Abstract

Pesticides are associated with excess risk of multiple myeloma, albeit inconclusively. We included 678 men (30-94 years) from a well-characterized prospective cohort of restricted-use pesticide applicators to assess the risk of monoclonal gammopathy of undetermined significance (MGUS). Serum samples from all subjects were analyzed by electrophoresis performed on agarose gel; samples with a discrete or localized band were subjected to immunofixation. Age-adjusted prevalence estimates of MGUS were compared with MGUS prevalence in 9469 men from Minnesota. Associations between pesticide exposures and MGUS prevalence were assessed by logistic regression models adjusted for age and education level. Among study participants older than 50 years (n = 555), 38 were found to have MGUS, yielding a prevalence of 6.8% (95% CI, 5.0%-9.3%). Compared with men from Minnesota, the age-adjusted prevalence of MGUS was 1.9-fold (95% CI, 1.3- to 2.7-fold) higher among male pesticide applicators. Among applicators, a 5.6-fold (95% CI, 1.9- to 16.6-fold), 3.9-fold (95% CI, 1.5- to 10.0-fold), and 2.4-fold (95% CI, 1.1- to 5.3-fold) increased risk of MGUS prevalence was observed among users of the chlorinated insecticide dieldrin, the fumigant mixture carbon-tetrachloride/carbon disulfide, and the fungicide chlorothalonil, respectively. In summary, the prevalence of MGUS among pesticide applicators was twice that in a population-based sample of men from Minnesota, adding support to the hypothesis that specific pesticides are causatively linked to myelomagenesis.

Introduction

Multiple myeloma is a clonal neoplasm of differentiated B cells (plasma cells) characterized by an overproduction of monoclonal immunoglobulins with evidence of hypercalcemia, renal insufficiency, anemia, or bone lesions.1,2  According to the American Cancer Society, almost 19 900 new multiple myeloma cases and 10 700 multiple myeloma deaths are expected in the United State during 2008.3  Multiple myeloma is usually preceded by the premalignant plasma cell disorder, monoclonal gammopathy of undetermined significance (MGUS). MGUS is defined by a serum monoclonal immunoglobulin concentration less than 3 g/dL; a proportion of plasma cells in the bone marrow less than 10%1 ; and the absence of lytic bone lesions, anemia, hypercalcemia, or renal insufficiency related to the proliferation of monoclonal plasma cells. On average, MGUS progresses to multiple myeloma at a rate of 1% per year.4 

Although the cause of MGUS and multiple myeloma remain largely unclear, previous cohort5-12  and case-control studies13-26  have reported an elevated risk of multiple myeloma among farmers and other agricultural workers. More specifically, pesticides (ie, insecticides, herbicides, fungicides) have been hypothesized as the basis for these associations.27-30  However, most prior investigations have been hampered by small numbers and limited exposure assessment.31  In the US Agricultural Health Study, in a prospective cohort of 57 310 private and commercial licensed applicators of restricted use of pesticides in Iowa and North Carolina, we found a 1.34-fold (95% confidence interval [CI], 0.97-1.81) excess risk of multiple myeloma among pesticide applicators compared with population rates in Iowa and North Carolina.32  Several pesticides widely used on farms and in homes and gardens by the general public were associated with increased multiple myeloma risk in previous analyses coming from this cohort.33-36  Currently, however, it is unclear whether the observed increased risk of multiple myeloma among persons exposed to pesticides might reflect a higher prevalence of MGUS or an increase in the rate of progression from MGUS to multiple myeloma.

We have conducted the first population-based study of MGUS in relation to pesticide exposure in a sample of 678 male pesticide applicators. The aims of our study were to estimate the age-specific prevalence of MGUS among pesticide applicators and to compare the prevalence to that in the general population as determined in a population-based screening study in Olmsted County, Minnesota.37  In addition, we assessed the prevalence of MGUS in relation to specific pesticide reportedly used by these farmers.

Methods

Study subjects

The Agricultural Health Study is a prospective cohort study of 57 310 private and commercial applicators licensed to apply restricted-use pesticides who lived in Iowa or North Carolina and who were enrolled between 1993 and 1997.38  Applicators completed a self-administered questionnaire at enrollment. Comprehensive occupational exposure information was obtained for 22 frequently used pesticides, and ever/never use was obtained for 28 additional pesticides for which more detailed exposure data were obtained in a take-home questionnaire. Detailed information included mean annual days of use of the individual pesticides, years of use, use of personal protective equipment while applying pesticides, pesticide application methods, how frequently the applicator mixed pesticides, and whether pesticide equipment was personally repaired by the study subject. For all participants, information was obtained on smoking and alcohol use, cancer history of first-degree relatives, and other basic demographic and health information.38  Occupational exposures, medical histories, and lifestyle factors were updated at a 5-year follow-up interview. All questionnaires may be accessed at http://www.aghealth.org/questionnaires.html. Cancer incidence, mortality, and changes in address are monitored annually.38 

A stratified random sample (based on lifetime organophosphate use) of 685 male study subjects, who completed all 3 phases of the Agricultural Health Study, were enrolled into a neurobehavioral study nested within the cohort and provided serum for analysis. For Iowa and North Carolina study subjects, phlebotomy was performed in 2006-2007 and 2008, respectively. Because of the low prevalence of women among the applicators in the cohort (2%), women were excluded from this study. On the basis of diagnostic criteria for MGUS,1  persons with a prior history of a lymphoproliferative malignancy (ie, multiple myeloma or lymphoma) were excluded (n = 7) in our study. Thus, a total of 678 study subjects were included in this investigation. All participants provided signed informed consent at time of blood draw, in accordance with the Declaration of Helsinki, and the protocol was approved by the Institutional Review Boards of the National Institutes of Health and its contractors.

