exposed | 23 | 0.94 (0.58–1.52)2 | 13 | 1.02 (0.54–1.02)2 |
, 0.05 p<0.1,
p 0.05
Adjusted for age at enrollment, education, research site, cigarette smoking, alcohol consumption, family history of cancers, and 5 pesticides correlated with paraquat
Adjusted for age at enrollment, education, research site, cigarette smoking, alcohol consumption, and family history of cancers. The 5 pesticides correlated with paraquat were not included in the adjustment because the mean parameter was either invalid or at a limit of its range for some observations due to the small number of cancer cases
The RRs for selected cancers by the LE and IWLE paraquat exposure metrics among 24,665 subjects who returned a take-homequestionnaire are presented in Table 3. Compared to participants with no reported paraquat use, those in the middle tertile and highest tertileof paraquat exposure level had an increased risk of NHL. However the RRs were not statistically significant (RR=1.18 [0.51–2.71], RR [95% CI] = 1.74 [0.69–4.42] for LE; and RR=1.25 [0.57–2.78], 1.86 [0.68–5.11] for IWLE, respectively), nor were the exposure-response trends for either LE or IWLE (p-trend=0.232; 0.197, respectively). However, the number of NHL cases in the exposed groups was small due to the limited number of follow-up years. In analyses using the low exposed group as the reference, the increased risks for NHL (RR=1.23 [0.40–3.74] for highest tertile of LE; 1.15 [0.36–3.62] for highest tertile of IWLE) were not statistically significant. The RRs for prostate, lung and colon cancers were not consistent across exposure categories. Paraquat LE or IWLE was not associated with any other cancers for which we had sufficient numbers of cases to explore risk.
Table 3.
Rate ratios for selected cancers by lifetime exposure-days (LE) and intensity weighted lifetime exposure-days (IWLE) to paraquat among AHS applicators with non-exposed and low exposed as referent groups, among 24,665 subjects who returned a take-homequestionnaire
| Lifetime exposure-days (LE) | Intensity weighted lifetime exposure-days (IWLE) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| N of cases | Non-exposed referent 1 RRLE(95% CI)1 | Low-exposed referent RRLE(95% CI)1 | N of cases3 | Non-exposed referent RRIWLE(95% CI)1 | Low-exposed referent 1 RRIWLE(95% CI)1 | ||||||
| All Cancer | |||||||||||
| Non-exposed | 1459 | 1.00 (reference) | Non-exposed | 1459 | 1.00 (reference) | ||||||
| 0.1–7 | 69 | 1.07(0.84–1.37) | 1.00 (reference) | 0.1–36 | 88 | 1.04 (0.84–1.29) | 1.00 (reference) | ||||
| 7.1–20.0 | 101 | 0.94(0.76–1.15) | 0.87 (0.63–1.19) | 36.1–154 | 90 | 0.99 (0.80–1.24) | 0.94 (0.70–1.27) | ||||
| 20.1+ | 110 | 0.98(0.80–1.21) | 0.91 (0.66–1.26) | 154.1+ | 90 | 0.95 (0.76–1.20) | 0.91 (0.66–1.24) | ||||
| P-trend§ | 0.757 | 0.653 | P-trend§ | 0.805 | 0.550 | ||||||
| Prostate | |||||||||||
| Non-exposed | 625 | 1.00 (reference) | Non-exposed | 625 | 1.00 (reference) | ||||||
| 0.1–7 | 35 | 1.27(0.90–1.80) | 1.00 (reference) | 0.1–36 | 42 | 1.16(0.84–1.60) | 1.00 (reference) | ||||
| 7.1–20.0 | 38 | 0.85(0.61–1.19) | 0.65 (0.41–1.04) | 36.1–154 | 34 | 0.90(0.63–1.28) | 0.76(0.48–1.21) | ||||
| 20.1+ | 38 | 0.83(0.59–1.18) | 0.61 (0.35–1.02) | 154.1+ | 32 | 0.83(0.57–1.20) | 0.68(0.41–1.11) | ||||
