Center for Food Safety & Applied Nutrition Office of Premarket Approval September 1995 (Effective June 18, 2001, Office of Premarket Approval is now Office of Food Additive Safety. See updated contact information) |
The examples below show how OPA reviewers would determine the EDI for direct additives or GRAS ingredients based on the uses proposed by a petitioner and for contaminants based on experimentally determined concentrations in food. The examples were selected to illustrate the diverse nature of the problems that OPA is asked to address. First, two simple examples (single-use additives) are used to highlight inherent conservatisms in OPA's estimates due to (1) the assumptions necessary for estimating eventual consumer use of the additive (broadest possible food groups considered and complete replacement of other technologically-equivalent additives by the new additive) and (2) the use of the maximum possible concentration of the additive in food. When the use of conservatisms leads to an EDI that is above the acceptable daily intake (determined toxicologically) the petitioner can submit additional data to allow a refinement (and perhaps a lowering) of the estimate. The responsibility for providing data for reevaluation remains with the petitioner. Second, examples are given of more complex cases (multiple-use additives), where the additive may be used in multiple food groups and additional assumptions are necessary to arrive at a reasonable estimate of probable exposure. Third, examples of exposure estimates pertaining to a secondary direct additive and to contaminant problems are presented. Finally, two special cases, requiring information or approaches not generally used, are provided.
Exposure to a Volatile Antimicrobial
A company petitioned OPA to affirm as GRAS a
volatile antimicrobial (VAM) agent for general
use in baked goods. This VAM was already
affirmed as GRAS for use as an antimicrobial in
one sub-category of baked goods. The company
submitted data demonstrating that the VAM
content of a baked good after storage with VAM
vapor was between 0.4 and 1.5%, but requested a
regulatory limit of 2%. For estimating exposure
to the VAM, use of this proposed regulatory limit
would be highly conservative. However, since the
submitted data showed that 1.5% is the likely
maximum residual level of the VAM, the use of this
technical limit was considered sufficiently
conservative for this EDI.
The baked goods food category (21 CFR 170.3(n)(1)) is
quite large, encompassing all breads, cakes and
pies, crackers, pancakes, ice cream cones, and
more. The chronic daily intake for someone
(belonging to the all ages group) consuming
baked goods at the mean level is approximately 140 grams/person/day
(g/p/d), according to the databases available to OPA. The
intake at the 90th percentile level of consumption is over 230 g/p/d.
OPA cannot predict the percentage of the baked
goods market eventually captured by the petitioned product. If it
is assumed that all baked goods will be treated with this
antimicrobial, which is an extremely conservative assumption on
both technical and market grounds, and that absorption into
the food will be at the 1.5% level, the chronic mean EDI for
this material would be 2.1 g/p/d (0.015 x 140 g/p/d).
This EDI is exaggerated because of the assumptions that all baked
goods consumed will have been treated with the additive and because
of the use of the intake from the broad baked goods subcategory
(which includes home-prepared products). Only consumers who
seek out products using this antimicrobial, a highly unlikely
occurrence, would be expected to be exposed to the additive at levels
comparable to that calculated. Refinements to this estimate could
be made if the petitioner were to restrict the
types of baked goods treated with the VAM.
Estimated Daily Intake for an Amino Acid in Sweet
Pickles
A petitioner sought approval for the use of an
amino acid as a flavor enhancer for sweetened
pickles. The petitioner proposed to use the amino
acid as a component of pickling spice at a level
of up to 1%. Data from a trade association
demonstrating that 4 pounds of brine are used to
pickle 6 pounds of vegetable and a statement
that the brine contains from 0.025 to 0.05% spice
were supplied. The trade association also
estimated that 50% of pickles are sweetened and
that Americans consume 9 pounds of pickles per
capita per year. This would result in 4.5 pounds
per capita per year disappearance of sweetened
pickles, or 5.5 grams per person per day.
For the estimate of chronic daily exposure to the
amino acid, data from a food intake survey were
used. Mean intake (eaters-only) of pickles (and
related products) was 7 grams per person per day
(g/p/d). Thus, the survey information is in good
agreement with the disappearance data. The EDI
for eaters of sweetened pickles was determined as follows:
Probable Exposure for an Emulsifier
A petitioner sought an amendment to a current
regulation which would expand the number of uses
of an emulsifier to include its use in "imitation"
seafood. The emulsifier was already regulated
for use in baked goods, baking mixes, dairy product analogs, frozen
dairy desserts and mixes, and whipped milk products (as an
emulsifier), biscuit mixes (as a texturizer), and
certain fruits (as a coating component) at levels
limited only by good manufacturing practices.
