Skip to main content
         

Vol. 10 Issue 2, Spring 2005

Agricultural Health Study
The Ribbon 

In 1993 enrollment began for the largest health study of the US farming community, the Agricultural Health Study (AHS). Twelve years later, data from the 89,658 enrollees in Iowa and North Carolina are beginning to generate the most comprehensive assessments of the link between individual pesticides and cancer risk ever published. The risk of various cancers is just one set of health endpoints monitored in this ground-breaking study; other health problems addressed include diseases of the nervous system, as well as respiratory and reproductive health. In The Ribbon we will focus on the work being done in this study to better understand pesticide exposures and any cancer-related outcomes. (Dr. Suzanne Snedeker specifically addresses the AHS work on pesticide exposure and breast cancer risk, in her Research Commentary on page 5.) We will also highlight the work being done as part of the AHS to help understand and reduce pesticide exposure.

There have been various epidemiological attempts to better understand the relationship between pesticide exposure and cancer risk. These, together with laboratory work, have presented a patchwork of data linking some pesticides with increased risk of some cancers. The large scale and comprehensive design of the AHS are meant to specifically address the weaknesses and gaps in prior research.

Enlisting the long-term support of a large portion of the farming communities of two states, Iowa and North Carolina - both with strong agricultural sectors with diverse agricultural methods and products - enables researchers to employ the strengths of prospective cohort studies. The chemical exposure and lifestyle information that is collected from these participants is the most detailed ever: participants responded to about 250 questions in the initial survey. Fifty pesticides were selected based on their widespread use and/or previous studies indicated their potential association with health risks, and 30 more were added, based on participant "write-ins." Follow-up surveys every five years provide scientists with updated information. Including spouses in the research is an important aspect that is providing data on women and pesticide use and exposure never before collected. There is also a small percentage of female certified (North Carolina) and licensed (Iowa) pesticide applicators included in the study. Overall, two-thirds of participants are applicators and one-third are spouses. With children also registered, the understanding of the health of farm families is being greatly enhanced by the AHS.

Cancer findings: early results

Previous research indicates that farmers experience some cancers to a lesser extent than the general population and some cancers to a greater extent. Fewer farmers die (results from mortality studies) from lung, esophagus, bladder and colon cancers, possibly the result of smoking less, eating a healthier diet and getting more physical exercise than the average American. But studies also suggest that farmers as a group experience higher rates of Hodgkin's disease, leukemia, multiple myeloma, non-Hodgkins lymphoma, and cancers of the lip, stomach, prostate, skin, brain, and connective tissue.

The AHS is now at the stage at which disease rates can be assessed, and the study is producing some important cancer findings. These findings reflect about five years of follow-up, following several years of enrollment and data collection (see box, "Study Timeline"). The two ways in which cancer rates are analyzed in the AHS are: 1) the comparison of cancer rates amongst enrollees with those of the general population, and 2) the comparison of cancer rates of those using specific pesticides with those who do not. A major finding - one that correlates with previously existing data - is that this population of farmers experiences lower overall cancer rates than the general population. The rates found thus far of all cancers, except prostate cancer, were lower. However, early findings do suggest an association between use of specific pesticides and an increased risk of specific cancers. It is important to note that discovering these associations does not mean a causal relationship has been found. The AHS employs a rigorous strategy of criteria for causal inference (see table, "Causal Inference in AHS: What are we looking for?" and more detail below).

These early findings of association between specific pesticide exposures and increased cancer risk include: prostate cancer and methyl bromide; immune/blood cancers and alachlor; lung cancer and chlorpyrifos and some evidence of increased risk of breast cancer associated with the use of several pesticides (on this topic please see Research Commentary is this issue).

Prostate cancer and pesticide exposure. This analysis found that the risk of prostate cancer was 14% higher among pesticide applicators than in the general populations of Iowa and North Carolina. Of the 45 pesticides analyzed for a relationship with prostate cancer, several showed an increased risk in men over 50 and in those with a family history of this disease. Researchers will be examining whether there exists a genetic-environment interaction in which men with a family history of prostate cancer are especially susceptible to carcinogenic effects of these pesticides. The pesticides that showed this relationship include chlorpyrifos, coumaphos, fonofos, phorate, permethrin and butylate.

One of the 45 pesticides tested, methyl bromide showed a "dose-response trend." In other words, the risk of prostate cancer increased with increased exposure to this soil fumigant. The highest exposure categories had a 3.5 times greater risk of prostate cancer. Fortunately, use of this soil fumigant, which is considered to be a potential occupational carcinogen by the National Institute for Occupational Safety and Health, is declining. This year the AHS will determine whether these initial findings on prostate cancer can be replicated.

