Authors
Reynolds P, Hurley SE, Goldberg DE, Yerabati S, Gunier RB, Hertz A, Anton-Culver H, Bernstein L, Deapen D, Horn-Ross PL, Peel D, Pinder R, Ross RK, West D, Wright WE, Ziogas A.

Title: Residential proximity to agricultural pesticide use and incidence of breast cancer in the California Teachers cohort

Source: Environmental Research 96: 206-218, 2004.

Summary:

Increases in breast cancer incidence rates over the past few decades has fueled the hypothesis that environmental contaminants such as pesticides, commonly used in agriculture, may contribute to its etiology. California has some of the highest breast cancer incidence rates in the U.S., and is also the largest agricultural state. A retrospective cohort study was conducted to examine the relationship between residential proximity to agricultural pesticide use and breast cancer.

The study participants were drawn from the California Teachers Study cohort. This cohort began in 1995 and is comprised of active and retired female teachers working in the kindergarten through the community college level. The cohort consists of about 133,479 women, and represents about 40% of the women who were initially invited to participate. Extensive information on breast cancer risk factors exists for this cohort, and it is linked annually with the California Cancer Registry. Of the 133,479 women in the cohort, 114,835 participated in this study. A case was defined as a women diagnosed with invasive breast cancer following her completion of the baseline questionnaire in 1995 through December 31, 1999.

Exposures were assessed using the California Pesticide Use Reporting (PUR) system in conjunction with the Public Land Survey System (PLSS). PUR mandates the reporting of all agricultural pesticides used in the state, and data from the years 1993-1995 was used for this study. PUR contains information on active ingredient, quantity applied, acres treated, crop treated and location in square mile sections for all agricultural pesticide applications in the state. The pesticides were combined into six groups according to their carcinogenic potential and mechanism of toxicity. The six toxicological groups were: probable or likely human carcinogens, possible or suggestive human carcinogens, mammary carcinogens, endocrine disruptors, anticholinesterases, and organochlorines. Pesticides were considered to be endocrine disruptors if they had the potential to cause mammary cell proliferation by directly or indirectly increasing estrogenic effects. Exposures for five selected individual pesticides (simazine, diuron, oryzalin, propargite, and methyl bromide) were also evaluated.

A geographic information system (GIS) was used to identify specific areas corresponding to the detailed information available regarding pesticide use. Individual exposures were evaluated by considering the average amount of applied pesticide within a half-mile radius of the subject's residence. The degree of pesticide exposure was reported by pounds of pesticide use per square mile (lb/mi2). Each grouping/individual pesticide was analyzed separately with adjustments made for a variety of other breast cancer risk factors. Quartiles of exposure were also examined to examine a potential dose response effect.

Among the 114,835 women participating in the study 1, 552 invasive breast cancer cases were identified. Overall, no association was observed between residential proximity to recent agricultural pesticide use and invasive breast cancer incidence for any of the six toxicological groups. Adjustment for potential confounders did not change the risk estimates.

A considerable strength of the study was that the investigators categorized pesticides by toxicological mechanism. It is important to recognize that pesticides have a wide range of chemical and toxicological properties. Total risk may be underestimated by if all pesticides are considered together in a single group as some chemicals may be more relevant to breast cancer development than others. Endocrine disruption has been hypothesized to be an important mechanism in the development of breast cancer, examining pesticides with endocrine disrupting properties separately was therefore a methodological strength of this study as compared to previous epidemiological investigations.

A limitation of the study is the fact that small numbers of women were found to have high exposures to pesticides. As higher levels of exposure were further subdivided into quartiles, this led to the problem of small numbers in many of the analysis groups, which may have contributed to the lack of significant findings. Another limitation was that sufficient time to allow for the latency of the disease was not provided, as exposures were assessed for 1993-1995 and the period of follow up for the cohort was 1996-1999. Assuming a latency period of 5 years it is possible that breast cancer cases resulting from the measured exposures may not have had enough time to develop. Exposure classifications were also based on reported pesticide use and not on environmental or biological measurements.

Future studies should build on the innovative use of GIS to assess environmental exposures. Improvements in the assessment methodology may include the incorporation of meteorological data and dispersion models in order to better estimate the degradation and movement of pesticide residues over time.