To estimate exposure from the regulated uses,
OPA relied on typical use levels supplied in the
earlier petition. The new petitioner supplied an
estimated "typical" level for the proposed use in imitation seafood.
As a part of the original petition review, OPA had
estimated exposure to the emulsifier from the
then-proposed uses. The broad range of uses led
reviewers to use food-frequency information for
the total-sample rather than for eaters-only
because it was deemed likely that the whole
population would consume foods that had been
treated. In developing the new exposure
estimate that would include the proposed use,
the food sub-category "crustacean shellfish"
was selected because the imitation seafood is
intended to replace crab and lobster meat.
The 14-day average food intakes and resultant estimation of
chronic exposure to the emulsifier of 395 mg/p/d are
outlined in the Table 3 below.
It can be seen that the petitioner's request
constitutes an increase of less than 1% (3.3
mg/p/d) to the total mean EDI. The conservatisms
built into this exposure estimate are as follows:
complete replacement of other technologically-equivalent
additives by the emulsifier and use of
the emulsifier at the maximum level suggested by
petitioners as typical in all regulated food groups.
As discussed above, a crude estimate of the 90th
percentile exposure to a substance can be made,
assuming that it is approximately 2-fold higher
than the mean. For this emulsifier, using this
procedure, the 90th percentile exposure would be
approximately 800 mg/p/d. Further evaluation of
the exposure estimation at the 90th percentile
for this material was accomplished by the use of
a Monte Carlo analysis. In this procedure,
log-normal distributions of the intakes of each of
the foods were combined with a triangular distribution of use
levels to obtain distributions of incremental exposures to the
emulsifier from each food. These exposure
distributions were then combined to derive the
distribution of overall exposure to the emulsifier.
The 90th percentile exposure in this example is 650 mg/p/d.
Continuing the analysis of this exposure
calculation, it was noted that the petitioner
included a per-capita estimate of the increase in
exposure to the emulsifier that would occur as a
result of the regulation of the proposed use.
The exposure was based on market data for
imitation seafood products currently available
commercially. The petitioner assumed: 150 million
pounds of seafood product, a US population of
240 million people, 55% imitation seafood in
products, and an emulsifier level of 0.3%. The
resulting increase in per-capita exposure to the
emulsifier would be 1.3 mg/person/day (mg/p/d).
OPA's estimate of a 3.3 mg/p/d increase is in good
agreement with this figure, given that the two
estimates were obtained by approaches that were independent of
each other, and provides assurance that OPA's estimate is reasonable.
As previously discussed, a secondary direct
additive is not intended to remain in the food
after its technical effect has been accomplished
or to have any intended effect in the food should
residues remain. Therefore, the use level
information supplied by the petitioner is not
pertinent to preparing the EDI unless, as a
worst-case scenario, the assumption that all of
the additive will be carried over to or remain in
the treated food is being employed. Such an
approach, however, is rarely used for secondary directs.
Generally, the concentration data used for
preparing an EDI for a secondary direct additive
is obtained by analyzing the treated food for
residues of the additive. This information is
then used in conjunction with the appropriate
food intake data to obtain the EDI. There are
secondary direct additives, however, for which
this procedure is not sufficient to determine the
EDI. An ion exchange resin, used as the support
for an immobilized preparation of glucose
isomerase in the preparation of high fructose
corn syrup, presents such a case, as shown in the
next example.
Exposure to Enzyme Immobilizing Agents and Their
Residues
A company petitioned OPA to authorize the use of
its ion exchange resin as a fixing agent for the
immobilization of glucose isomerase preparations.
The company had calculated daily intakes for the
resin based on assumptions concerning the useful
lifetime of the resin in the preparation of high
fructose corn syrup (HFCS). These were:
The resulting exposure to the resin is 87 µg/p/d.
This EDI reflects the typical conservatisms: all
HFCS entering the marketplace will be produced
using this preparation; all column degradation
leachate will remain in the HFCS (the company
supplied no data concerning the removal of
leachate during processing, but merely stated,
without supporting data, that greater than 90%
was removed); and the resin will always be used at
the maximum enzyme preparation-to-syrup ratio.