| P-trend§ | 0.303 | 0.069 | P-trend§ | 0.396 | 0.116 | ||||||
| Lung | |||||||||||
| Non-exposed | 137 | 1.00 (reference) | Non-exposed | 137 | 1.00 (reference) | ||||||
| 0.1–24.5 | 21 | 1.08(0.67–1.74) | 1.00 (reference) | 0.1–36 | 19 | 1.02(0.62–1.69) | 1.00 (reference) | ||||
| 24.5+ | 11 | 0.83(0.43–1.59) | 0.76 (0.35–1.62) | 154.1+ | 10 | 0.92(0.46–1.83) | 0.91(0.41–2.05) | ||||
| P-trend§ | 0.659 | - | P-trend§ | 0.799 | - | ||||||
| Colon | |||||||||||
| Non-exposed | 112 | 1.00 (reference) | Non-exposed | 112 | 1.00 (reference) | ||||||
| 0.1–20.0 | 18 | 1.68(0.99–2.84) | 1.00 (reference) | 0.1–36 | 15 | 1.42(0.80–2.50) | 1.00 (reference) | ||||
| 20.1+ | 6 | 0.96(0.41–2.28) | 0.56 (0.22–1.42)2 | 154.1+ | 7 | 1.35(0.60–3.05) | 0.95(0.38–2.40)2 | ||||
| P-trend§ | 0.481 | - | P-trend§ | 0.238 | |||||||
| Non-Hodgkin's lymphoma | |||||||||||
| Non-exposed | 47 | 1.00 (reference) | Non-exposed | 47 | 1.00 (reference) | ||||||
| 0.1–20.0 | 7 | 1.18(0.51–2.71) | 1.00 (reference) | 0.1–154 | 8 | 1.25(0.57–2.78) | 1.00 (reference) | ||||
| 20.1+ | 6 | 1.74 (0.69–4.42) | 1.23 (0.40–3.74)2 | 154.1+ | 5 | 1.86(0.68–5.11) | 1.15(0.36–3.62)2 | ||||
| P-trend§ | 0.232 | - | P-trend§ | 0.197 | - | ||||||
P-trend§, P-values for trend among the subjects with information of life duration.
Adjusted for age at enrollment, education, research site, cigarette smoking, alcohol consumption, family history of cancers, and 5 pesticides correlated with paraquat
Adjusted for age at enrollment, education, research site, cigarette smoking, alcohol consumption, and family history of cancers. The 5 pesticides correlated with paraquat were not included in the adjustment because the mean parameter was either invalid or at a limit of its range for some observations due to the small number of cancer cases.
Numbers do not always sum to those of lifetime exposure-days due to missing data
DISCUSSION
In this study, NHL risk was significantly associated with the ever-use of paraquat. Monotonic but non-significant trends of increasing risk were observed for increasing lifetime exposure days and intensity weighted lifetime exposure days using non-exposed for comparison, but not when using low exposed as the referent. None of the exposure-response trends were statistically significant.
Previous studies of workers employed in the production of paraquat or farmers living in coffee growing areas where paraquat is extensively used were found to be at higher risk of developing skin melanoma, and cancer of the rectum, bladder, ovary, cervix, and lung (Wesseling et al., 1996), but not non-Hodgkin’s lymphoma. We found no excesses for the cancers noted in earlier studies. Our knowledge about the possible mechanism of action of paraquat in human carcinogenesis is limited. The occurrence of chromosomal aberrations and gene mutations in human lymphocytes exposed to paraquat in vitro is interesting (Salam et al., 1993; Kuo et al., 1993; Rios et al., 1995; Ribas et al., 1997/98), given the suggestion of an association with NHL in our data. Further investigation of the association between paraquat and NHL is warranted as more cases occur in the AHS cohort and elsewhere.