Collection of biologic samples

Venous blood was collected from the antecubital vein with the use of standard aseptic techniques. Blood was processed in the field and stored in a secure −20°C freezer on the testing site.

Laboratory tests

All serum samples were processed and analyzed for MGUS in an identical fashion and in the same laboratory (Mayo Clinic Protein Immunology Laboratory, Rochester, MN) as the population-based study of MGUS in Olmsted County, Minnesota, to which the rates of MGUS in the Agricultural Health Study are compared.37  Electrophoresis was performed on agarose gel (REP; Helena Laboratories, Beaumont, TX). The agarose strip was inspected by a technician and by 2 of the authors (R.A.K. and J.A.K.). Any serum with a discrete band or thought to have a localized band was subjected to immunofixation (Hydrasys and Hydragel; Sebia, Norcross, GA).39  MGUS was defined in accordance with previous definition, which was identical to the definition used in the Olmsted County prevalence study.37,39 

Statistical analysis

Age-specific prevalence rates for pesticide applicators and Olmsted County men were calculated by dividing the number of persons with MGUS in each age stratum by the number of subjects in that stratum. Associations of MGUS prevalence with pesticide exposures, demographics, and subject characteristics were assessed in logistic regression models (PROC LOGISTIC, SAS 9.1; SAS Institute, Cary, NC) adjusted for age and education level. All P values are 2-sided. For every significant association between a specific pesticide and MGUS, we evaluated the 5 most highly correlated pesticides with the pesticide of interest as a potential source of confounding. We also assessed the potential confounding effect of other pesticides that had a significant association with MGUS.

Results

The median age of study subjects was 60 years (range, 30-94 years). Ninety-five percent were white, 1.9% were African American, and the remaining 3.1% included all other racial groups and those with missing racial data. By design, 50% of the pesticide applicators were from North Carolina and 50% were from Iowa.

Prevalence of MGUS

MGUS was detected in the serum of 38 (5.6%) of the 678 participants, and the prevalence varied significantly by age (P < .001, χ2 test; Table 1). The age-specific prevalence rate of MGUS among pesticide applicators aged 50 to 59 years, 60 to 69 years, 70 to 79 years, and 80 years or older was 2.3%, 5.5%, 14.0%, and 16.7%, respectively. No MGUS cases were observed among the 123 study subjects younger than 50 years, but the prevalence of MGUS was 6.8% (95% CI, 5.0%-9.3%) among Agricultural Health Study subjects who were 50 years of age or older. The results were similar when we stratified the analyses by state (data not shown).

Table 1

MGUS prevalence (%) among 678 men in the Agricultural Health Study and 9469 men in Olmsted County, MN

Age group, yAgricultural Health Study Cohort
Olmsted County, MN37
P*
Total, nMGUS, nPrevalence (95% CI)Total, nMGUS, nPrevalence (95% CI)
< 50 123 NA  
50-59 214 2.3 (1.0-5.4) 4038 82 2.0 (1.6-2.5)  
60-69 182 10 5.5 (3.0-9.9) 2864 105 3.7 (3.0-4.4)  
70-79 129 18 14.0 (9.0-21.1) 1858 104 5.6 (4.6-6.7)  
> 80 30 16.7 (7.1-34.3) 709 59 8.3 (6.5-10.6)  
Total 555 38 6.8 (5.0-9.3) 9469 350 3.7 (3.3-4.1) < .001 
Age group, yAgricultural Health Study Cohort
Olmsted County, MN37
P*
Total, nMGUS, nPrevalence (95% CI)Total, nMGUS, nPrevalence (95% CI)
< 50 123 NA  
50-59 214 2.3 (1.0-5.4) 4038 82 2.0 (1.6-2.5)  
60-69 182 10 5.5 (3.0-9.9) 2864 105 3.7 (3.0-4.4)  
70-79 129 18 14.0 (9.0-21.1) 1858 104 5.6 (4.6-6.7)  
> 80 30 16.7 (7.1-34.3) 709 59 8.3 (6.5-10.6)  
Total 555 38 6.8 (5.0-9.3) 9469 350 3.7 (3.3-4.1) < .001 

CI indicates confidence interval; and NA, not applicable.

*

χ2P < .001; logistic regression adjusted for age yields an odds ratio = 1.9 (95% CI, 1.3-2.7) comparing MGUS positive in the Agricultural Health Study versus Olmsted County, MN.37 

Per 100 persons.