The procedures for estimating exposure to a
contaminant in a food generally parallel those
described above for direct additives. It is
usually imperative, however, to examine the
nature of the concern for the contaminant in
question. As mentioned previously, there are two
general contaminant cases: contaminants in food
additives (that require premarket approval) and
contaminants in food (or food substances that do
not require premarket approval). In the former
case, the burden of proof of safety is on the
petitioner. OPA could rely on a specification
maximum for preparing the EDI and determining
whether a safety concern exists. The petitioner
may be able to supply additional data on the
actual level of contaminant in an additive to
allow an adjustment of the EDI, if needed. In the
case of a contaminant in food, the burden of
proof of hazard is often on FDA. It may be
necessary to prove that the presence of a
contaminant in a food or food substance may
render it injurious to health. The concentration
chosen for use in calculating an intake estimate
would depend on the nature of the risk. The
concentration used for a contaminant posing an
acute risk would probably be at the maximum of
the observed experimental levels. If the risk is
due to chronic exposures, however, a mean
concentration level, representing the likely
lifetime level of exposure would be more
appropriate. Again, each contaminant must be
dealt with on a case-by-case basis.
For the simple case (no acute risk) where there is
a safety concern due to the presence of a
contaminant in a single product or class of
products, it is sufficient to use the mean
concentration of the contaminant in the food and
the intake of the food in order to estimate the
exposure to the contaminant. The next example
illustrates the case of a naturally-occurring
contaminant in an additive with a limited use.
More complex cases would require the use of the
same assumptions (substitution, etc.) described
above for multiple-use food additives.
Exposure to Residues in a Refined Food Oil
A company submitted a petition for GRAS
affirmation of a refined food oil for use in
candies and confections (chocolate products) at
levels from 5-30%. The company reported that
the crude oil contains 3-7% unsaponifiable
matter that could be reduced to a specified
maximum of 1.5% upon refining. The unsaponifiable
material remaining after refining consisted
primarily of triterpene alcohols and sterols,
including alpha-spinasterol. The presence of
alpha-spinasterol was unique to this oil among the
edible oils. Therefore, it was necessary to
estimate exposure to this natural contaminant
as part of the safety evaluation for this petition.
The exposure to the oil from its use in chocolate
products was first evaluated. The chronic EDI
was calculated to be 2.2 g/p/d (all ages). The
company submitted data showing that the typical
residual sterol level in the oil was approximately
0.05%, 40% of which is alpha-spinasterol. Therefore,
use of the oil in chocolate would result in a
total potential exposure to sterols and methyl
sterols of approximately 1.1 mg/p/d. This
exposure was determined by multiplying the exposure
to the oil by the level of the sterol in the oil as follows:
0.05% sterol in oil x 2.2 g oil/p/d = 1.1 mg sterol/p/d
The probable exposure to alpha-spinasterol would be
40% of that for all sterols, i.e., 0.4 mg/p/d.
Mercury Contamination in Shark
This example illustrates some assumptions that
might be made when none of the available food
intake databases contain information sufficient
to allow a straightforward estimation of exposure.
A state's health officials had collected 25
samples of shark meat from retail fish outlets.
The species collected were not known, nor were
the locations of the catch sites. Analysis for
methylmercury showed an average level of 1.48
ppm with a range of 0.35 to 3.9 ppm. Based on
these results, the state's health department
advised that adults should eat shark no more
than once a week. Children and women of
childbearing age were advised to eat shark no
more than once a month. Following this health
advisory bulletin FDA was asked to prepare an
estimate of methylmercury intake that might
result from consumption of shark.
Mercury values obtained by FDA from domestic
and import compliance and surveillance samples
were used as the basis for establishing the
concentration to use in the exposure assessment. For 113 samples,
the average mercury concentration was 0.96 ppm (+/- 0.61 SD).
These samples were taken from those shark meats
used as food, and so should reflect levels of
methylmercury actually ingested.
Preparation of an estimate of shark consumption
using a straight per capita disappearance figure
(ca. 0.035 g/person/day)(35)
would have given an inappropriately low estimate of likely shark
intake for actual consumers of shark.
Shark is an infrequently consumed food. Consequently, the consumption
information presented in dietary surveys such as the USDA
3-day survey and the MRCA 14-day survey can be
expected to overestimate, to varying degrees,
the average daily intake of shark consumed by
the individual(36).
Of these surveys, the MRCA survey, due to its longer time
base, is more likely to capture the consumption of infrequently
consumed foods . However, due to the extremely
low consumption rate for shark, even the MRCA
survey would likely overestimate chronic
exposure to shark for eaters of shark in the
general population. Therefore, to more
accurately estimate chronic shark intake, an
average daily intake was first calculated using a
modified per capita approach, in which the amount
of shark disappearing into the U.S. food supply
was divided by the population of shark eaters as
determined using percentage of eaters
information from the state's seafood consumption survey.
According to the state's survey, approximately
1% of the general population are eaters of shark.
Thus, of a total estimated U.S. population of 260
million people (1994), there might be 2.6 million
people nationwide who are consumers of shark meat.