Several limitations of this study should be noted. Despite being the largest cohort study of the association between NHL and paraquat to date, the number of exposed incident cancers is small and the follow-up interval is relatively short at this time. Pesticide exposure patterns can be complex and the use of several pesticides may be correlated. While we were able to control for the use of other pesticides in our analyses and saw no change in cancer risk estimates with these adjustments, the precision of our estimates was limited. Although farmers can provide considerable detail on their pesticide use (Blair & Zahm 1995) and the reliability of reporting in the study is quite good (Blair 2002; Hoppin 2002), some misclassification of exposures also undoubtedly occurred. In a prospective study, such misclassification of exposure is likely to be nondifferential, which would tend to bias estimates of relative risk toward the null.
The strengths of this study include its prospective design, thereby minimizing the potential for recall bias and ensuring temporal relevance of any associations uncovered between paraquat exposure and cancer, and detailed information on the use of many different pesticides. Availability of comprehensive information collected about other jobs, medical conditions and lifestyle factors also allowed adjustment for many potential confounders.
Continued updating of paraquat exposure and accumulation of more incident cases will allow for future re-evaluation.
Acknowledgments
This research was supported, in part, by the Intramural Research Program of the National Institutes of Health (Division of Cancer Epidemiology and Genetics of the National Cancer Institute and the National Institute of Environmental Health Sciences). We are indebted to the participants in the Agricultural Health Study for their contribution to this study.
References
- 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]
- Blair A, Tarone R, Sandler DP, Lynch CF, Rowland A, Wintersteen W, Steen WC, Samanic C, Dosemeci M, Alavanja MC. Reliability of reporting on life-style and agricultural factors by a sample of participants in the Agricultural Health Study from Iowa. Epidemiology. 2002;13:94–9. 26. doi: 10.1097/00001648-200201000-00015. [DOI] [PubMed] [Google Scholar]
- Blair A, Zahm SH. Agricultural exposures and cancer. Environ Health Perspect. 1995;103:205–8. doi: 10.1289/ehp.95103s8205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bromilow RH. Paraquat and sustainable agriculture. Pest Manag Sci. 2004;60:340–9. doi: 10.1002/ps.823. [DOI] [PubMed] [Google Scholar]
- Dosemeci M, Alavanja MC, Rowland AS, Mage D, Zahm SH, Rothman N, Lubin JH, Hoppin JA, Sandler DP, Blair A. A quantitative approach for estimating exposure to pesticides in the Agricultural Health Study. Ann Occup Hyg. 2002;46:245–60. doi: 10.1093/annhyg/mef011. [DOI] [PubMed] [Google Scholar]
- Dusinska M, Kovacikova Z, Vallova B, Collins A. Responses of alveolar macrophages and epithelial type II cells to oxidative DNA damage caused by paraquat. Carcinogenesis. 1998;19:809–12. doi: 10.1093/carcin/19.5.809. [DOI] [PubMed] [Google Scholar]
- Hoppin JA, Umbach DM, London SJ, Alavanja MC, Sandler DP. Chemical predictors of wheeze among farmer pesticide applicators in the Agricultural Health Study. Am J Respir Crit Care Med. 2002;165:683–9. doi: 10.1164/ajrccm.165.5.2106074. [DOI] [PubMed] [Google Scholar]
- Hoppin JA, Yucel F, Dosemeci M, Sandler DP. Accuracy of self-reported pesticide use duration information from licensed pesticide applicators in the Agricultural Health Study. J Expo Anal Environ Epidemiol. 2002;12:313–8. doi: 10.1038/sj.jea.7500232. [DOI] [PubMed] [Google Scholar]
- Jaworska A, Rosiek O. Paraquat increases superoxide dismutase activity and radiation resistance in two mouse lymphoma L5178Y cell strains of different radiosensitivities. Int J Radiat Biol. 1991;60:899–906. [PubMed] [Google Scholar]
- Jee SH, Kuo HW, Su WP, Chang CH, Sun CC, Wang JD. Photodamage and skin cancer among paraquat workers. Int J Dermatol. 1995;34:466–9. doi: 10.1111/j.1365-4362.1995.tb00611.x. [DOI] [PubMed] [Google Scholar]
- Kuo ML, Lin JK. The genotoxicity of the waste water discharged from paraquat manufacturing and its pyridyl components. Mutat Res. 1993;300:223–9. doi: 10.1016/0165-1218(93)90054-h. [DOI] [PubMed] [Google Scholar]
- Lin JL. Acute paraquat poisoning. Clin Med (Taiwan) 1990;26:352–358. [Google Scholar]
- Percy C, Van Holten V, Muir C. International classification of diseases for oncology. 2 Geneva (Switzerland): World Health Organization; 1990. [Google Scholar]
- Pesticide Action Network (PAN) International. Position paper of PAN International on paraquat. 2005:1–2. http://www.panna.org/resources/panups/PANParaquatPPaper.pdf.