Older than 50 years of age.

The age-adjusted prevalence of MGUS was significantly higher (P < .001) in this study of Agricultural Health Study participants compared with the rate 3.7% (95% CI, 3.3%-4.1%) observed among 9469 Olmsted County men.33  The age-adjusted prevalence ratio was 1.9 (95% CI, 1.3-2.7; Table 1).

Selected demographic factors and risk of MGUS prevalence

We explored age-specific prevalence patterns by demographic factors within the study. The increase in MGUS prevalence with age is shown in Table 2, and we found those with more than 12 years of formal education to have 60% nonsignificantly reduced prevalence of MGUS compared with those who did not graduate from high school. This inverse association was not explained by age disparities or any other demographic variable. There were too few African Americans in our cohort to evaluate racial differences in prevalence.

Table 2

Risk of MGUS among male Agricultural Health Study participants stratified by selected demographic and lifestyle factors

FactorTotal, nMGUS, nOR (95% CI)*
Age, y    
    < 65 452 13 1.0 (ref) 
    66-74 141 13 3.4 (1.6-7.6) 
    > 75 85 12 5.6 (2.4-12.6) 
Race    
    White 645 36 1.0 (ref) 
    Nonwhite 17 1.3 (0.2-10.4) 
    Missing 16 1.0 (0.1-8.3) 
State    
    Iowa 338 18 1.0 (ref) 
    North Carolina 340 20 1.0 (0.5-1.9) 
Education    
    < 12 37 1.0 (ref) 
    12 295 21 0.7 (0.2-2.1) 
    > 12 320 11 0.4 (0.1-1.1) 
    Other/missing 26 0.3 (0.1-2.8) 
Smoking    
    Never 374 22 1.0 (ref) 
    Current 223 16 1.1 (0.5-2.1) 
First-degree relative with cancer    
    No 296 16 1.0 (ref) 
    Yes 360 22 1.0 (0.5-2.0) 
FactorTotal, nMGUS, nOR (95% CI)*
Age, y    
    < 65 452 13 1.0 (ref) 
    66-74 141 13 3.4 (1.6-7.6) 
    > 75 85 12 5.6 (2.4-12.6) 
Race    
    White 645 36 1.0 (ref) 
    Nonwhite 17 1.3 (0.2-10.4) 
    Missing 16 1.0 (0.1-8.3) 
State    
    Iowa 338 18 1.0 (ref) 
    North Carolina 340 20 1.0 (0.5-1.9) 
Education    
    < 12 37 1.0 (ref) 
    12 295 21 0.7 (0.2-2.1) 
    > 12 320 11 0.4 (0.1-1.1) 
    Other/missing 26 0.3 (0.1-2.8) 
Smoking    
    Never 374 22 1.0 (ref) 
    Current 223 16 1.1 (0.5-2.1) 
First-degree relative with cancer    
    No 296 16 1.0 (ref) 
    Yes 360 22 1.0 (0.5-2.0) 

OR indicates odds ratio; CI, confidence interval; and NA, not applicable.

*

Estimates are adjusted for age.

Specific pesticide exposures and risk of MGUS prevalence

To improve our understanding on the observed increased risk of MGUS prevalence among pesticide users, we evaluated the potential association for 50 specific pesticides for which we have usage data in the Agricultural Health Study.

As shown in Table 3, we found a 5.6-fold (95% CI, 1.9- to 16.6-fold), 3.9-fold (95% CI, 1.5- to 10.0-fold), and 2.6-fold (95% CI, 1.2- to 5.7-fold) significantly increased risk of MGUS prevalence among users of the chlorinated insecticide dieldrin, the fumigant mixture carbon-tetrachloride/carbon disulfide, and the fungicide chlorthalonil, respectively. Several other insecticides, herbicides, and fungicides were associated with MGUS, however, not significantly (Table 3).

Table 3

Specific pesticide use at enrollment and risk of MGUS IN 2008 among 678 male applicators in the Agricultural Health Study

