Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 28.
Published in final edited form as: Int J Hyg Environ Health. 2025 Jun 28;268:114615. doi: 10.1016/j.ijheh.2025.114615

Carbaryl use and incident cancer in the Agricultural Health Study: an updated analysis

Patricia A Erickson 1,4, Gabriella Andreotti 1, Richard V Remigio 1, Mary H Ward 1, Christine G Parks 2, Dale P Sandler 2, Paul S Albert 3, Kathryn Hughes Barry 4,5, Jonathan N Hofmann 1, Laura E Beane Freeman 1
PMCID: PMC12306554  NIHMSID: NIHMS2094247  PMID: 40582234

Abstract

Carbaryl is a common carbamate insecticide in the United States (USA). Previous epidemiologic investigations, including within the Agricultural Health Study (AHS), have suggested potential associations between carbaryl use and cancer risk. The AHS is a prospective cohort study of licensed pesticide applicators in North Carolina (NC) and Iowa (IA), USA. Information on lifetime pesticide use was reported at enrollment (1993-1997) and follow-up (1999-2005). We evaluated cancer risks associated with ever- and intensity-weighted lifetime days (IWLD) of carbaryl use. Among 52,625 applicators, 8,713 incident cancer cases were identified from linkages with state cancer registries through 2014 (NC) or 2017 (IA). We used Poisson regression to estimate rate ratios (RR) and 95% confidence intervals (CI), controlling for confounders, and evaluated lagged exposures. Approximately 51% of applicators reported using carbaryl. Increasing IWLD of carbaryl use was associated with increased stomach cancer risk (third tertile vs. never use; RRT3=2.07, 95% CI: 1.05-4.07, p-trend=0.02), persisting when exposure was lagged by 5-years (RRT3=2.20, 95% CI: 1.12-4.33). We noted elevated risks of esophageal (RR=1.52, 95% CI: 1.01-2.27) and tongue (RR=1.91, 95% CI: 0.95-3.81) cancers with ever-use. There was an increased risk of aggressive prostate cancer when carbaryl exposure was lagged by 30 years (RRlag30q4=1.56, 95% CI: 1.18-2.08, p-trend=0.002). This is the largest and most comprehensive prospective evaluation of carbaryl and cancer risk to date. We provide novel evidence of associations between carbaryl exposure and specific cancers. There is a need for additional studies to confirm these findings and to elucidate the biological mechanisms underlying the observed associations.

Keywords: cancer, carbaryl, pesticides, agriculture, occupational exposure

Graphical Abstract

graphic file with name nihms-2094247-f0001.jpg

Introduction:

Carbaryl is a N-methyl carbamate insecticide and acts as an acetylcholinesterase inhibitor (United States Environmental Protection Agency, 2008); exposure can lead to an accumulation of acetylcholine in the synaptic cleft, resulting in neurotoxic effects (Abreu-Villaça & Levin, 2017). It was first registered for use on cotton crops (United States Environmental Protection Agency, 2008), but remains widely used in the United States (US) with registrations for both agricultural and residential uses, including applications to fruits, vegetables, grains, and lawn and garden care (United States Environmental Protection Agency, 2008). Based on the most recent usage information for the US, more than 700,000 pounds of carbaryl were applied agriculturally each year between 2013 and 2017, while more than 2 million pounds were applied in non-agricultural settings (including applications to buildings, turf, pastures, and roadways) (United States Environmental Protection Agency, 2020). Carbaryl continues to be used in Canada and Australia, though with more limited use than the United States (Australian Government, 2024; Health Canada, 2016). The European Union banned all uses of carbaryl in 2006, in part because of concerns for “potential carcinogenic properties of the active substance” (Euopean Comission, 2007).

Carbaryl was reviewed by the International Agency for Research on Cancer (IARC) in 1987, when it was categorized as a group 3 carcinogen or “not classifiable as to its carcinogenicity in humans”, as based on insufficient data from human studies and inadequate evidence from experimental studies (IARC, 1987). In 2004, the US Environmental Protection Agency (EPA) classified carbaryl as “likely to be carcinogenic in humans,” based primarily on experimental studies that observed hemangiosarcoma in mice (United States Environmental Protection Agency, 2008, 2021 ). The most recent US EPA’s Human Health Risk Assessment raised concern for mutagenic effects based on in-vitro studies, in which carbaryl produced chromosomal aberrations and affected mitosis (United States Environmental Protection Agency, 2021 ). Other in-vitro studies have suggested carbaryl may be genotoxic and may act as an endocrine disruptor by activating estrogen receptors (Önfelt & Klasterska, 1984; Tange et al., 2016; United States Environmental Protection Agency, 2021). These properties of carbaryl make up several of the key characteristics of carcinogens identified by IARC in their hazard identification of potential carcinogens (Smith et al., 2016). Additionally, carbaryl is a known nitrosatable pesticide, as it is able to form N-nitroso carbaryl in the presence of nitrosating agents derived from ingested nitrate and nitrite (IARC, 1983). Several nitrosatable compounds (NOC), including carbaryl, have been shown to be carcinogenic in experimental studies (IARC, 1983) and NOCs have been associated with an increased risk of several cancers (IARC, 2010; Ward et al., 2018).

Previous epidemiologic studies, though limited, provide some evidence of an association between carbaryl and increased risk of specific cancers. For example, a 2007 evaluation of carbaryl and incident cancer in the Agricultural Health Study (AHS) observed an increased risk of melanoma, as did a later evaluation in the AHS (Dennis et al., 2010; Mahajan et al., 2007). Additionally, an evaluation from the agricultural cohort AGRICAN reported an increased risk of central nervous system tumors (e.g., gliomas) (Piel et al., 2019). However, for other cancer sites, such as non-Hodgkin Lymphoma (NHL), results have been inconsistent and conflicting (Alavanja et al., 2014; De Roos et al., 2021; Koutros et al., 2019; Leon et al., 2019).

In this study, we updated the 2007 evaluation of carbaryl use and incident cancer in the AHS. Our analysis included a total of 4,629 incident carbaryl-exposed cancer cases, with an additional 3,338 incident exposed cases, almost 14 additional years of follow-up, and updated exposure information as compared to the previous evaluation (Mahajan et al., 2007).

Methods:

Study Population:

The AHS is a prospective cohort study of private and commercial pesticide applicators and their spouses in Iowa (IA) and North Carolina (NC) (Alavanja et al., 1996). The AHS has been described previously; briefly, 57,310 applicators seeking a license to apply restricted-use pesticides were enrolled between 1993 and 1997 (Alavanja et al., 1996). The study was approved by the National Cancer Institute (NCI) Special Studies Institutional Review Board (IRB) and the IRBs of other participating institutions. All participants implied consent by completing the enrollment questionnaire, consistent with IRB rules in effect at the time.

Exposure Assessment:

At enrollment, all pesticide applicators completed a self-administered questionnaire that collected information on demographics, personal and family medical history, and detailed pesticide use information, including lifetime use of carbaryl and 49 other pesticides. For 22 of these pesticides, applicators reported additional details related to application practices and history (i.e., years and days/year of personal pesticide mixing or application) for 22 of these pesticides. For the remaining 28 pesticides, including carbaryl, this detailed information was collected from a take-home questionnaire (also administered at enrollment), which was completed by 44% of applicators. In a follow-up interview administered from 1999-2005, updated exposure information, including frequency, duration, and other factors related to exposure (such as use of personal protective equipment and the application method) was collected from approximately 67% of the enrolled pesticide applicators. Questionnaires are available at https://aghealth.nih.gov/collaboration/questionnaires.html. For participants who did not complete the follow-up interview, missing exposure data was imputed using data-driven multiple imputation, which has been previously described (Heltshe et al., 2012). Briefly, logistic regression and stratified sampling were used to impute pesticide exposures based on a number of variables from the enrollment questionnaire. The final model included demographic variables, farm characteristics, such as whether the applicator was a farm owner and the size of the farm, factors associated with pesticide use (including duration and frequency, application methods, applicator type, the use of personal protective equipment), whether crops or animals were raised in the last 12 months, and history of specific medical conditions (Heltshe et al., 2012). Prior analyses have shown that this imputation method yields similar cancer risk estimates for participants with complete follow-up data and those with both complete and imputed data (Andreotti et al., 2019).

Based on this information, three exposure metrics were defined: (1) ever vs. never use, (2) cumulative lifetime exposure days, calculated as the number of days used per year multiplied by number of years used, and (3) intensity-weighted lifetime days, calculated as cumulative lifetime days multiplied by an intensity score. The intensity score, or weighting factor, is based on a previously derived algorithm that integrates both literature-based evidence and information provided by pesticide applicators. This included details such as whether the applicator mixed or applied pesticides, repaired pesticide-related equipment, and used personal protective equipment. The initial weighting factors were informed by literature on the relative contribution of these activities to pesticide exposures (Dosemeci, 2002). These weights were subsequently refined using field studies within the AHS (Hines et al., 2007; Hines et al., 2008; Thomas et al., 2010), leading to the updated the intensity-weighting algorithm used in this analysis (Coble et al., 2011).