According to recent (1993) National Marine
Fisheries Service
(NMFS)(37).
figures, 8.9 x 106 lbs
of shark fillets and steaks were produced in the
U.S. in 1993. Therefore, the average daily intake of shark is:
This average was then used (as noted below) to
create a distribution of intakes for shark meat
by assuming a standard deviation equal to the mean intake and
assuming that the data were distributed in a log-normal
fashion(38).
According to the NMFS, methylmercury comprises
90-100% of total mercury in most fish. For the
purpose of this estimate, it was assumed that
mercury in shark is 100% methylmercury. To
estimate the intake of methylmercury resulting
from consumption of shark, log-normal distributions based on the
mean shark intake and the mean methylmercury concentration in shark
(and their associated standard deviations) were used in a Monte Carlo
simulation(38).
That is, exposure was calculated assuming that 4 +/- 4 g
shark is the mean daily consumption for eaters
and that shark is contaminated with
methylmercury at a level of 0.96 +/- 0.61 ppm. The
log-normal distributions of these data sets were
combined to give a methylmercury intake
distribution for shark eaters. The results of
this calculation are as follows:
A possible subset of shark eaters consists of
individuals who consume shark on a regular and
frequent basis, perhaps eating one shark meal
per week. This is the group to which the state's
health advisory was directed. Assuming that
such persons eat a serving of shark equal in
weight to the USDA mean portion size of 145 g for
fish consumption, the mean shark intake for these
regular and frequent consumers of shark would be
21 g/person/day. This scenario reflects the
consumption behavior of slightly less than 1% of
shark eaters (i.e., consumption at the 99.2th
percentile of shark eaters). If the shark were
contaminated with methylmercury at the mean
observed level of 0.96 ppm, methylmercury intake
would be 20 µg/day. This is below the tolerable
daily intake (TDI) of 30 µg/p/d, established through animal
studies(39).
The above examples point out some of the
considerations routinely used by OPA in
preparing an EDI. There are many cases, however,
where additional, or entirely different,
assumptions must be made concerning the factors
taken into account for evaluating exposure. Two
of these cases will be discussed: estimating exposure to
components of tablets (e.g., vitamin and mineral supplements),
and the use of dietary scenarios for estimating exposure.
Potential Exposure to Tabletting Agents
Numerous companies have petitioned to have
tabletting agents regulated for use in the
preparation of vitamin and mineral supplements.
Because the pattern of vitamin and mineral
consumption is different from the pattern of
intake of typical food, it has been necessary to
evaluate exposure to supplements using a novel
approach. FDA personnel determined, through a
telephone survey in 1980, that the average
intake of vitamin or supplement pills in the
United States was one tablet per day (J. Am. Diet.
Assoc., 1985, 1585-90). In recent years, however,
the daily intake of vitamin/supplement pills by
many individuals appears to have increased
substantially. Using 600 mg as the weight of a
typical supplement tablet, it is possible to
readily evaluate exposure to materials used in
the tablet formulation. For example, one company
petitioned for permission to use a fatty acid
ester as a formulation aid in tablets at a level
of 1-3%. Using the typical vitamin tablet size of
600 mg, the maximum level of the fatty acid ester
per tablet would be 18 mg, and the EDI from a
single tablet would be 18 mg/p/d. The company
petitioning for the fatty acid ester stated that
a maximal user of vitamins might ingest six
tablets a day. Therefore, the maximum exposure
to this fatty acid ester from its use in tablets
would be 108 mg/p/d. These estimates are highly
conservative in that they assume that all
vitamins consumed on a daily basis would be
formulated with the fatty acid ester.
Dietary scenarios
When specific food intake or substance
concentration data are not available, a dietary
scenario may be developed to complete an
exposure analysis. For example, CFSAN may be
asked to determine the safety of one lot of a
rarely eaten food product that is being
considered for regulatory action because of
contaminants that may pose a health risk. In this
instance, a scenario can be devised to estimate
the amount of food consumed per eating occasion
(or per day) and substance concentration can be
used that has been either experimentally
determined or chosen to be representative of the
lot. The exposure thus determined may then be
used to evaluate the risk of ingestion of the
substance in the food (based on chronic or acute effects).
Two topical examples demonstrate the use of
dietary scenario assumptions. Dioxins, a family
of related chlorinated compounds associated with
the bleaching of wood and paper products, have
been determined to be carcinogenic at extremely
low levels. Fish caught downstream from paper
mills were found to contain elevated levels of
the dioxins. Because fish intake data were not
available at the time for the target population
of subsistence fishermen, it was assumed that a
fisherman's normal intake of fish protein would
equal that of the average person's red meat,
chicken, and fish intake. Thus, red meat and
chicken intakes from a national food consumption
survey were used with dioxin levels of fish
caught near paper mills to model dioxin exposure
for subsistence fishermen.