Pesticide/categoryExposedTotal, nMGUS, nOR (95% CI)*
Chlorinated insecticides     
    Dieldrin Never 649 31 1.0 (reference) 
 Ever 20 5.6 (1.9-16.6) 
    Lindane Never 539 26 1.0 (reference) 
 Ever 138 12 1.9 (0.9-3.9) 
    Aldrin Never 546 24 1.0 (reference) 
 Ever 124 13 1.7 (0.8-3.6) 
    Toxaphene Never 525 24 1.0 (reference) 
 Ever 150 14 1.6 (0.8-3.2) 
    DDT Never 468 18 1.0 (reference) 
 Ever 205 20 1.4 (0.7-3.1) 
    Chlordane Never 486 24 1.0 (reference) 
 Ever 185 14 1.3 (0.6-2.6) 
Herbicides     
    2,4-D Never 139 1.0 (reference) 
 Ever 537 33 1.8 (0.7-4.8) 
    Pendimethalin Never 373 17 1.0 (reference) 
 Ever 305 21 1.7 (0.8-3.3) 
    Imazethapyr Never 478 25 1.0 (reference) 
 Ever 192 13 1.6 (0.8-3.4) 
    Cyanazine Never 370 18 1.0 (reference) 
 Ever 301 20 1.4 (0.7-2.8) 
    Alachlor Never 217 11 1.0 (reference) 
 Ever 453 27 1.3 (0.6-2.8) 
    Butylate Never 479 27 1.0 (reference) 
 Ever 199 11 1.1 (0.5-2.4) 
    Atrazine Never 151 1.0 (reference) 
 Ever 525 30 1.1 (0.5-2.5) 
    Trifluralin Never 390 22 1.0 (reference) 
 Ever 279 16 1.1 (0.5-2.1) 
    Chlorimuron-ethyl Never 442 24 1.0 (reference) 
 Ever 236 14 1.1 (0.5-2.2) 
    Metribuzin Never 456 25 1.0 (reference) 
 Ever 222 13 1.0 (0.5-2.1) 
    Dicamba Never 352 21 1.0 (reference) 
 Ever 316 17 0.9 (0.5-1.8) 
    Metolachlor Never 316 21 1.0 (reference) 
 Ever 356 17 0.8 (0.4-1.5) 
    Glyphosate Never 108 11 1.0 (reference) 
 Ever 570 27 0.5 (0.2-1.0) 
Fumigants     
    Carbon-tetrachloride/carbon disulfide mix Never 632 31 1.0 (reference) 
 Ever 41 3.9 (1.5-10.0) 
Fungicides     
    Chlorthalonil Never 561 28 1.0 (reference) 
 Ever 115 10 2.4 (1.1-5.3) 
    Captan Never 558 29 1.0 (reference) 
 Ever 115 1.9 (0.8-4.2) 
    Metalaxyl Never 476 25 1.0 (reference) 
 Ever 202 13 1.4 (0.7-2.9) 
Insecticides     
    Diazinon Never 435 19 1.0 (reference) 
 Ever 242 19 1.8 (0.9-3.6) 
    Chlorpyrifos Never 305 16 1.0 (reference) 
 Ever 367 21 1.3 (0.7-2.7) 
    Phorate Never 457 23 1.0 (reference) 
 Ever 220 15 1.3 (0.6-2.5) 
    Carbofuran Never 405 20 1.0 (reference) 
 Ever 266 17 1.2 (0.6-2.5) 
    Fonofos Never 489 26 1.0 (reference) 
 Ever 181 11 1.2 (0.5-2.4) 
    Carbaryl Never 281 15 1.0 (reference) 
 Ever 396 23 1.0 (0.5-2.0) 
    Malathion Never 161 11 1.0 (reference) 
 Ever 516 27 0.7 (0.3-1.5) 
    Terbufos Never 345 22 1.0 (reference) 
 Ever 326 15 0.7 (0.3-1.4) 
Pesticide/categoryExposedTotal, nMGUS, nOR (95% CI)*
Chlorinated insecticides     
    Dieldrin Never 649 31 1.0 (reference) 
 Ever 20 5.6 (1.9-16.6) 
    Lindane Never 539 26 1.0 (reference) 
 Ever 138 12 1.9 (0.9-3.9) 
    Aldrin Never 546 24 1.0 (reference) 
 Ever 124 13 1.7 (0.8-3.6) 
    Toxaphene Never 525 24 1.0 (reference) 
 Ever 150 14 1.6 (0.8-3.2) 
    DDT Never 468 18 1.0 (reference) 
 Ever 205 20 1.4 (0.7-3.1) 
    Chlordane Never 486 24 1.0 (reference) 
 Ever 185 14 1.3 (0.6-2.6) 
Herbicides     
    2,4-D Never 139 1.0 (reference) 
 Ever 537 33 1.8 (0.7-4.8) 
    Pendimethalin Never 373 17 1.0 (reference) 
 Ever 305 21 1.7 (0.8-3.3) 
    Imazethapyr Never 478 25 1.0 (reference) 
 Ever 192 13 1.6 (0.8-3.4) 
    Cyanazine Never 370 18 1.0 (reference) 
 Ever 301 20 1.4 (0.7-2.8) 
    Alachlor Never 217 11 1.0 (reference) 
 Ever 453 27 1.3 (0.6-2.8) 
    Butylate Never 479 27 1.0 (reference) 
 Ever 199 11 1.1 (0.5-2.4) 
    Atrazine Never 151 1.0 (reference) 
 Ever 525 30 1.1 (0.5-2.5) 
    Trifluralin Never 390 22 1.0 (reference) 
 Ever 279 16 1.1 (0.5-2.1) 
    Chlorimuron-ethyl Never 442 24 1.0 (reference) 
 Ever 236 14 1.1 (0.5-2.2) 
    Metribuzin Never 456 25 1.0 (reference) 
 Ever 222 13 1.0 (0.5-2.1) 
    Dicamba Never 352 21 1.0 (reference) 
 Ever 316 17 0.9 (0.5-1.8) 
    Metolachlor Never 316 21 1.0 (reference) 
 Ever 356 17 0.8 (0.4-1.5) 
    Glyphosate Never 108 11 1.0 (reference) 
 Ever 570 27 0.5 (0.2-1.0) 
Fumigants     
    Carbon-tetrachloride/carbon disulfide mix Never 632 31 1.0 (reference) 
 Ever 41 3.9 (1.5-10.0) 
Fungicides     
    Chlorthalonil Never 561 28 1.0 (reference) 
 Ever 115 10 2.4 (1.1-5.3) 
    Captan Never 558 29 1.0 (reference) 
 Ever 115 1.9 (0.8-4.2) 
    Metalaxyl Never 476 25 1.0 (reference) 
 Ever 202 13 1.4 (0.7-2.9) 
Insecticides     
    Diazinon Never 435 19 1.0 (reference) 
 Ever 242 19 1.8 (0.9-3.6) 
    Chlorpyrifos Never 305 16 1.0 (reference) 
 Ever 367 21 1.3 (0.7-2.7) 
    Phorate Never 457 23 1.0 (reference) 
 Ever 220 15 1.3 (0.6-2.5) 
    Carbofuran Never 405 20 1.0 (reference) 
 Ever 266 17 1.2 (0.6-2.5) 
    Fonofos Never 489 26 1.0 (reference) 
 Ever 181 11 1.2 (0.5-2.4) 
    Carbaryl Never 281 15 1.0 (reference) 
 Ever 396 23 1.0 (0.5-2.0) 
    Malathion Never 161 11 1.0 (reference) 
 Ever 516 27 0.7 (0.3-1.5) 
    Terbufos Never 345 22 1.0 (reference) 
 Ever 326 15 0.7 (0.3-1.4) 