Case Ascertainment and Classification:

Participant vital status was obtained from the National Death Index while incident cancer cases were ascertained from regular linkages to population-based cancer registries in Iowa and North Carolina. Cancer sites were classified according to the International Classification of Diseases for Oncology, 3rd Revision (ICD-O-3) (Fritz et al., 2013). The Surveillance, Epidemiology, and End Results Program Lymphoma Subtype Recodes (2008) was used for classification of subtypes of lymphoid neoplasms (http://seer.cancer.gov/lymphomarecode/lymphoma-orig.html). Additionally, for this analysis, we defined aggressive prostate cancer as having one or more of the following characteristics: Gleason score of ≥ 8, poorly differentiated, a diagnosed grade or stage at or above 3, or fatal (i.e., underlying cause of death was prostate cancer).

Statistical Methods and Analyses:

For our evaluation, we excluded participants who, at enrollment, had a history of cancer (other than non-melanoma skin cancer; n=1,096), did not live in IA or NC at enrollment (n=341), or were missing information on carbaryl use (n=3,248), leaving n=52,625 applicators and n=8,713 total cancer cases for our analysis.

This analysis incorporates time-varying covariates, including cumulative exposure, calendar year of follow-up, and attained age from cohort participants. These time-dependent covariates were considered constant over fixed 2-year intervals from enrollment until a cancer diagnosis, death, movement out of state of end of follow-up period (IA: December 31, 2017; NC: December 31, 2014), whichever occurred first. We estimated person-time within each interval, calculating time-dependent covariate values at the start of each interval, which were incorporated into Poisson regression models. We used Poisson regression to estimate incidence rate ratios (RR) and 95% confidence intervals (CI) for each category of carbaryl use. In exposure-response analyses, the Wald test was used to estimate linear trends (p-trend) using the median of each exposure category as a continuous variable. The PROC MIANALYZE procedure was used to obtain the appropriate variance for the imputed exposure data.

Based on the distribution of exposure among all cancer cases, we categorized cumulative lifetime days and intensity-weighted lifetime days of carbaryl exposure into quartiles, tertiles, or the median, such that there were at least 5 exposed cases per category. For each analysis, the referent group comprised those who reported no use of carbaryl. To address issues related to cancer latency, we evaluated lagged exposures in 5-year intervals, excluding the most recent years of exposure (ranging from 5 to 30 years), prior to cancer diagnosis or other censoring event (death, relocation, or end of follow-up).

All models were adjusted for attained age (continuous), state of residence (IA or NC), applicator type (private or commercial), race (White, Black and Other, missing), sex (male or female), education (high school or less, beyond high school, other education, missing), and cigarette smoking status and pack-years (never, former/pack-years tertiles 1-3, current pack-years tertile 1-3, missing). For some cancer sites, adjustments for specific variables were not included due to model convergence issues, which resulted primarily from small sample sizes in certain strata. These limitations made reliable parameter estimation difficult, and therefore, we excluded those adjustments to maintain model stability. For several cancer sites, we performed sensitivity analyses in which we considered potential site-specific confounders (i.e., melanoma: sun exposure, esophageal cancer: alcohol consumption and other tobacco products, thyroid cancer: family history, body mass index (BMI) and we evaluated effect modification by factors that may increase risk among specific subgroups of the population (i.e., stomach cancer: drinking water nitrate exposures). We also examined the potential for confounding from the pesticides most highly correlated with carbaryl as based on continuous IWLD, Pearson correlation coefficients were ≥ 0.30 for carbaryl and malathion, diazinon, pendimethalin, chlorimuron-ethyl, metribuzin, and chlordane. The inclusion of these pesticides did not substantially change (≤10%) the risk estimates and therefore they were not included in any of the final models.

All analyses were conducted with SAS version 9.4 (Cary, NC). All tests were two-sided and considered to be statistically significant at an alpha of 0.05.

Results:

Among 52,625 participants, 26,579 (51%) reported ever-use of carbaryl (Table 1). This cohort had a mean attained age of 65 years; most participants were male (97%), white (97%), private applicators (91%), and from IA (66%). Applicators reporting use of carbaryl were older than non-users and more likely to be from NC. A total of 8,713 incident cancers were diagnosed, of which 4,629 were among carbaryl users. The median value of lifetime days of carbaryl use was 25, whereas the median value for intensity-weighted lifetime days was 992. The cancer sites with the greatest numbers of exposed cases, as based on ever use, were prostate cancer (n=1,773), lung (n=457), non-Hodgkin lymphoma (n=371), and colon cancer (n=297).

Table 1.

Select characteristics of Agricultural Health Study population by carbaryl exposure status from enrollment questionnaire data (1993-1997)

Overall Unexposeda Exposeda p-value
Characteristicsb N, % N, % N, %
Total 52,625 (100) 26,046 (49.5) 26,579 (50.5)
Attained Age (mean, SD) 65 (12) 63.7 (12.2) 66.2 (11.8) < .001
Incident Cancer Cases
  No 43,912 (83.4) 21,962 (84.4) 21,950 (82.5) < .001
  Yes 8,713 (16.6) 4,084 (15.7) 4,629 (17.4)
Race
  White 50,913 (96.7) 25,179 (96.8) 25,734 (96.7) 0.09
  Black and Other 1,143 (2.2) 598 (2.3) 545 (2.0)
  Missing 569 (1.1) 269 (1.0) 300 (1.1)
Sex
  Male 51,226 (97.3) 25,481 (97.9) 25,745 (96.8) < .001
  Female 1,399 (2.7) 565 (2.2) 834 (3.1)
State of Residence
  Iowa 34,898 (66.3) 21,790 (83.7) 13,108 (49.3) < .001
  North Carolina 17,727 (33.7) 4,256 (16.4) 13,471 (50.6)
Applicator Type
  Private 48,013 (91.2) 2,3297 (89.5) 24,716 (92.9) < .001
  Commercial 4,612 (8.8) 2,749 (10.6) 1,863 (7.0)
Smoking Status & Packyears
  Never 27,603 (52.5) 14,956 (57.5) 12,647 (47.5) < .001
  Former, Tertile 1 (<3.75) 5,196 (9.9) 2,514 (9.7) 2,682 (10.1)
  Former, Tertile 2 (3.75–≤15) 4,694 (8.9) 2,044 (7.9) 2,650 (10.0)
  Former, Tertile 3 (>15) 4,384 (8.3) 1,619 (6.2) 2,765 (10.4)
  Current, Tertile 1 (<3.75) 2,955 (5.6) 1,597 (6.1) 1,358 (5.1)
  Current, Tertile 2 (3.75–≤15) 2,709 (5.1) 1,243 (4.8) 1,466 (5.5)
  Current, Tertile 3 (>15) 2,783 (5.3) 1,098 (4.2) 1,685 (6.3)
  Missing 2,301 (4.4) 975 (3.7) 1,326 (5.0)
Education Level
  High School or less 28,661 (54.5) 15,107 (58.1) 13,554 (50.9) < .001
  Beyond high school 22,174 (42.1) 10,136 (39) 12,038 (45.2)
  Other Education 119 (0.2) 63 (0.2) 56 (0.2)
  Missing 1,671 (3.2) 740 (2.8) 931 (3.5)
Family History of Cancer
  No 29,875 (56.8) 15,635 (60.1) 14,240 (53.5) < .001
  Yes 20,665 (39.3) 9,521 (36.6) 11,144 (41.9)
  Missing 2,085 (4) 890 (3.4) 1,195 (4.5)
Body Mass Index (kg/m2)
  <25 10,139 (19.3) 5,090 (19.6) 5,049 (19.0) 0.005
  25-<30 19,577 (37.2) 9,802 (37.7) 9,775 (36.7)
  30+ 9,026 (17.2) 4,451 (17.1) 4,575 (17.2)
  Missing 13,883 (26.4) 6,703 (25.8) 7,180 (27.0)
Alcohol Consumption
  Never 16,2,84 (30.9) 6,541 (25.1) 9,743 (36.6) < .001
  Ever 34,688 (65.9) 18,652 (71.7) 16,036 (60.3)
  Missing 1,653 (3.1) 827 (3.2) 826 (3.1)
Open in a new tab
a

Carbaryl exposure information from enrollment and follow-up questionnaires

b

Demographic information from enrollment questionnaire

Overall, there was no association between all cancers combined and either ever-use (RR=0.98, 95% CI: 0.94-1.03, Table 2) or increasing intensity weighted lifetime days of carbaryl use (RRQ4=0.95, 95% CI: 0.86-1.05, p-trend=0.26) (Table 3).

