Introduction to: Pesticides in Children's
Foods Methods:
Updating the CU Scoring Scheme
Pesticides in Children's Foods1
Updating the CU Scoring Scheme
You can't control pesticide use by the agricultural industry...
...but you can remove the pesticides!
Food and Water Ozonator
|
Our 1999 report lays out
in detail our methodology for calculating the CU “Toxicity Index” for
various foods. In brief, the TI for a given food is based on
frequency
of detection, mean concentration and toxicity of each residue
found in each food. Toxicity indices for individual
pesticides are based on a scoring scheme we developed that
combines several measures of acute and chronic toxicity.
For this
report, we have updated our scoring scheme, in two respects:
(1)
Impacts of updated EPA risk assessments. As the EPA
reviews new toxicity data on individual pesticides under the
Food Quality Protection Act (FQPA), it publishes updated risk
assessments.
Based on these reviews, the EPA has, for example, revised its
Reference Doses (RfDs) for certain of the organophosphate insecticides.
We have incorporated
changes in the EPA’s toxicity database into our database.
Tables 2 and 3 of this report present the basis for our Acute
and Chronic Toxicity Indices for each pesticide active ingredient.
These tables, updated from the comparable tables in last year’s
report, include the most recent EPA toxicity factors.
A change in the EPA RfD can significantly change our Chronic
Toxicity Index for a pesticide. For example, within the past
year, EPA has lowered the chronic RfDs for chlorpyrifos-methyl
and pirimiphos-methyl 100-fold, lowered the chronic RfD for
chlorpyrifos 30-fold, and lowered chronic RfDs for methamidophos,
mevinphos and several other
active ingredients 10-fold
or more. Since the chronic RfD is a component of our toxicity
index for a chemical, changes in EPA’s RfD values impact
our scores—affecting
both the overall TI score and the relative share of the total
TI attributable to these specific residues, for any food that
contains these residues.
The most striking effect of revisions in EPA RfDs on scores
for PDP foods involves wheat grain, tested in 1995, 1996
and 1997. In our report last year, wheat received comparatively
low TI scores of 18, 29 and 32 for those three years, respectively.
But the primary residue found in wheat is chlorpyrifos- methyl,
which is used post-harvest, to control
insect losses in stored grain. With the EPA’s revision
of the RfD for chlorpyrifos-methyl (lowering it 100-fold),
scores for wheat shot up to 747, 894 and 777 for the three
years PDP tested wheat. If wheat had had scores that
high last year, we would likely have flagged chlorpyrifos-methyl
residues on wheat as an important problem. Residues in stored
grain may not represent residues in processed wheat foods eaten
by consumers, but scores this high would certainly have led
us to call for additional testing of wheat-based
foods.
However, shortly after the EPA published its revised risk assessment
for chlorpyrifos-methyl, its manufacturer asked the agency
to voluntarily cancel this use (i.e., to remove chlorpyrifos-methyl
from the market for application to stored grain). EPA and
the manufacturer recognized right away that they had a potential
dietary exposure problem, given
the revised risk assessment, and they took swift action to
resolve that problem.
Changes in other EPA RfDs affected our previously-published
scores for several other foods, though less dramatically
than was the case for wheat. Foods in which chlorpyrifos
residues were frequently detected generally have higher scores
with our recalculated TIs; this is especially notable for
apples from New Zealand and tomatoes from Mexico, for example.
Foods
that contained frequent methamidophos residues, such
as U.S.-grown fresh green beans and tomatoes, saw their scores
increase as well. A changed TI for mevinphos tripled the score
for Mexican broccoli (1994). However, the impacts of new EPA
RfDs on scores for most foods were modest, and did little
to change the overall picture revealed in our
1999 analysis.
(2) Changes in scoring for endocrine disruption. We made an
additional change in our scoring scheme based on a routine
review of our methods, in which we considered, among other
things, comments received from various analysts in response
to our 1999 report. Our Toxicity Index incorporates several
dimensions of a chemical’s toxicity. It combines an
index of acute toxicity, based on the LD50, and an index
of chronic toxicity, based on multiple parameters. For last
year’s report, we used
a chronic toxicity index that included four factors. The first
factor was derived from the EPA RfD, a general measure of
chronic toxicity. Two factors
addressed carcinogenicity: one based on the quantity and quality
of evidence, as reflected in the EPA’s carcinogen classification
for the chemical, and one based on carcinogenic potency, reflected
in the EPA’s Q1* value. Finally, we included a factor
for endocrine disruption. Pesticides identified as potential
endocrine disrupters
by Colborn et al. in a ground-breaking 1993 paper (cited in
our 1999 report), based on observed effects in wildlife, received
an additional factor of three in the RfD component of our 1999
scoring scheme.