As described in “Methods,” a total of 50 specific pesticides were evaluated in this study. Unless there was a significant association with MGUS/MM, we only provide information for specific pesticides with more than 10 MGUS cases in the “ever” exposed category.

OR indicates odds ratio; and CI, confidence interval.

Underlined, italicized entries are statistically significant (P < 0.05).

*

Estimates are adjusted for age and education level.

The excess risk of MGUS prevalence with dieldrin and chlorthalonil use was not attenuated and remained significant after adjusting for the influence of the use of other pesticides with a potential for confounding (data not shown). Similarly, the excess risk of MGUS prevalence with the use of the mixture carbon-tetrachloride/carbon disulfide was virtually the same when we controlled for other pesticides; however, the 95% CIs on the adjusted risks estimates sometimes included one. Finally, in exploratory subanalysis, we assessed risk of MGUS prevalence by lifetime days (below/above median) of exposure for specific pesticides, but in general numbers were too small to be informative.

Discussion

Beyond age, sex, race, and a positive family history of MGUS/multiple myeloma, no consistent extrinsic risk factors have been clearly linked to multiple myeloma.5-26,40,41  Previous studies from around the world suggest that agricultural work is associated with multiple myeloma risk, but specific agents have not beenidentified.5-26,40  Compared with population rates in Iowa and North Carolina, a 1.3-fold excess risk of multiple myeloma was observed in the Agricultural Health Study.32  Although there is no clear explanation for this excess, pesticide-specific analyses from the Agricultural Health Study have suggested associations with a few pesticides. In our survey of MGUS among a subset of Agricultural Health Study pesticide applicators ages 50 years and older, we found a prevalence of 6.8%, which is twice that of a group of a population-based sample of men from Minnesota of comparable age.37 

Our analyses point to possible links with dieldrin, a chlorinated insecticide, which had a significant 5-fold excess risk of MGUS prevalence, and carbon tetrachloride/carbon disulfide mix, a fumigant, which had a significant 4-fold risk of MGUS prevalence. Several other chlorinated insecticides were associated, but not significantly, with MGUS (Table 3). Most of these chlorinated compounds have been taken off the market in the United States and most other developed nations, but their use persists in many developing nations. With a few exceptions, organochlorine pesticides are fat soluble and persist in adipose tissue of many persons.42 

We also found a 3-fold significant excess of MGUS prevalence among users of chlorthalonil. Chlorthalonil is a fungicide with broad applications to fruits and vegetables. This pesticide has not been evaluated for the association with multiple myeloma in the Agricultural Health Study because of relatively small numbers of exposed cases.

Permethrin use has been linked to multiple myeloma in the Agricultural Health Study32  and a nonsignificantly increased risk of multiple myeloma was seen among study participants using the widely used herbicides atrazine35  and glyphosate33  and the insecticide chlorpyrifos.34  In the present study, we had only 4 MGUS-positive exposed cases, and we could not reliably evaluate the association between MGUS and permethrin exposure. Among users of chlorpyrifos and atrazine we observed a nonsignificant elevated risk of MGUS, whereas for users of glyphosate we found a nonsignificant decreased risk of MGUS. Chlorpyrifos, an organophosphorothioate insecticide, was previously used in home and garden applications; however, its use is now restricted to agriculture.34  Atrazine is used primarily on corn and soybean to control broadleaf and grassy weeds, and it is one of the most heavily used agricultural pesticides in the United States.43 