- Ribas G, Surralles J, Carbonell E, Xamena N, Creus A, Marcos R. Genotoxic evaluation of the herbicide paraquat in cultured human lymphocytes. Teratogenesis Carcinog Mutagen. 1997/98;17:339–347. [PubMed] [Google Scholar]
- Rios AC, Salvadori DM, Oliveira SV, Ribeiro LR. The action of the herbicide paraquat on somatic and germ cells of mice. Mutat Res. 1995;328:113–8. doi: 10.1016/0027-5107(94)00199-f. [DOI] [PubMed] [Google Scholar]
- Ross WE, Block ER, Chang RY. Paraquat-induced DNA damage in mammalian cells. Biochem Biophys Res Commun. 1979;91:1302–8. doi: 10.1016/0006-291x(79)91208-7. [DOI] [PubMed] [Google Scholar]
- Salam AZ, Hussein EHA, El-Itriby HA, Anwar WA, Mansour SA. The mutagenicity of gramoxone(paraquat) on different eukaryotic systems. Mutat Res. 1993;319:89–101.36. doi: 10.1016/0165-1218(93)90067-n. [DOI] [PubMed] [Google Scholar]
- U.S. Dept. of Agriculture/National Agricultural Statistics Service. Agricultural Chemical Usage, Vegetables, 2004 Summary. Agricultural Statistics Board, National Agricultural Statistics Service, U.S. Department of Agriculture; 2005. [Google Scholar]
- Uversky VN. Neurotoxicant-induced animal models of Parkinson’s disease: understanding the role of rotenone, maneb and paraquat in neurodegeneration. Cell Tissue Res. 2004;318:225–41. doi: 10.1007/s00441-004-0937-z. [DOI] [PubMed] [Google Scholar]
- Vale JA, Meredith TJ, Buckly BM. Paraquat poisoning: Clinical features and immediate general management. Hum Toxicol. 1987;6:41–47. doi: 10.1177/096032718700600107. [DOI] [PubMed] [Google Scholar]
- Wang JD, Li WE, Hu FC, Hu KH. Occupational risk and the development of premalignant skin lesions among paraquat manufacturers. Br J Ind Med. 1987;44:196–200. doi: 10.1136/oem.44.3.196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weidauer E, Lehmann T, Ramisch A, Rohrdanz E, Foth H. Response of rat alveolar type II cells and human lung tumor cells towards oxidative stress induced by hydrogen peroxide and paraquat. Toxicol Lett. 2004;151:69–78. doi: 10.1016/j.toxlet.2004.02.014. [DOI] [PubMed] [Google Scholar]
- Wesseling C, Ahlbom A, Antich D, Rodriguez AC, Castro R. Cancer in banana plantation workers in Costa Rica. Int J Epidemiol. 1996;25:1125–31. doi: 10.1093/ije/25.6.1125. [DOI] [PubMed] [Google Scholar]
- Wesseling C, Antich D, Hogstedt C, Rodriguez AC, Ahlbom A. Geographical differences of cancer incidence in Costa Rica in relation to environmental and occupational pesticide exposure. Int J Epidemiol. 1999;28:365–74. doi: 10.1093/ije/28.3.365. [DOI] [PubMed] [Google Scholar]
- Wesseling C, van Wendel de Joode B, Ruepert C, Leon C, Monge P, Hermosillo H, Partanen TJ. Paraquat in developing countries. Int J Occup Environ Health. 2001;7:275–86. doi: 10.1179/107735201800339209. [DOI] [PubMed] [Google Scholar]