The most common cancer in children is leukemia, which constitutes roughly a third of all
childhood cancers. Little is known about the causes of cancer in children, but a number of
environmental agents, including pesticides, are under investigation for their suspected link.
Reviews of childhood cancer studies provide compelling evidence that pesticide exposure
potentially predisposes children to a number of cancers. These studies are almost entirely
case-control studies looking at home and garden pesticide use and exposures related to
parental occupation (e.g., toxic chemical residues on clothing). However, information on
residential proximity to agricultural exposures is lacking. The studies are also limited by
small numbers of cases, nonspecific pesticide information, and potential case-response bias.
These limitations illustrate the need for better risk assessment of agricultural exposures in children.
Ambient exposure—dispersion of applied pesticides into unintended areas—also needs to be
addressed. Pesticide drift can produce exposure in surrounding areas and varies depending on
the agents of use and weather conditions at the time of application. Attempts to assess these health risks in California have involved Geographic Information System (GIS) technology,
which is a way of incorporating information on environmental exposures with the incidence
of disease on a community and state level.
California has a unique data resource—since 1991, a pesticide use reporting system has
required full-use reporting of all pesticide applications in the state. This system provides a wealth of information, including location, active ingredient(s), number of pounds applied,
number of acres, date applied, type of crop, and application method. All of this information is recorded at the geographic detail of approximately one square mile and provides sufficient
power to detect moderate increases in cancer rates. Researchers are able to quantify pesticide Indianapolis, IN
3
October 21, 2008
levels by census block groups in the state. The reporting system is limited by its inability to capture the timing of exposure, homogeneity or heterogeneity of disease, and population
subgroups.
Of the annual average of nearly 160 million pounds of pesticides applied in the state of
California, over 12 million pounds are probable carcinogens and nearly 10 million additional
pounds are possible carcinogens. Socioeconomic status and ethnicity are strong predictors of
potential exposure to these pesticides—Hispanic children and those of low socioeconomic
status have a much higher probability of living in neighborhoods with high pesticide use. As
the exact risk of exposure remains unknown, a precautionary approach to the use of these
chemicals needs to be maintained.
A large case-control study is being conducted in a 35-county area of Northern and Central California, which involves extensive interviews with parents, household surveys, and
environmental sampling. The study looks at GIS-generated estimates of exposure in
conjunction with self-reported exposure risk, use in households, parents’ occupations, and an analysis of chemicals found in home dust samples. However, difficulties include measuring
risk factors, disease latency, time windows of exposure, and degree of statistical uncertainty.
Transdisciplinary efforts from epidemiologists, clinicians, geneticists, toxicologists,
statisticians, geographers, health educators, and advocates are necessary to overcome these
barriers.
DR. LAURA BEANE FREEMAN:
EVALUATION OF AGRICULTURAL EXPOSURES: THE AGRICULTURAL
HEALTH STUDY AND THE AGRICULTURAL COHORT CONSORTIUM