Table 2:

Adjusted Rate Ratios and 95% Confidence Intervals for Ever vs. Never Use of Carbaryl and Select Cancer Sites among AHS Applicators

Cancer Sitea Exposed
Cases		 Unexposed
Cases		 RR
(95% CI)
All cancers combined 4629 4084 0.98 (0.94,1.03)
Oral Cavity 111 108 0.99 (0.74,1.33)
Lipb 32 38 0.89 (0.53,1.50)
Tonguec 24 14 1.91 (0.95,3.81)
Pharynx 26 28 0.73 (0.41,1.32)
Esophagus 72 48 1.52 (1.01,2.27)
Stomachb 66 50 1.13 (0.75,1.82)
Small Intestineb 25 27 1.02 (0.57,1.84)
Colon 297 292 0.88 (0.73,1.05)
Rectum 136 127 0.93 (0.70,1.22)
Liverb 38 24 1.01 (0.56,1.80)
Gall & Biliaryd 18 18 0.82 (0.40,1.69)
Pancreas 105 92 0.95 (0.70,1.30)
Larynxb 45 32 1.25 (0.77,2.05)
Lung 457 330 0.83 (0.70,0.97)
Melanoma 236 196 1.13 (0.92,1.40)
Prostatee 1773 1582 1.04 (0.96,1.12)
Aggressive Prostatee, f 384 347 0.99 (0.88,1.11)
Testise 21 31 0.85 (0.46,1.57)
Kidney 167 154 0.93 (0.72,1.19)
Bladder 233 204 0.90 (0.73,1.11)
Braind 44 53 0.88 (0.57,1.37)
Thyroid 31 61 0.49 (0.30,0.78)
Hodgkin Lymphoma g 13 20 0.87 (0.41,1.84)
NHL 371 356 0.96 (0.82,1.13)
A. NHL, B-Cell 347 325 0.98 (0.83,1.16)
  A.1. CLL / SLL 84 80 1.09 (0.77,1.53)
  A.2. Follicular Lymphoma 48 39 1.21 (0.76,1.91)
 B. NHL, T-cellb 12 21 0.63 (0.29,1.36)
Leukemia 76 52 1.10 (0.74,1.64)
 A. Myeloid Leukemia 62 45 1.06 (0.69,1.62)
Open in a new tab

Note: RR, Risk Ratio. 95% CI, 95% Confidence Interval. NHL, Non-Hodgkin Lymphoma. CLL/SLL, Chronic Lymphocytic Leukemia / Small Lymphocytic Leukemia.

a

Unless otherwise noted, models were adjusted for attained age, state of residence, sex, race, education, smoking status (never, formerT1-T3, currentT1-T3), and applicator type. Other models excluded specific variables to preserve stability due to model convergence

b

Adjusted for attained age, state of residence, race, education, smoking status (never, former T1-T3, currentT1-T3), and applicator type

c

Adjusted for attained age, state of residence, education, smoking status (never, former T1-T3, currentT1-T3), and applicator type.

d

Adjusted for attained age, state of residence, sex, education, smoking status (never, former T1-T3, currentT1-T3), and applicator type.

e

Male applicators only

f

Aggressive prostate cancer defined as: total Gleason score of 8, poorly differentiated, a diagnosed grade or stage at or above 3, or fatal (i.e., underlying cause of death was prostate cancer).

g

Adjusted for attained age, state of residence, smoking status (never, formerT1-T3, current T1-T3), and applicator type.

Table 3:

Adjusted Rate Ratios and 95% Confidence Intervals for Intensity-Weighted Lifetime Days of Carbaryl Use and Select Cancer Sites among AHS Applicators

Cancer Sitea Carbaryl
Use		 Exposed
Cases		 RR &
95% CI		 p-trend
All cancers combined None 4084 ref. 0.26
Q1 581 1.02 (0.93, 1.12)
Q2 582 1.00 (0.92, 1.10)
Q3 581 1.04 (0.95, 1.14)
Q4 580 0.95 (0.86, 1.05)
Oral Cavity None 108 ref. 0.18
Q1 18 1.34 (0.81, 2.22)
Q2 19 1.26 (0.76, 2.11)
Q3 16 1.06 (0.58, 1.91)
Q4 9 0.65 (0.31, 1.37)
Lipb None 38 ref. 0.17
T1 7 0.90 (0.35, 2.34)
T2 8 1.47 (0.67, 3.22)
T3 6 1.37 (0.51, 3.72)
Pharynx None 28 ref. 0.22
M1 10 1.36 (0.65, 2.87)
M2 6 0.57 (0.20, 1.64)
Esophagusc None 48 ref. 0.34
Q1 9 1.57 (0.74, 3.32)
Q2 10 1.48 (0.70, 3.11)
Q3 11 1.65 (0.80, 3.41)
Q4 8 1.28 (0.55, 2.99)
Stomachb None 50 ref. 0.02
T1 6 0.68 (0.27, 1.74)
T2 9 0.98 (0.44, 2.17)
T3 18 2.07 (1.05, 4.07)
Colon None 292 ref. 0.82
Q1 37 0.99 (0.66, 1.49)
Q2 40 0.98 (0.68, 1.40)
Q3 31 0.82 (0.56, 1.22)
Q4 44 1.03 (0.72, 1.48)
Rectum None 127 ref. 0.39
Q1 15 0.82 (0.45, 1.51)
Q2 14 0.81 (0.42, 1.56)
Q3 18 1.13 (0.67, 1.93)
Q4 13 0.72 (0.37, 1.39)
Gall & Biliaryc None 18 ref. 0.56
M1 5 0.77 (0.26, 2.29)
M2 5 0.68 (0.22, 2.11)
Pancreas None 92 ref. 0.74
Q1 13 0.88 (0.46, 1.70)
Q2 9 0.74 (0.36, 1.54)
Q3 17 1.33 (0.76, 2.31)
Q4 13 0.87 (0.45, 1.67)
Larynxb None 32 ref. 0.15
T1 5 1.08 (0.42, 2.77)
T2 7 1.01 (0.37, 2.74)
T3 10 1.82 (0.79, 4.20)
Lung None 330 ref. 0.74
Q1 37 0.75 (0.53, 1.08)
Q2 47 0.83 (0.59, 1.17)
Q3 60 0.84 (0.61, 1.16)
Q4 80 0.89 (0.66, 1.20)
Melanoma None 196 ref. 0.81
Q1 41 1.55 (1.08, 2.20)
Q2 27 1.05 (0.69, 1.60)
Q3 27 1.18 (0.75, 1.85)
Q4 29 1.13 (0.72, 1.78)
Total Prostate Cancerd None 1582 ref. 0.17
Q1 245 1.11 (0.96, 1.28)
Q2 231 1.07 (0.92, 1.25)
Q3 230 1.14 (0.98, 1.33)
Q4 195 0.91 (0.76, 1.08)
Aggressive Prostate Cancerd, e None 347 ref. 0.50
Q1 57 1.16 (0.94, 1.43)
Q2 45 0.90 (0.72, 1.14)
Q3 44 0.92 (0.70, 1.20)
Q4 58 1.08 (0.86, 1.27)
Kidneyc None 154 ref. 0.53
Q1 18 0.89 (0.54, 1.49)
Q2 21 0.90 (0.54, 1.49)
Q3 19 0.86 (0.51, 1.43)
Q4 20 0.86 (0.49, 1.51)
Bladder None 204 ref. 0.28
Q1 32 1.07 (0.73, 1.56)
Q2 44 1.29 (0.90, 1.86)
Q3 37 1.15 (0.79, 1.67)
Q4 30 0.84 (0.54, 1.28)
Brainc None 53 ref. 0.95
M1 8 0.62 (0.29, 1.32)
M2 11 0.95 (0.44, 2.05)
Thyroid None 61 ref. 0.11
M1 7 0.48 (0.22, 1.04)
M2 7 0.45 (0.19, 1.06)
NHL None 356 ref. 0.66
Q1 47 0.95 (0.70, 1.30)
Q2 48 0.96 (0.67, 1.36)
Q3 37 0.86 (0.60, 1.23)
Q4 46 1.07 (0.75, 1.52)
 A. NHL, B-Cell None 325 ref. 0.66
Q1 45 0.99 (0.72, 1.34)
Q2 46 0.99 (0.70, 1.41)
Q3 34 0.86 (0.59, 1.24)
 A.1. CLL / SLL None 80 ref. 0.77
T1 19 1.24 (0.73, 2.10)
T2 10 0.71 (0.34, 1.50)
T3 10 0.92 (0.43, 1.96)
 A.2. Follicular Lymphomaf None 39 ref. 0.19
M1 10 1.24 (0.63, 2.44)
M2 11 1.72 (0.79, 3.74)
Leukemia b None 52 ref. 0.21
Q1 7 0.75 (0.29, 1.91)
Q2 14 1.76 (0.93, 3.31)
Q3 8 1.23 (0.56, 2.74)
Q4 14 1.71 (0.80, 3.68)
 A. Myeloid Leukemiab None 45 ref. 0.67
Q1 6 0.71 (0.25, 2.01)
Q2 11 1.67 (0.84, 3.30)
Q3 6 1.05 (0.42, 2.64)
Q4 8 0.97 (0.36, 2.62)
Open in a new tab