The Agricultural Health Study has several important strengths. It is the largest study to date of private and commercial applicators licensed to apply restricted-use pesticides. Recall bias was minimized because exposure information was collected before blood draw and serum protein analysis. Furthermore, it has been shown that farmers provide accurate and reliable information and considerable detail about their pesticide application history.44 

Our study also has some limitations. For example, we did not have a control group from Iowa and North Carolina. Instead, we chose Olmsted County in Minnesota as a comparison group, given that the Mayo Clinic had available data from the largest population-based MGUS screening study to date. Although a control group from Iowa and North Carolina might have been better, we think that Olmsted County in Minnesota is a reasonable control group. In fact, the difference between the 2 groups (ie, Olmsted County and the Agricultural Health Study) are not particularly different by education attainment (90.3% of the Agricultural Health Study cohort have graduated from high school and 40.7% have some education beyond high school; in Olmsted County 91.1% were high school graduates or higher). The populations of both groups are predominantly white, and the age structure is also similar as shown in Table 2. Two-thirds of the Agricultural Health Study participants are from the Midwest (ie, Iowa), as are all the Olmsted County participants. We evaluated only the male population in both locations. In the present study, only 2 demographic variables were associated with a higher prevalence of MGUS: age and education status. We used both of these as adjustment variables in our multivariate analyses. We think, therefore, that our evaluation of the prevalence of MGUS in each of the populations provides a meaningful comparison and that the analyses by individual pesticides suggest pesticides may play an important role in causing the excess in the Agricultural Health Study cohort. Finally, because we explored the potential association between MGUS and 50 specific pesticides for which we have usage data in the Agricultural Health Study, one has to interpret detected associations for specific pesticides with caution. Future larger studies are needed to replicate our findings.

In summary, several million Americans use pesticides for which we have found an association with MGUS in the Agricultural Health Study. Some of these same chemicals have been associated with excess multiple myeloma risk. Importantly, a recent investigation based on 77 469 healthy adults enrolled in a US nationwide population-based prospective cancer screening trial identified 71 persons who developed multiple myeloma during the course of the study. With the use of serially collected prediagnostic serum samples obtained up to almost 10 years before multiple myeloma diagnosis, all multiple myeloma cases were found to be preceded by MGUS.45  This finding establishes a key role for MGUS in the pathway to multiple myeloma. In turn, it suggests that our present observation of pesticide exposure being associated with excess MGUS risk might be an underlying explanation of the previously observed excess multiple myeloma risk among persons exposed to pesticides.32-36  Future studies are needed to improve our knowledge on the role of pesticide exposure in the pathogenesis of MGUS, as well as the potential role in progression from MGUS to multiple myeloma. Identifying specific exposures responsible for myelomagenesis will be important to better understand chemical carcinogenesis in humans and to reduce the risk of disease by taking appropriate public health action.

The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Acknowledgments

We thank Dr Dale Sandler, National Institute of Environmental Health Sciences (NIEHS); Drs. Aaron Blair, Joseph Coble, and Jay Lubin, National Cancer Institute (NCI); and Dr L. Joseph Melton III, Mayo Clinic, for scientific input; Dr Fredric Gerr and Ms Sarah Starks, University of Iowa, for providing us with blood samples from their Neurobehavioral Study that was nested within the Agricultural Cohort Study cohort; and Mr Stuart Long at Westat, and Mr Joe Barker at IMS, Inc, for computer programming.

This work was supported by NCI (research grants CA 62242 and CA 107476); the Intramural Research Program of the NIH, NCI, Division of Cancer Epidemiology and Genetics (DCEG); the NIEHS (Z01-ES049030); and the NCI (Z01-CP010119).

National Institutes of Health

Authorship

Contribution: O.L. and M.C.A. initiated this work and wrote the report. All authors were involved in the design of the study; obtained and analyzed data; were involved in the interpretation of the results; read, gave comments, and approved the final version of the manuscript; had full access to the data in the study; and take responsibility for the integrity of the data and the accuracy of the data analysis.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Ola Landgren, National Cancer Institute, National Institutes of Health, Center for Cancer Research, Medical Oncology Branch, 9000 Rockville Pike, Bldg 10/Rm 13N240, Bethesda, MD, 20892; e-mail: landgreo@mail.nih.gov.

References

1
International Myeloma Working Group
Criteria for the classification of monoclonal gammopathies, multiple myeloma and related disorders: a report of the International Myeloma Working Group.
Br J Haematol
2003
, vol. 
121
 (pg. 
749
-
757
)
2
Kyle
 
RA
Rajkumar
 
SV
Multiple myeloma.
N Engl J Med
2004
, vol. 
351
 (pg. 
1860
-
1873
)
3
American Cancer Society
Cancer Facts and Figures 2008.
2008
Atlanta, GA
American Cancer Society
4
Kyle
 
RA
Therneau
 
TM
Rajkumar
 
SV
, et al. 
A long-term study of prognosis in monoclonal gammopathy of undetermined significance.
N Engl J Med
2002
, vol. 
346
 (pg. 
564
-
569
)
5
Alavanja
 