Immune/blood cancers and alachlor. About half of the applicators in the AHS reported using the herbicide alachlor, which is classified by the US Environmental Protection Agency (EPA) as a probable human carcinogen. The AHS analysis did not find a relationship between alachlor exposure and those cancers in which previous animal studies had shown one (lung tumors in mice, and thyroid gland, stomach and nasal passage tumors in the rat models). But it did show a relationship between alachlor exposure and risk of blood and immune system cancers. These cancers include leukemia, multiple myeloma and non-Hodgkin's lymphoma collectively. Increased risk was observed with increased exposure.

Chlorpyrifos and lung cancer. In an AHS analysis of exposure to chlorpyrifos, one of the most widely used insecticides in the US, researchers found an association between use of this pesticide and the incidence of lung cancer, but not with any other cancer. In the highest exposure group, there was a 2.18 relative risk of lung cancer. This finding is of particular interest because farmers in general, and also those within the AHS cohort, have lower rates of lung cancer than the general population, probably because of lower smoking rates. (The chlorpyrifos-lung cancer association in this analysis was determined controlling for all other known cancer risk factors, including smoking.) In addition to its agricultural uses, chlorpyrifos was widely used in US households until 2000, and exposure to chlorpyrifos is the focus of an intensive exposure study within the AHS (see below).

Dr. Michael Alavanja of the National Cancer Institute (NCI) is the Director of the AHS. He describes the important junction at which the study has arrived:

"The AHS will now be generating ten to fifteen cancer etiology papers per year, and this will continue for the next four years (2005-2008). In all cases we will attempt to prevent false positive conclusions by limiting the interpretation of our findings until we observe dose-response findings in both Iowa and North Carolina and we can replicate the finding at two points in time. We are also planning molecular epidemiology to try to understand the mechanism of action of any positive finding."

The table below describes this approach: ideally all boxes will be "filled in" as shown. However, Dr. Alavanja cautions, initial dose-response findings in both states, across pesticide applicator license type, replicated in time, would suggest a causal relationship, even in the absence of biological evidence.

New insights regarding pesticide exposure

One of the major challenges in studying the effects of pesticide exposure on human health is the difficulty in precisely assessing real-life exposures. Pesticides are widely used without a detailed understanding of all the various ways they might enter the body and in what quantities they do so. As part of its study of the Iowa and North Carolina farming populations, the AHS has prioritized the improvement of scientific methods to better understand pesticide exposure. This work has revealed much information that enables scientists to more accurately determine potential health effects of pesticide exposure, and also assists in the development of better controls and practices to decrease exposure.

For example, early on in the study members of the research team looked closely at the characteristics of people who self-reported a "high pesticide exposure event." Closely examining these 14% of the study population enabled the epidemiologists to determine what home and farm features or practices are associated with incidents or experiences leading to unusually high personal exposures. These characteristics ranged from how, where and when work clothing was laundered and the types of pesticides being used, to whether the family was experiencing financial stress. After taking into account education and total number of lifetime applications made, researchers saw that women had significantly fewer events than men who applied pesticides. Job characteristics more common among those who reported a high pesticide exposure event included repairing pesticide application equipment oneself, and having first used pesticides more than 10 years ago. While the demographic, work practice and job characteristics identified in this investigation were not necessarily the cause of the high pesticide exposure, identifying these factors is a first step in the eventual prevention of these potentially hazardous events. Computer modeling - used as a tool in the AHS for exposure estimates and analyses, in addition to biological measurements - showed that following all pesticide label requirements could prevent many of these high exposure events.

Other analyses looked at exposure hazards of families of pesticide applicators. The design of the AHS takes into account that farmers "living where they work" presents potential exposures for spouses and children. Researchers found that 21% of homes are within 50 yards of the pesticide mixing area, that 27% of applicators store pesticides in their homes (including attached garage or basements), and that most pesticide-contaminated clothing is washed in the same machine as other laundry. At least half of the wives reported working the fields, 40% reported mixing or applying pesticides, and more than half of the children over age 11 did farm chores, some of which may have put them into contact with pesticides.

Researchers are taking this farm/home and work practice information and developing ways to determine the health implications of the resulting exposures. For example, scientists at the EPA conducted five-day continuous sampling of a group of volunteers in the AHS population, in order to gain a detailed understanding of exposure to two pesticides, 2, 4-D and chlorpyrifos. Biological samples from volunteering spouses and children were also collected. Exposure to these two pesticides is also being analyzed in a project that compares average exposures between different sub-populations in the AHS, for example between the two states, between male and female, and between farmers and commercial applicators. All of these data will enable a more complete analysis of exposures to these two widely used pesticides and any subsequent health outcomes.