Note: RR, Rate Ratio. 95% CI = 95% Confidence Interval. NHL, Non-Hodgkin Lymphoma. CLL/SLL, Chronic Lymphocytic Leukemia / Small Lymphocytic Leukemia; Cutpoints for carbaryl exposure based on IWLD among cancer cases. Q0: 0 (no reported use), Q1: >0 - ≤ 302.93, Q2: > 302.93 - ≤ 991.64, Q3: > 991.64 - ≤3875, Q4: >3875. Tertiles: T0: 0, T1: >0 - ≤434, T2: > 434 - ≤ 2254, T3: > 2254. Median: M0: 0, M1: ≤ 991.64, M2: > 991.64. Linear p-trend trend based on Wald test by using median of each exposure category as a continuous variable. The reported frequencies for the referent group include unexposed cancer cases and do not contribute to the combined total number of cases among exposed participants.

a

Unless otherwise noted, models were adjusted for attained age, state of residence, sex, race, education, smoking status (never, formerT1-T3, currentT1-T3), and applicator type. Other models excluded specific variables to preserve stability due to model convergence.

b

Adjusted for attained age, state of residence, race, education, smoking status (never, former T1-T3, currentT1-T3), and applicator type.

c

Adjusted for attained age, state of residence, sex, education, smoking status (never, former T1-T3, currentT1-T3), and applicator type.

d

Male applicators only

e

Aggressive prostate cancer defined as: total Gleason score of 8, poorly differentiated, a diagnosed grade or stage at or above 3, or fatal (i.e., underlying cause of death was prostate cancer).

f

Adjusted for attained age, state of residence, education, smoking status (never, former T1-T3, currentT1-T3), and applicator type.

Although there was no association with ever-use, higher intensity-weighted lifetime days of use of carbaryl was associated with stomach cancer (RRT3=2.07, 95% CI: 1.05-4.07, p-trend=0.02) (Table 3). This association remained significantly elevated when carbaryl exposure was lagged by 5 years (Table 4); we were unable to evaluate further lag periods because of a small number of exposed cases. In stratified analyses, we found no evidence of interaction of ever use of carbaryl with nitrate from drinking water and stomach cancer (data not shown); however, we were only able to evaluate interactions with ever-use due to a small number of participants with available nitrate information. Additionally, ever-use of carbaryl was associated with an increased risk of esophageal cancer (RR=1.52, 95% CI: 1.01-2.27) (Table 2), though there was no evidence of an exposure-response in IWLD (RRQ4=1.28, 95% CI: 0.55-2.99, p-trend=0.34) (Table 3). These findings were unchanged in sensitivity analyses where we further adjusted for potential confounders of esophageal cancer, such as alcohol and smokeless tobacco (data not shown). Finally, we saw a suggestive association between ever-use of carbaryl and tongue cancer (RR=1.91, 95% CI: 0.95-3.81) (Table 2), though we were unable to conduct exposure-response analyses due to a small number of exposed cases (n=8).

Table 4:

Adjusted Rate Ratios and 95% Confidence Intervals in Relation to Lagged Intensity-Weighted Lifetime Days of Carbaryl Use and Select Cancer Sites among AHS Applicators

Lag 5 Lag 15 Lag 30
Cancer Sitea Carbaryl
Use		 Exposed
Cases		 RR
(95%
CI)		 p-trend Exposed
Cases		 RR
(95%
CI)		 p-trend Exposed
Cases		 RR
(95%
CI)		 p-trend
All cancers combined None 4172 ref. 0.28 4606 ref. 0.35 5602 ref. 0.49
Q1 560 1.05 (0.96, 1.15) 458 1.06 (0.96, 1.17) 204 1.18 (1.02, 1.35)
Q2 559 0.99 (0.90, 1.08) 445 0.98 (0.88, 1.08) 201 0.85 (0.74, 0.98)
Q3 559 1.07 (0.97, 1.18) 449 1.04 (0.94, 1.15) 200 0.99 (0.86, 1.14)
Q4 558 0.95 (0.86, 1.05) 450 0.95 (0.86, 1.06) 201 0.95 (0.82, 1.10)
Oral Cavity None 111 ref. 0.27 129 ref. 0.10 158 - -
Q1 17 1.34 (0.80, 2.25) 15 1.38 (0.80, 2.37) * -
Q2 20 1.35 (0.82, 2.23) 9 0.77 (0.39, 1.52) * -
Q3 11 0.73 (0.35, 1.49) 11 1.03 (0.54, 1.99) * -
Q4 10 0.73 (0.36, 1.48) 6 0.71 (0.36, 1.42) * -
Pharynx None 28 ref. 0.26 31 ref. 0.60 39 - -
M1 11 1.67 (0.82, 3.44) 8 1.63 (0.73, 3.62) * -
M2 5 0.52 (0.17, 1.60) 5 0.71 (0.23, 2.19) * -
Esophagusb None 48 ref. 0.66 55 ref. 0.69 70 - -
Q1 9 1.66 (0.80, 3.48) 8 1.90 (0.84, 4.26) * -
Q2 10 1.54 (0.73, 3.25) 10 1.72 (0.82, 3.60) 8 -
Q3 11 1.82 (0.89, 3.19) 6 1.25 (0.52, 3.00) * -
Q4 8 1.38 (0.59, 1.08) 7 1.34 (0.56, 3.20) * -
Stomachc None 51 ref. 0.01 54 - - 74 - -
T1 6 0.84 (0.35, 2.21) * - * -
T2 8 0.88 (0.38, 2.03) 13 - * -
T3 18 2.20 (1.12, 4.33) 13 - 5 -
Colon None 298 ref. 0.63 322 ref. 0.72 399 ref. 0.41
Q1 36 0.98 (0.67, 1.43) 30 1.00 (0.68, 1.46) 11 0.89 (0.49, 1.63)
Q2 37 0.94 (0.67, 1.34) 25 0.81 (0.53, 1.23) 8 0.47 (0.23, 0.95)
Q3 29 0.81 (0.54, 1.20) 33 1.14 (0.78, 1.66) 14 0.96 (0.56, 1.66)
Q4 44 1.08 (0.75, 1.54) 34 1.05 (0.71, 1.56) 12 0.77 (0.42, 1.40)
Rectum None 133 ref. 0.45 150 ref. 0.40 170 - -
Q1 14 0.87 (0.50, 1.53) 7 0.57 (0.27, 1.22) 5 -
Q2 10 0.64 (0.32, 1.31) 7 0.48 (0.21, 1.07) 5 -
Q3 18 1.16 (0.68, 1.99) 14 1.08 (0.61, 1.93) * -
Q4 12 0.73 (0.38, 1.41) 9 0.67 (0.33, 1.38) * -
Pancreas None 95 ref. 0.82 104 ref. 0.95 124 - -
Q1 11 0.81 (0.41, 1.61) 11 1.08 (0.58, 2.21) 5 -
Q2 9 0.79 (0.38, 1.66) 6 0.59 (0.26, 1.35) * -
Q3 16 1.28 (0.73, 2.24) 11 1.06 (0.55, 2.04) * -
Q4 13 0.89 (0.46, 1.71) 12 0.99 (0.51, 1.91) 8 -
Larynxc None 32 ref. 0.12 39 - - 47 - -
T1 6 1.31 (0.54, 3.16) 5 - * -
T2 6 0.87 (0.29, 2.67) * - * -
T3 10 1.94 (0.84, 4.48) 7 - * -
Lung None 339 ref. 0.67 382 ref. 0.70 482 ref. 0.21
Q1 34 0.74 (0.51, 1.08) 27 0.83 (0.56, 1.23) 9 0.61 (0.32, 1.18)
Q2 47 0.85 (0.62, 1.17) 36 0.83 (0.58, 1.18) 16 0.67 (0.41, 1.11)
Q3 59 0.92 (0.67, 1.25) 50 0.99 (0.72, 1.35) 26 1.04 (0.69, 1.55)
Q4 75 0.89 (0.67, 1.19) 59 0.92 (0.68, 1.24) 21 0.73 (0.46, 1.14)
Melanoma None 198 ref. 0.97 226 ref. 0.45 275 ref. 0.89
Q1 41 1.64 (1.15, 2.34) 29 1.45 (0.96, 2.18) 12 1.52 (0.85, 2.72)
Q2 26 1.07 (0.69, 1.66) 27 1.29 (0.86, 1.95) 18 1.72 (1.06, 2.80)
Q3 29 1.36 (0.89, 2.07) 20 1.04 (0.64, 1.67) 6 0.69 (0.30, 1.57)
Q4 26 1.07 (0.67, 1.71) 18 0.87 (0.52, 1.47) 9 1.02 (0.51, 2.03)
Prostate Cancerd None 1617 ref. 0.15 1781 ref. 0.30 2163 ref. 0.80
Q1 235 1.12 (0.37, 1.29) 194 1.11 (0.95, 1.29) 86 1.24 (1.00, 1.54)
Q2 224 1.04 (0.90, 1.21) 179 1.03 (0.88, 1.20) 83 0.90 (0.72, 1.13)
Q3 219 1.14 (0.98, 1.33) 172 1.06 (0.89, 1.25) 76 1.01 (0.80, 1.28)
Q4 188 0.90 (0.76, 1.07) 157 0.92 (0.77, 1.10) 75 0.97 (0.77, 1.24)
Aggressive Prostate Cancerd,e None 352 ref. 0.67 388 ref. 0.49 460 ref. 0.002
Q1 58 1.23 (1.00, 1.51) 46 1.12 (0.89, 1.40) 16 1.00 (0.70, 1.42)
Q2 42 0.83 (0.66, 1.05) 35 0.86 (0.66, 1.10) 23 1.07 (0.79, 1.45)
Q3 44 0.94 (0.73, 1.21) 35 0.88 (0.67, 1.14) 23 1.30 (0.95, 1.76)
Q4 55 1.05 (0.83, 1.33) 47 1.09 (0.85, 1.38) 29 1.56 (1.18, 2.08)
Kidneyb None 156 ref. 0.70 166 ref. 0.73 205 - -
Q1 16 0.93 (0.55, 1.59) 16 1.10 (0.66, 1.84) 8 -
Q2 21 0.94 (0.57, 1.57) 17 1.06 (0.63, 1.78) 7 -
Q3 20 1.00 (0.60, 1.65) 18 1.19 (0.72, 1.98) * -
Q4 19 0.88 (0.50, 1.55) 15 0.91 (0.51, 1.62) 11 -
Bladder None 206 ref. 0.23 241 ref. 0.20 294 ref. 0.87
Q1 32 1.10 (0.75, 1.61) 26 1.05 (0.69, 1.58) 17 1.64 (1.00, 2.68)
Q2 44 1.32 (0.92, 1.90) 35 1.28 (0.89, 1.84) 12 0.84 (0.47, 1.50)
Q3 37 1.23 (0.84, 1.79) 21 0.78 (0.49, 1.24) 9 0.70 (0.36, 1.37)
Q4 28 0.81 (0.52, 1.26) 24 0.77 (0.49, 1.22) 15 1.09 (0.63, 1.88)
Brainb None 55 ref. 0.89 58 ref. 0.83 64 - -
M1 7 0.62 (0.28, 1.37) 5 0.53 (0.21, 1.34) * -
M2 10 0.89 (0.40, 1.95) 9 1.03 (0.47, 2.27) * -
Thyroid None 61 ref. 0.10 62 ref. 0.18 71 - -
M1 8 0.54 (0.26, 1.16) 8 0.79 (0.37, 1.67) * -
M2 6 0.43 (0.17, 1.06) 5 0.51 (0.19, 1.35) * -
NHL None 359 ref. 0.56 384 ref. 0.55 469 ref. 0.41
Q1 50 1.05 (0.77, 1.43) 42 1.12 (0.81, 1.55) 21 1.39 (0.89, 2.16)
Q2 44 0.92 (0.65, 1.31) 39 1.01 (0.71, 1.43) 17 0.87 (0.54, 1.42)
Q3 36 0.92 (0.64, 1.32) 33 1.03 (0.71, 1.49) 15 0.97 (0.57, 1.64)
Q4 45 1.11 (0.78, 1.58) 36 1.13 (0.78, 1.65) 12 0.79 (0.44, 1.43)
A. NHL, B-Cell None 328 ref. 0.47 348 ref. 0.34 429 ref. 0.43
Q1 48 1.10 (0.80, 1.50) 42 1.23 (0.89, 1.70) 21 1.50 (0.97, 2.33)
Q2 42 0.95 (0.67, 1.36) 35 0.99 (0.69, 1.43) 17 0.95 (0.58, 1.54)
Q3 32 0.89 (0.60, 1.30) 33 1.14 (0.78, 1.65) 15 1.05 (0.62, 1.78)
Q4 43 1.15 (0.80, 1.65) 35 1.22 (0.83, 1.80) 11 0.78 (0.42, 1.45)
 A.1. CLL/SLL None 81 ref. 0.48 85 ref. 0.67 105 - -
T1 19 1.33 (0.78, 2.26) 15 1.28 (0.72, 2.27) 7 -
T2 9 0.72 (0.34, 1.50) 9 0.85 (0.40, 1.80) * -
T3 10 0.98 (0.46, 2.08) 10 1.36 (0.65, 2.85) * -
 A.2. Follicular Lymphomaf None 39 ref. 0.19 40 ref. 0.07 52 - -
M1 10 1.31 (0.66, 2.59) 11 1.54 (0.79, 2.98) 7 -
M2 11 1.89 (0.87, 4.11) 9 1.81 (0.86, 3.84) * -
Leukemia c None 54 ref. 0.34 62 ref. 0.86 79 - -
Q1 7 0.89 (0.39, 2.07) 6 0.85 (0.34, 2.12) 5 -
Q2 14 1.80 (0.97, 3.33) 11 1.81 (0.94, 3.48) 5 -
Q3 7 1.11 (0.49, 2.53) 9 1.57 (0.75, 3.28) * -
Q4 13 1.57 (0.72, 3.40) 7 1.13 (0.47, 2.67) 6 -
A. Myeloid Leukemiac None 47 ref. 0.58 54 - - 65 - -
Q1 6 0.87 (0.35, 2.16) 5 - 5 -
Q2 11 1.70 (0.87, 3.32) 9 - * -
Q3 5 0.89 (0.34, 2.33) 6 - * -
Q4 7 0.80 (0.28, 2.29) * - * -
Open in a new tab