MC
Blair
 
A
Merkle
 
S
Teske
 
J
Eaton
 
B
Mortality among agricultural extension agents.
Am J Ind Med
1988
, vol. 
14
 (pg. 
167
-
176
)
6
Baris
 
D
Silverman
 
DT
Brown
 
LM
, et al. 
Occupation, pesticide exposure and risk of multiple myeloma.
Scand J Work Environ Health
2004
, vol. 
30
 (pg. 
215
-
222
)
7
Cerhan
 
JR
Cantor
 
KP
Williamson
 
K
Lynch
 
CF
Torner
 
JC
Burmeister
 
LF
Cancer mortality among Iowa farmers: recent results, time trends, and lifestyle factors (United States).
Cancer Causes Control
1998
, vol. 
9
 (pg. 
311
-
319
)
8
Lee
 
E
Burnett
 
CA
Lalich
 
N
Cameron
 
LL
Sestito
 
JP
Proportionate mortality of crop and livestock farmers in the United States, 1984-1993.
Am J Ind Med
2002
, vol. 
42
 (pg. 
410
-
420
)
9
Nandakumar
 
A
English
 
DR
Dougan
 
LE
Armstrong
 
BK
Incidence and outcome of multiple myeloma in Western Australia, 1960 to 1984.
Aust N Z J Med
1988
, vol. 
18
 (pg. 
774
-
779
)
10
Pukkala
 
E
Notkola
 
V
Cancer incidence among Finnish farmers, 1979-93.
Cancer Causes Control
1997
, vol. 
8
 (pg. 
25
-
33
)
11
Stark
 
AD
Chang
 
HG
Fitzgerald
 
EF
Riccardi
 
K
Stone
 
RR
A retrospective cohort study of cancer incidence among New York State Farm Bureau members.
Arch Environ Health
1990
, vol. 
45
 (pg. 
155
-
162
)
12
Steineck
 
G
Wiklund
 
K
Multiple myeloma in Swedish agricultural workers.
Int J Epidemiol
1986
, vol. 
15
 (pg. 
321
-
325
)
13
Boffetta
 
P
Stellman
 
SD
Garfinkel
 
L
A case-control study of multiple myeloma nested in the American Cancer Society prospective study.
Int J Cancer
1989
, vol. 
43
 (pg. 
554
-
559
)
14
Brownson
 
RC
Reif
 
JS
Chang
 
JC
Davis
 
JR
Cancer risks among Missouri farmers.
Cancer
1989
, vol. 
64
 (pg. 
2381
-
2386
)
15
Burmeister
 
LF
Cancer in Iowa farmers: recent results.
Am J Ind Med
1990
, vol. 
18
 (pg. 
295
-
301
)
16
Cantor
 
KP
Blair
 
A
Farming and mortality from multiple myeloma: a case-control study with the use of death certificates.
J Natl Cancer Inst
1984
, vol. 
72
 (pg. 
251
-
255
)
17
Costantini
 
AS
Miligi
 
L
Kriebel
 
D
, et al. 
A multicenter case-control study in Italy on hematolymphopoietic neoplasms and occupation.
Epidemiology
2001
, vol. 
12
 (pg. 
78
-
87
)
18
Cuzick
 
J
De Stavola
 
B
Multiple myeloma–a case-control study.
Br J Cancer
1988
, vol. 
57
 (pg. 
516
-
520
)
19
Demers
 
PA
Vaughan
 
TL
Koepsell
 
TD
, et al. 
A case-control study of multiple myeloma and occupation.
Am J Ind Med
1993
, vol. 
23
 (pg. 
629
-
639
)
20
Eriksson
 
M
Karlsson
 
M
Occupational and other environmental factors and multiple myeloma: a population based case-control study.
Br J Ind Med
1992
, vol. 
49
 (pg. 
95
-
103
)
21
Figgs
 
LW
Dosemeci
 
M
Blair
 
A
Risk of multiple myeloma by occupation and industry among men and women: a 24-state death certificate study.
J Occup Med
1994
, vol. 
36
 (pg. 
1210
-
1221
)
22
Franceschi
 
S
Barbone
 
F
Bidoli
 
E
, et al. 
Cancer risk in farmers: results from a multi-site case-control study in north-eastern Italy.
Int J Cancer
1993
, vol. 
53
 (pg. 
740
-
745
)
23
Heineman
 
EF
Olsen
 
JH
Pottern
 
LM
Gomez
 
M
Raffn
 
E
Blair
 
A
Occupational risk factors for multiple myeloma among Danish men.
Cancer Causes Control
1992
, vol. 
3
 (pg. 
555
-
568
)
24
La Vecchia
 
C
Negri
 
E
D'Avanzo
 
B
Franceschi
 
S
Occupation and lymphoid neoplasms.
Br J Cancer
1989
, vol. 
60
 (pg. 
385
-
388
)
25
Milham
 
S
Leukemia and multiple myeloma in farmers.
Am J Epidemiol
1971
, vol. 
94
 (pg. 
507
-
510
)
26
Pearce
 
