9 Second, we did not have a direct measure of pesticide exposure, and were forced to use surrogates of pesticide usage. Even if pesticide usage could be quantified at a county level, we would not be able to convert usage data directly to exposure doses to individuals. Further, a variety of agricultural chemicals are used on the same crops, depending on climatic conditions and contemporary farming practices. Thus estimates of specific chemical pesticide exposures are extremely difficult to make. Given these limitations, we believe that our results are useful in putting forth the general hypothesis that pesticide usage, even in a region of low overall cancer mortality, may be associated with increased cancer risk. This risk, if established, may be attributable to former use of pesticide arsenicals or to more recent and widespread use of chlorinated hydrocarbons. If arsenicals are responsible factors, the association with excess cancer should diminish over time. association between lung cancer and arsenic exposure would be consistent with evidence from occupational arsenic exposure. On the other hand, chlorinated hydrocarbons, use of which greatly increased in the 1950's, would probably only begin to manifest a cancer effect on farming populations in the 1970's or later. An These various hypotheses need to be investigated in analytical studies of high risk groups, as in retrospective mortality or case-control studies. We believe that our 10 findings suggest an additional public health problem related to pesticide usage, namely cancer risk. The problem is one that involves a major segment of the population in many countries and may increase with world-wide demand for more agricultural output. We urge investigators to carry out analytical studies of cancer risk in agricultural groups as a follow-on to this ecological study. |