Note: RR, Rate Ratio. 95% CI, 95% Confidence Interval. NHL, Non-Hogkin Lymphoma. CLL/SLL, Chronic Lymphocytic Leukemia / Small Lymphocytic Leukemia; Cutpoints for carbaryl exposure based on IWLD among cancer cases. Q0: 0 (no reported use), Q1: >0 - ≤ 302.93, Q2: > 302.93 - ≤ 991.64, Q3: > 991.64 - ≤3875, Q4: >3875. Tertiles: T0: 0, T1: >0 - ≤434, T2: > 434 - ≤ 2254, T3: > 2254. Median: M0: 0, M1: ≤ 991.64, M2: > 991.64. Linear p-trend trend based on Wald test by using median of each exposure category as a continuous variable. 95% CI = 95% Confidence Interval.

*

Less than 5 participants.

a

Unless otherwise noted, models were adjusted for attained age, state of residence, sex, race, education, smoking status (never, formerT1-T3, currentT1-T3), and applicator type. Other models excluded specific variables to preserve stability due to model convergence

b

Adjusted for attained age, state of residence, sex, education, smoking status (never, formerT1-T3, currentT1-T3), and applicator type

c

Adjusted for attained age, state of residence, race, education, smoking status (never, formerT1-T3, currentT1-T3), and applicator type.

d

Male applicators only

e

Aggressive prostate cancer defined as: total Gleason score of 8, poorly differentiated, a diagnosed grade or stage at or above 3, or fatal (i.e., underlying cause of death was prostate cancer).

f

Adjusted for attained age, state of residence, education, smoking status (never, former T1-T3, currentT1-T3), and applicator type.

Neither ever-use (RR=1.04, 95% CI: 0.96-1.12) nor increasing IWLD of use of carbaryl (RRQ4=0.91, 0.76-1.08, p-trend=0.17) was associated with total prostate cancer. Likewise, there was no association with either ever-use (RR=0.99, 95% CI: 0.88-1.11) or increasing IWLD use of carbaryl (RRQ4=1.08, 95% CI: 0.86-1.27, p-trend=0.50) and aggressive prostate cancer. However, when carbaryl exposure was lagged by 30 years, there was a significantly elevated risk of aggressive prostate cancer (RR Q4=1.56, 95% CI: 1.18-2.07, p-trend=0.002) (Table 4).

Thyroid cancer was inversely associated with both ever-use of carbaryl (RR=0.49, 95% CI: 0.30-0.78) (Table 2) and increasing intensity-weighted lifetime days of use (RRm2=0.45, 95% CI: 0.19-1.06, p-trend=0.11), though the number of carbaryl-exposed cases was small (n=14) (Table 3). The inverse associations remained in sensitivity analyses, where we further adjusted for other potential confounders, including family history of cancer, BMI, and pesticides previously associated with thyroid cancer in the AHS (metalaxyl, lindane, and chlorimuron-ethyl).

For melanoma, there was limited evidence of an association with carbaryl use (ever-use RR=1.13, 95% CI: 0.92-1.40, Table 2) and the highest quartile of intensity-weighted lifetime days (RRQ4=1.13, 95% CI: 0.72-1.78, p-trend=0.81) (Table 3). These results were unchanged in sensitivity analyses (data not shown), which adjusted for melanoma-specific risk factors, including skin reactivity, hours of sun exposure, pesticides previously associated with melanoma in the AHS (maneb/mancozeb, parathion, malathion, acetochlor), and additional factors adjusted for in previous analyses (i.e., hair color, skin reactivity). We assessed potential effect modification by stratifying on sun exposure; results remained unchanged when stratified (data not shown).