N
Reif
 
JS
Epidemiologic studies of cancer in agricultural workers.
Am J Ind Med
1990
, vol. 
18
 (pg. 
133
-
148
)
27
Blair
 
A
Dosemeci
 
M
Heineman
 
EF
Cancer and other causes of death among male and female farmers from twenty-three states.
Am J Ind Med
1993
, vol. 
23
 (pg. 
729
-
742
)
28
Blair
 
A
Zahm
 
SH
Pearce
 
NE
Heineman
 
EF
Fraumeni
 
JF
Clues to cancer etiology from studies of farmers.
Scand J Work Environ Health
1992
, vol. 
18
 (pg. 
209
-
215
)
29
Brown
 
LM
Burmeister
 
LF
Everett
 
GD
Blair
 
A
Pesticide exposures and multiple myeloma in Iowa men.
Cancer Causes Control
1993
, vol. 
4
 (pg. 
153
-
156
)
30
Khuder
 
SA
Mutgi
 
AB
Meta-analyses of multiple myeloma and farming.
Am J Ind Med
1997
, vol. 
32
 (pg. 
510
-
516
)
31
Alavanja
 
MC
Hoppin
 
JA
Kamel
 
F
Health effects of chronic pesticide exposure: cancer and neurotoxicity.
Annu Rev Public Health
2004
, vol. 
25
 (pg. 
155
-
197
)
32
Alavanja
 
MC
Sandler
 
DP
Lynch
 
CF
, et al. 
Cancer incidence in the agricultural health study.
Scand J Work Environ Health
2005
, vol. 
31
 
suppl 1
(pg. 
39
-
45
discussion 5–7
33
De Roos
 
AJ
Blair
 
A
Rusiecki
 
JA
, et al. 
Cancer incidence among glyphosate-exposed pesticide applicators in the Agricultural Health Study.
Environ Health Perspect
2005
, vol. 
113
 (pg. 
49
-
54
)
34
Lee
 
WJ
Blair
 
A
Hoppin
 
JA
, et al. 
Cancer incidence among pesticide applicators exposed to chlorpyrifos in the Agricultural Health Study.
J Natl Cancer Inst
2004
, vol. 
96
 (pg. 
1781
-
1789
)
35
Rusiecki
 
JA
De Roos
 
A
Lee
 
WJ
, et al. 
Cancer incidence among pesticide applicators exposed to atrazine in the Agricultural Health Study.
J Natl Cancer Inst
2004
, vol. 
96
 (pg. 
1375
-
1382
)
36
Rusiecki
 
JA
Patel
 
R
Koutros
 
S
, et al. 
Cancer incidence among pesticide applicators exposed to permethrin in the Agricultural Health Study.
Environ Health Perspect
2009
, vol. 
117
 (pg. 
581
-
586
)
37
Kyle
 
RA
Therneau
 
TM
Rajkumar
 
SV
, et al. 
Prevalence of monoclonal gammopathy of undetermined significance.
N Engl J Med
2006
, vol. 
354
 (pg. 
1362
-
1369
)
38
Alavanja
 
MC
Sandler
 
DP
McMaster
 
SB
, et al. 
The Agricultural Health Study.
Environ Health Perspect
1996
, vol. 
104
 (pg. 
362
-
369
)
39
Kyle
 
RA
Katzmann
 
JA
Lust
 
JA
Dispenzieri
 
A
Rose
 
NR
Hamilton
 
RG
Detrik
 
B
Immunochemical characterization of immunoglobulins.
Manual of Clinical Laboratory Immunology
2002
Washington, DC
ASM Press
(pg. 
71
-
91
)
40
De Roos
 
AJ
Baris
 
D
Weiss
 
NS
Herrington
 
LJ
Schottenfeld
 
D
Fraumeni
 
JF
Multiple myeloma.
Cancer Epidemiology and Prevention
2006
New York, NY
Oxford University Press
(pg. 
946
-
970
)
41
Landgren
 
O
Kristinsson
 
SY
Goldin
 
LR
, et al. 
Risk of plasma-cell and lymphoproliferative disorders among 14,621 first-degree relatives of 4458 patients with monoclonal gammopathy of undetermined significance (MGUS) in Sweden.
Blood
 
Prepublished on January 30, 2009, as DOI 10.1182/blood-2008-12-191676
42
Centers for Disease Control and Prevention
Third National Report on Human Exposure to Environmental Chemicals.
2005
Atlanta, GA
CDC
43
US Environmental Protection Agency
Pesticides: Topical and Chemical Fact Sheets.
2008
Accessed April 2009, vol. 
Vol 2008
 
Washington, DC
EPA
 
44
Blair
 
A
Tarone
 
R
Sandler
 
D
, et al. 
Reliability of reporting on life-style and agricultural factors by a sample of participants in the Agricultural Health Study from Iowa.
Epidemiology
2002
, vol. 
13
 (pg. 
94
-
99
)
45
Landgren
 
O
Kyle
 
RA
Pfeiffer
 
RM
, et al. 
Monoclonal gammopathy of undetermined significance (MGUS) precedes multiple myeloma: a prospective study.
Blood