Studies like these both refine exposure information for the AHS specifically, and develop methodologies for future pesticide exposure research. AHS computer models predicting pesticide exposure have already been found to be an effective predictor of exposure in a non-AHS population. AHS efforts also include comparisons between African-American farmers' and white farmers' pesticide use practices in North Carolina, and analyses of the use protective equipment and application method. Comparisons of work practices and attitudes toward risk between the two states are proving to be instructive. All of this research also clearly contributes to improved health and safety education and communication tailored to the farming community.

Keeping in touch with enrollees and communicating widely: current projects and future possibilities

The AHS has a strong, multi-faceted communication component. Ongoing communication with enrollees is a high priority. Two state-based entities, the University of Iowa College of Public Health and the Battelle Centers for Public Health Research and Evaluation in North Carolina, responsible for carrying out the study in their respective states, keep study participants closely informed of AHS progress. Julia Storm, an Agromedicine Information Specialist at North Carolina State University who has collaborated with the AHS on producing educational materials, says of the two state offices: "they are doing an excellent job putting the information that comes out of the study into a practical light for the farm families enrolled."

Storm and collaborators recently completed a project to communicate AHS study data to members of the cohort and other important audiences: those who may not be involved in the study, but for whom the information is critically important for improving health and safety in the broader agricultural community. This includes all farmers and farm families, pesticide safety educators, health professionals, and policy makers. The project, funded by the National Institute for Occupational Safety and Health through the Southern Coastal Agromedicine Center of the North Carolina Agromedicine Institute, produced three detailed, full-color, fold-out brochures about the study that are now available (these were the source of much information for this article; see ordering information on back cover). Julia Storm has also developed a PowerPoint presentation entitled "Understanding the Agricultural Health Study" that will soon be available for educators to use (if interested, please contact Storm at julia_storm@ncsu.edu). In addition, she assisted collaborators at the Eastern Area Health Education Center associated with the East Carolina University School of Medicine in developing an online continuing medical education course on pesticides and health that will be available in summer of 2005.

In 2004, Dr. Amy Brown, an Associate Professor at the University of Maryland, formed the AHS Risk Communication Work Group together with Dr. Snedeker and Ms. Storm, to offer a proposal to the AHS National Advisory Panel (NAP), of which Dr. Brown is a member. The Work Group proposed strategies to further enhance and expand AHS risk communication efforts to pesticide applicators across the country (in addition to those already receiving information in Iowa and North Carolina) as well as to health care providers. Information from AHS developed for lay audiences, such as this article and the outreach materials developed by Storm, is currently shared through the American Association of Pesticide Safety Educators (AAPSE) listserv and thus reaches the state pesticide educators responsible for the majority of pesticide applicator training in the US. The proposal advocates strengthening the relationship between AHS and AAPSE and offers seven additional recommendations to increase the understanding of risks associated with pesticide application and of measures applicators can take to minimize their risks. Dr. Brown presented the proposal at the annual meeting of the NAP in March 2005, and it is currently under review by the Panel.

The web site for the AHS (www.aghealth.org) contains information for enrollees and others, in the format of Frequently Asked Questions, as well as agricultural health resources for both states and national links. Much of the web site, while geared toward a scientific readership, is accessible to lay persons wanting to read AHS related journal articles, which are all available in abstract or full text format. Watch the web site for the wealth of information expected to be published in the coming years, during AHS's intensive five-year period (2004-2008) of assessing disease rates and risk factors.

The Agricultural Health Study is supported by:

Iowa Office of the Agricultural Health Study: 1-800-217-1954
North Carolina Office of the Agricultural Health Study: 1-800-424-7883

Ordering Information:
Understanding the Agricultural Health Study
A summary of the first 12 years of a long-term study of Iowa and North Carolina pesticide applicators and farm families

To order the 3-part publication series, designed for the agricultural community and Extension educators, contact:
NC Cooperative Extension Publications, Department of Communication Services, NC State University:
Telephone: 919-513-3112
Fax: 919-515-8602
E-mail: Rhonda_Thrower@ncsu.edu
Refer to Publication #'s AG-Med-24, AG-Med-25, AG-Med-26
Note: Publications are free, but shipping is charged for delivery outside of North Carolina. Call for information regarding shipping charges based on quantity desired and delivery destination.

Back to the top

Last Update 05.06.05