In our analyses, NHL was not associated with ever-use of carbaryl (RR=0.97, 95% CI: 0.83-1.14) (Table 2) or increasing carbaryl use (Table 3). The risk of follicular lymphoma (FL) was elevated, though not statistically significant, for above the median IWLD (RRm2 = 1.72, 95% CI: 0.79-3.74, p-trend=0.18). For lifetime days of use, a similar pattern for FL was observed, which achieved statistical significance (RRm2=2.16, 95% CI: 1.01-4.60, p-trend=0.04, Supplemental Table 1).

For the remaining cancer sites included in this evaluation, there were no notable patterns of associations with carbaryl use. Results for cumulative lifetime days were generally similar in magnitude and direction to the results of the intensity-weighted lifetime days of exposure (Supplemental Table 1).

Discussion:

In this updated analysis of carbaryl use among AHS applicators, we provide novel evidence of cancer at several sites. We observed an elevated risk of stomach cancer in association with increasing carbaryl use, which persisted when carbaryl exposure was lagged by 5 years. Additionally, both esophageal and tongue cancers were positively associated with ever use of carbaryl. We also noted an increased risk of aggressive prostate cancer when carbaryl exposure was lagged by 30 years and an inverse association between carbaryl use and thyroid cancer.

Although previous studies have suggested an increased risk of stomach cancer among agricultural workers (Acquavella et al., 1998; Shah et al., 2020), few have focused on carbaryl specifically. Our evaluation is, to the best of our knowledge, the first to report an association between carbaryl and stomach cancer. An early evaluation in the AHS of carbaryl was unable to examine risk of stomach cancer as there were too few exposed cases (Mahajan et al., 2007). A previous population-based case-control study with 9 exposed cases reported no association between carbaryl and stomach cancer (Lee et al., 2004). However, in that case-control study, proxy respondents (mostly spouses, children, or parents) were the source of exposure information for 80% of stomach cancer cases. These proxy respondents were more likely to report ‘don’t know’ and less likely to report use of specific pesticides as compared to the cases themselves, potentially resulting in misclassification of exposure (Lee et al., 2004). In contrast, in the current evaluation, all participants provided self-reported information regarding pesticide use.

While the mechanisms linking carbaryl to stomach cancer are unknown, one possibility involves nitrosation. Carbaryl is able to form N-nitroso carbaryl in the presence of nitrosating agents derived from ingested nitrate and nitrite; several N-nitroso compounds (NOC) have been shown to be carcinogenic in experimental studies (IARC, 1983). NOCs have also been associated with an increased risk of several cancers (IARC, 2010; Ward et al., 2018). Studies have shown N-nitroso-carbaryl can form in the stomach when carbaryl reacts with ingested nitrate and nitrite from drinking water and food sources (Elespuru & Lijinsky, 1973; IARC, 1983) and high doses of this compound have been shown to induce gastric cancer in rats (Eisenbrand et al., 1976). Given this, we attempted to evaluate potential effect modification of drinking water nitrate on the association between stomach cancer and carbaryl exposure in our evaluation by using data on nitrate in drinking water from the AHS (Manley et al., 2022). However, only a limited number of cases had information on drinking water nitrate, which precluded exposure-response analyses. Based on the limited number of cases, we saw no effect modification by nitrate.

Although there was no association with aggressive prostate cancer in unlagged analyses in our evaluation, there was a significant exposure-response trend when carbaryl exposure was lagged by 30 years. Prior studies focused on carbaryl and aggressive prostate cancer are limited: no association with aggressive prostate cancer was reported in a previous AHS analysis, though lagged exposures were not evaluated (Koutros et al., 2013). While the mechanisms through which carbaryl may influence prostate cancer risk are not understood, experimental studies have suggested carbaryl may be genotoxic and potentially act as an endocrine disruptor;(Önfelt & Klasterska, 1984; Renglin et al., 1999; Tange et al., 2016; United States Environmental Protection Agency, 2021 ) both of which have been implicated in prostate cancer carcinogenesis (Corti et al., 2022; Saramaki & Visakorpi, 2007).

Our findings of an inverse association between thyroid cancer and carbaryl are consistent with a previous analysis in the AHS (Lerro et al., 2021). Additionally, a previous analysis in the Biomarkers of Exposure and Effect in Agriculture, a sub-study of the AHS, noted a suggestive inverse association between the highest quartile of exposure to carbaryl and levels of the thyroid hormone T3 (total triiodothyronine, p-trend=0.08) (Lerro et al., 2021), (Lerro et al., 2018); though the role of thyroid hormones in the development of thyroid cancer is not yet completely understood (Wang et al., 2022). Further studies are warranted to better understand potential associations between carbaryl exposure, thyroid hormones, and risk of thyroid cancer.

In contrast to two prior evaluations in the AHS that reported an increased risk of melanoma with increasing carbaryl use (Dennis et al., 2010; Mahajan et al., 2007), we found limited evidence linking carbaryl use to melanoma risk. This lack of association was consistent for both exposure metrics, lagged analyses, and after further adjustment for additional risk factors that were also considered in previous evaluations (such as hours of sun exposure, skin reactivity, and hair color) (Mahajan et al., 2007). The differences in results may be explained by our study’s inclusion of more than twice as many exposed cases, the incorporation of additional follow-up time, and updated exposure information, which together allowed for a more robust evaluation as compared to the two prior AHS evaluations (Dennis et al., 2010; Mahajan et al., 2007). To our knowledge, no other studies have reported an association between carbaryl and melanoma. However, a pooled analysis of two case-control studies from Brazil and Italy observed an increased risk of melanoma among those ever-exposed to carbamate pesticides (n=10 exposed cases; OR=4.15, 95% CI: 0.81-21.1), but did not report results of individual carbamate insecticides (Fortes et al., 2016).

The lack of association observed between NHL and carbaryl is generally consistent with previous AHS evaluations (Alavanja et al., 2014; Mahajan et al., 2007). Similarly, analyses from both AGRICOH, a consortium comprised of three cohort studies (including the AHS), and the North American Pooled Project (NAPP) a pooled analysis of case-control studies in the US and Canada, noted no associations between ever exposure to carbaryl and risk of NHL (Koutros et al., 2019; Leon et al., 2019). However, in an evaluation from the InterLymph Consortium, which pooled nine case-control studies from the US and Europe (none of which were part of the NAPP), ever-use of carbaryl was associated with an increased risk of NHL (OR=1.34, 95% CI: 1.03-1.75), particularly for longer use and when exposure was lagged by at least 10 years (De Roos et al., 2021). Moreover, in both the NAPP and InterLymph Consortium, exposure to carbaryl was associated with follicular lymphoma (FL)(De Roos et al., 2021; Koutros et al., 2019), while the NAPP also observed an association with small-cell lymphocytic lymphoma (another NHL sub-type) (Koutros et al., 2019). In the current evaluation, there was an elevated, though not significant, risk of FL with IWLD of carbaryl use, which achieved statistical significance in analyses of lifetime days.

There are several factors to consider when comparing results for NHL in this analysis to those from the previous studies. Previous associations were based on ever-use or years of carbaryl use, while we used both ever vs. never use and IWLD, a more detailed exposure metric. While ever vs. never use is useful for initial assessments, cumulative exposure metrics provide deeper insight into potential associations (Leon et al., 2019). The evolution of the definition of NHL should also be considered when comparing our results to previous studies. The definition of NHL used in our evaluation includes both multiple myeloma (MM) and chronic lymphocytic lymphoma (CLL) (Morton et al., 2007). However, many of the cases included in the NAPP and InterLymph analyses were identified between 1979 and 1986 and were based on the working definition at that time, in which CLL and MM were not included (National Cancer Institute, 1982). In a sensitivity analysis in which we excluded CLL and MM cases from our NHL analyses, our results remained unchanged, suggesting the definition change does not explain the difference in findings.

This prospective cohort study has several strengths. First, we expanded upon a previous analysis of carbaryl use and incident cancer risk by including up to 14 additional years of follow-up and more than 4,000 additional carbaryl-exposed cancer cases. We also incorporated additional pesticide use information from a follow-up questionnaire that was completed by more than 60% of the cohort. For those who did not complete the follow-up questionnaire, we were able to impute carbaryl use and a previous study has shown there is no evidence of bias from incorporating this imputed data (Andreotti et al., 2019). Additionally, the exposure information was collected prior to cancer diagnosis; therefore, any misclassification of exposure would most likely be non-differential and which would generally be expected to bias the risk estimates toward the null. Although the exposure information was self-reported, which can be a limitation, it has been shown previously that pesticide applicators reliably report on their pesticide usage (Blair et al., 2002; Hoppin et al., 2002). Likewise, while pesticide applicators are exposed to more than one pesticide, we were able to consider the use of correlated pesticides, which did not significantly impact the risk estimates.