We published our detailed methodology in our 1999 report, seeking
critical comments from other analysts. Several comments suggested
that the current scientific basis for classifying chemicals
as endocrine disrupters is relatively weak, and some analysts
questioned whether inclusion of this factor in our scoring
scheme was fully justified. We agree that there is far less
scientific support for a consensus over whether a chemical
is likely to have endocrine effects in humans than there
is for other components of our chronic toxicity index. While
we believe endocrine effects of pesticides are important
and deserve to be reflected in any overall toxicity index,
we decided to remove the endocrine disruption factor from
our scoring system this year. We will consider restoring
it in the future, when the
scientific community has had more time to develop a consensus
approach to identifying, and ranking the relative potency of,
hormonally active agents. For now, we are content to compare
the relative risks of pesticide residues without including
this aspect of toxicity.
Removing the ED factor from our chronic toxicity index reduces
a primary component of the chronic TI for any chemical we
had classed as an ED, to one-third of its previous value.
Since our overall chronic TI for a chemical is the sum of
a carcinogenicity component and an RfD/ED component, the
effect of removing the ED factor varies. It reduces the
chronic TI for a non- carcinogenic pesticide by two-thirds,
but has less effect on the chronic TI for a chemical that is
classed by EPA as a carcinogen. Likewise, the overall TI
for a chemical in our scoring scheme is a weighted sum of
chronic
and acute toxicity indices (with chronic weighted twice and
acute once). Removing the ED factor has a greater effect on
the overall TI score for a chemical that is very toxic chronically
but not very acutely toxic, and a smaller effect on a chemical
with the opposite toxicity attributes.
In sum, while the impact varies from chemical to chemical,
the net effect of taking out the ED component is to reduce
the toxicity indices for chemicals we had classed as endocrine
disrupters, and to reduce the TI scores for foods that contain
residues of those chemicals.
The most striking example of these effects involves scores
for foods containing the organophosphate insecticide methyl
parathion. The EPA RfD for methyl parathion is extremely low,
and it was also listed as an endocrine disrupter by Colborn
et al. Consequently, our TI for this chemical was very high,
and foods that contained residues of methyl parathion had
very high TI scores. Removing the ED factor from our TI for
methyl parathion lowered the scores substantially for several
foods.
Most notably, scores for fresh peaches were the highest for
any food in our 1999 report, at 4,390, 5,376, and 4,848 for
U.S.- grown peaches tested by the PDP in 1994, 1995 and 1996,
respectively. Methyl parathion accounted for more than 90
percent of those total scores each
year. Without the ED factor, the recalculated scores for peaches
are 1,640, 2,050, and 1,773, respectively. These scores are
still very high—with all scores for 1994-1997 recalculated
(see Table 4), peaches remain one of the two foods with the
highest
scores (along with winter squash, which contains dieldrin residues).
U.S. peaches also still have much greater TI scores than imported
peaches.
Several other foods that contain methyl parathion residues
and had high TI scores in our 1999 report have somewhat lower
recalculated scores. They include U.S.-grown apples and pears,
U.S. grapes for 1994 only, and frozen U.S. green beans (1997).
However, the recalculated scores for these foods are still
among the highest of the PDP-tested foods. Even with the
ED factor removed, methyl parathion still has an extremely
low chronic RfD, and for that reason it contributes substantially
to the TI values for foods that contain it. Last August,
the EPA announced a ban on uses of methyl parathion on most
foods
that the PDP data showed have residue
problems, a testament to both the chemical’s high toxicity
the agency’s justified decision to make it a priority
for risk-reducing tolerance revisions.
In most other cases, removing the ED factor had only modest
effects on the relative scores for different foods, because
individual pesticides account for only part of the overall
TI for any given food.
Summary of effects of revised scoring scheme. The combined
effects of all the changes in our scoring scheme produced
few substantive changes in the results of prior years’ analyses.
Some changes for individual chemicals (e.g. a lower EPA RfD)
led to an increased score, while other changes (such as removing
the ED factor) led to a reduced score. Most PDP foods contain
multiple pesticide residues, and the changes
in our scoring scheme raised the score for some residues, and
lowered the score for others. The net effect for most foods
was a modest change in absolute numerical score, but there
were few changes in relative scores for different foods. Except
for the examples noted above, we regard the changes in scores
as not very meaningful.
In summary, none of our 1999 conclusions, based on the relative
scores of various foods and the roles of particular pesticides
in driving the scores for specific foods, need revision.
The overall picture revealed by our analysis of the 1994-97
PDP data remains essentially as we described it last year,
with recalculated scores for those prior years. Foods that
had high scores in last year’s report still have high
scores, and those that had low scores still have low scores.
A few foods that had “medium” scores
last year have moved up into the low end of the “high-score” group
(e.g., Mexican broccoli in 1994, U.S. and Canadian carrots
in 1996, U.S. potatoes in 1995 and 1996), but these changes
are still modest. While the importance of some residues has
increased because of revised (lower) EPA estimates
of the safe dose, those chemicals were generally already on
our list of “risk drivers.” The revised scoring
scheme has shifted the relative importance of some chemicals
up or down slightly, but the list is essentially the same as
it was last year.
This report, therefore, focuses primarily on highlights of
our analysis of the 1998 PDP data.1
Goto the Next Section: Highlights
of Results of the 1998 PDP Analysis |
Search Our Site