In conclusion, this study provides novel evidence of an association between carbaryl use and stomach cancer and evidence of elevated risks of esophageal and tongue cancers. Future studies are needed to confirm these findings and elucidate the biological mechanisms underlying these observed associations.

Supplementary Material

1

Acknowledgements:

Data in this analysis are based on the AHS data releases P1REL2022, P2REL2022, P3REL2022. This work was supported by the intramural research program of the National Institutes of Health, the National Institute of Environmental Health Sciences (Z01-ES049030) and National Cancer Institute (Z01-CP010119).

Sources of support:

This research was supported by the Intramural Research Program of the National Institutes of Health/National Institute of Environmental Health Sciences (Z01-ES049030) and National Cancer Institute (Z01-CP010119). This project was also supported by additional funding from the National Cancer Institute (K07 CA230182 and 3K07CA230182-05S1 to K.H.B.) and the Maryland Department of Health’s Cigarette Restitution Fund Program – CH-649-CRF (salary support for K.H.B.).

Footnotes

Disclosure: None of the authors have any conflicts of interest to disclose.

Ethics Statement: The study was approved by the National Cancer Institute (NCI) Special Studies Institutional Review Board (IRB) and the IRBs of other participating institutions. All participants implied consent by completing the enrollment questionnaire, consistent with IRB rules in effect at the time.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Data Availability:

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References:

  1. Abreu-Villaça Y, & Levin ED (2017). Developmental neurotoxicity of succeeding generations of insecticides. Environ Int, 99, 55–77. 10.1016/j.envint.2016.11.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Acquavella J, Olsen G, Cole P, Ireland B, Kaneene J, Schuman S, & Holden L (1998). Cancer among farmers: a meta-analysis. Ann Epidemiol, 8(1), 64–74. 10.1016/s1047-2797(97)00120-8 [DOI] [PubMed] [Google Scholar]
  3. Alavanja MC, Hofmann JN, Lynch CF, Hines CJ, Barry KH, Barker J, Buckman DW, Thomas K, Sandler DP, Hoppin JA, Koutros S, Andreotti G, Lubin JH, Blair A, & Beane Freeman LE (2014). Non-hodgkin lymphoma risk and insecticide, fungicide and fumigant use in the agricultural health study. PLoS One, 9(10), e109332. 10.1371/journal.pone.0109332 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Alavanja MC, Sandler DP, McMaster SB, Zahm SH, McDonnell CJ, Lynch CF, Pennybacker M, Rothman N, Dosemeci M, Bond AE, & Blair A (1996). The Agricultural Health Study. Environ Health Perspect, 104(4), 362–369. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1469343/pdf/envhper00335-0022.pdf [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Andreotti G, Lubin JH, Koutros S, Hofmann JN, Sandler DP, Lerro CC, Parks CG, Silverman DT, & Beane Freeman LE (2019). Response to Sheppard and Shaffer. J Natl Cancer Inst, 111(2), 216–218. 10.1093/jnci/djy201 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Australian Government. (2024). Carbaryl chemical review. Retrieved 6/24/2024 from https://www.apvma.gov.au/chemicals-and-products/chemical-review/listing/carbaryl
  7. Blair A, Tarone R, Sandler D, Lynch CF, Rowland A, Wintersteen W, Steen WC, Samanic C, Dosemeci M, & Alavanja MCR (2002). Reliability of reporting on life-style and agricultural factors by a sample of participants in the Agricultural Health Study from Iowa. Epidemiology, 13(1), 94–99. [DOI] [PubMed] [Google Scholar]
  8. Coble J, Thomas KW, Hines CJ, Hoppin JA, Dosemeci M, Curwin B, Lubin JH, Beane Freeman LE, Blair A, Sandler DP, & Alavanja MC (2011). An updated algorithm for estimation of pesticide exposure intensity in the agricultural health study. Int J Environ Res Public Health, 8(12), 4608–4622. 10.3390/ijerph8124608 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Corti M, Lorenzetti S, Ubaldi A, Zilli R, & Marcoccia D (2022). Endocrine disruptors and prostate cancer. Int J Mol Sci, 23(3). 10.3390/ijms23031216 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. De Roos AJ, Schinasi LH, Miligi L, Cerhan JR, Bhatti P, Mannetje A, Baris D, Benavente Y, Benke G, Clavel J, Casabonne D, Fritschi L, Hofmann JN, Huynh T, Monnereau A, Piro S, Slager SL, Vajdic CM, Wang SS, Zhang Y, Bernstein L, & Cocco P (2021). Occupational Insecticide Exposure and Risk of Non-Hodgkin Lymphoma: A Pooled Case-Control Study from the InterLymph Consortium. Int J Cancer, 149(10), 1768–1786. 10.1002/ijc.33740 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dennis LK, Lynch CF, Sandler DP, & Alavanja MC (2010). Pesticide use and cutaneous melanoma in pesticide applicators in the agricultural heath study. Environ Health Perspect, 118(6), 812–817. 10.1289/ehp.0901518 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dosemeci M. (2002). A Quantitative Approach for Estimating Exposure to Pesticides in the Agricultural Health Study. The Annals of Occupational Hygiene, 46(2), 245–260. 10.1093/annhyg/mef011 [DOI] [PubMed] [Google Scholar]
  13. Eisenbrand G, Schmähl D, & Preussmann R (1976). Carcinogenicity in rats of high oral doses of N-nitroso-carbaryl, a nitrosated pesticide. Cancer Lett, 1(5), 281–284. 10.1016/s0304-3835(75)97774-5 [DOI] [PubMed] [Google Scholar]
  14. Elespuru RK, & Lijinsky W (1973). The formation of carcinogenic nitroso compounds from nitrite and some types of agricultural chemicals. Food Cosmet Toxicol, 11(5), 807–817. 10.1016/0015-6264(73)90139-9 [DOI] [PubMed] [Google Scholar]
  15. Euopean Comission. (2007). Article 5 of the Rotterdam Convention - Carbaryl. https://www.pic.int/Portals/5/download.aspx?d=UNEP-FAO-RC-FRA-NOTIF-Carbaryl-63252-EU-20071123.En.pdf
  16. Fortes C, Mastroeni S, Segatto MM, Hohmann C, Miligi L, Bakos L, & Bonamigo R (2016). Occupational exposure to pesticides with occupational sun exposure increases the risk for cutaneous melanoma. J Occup Environ Med, 58(4), 370–375. 10.1097/JOM.0000000000000665 [DOI] [PubMed] [Google Scholar]
  17. Fritz A, Percy C, Jack A, Shanmugaratnam K, Sobin L, Parkin DM, & Whelan S (2013). International Classification of Diseases for Oncology (Fritz A, Percy C, Jack A, Shanmugaratnam K, Sobin L, Parkin DM, & Whelan S, Eds. 3rd ed.). World Health Organization. [Google Scholar]
  18. Health Canada. (2016). Re-evaluation Decision: Carbaryl. https://www.canada.ca/content/dam/hc-sc/migration/hc-sc/cps-spc/alt_formats/pdf/pubs/pest/_decisions/rvd2016-02/rvd2016-02-eng.pdf
  19. Heltshe SL, Lubin JH, Koutros S, Coble JB, Ji BT, Alavanja MC, Blair A, Sandler DP, Hines CJ, Thomas KW, Barker J, Andreotti G, Hoppin JA, & Beane Freeman LE (2012). Using multiple imputation to assign pesticide use for nonresponders in the follow-up questionnaire in the Agricultural Health Study. J Expo Sci Environ Epidemiol, 22(4), 409–416. 10.1038/jes.2012.31 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hines CJ, Deddens JA, Coble J, & Alavanja MCR (2007). Fungicide Application Practices and Personal Protective Equipment Use Among Orchard Farmers in the Agricultural Health Study. Journal of Agricultural Safety and Health, 13(2), 205–223. 10.13031/2013.22620 [DOI] [PubMed] [Google Scholar]
  21. Hines CJ, Deddens JA, Jaycox LB, Andrews RN, Striley CAF, & Alavanja MCR (2008). Captan Exposure and Evaluation of a Pesticide Exposure Algorithm among Orchard Pesticide Applicators in the Agricultural Health Study. The Annals of Occupational Hygiene, 52(3), 153–166. 10.1093/annhyg/men001 [DOI] [PubMed] [Google Scholar]
  22. Hoppin JA, Yucel F, Dosemeci M, & Sandler DP (2002). Accuracy of self-reported pesticide use duration information from licensed pesticide applicators in the Agricultural Health Study. J Expo Anal Environ Epidemiol, 12(5), 313–318. 10.1038/sj.jea.7500232 [DOI] [PubMed] [Google Scholar]
  23. IARC. (1983). IARC monographs on the evaluation of carcinogenic risks to humans. Miscellaneous Pesticides (IARC Monogr Eval Carcinog Risks Hum, Issue. [Google Scholar]
  24. IARC. (1987). Overall Evaluations of Carcinogenicity: An Updating of IARC Monographs Volumes 1 to 42. (IARC Monogr Eval Carcinog Risks Hum, Issue. [PubMed] [Google Scholar]
  25. IARC. (2010). IARC monographs on the evaluation of carcinogenic risks to humans. Ingested nitrate and nitrite, and cyanobacterial peptide toxins (1017-1606 (Print) 1017-1606). (IARC Monogr Eval Carcinog Risks Hum, Issue. [PMC free article] [PubMed] [Google Scholar]
  26. 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 (2013). Risk of total and aggressive prostate cancer and pesticide use in the Agricultural Health Study. Am J Epidemiol, 177(1), 59–74. 10.1093/aje/kws225 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Koutros S, Harris SA, Spinelli JJ, Blair A, McLaughlin JR, Zahm SH, Kim S, Albert PS, Kachuri L, Pahwa M, Cantor KP, Weisenburger DD, Pahwa P, Pardo LA, Dosman JA, Demers PA, & Beane Freeman LE (2019). Non-Hodgkin lymphoma risk and organophosphate and carbamate insecticide use in the north American pooled project. Environ Int, 127, 199–205. 10.1016/j.envint.2019.03.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lee WJ, Lijinsky W, Heineman EF, Markin RS, Weisenburger DD, & Ward MH (2004). Agricultural pesticide use and adenocarcinomas of the stomach and oesophagus. Occup Environ Med, 61(9), 743–749. 10.1136/oem.2003.011858 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Leon ME, Schinasi LH, Lebailly P, Beane Freeman LE, Nordby KC, Ferro G, Monnereau A, Brouwer M, Tual S, Baldi I, Kjaerheim K, Hofmann JN, Kristensen P, Koutros S, Straif K, Kromhout H, & Schuz J (2019). Pesticide use and risk of non-Hodgkin lymphoid malignancies in agricultural cohorts from France, Norway and the USA: a pooled analysis from the AGRICOH consortium. Int J Epidemiol, 48(5), 1519–1535. 10.1093/ije/dyz017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lerro CC, Beane Freeman L, & DellaValle CT (2018). Occupational pesticide exposure and subclinical hypothyroidism among male pesticide applicators. Occup Environ Med, 75, 79–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lerro CC, Beane Freeman LE, DellaValle CT, Andreotti G, Hofmann JN, Koutros S, Parks CG, Shrestha S, Alavanja MCR, Blair A, Lubin JH, Sandler DP, & Ward MH (2021). Pesticide exposure and incident thyroid cancer among male pesticide applicators in agricultural health study. Environ Int, 146, 106187. 10.1016/j.envint.2020.106187 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Mahajan R, Blair A, Coble J, Lynch CF, Hoppin JA, Sandler DP, & Alavanja MC (2007). Carbaryl exposure and incident cancer in the Agricultural Health Study. Int J Cancer, 121(8), 1799–1805. 10.1002/ijc.22836 [DOI] [PubMed] [Google Scholar]
  33. Manley CK, Spaur M, Madrigal JM, Fisher JA, Jones RR, Parks CG, Hofmann JN, Sandler DP, Beane Freeman L, & Ward MH (2022). Drinking water sources and water quality in a prospective agricultural cohort. Environ Epidemiol, 6(3), e210. 10.1097/ee9.0000000000000210 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Morton LM, Turner JJ, Cerhan JR, Linet MS, Treseler PA, Clarke CA, Jack A, Cozen W, Maynadié M, Spinelli JJ, Costantini AS, Rüdiger T, Scarpa A, Zheng T, & Weisenburger DD (2007). Proposed classification of lymphoid neoplasms for epidemiologic research from the Pathology Working Group of the International Lymphoma Epidemiology Consortium (InterLymph). Blood, 110(2), 695–708. 10.1182/blood-2006-11-051672 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. National Cancer Institute. (1982). National Cancer Institute Sponsored Study of Classifications of Non-Hodgkin’s Lymphomas - Summary and Description of a Working Formulation for Clinical Usage. Cancer, 49(10), 2112–2135. 10.1002/1097-0142(19820515)49:10<2112::AID-CNCR2820491024>3.0.CO;2-2 [DOI] [PubMed] [Google Scholar]
  36. Onfelt A, & Klasterska I (1984). Sister-chromatid exchanges and thioguanine resistance in v79 Chinese hamster cells after treatment with the aneuploidy-inducing agent carbaryl +/− S9 mix. Mutation Research, 125, 269–274. 10.1016/0027-5107(84)90076-9 [DOI] [PubMed] [Google Scholar]
  37. Piel C, Pouchieu C, Migault L, Beziat B, Boulanger M, Bureau M, Carles C, Gruber A, Lecluse Y, Rondeau V, Schwall X, Tual S, Lebailly P, Baldi I, & group, A. (2019). Increased risk of central nervous system tumours with carbamate insecticide use in the prospective cohort AGRICAN. Int J Epidemiol, 48(2), 512–526. 10.1093/ije/dyy246 [DOI] [PubMed] [Google Scholar]
  38. Renglin A, Harmala-Brasken A-S, Eriksson JE, & Onfelt A (1999). Mitotic aberrations induced by carbaryl reflect tyrosine kinase inhibition with coincident up-regulation of serine/threonine protein phosphatase activity: implications for coordination of karyokinesis and cytokinesis. Mutagenesis, 14(3), 327–333. 10.1093/mutage/14.3.327 [DOI] [PubMed] [Google Scholar]
  39. Saramaki O, & Visakorpi T (2007). Chromosomal aberrations in prostate cancer. Front Biosci, 12, 3287–3301. 10.2741/2312 [DOI] [PubMed] [Google Scholar]
  40. Shah SC, Boffetta P, Johnson KC, Hu J, Palli D, Ferraroni M, Tsugane S, Hamada GS, Hidaka A, Zaridze D, Maximovich D, Vioque J, Navarrete-Munoz EM, Zhang Z-F, Mu L, Boccia S, Pastorino R, Kurtz RC, Rota M, Bonzi R, Negri E, La Vecchia C, Pelucchi C, & Hashim D (2020). Occupational exposures and odds of gastric cancer: a StoP project consortium pooled analysis. International Journal of Epidemiology, 49(2), 422–434. 10.1093/ije/dyz263 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Smith MT, Guyton KZ, Gibbons CF, Fritz JM, Portier CJ, Rusyn I, DeMarini DM, Caldwell JC, Kavlock RJ, Lambert PF, Hecht SS, Bucher JR, Stewart BW, Baan RA, Cogliano VJ, & Straif K (2016). Key Characteristics of Carcinogens as a Basis for Organizing Data on Mechanisms of Carcinogenesis. Environ Health Perspect, 124(6), 713–721. 10.1289/ehp.1509912 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Tange S, Fujimoto N, Uramaru N, Wong FF, Sugihara K, Ohta S, & Kitamura S (2016). In vitro metabolism of methiocarb and carbaryl in rats, and its effect on their estrogenic and antiandrogenic activities. Environ Toxicol Pharmacol, 41, 289–297. 10.1016/j.etap.2015.08.014 [DOI] [PubMed] [Google Scholar]
  43. Thomas KW, Dosemeci M, Coble JB, Hoppin JA, Sheldon LS, Chapa G, Croghan CW, Jones PA, Knott CE, Lynch CF, Sandler DP, Blair AE, & Alavanja MC (2010). Assessment of a pesticide exposure intensity algorithm in the agricultural health study. Journal of Exposure Science & Environmental Epidemiology, 20(6), 559–569. 10.1038/jes.2009.54 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. United States Environmental Protection Agency. (2008). Amended Reregistration Eligibility Decision (RED) for Carbaryl (EPA-738-R-08-010). https://archive.epa.gov/pesticides/reregistration/web/pdf/carbaryl-red-amended.pdf
  45. United States Environmental Protection Agency. (2020). Carbaryl: National and State Summary Use and Usage Matrix. https://www3.epa.gov/pesticides/nas/carbaryl/appendix-1-4.pdf
  46. United States Environmental Protection Agency. (2021. ). Carbaryl: Revised Draft Human Health Risk Assessment in Support of Registration Review. https://www.regulations.gov/document/EPA-HQ-OPP-2010-0230-0074
  47. Wang Z, Lin Y, Jiang Y, Fu R, Wang Y, & Zhang Q (2022). The associations between thyroid-related hormones and the risk of thyroid cancer: An overall and dose-response meta-analysis. Front Endocrinol (Lausanne), 13, 992566. 10.3389/fendo.2022.992566 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Ward M, Jones R, Brender J, De Kok T, Weyer P, Nolan B, Villanueva C, & Van Breda S (2018). Drinking Water Nitrate and Human Health: An Updated Review. Int J Environ Res Public Health, 15(7), 1557. 10.3390/ijerph15071557 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2094247-supplement-1.docx (30KB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

RESOURCES