The use of macronutrient substitutes, such as
fat replacers that could affect the nature of
some consumers' diets, serves as the second
example for which the use of a dietary scenario
is illuminating. The availability of fat replacers
might result in highly motivated individuals
substituting the replacer for all added fat in
their diet. The exposure estimate for a fat
replacer that would be available to the consumer
for home food preparation, including baking and
frying, can be based on considerations of food energy obtained from
fat intake. If the maintenance of an appropriate weight for a
man (19-50 yrs. old) is assumed to require an
average of 2900 calories, of which 40% are derived from fat intake
(not unusual for a substantial fraction of the US population), then
1160 calories a day are derived from fat. At 9
calories per gram of fat, fat intake in this diet
would amount to 129 grams per day. Approximately
50% of fat in the diet is estimated to be derived from
added triglycerides (lard, tallow, butter, margarine, vegetable oils,
shortening)(40).
Thus, "added fat" would amount to 64 grams per day. If
the fat replacer were intended to substitute for
"added fat" on a one-to-one basis, the EDI would
be 64 g/p/d. When the fat replacer is not caloric,
i.e., it is a non-metabolizable substance, the 64
grams of replacer represents a decrease of
about 576 (9 x 64) calories from the reference
daily intake of 2900 calories. Assuming that the
consumer would unconsciously eat, on average, 576 additional total
calories(41) to
compensate for the shortfall and continue to maintain a 2900
calorie diet, additional fat replacer would be
consumed as part of these extra calories,
presumably in the same proportion of calories
from fat in the total diet (40%). This would
result in an additional exposure to the replacer
of 13 g/p/d (576 total cal x 0.4 cal fat/total cal x
0.50 g added fat/g total fat x 0.111 g fat/cal fat),
bringing the total exposure to 77 grams per day
as an upper limit for an average man.
The use of such scenarios must be based on
assumptions that are reasonable reflections of
a consumer's actions. The assumptions must be
clearly delineated and defensible. If possible, a
second approach to estimating exposure should
be used to provide corroboration of the result
obtained using a dietary scenario.
There is no single method for estimating the
probable exposure to a substance found in or
added to food. This document has outlined cases
where a substance's concentration in specific
foods was known or could be predicted; therefore,
OPA used intakes for those specific foods to
estimate exposure. Special cases have been
considered, such as the use of dietary scenarios
to estimate exposure to fat replacers, where new
methodologies have been developed to estimate
exposure. Limitations caused by data
availability (or lack thereof), such as the
intakes of specific foods or the variability in
the concentrations of contaminants in different
foods, are often the determining factors in the
method used to estimate exposure. This document
is not an attempt to be an all-inclusive
consideration of OPA's deliberations in preparing
exposure estimates. Rather, it is intended to be
an overview of some of the general principles and assumptions
used, while emphasizing the case-by-case nature of the process.
Return to
Estimating Exposure to Direct Food Additives and Chemical Contaminants
in the Diet or go to the
footnotes for this document.
Hypertext updated by dms/hrw 2004-JAN-06 Single-Use Additives
The EDI for this amino acid resulting from the
proposed use is conservatively high due to the
assumptions that all of the pickling spice used is
transferred to the pickle, and that all sweetened
pickles will be treated with this flavor enhancer.
The magnitude of the overestimate caused by
using these conservative assumptions cannot be
evaluated with the available information.
Multiple-use Additives
Table 3
Estimated Daily Exposure to the Emulsifier
(2+ years, mean intake, total sample basis)
Food Category (use level)
Food Intake (g/p/d)
Emulsifier Intake (mg/p/d)
Fish, Seafood (0.3%)
Shellfish (crustacean)
1.1 3.3 Milk Products (0.5%)
Chocolate Milk
5.6 28
Flavored Milk
0.04 0.2
Eggnog
0.4 2.0
Sour Cream
1.1 5.5
Creams
1.4 7.0
Diet Milk Products
1.9 9.5
Cheese (0.5%)
Processed Cheese 6.6 33
Cottage Cheese 4.3 22
Cream Cheese 0.8 4
Diet Cheese Products 0.4 2
Other Cheese Products 1.7 8.5
Baked Goods (0.5%)
Breads, Rolls 39 200
Cakes 6.4 32
Ice Cream, Milk (0.5%)
Ice Cream 6.6 33 Dairy Products, Analogues (0.5%)
Toppings 0.35 1.8
Whiteners 0.64 3.2
Total (simple sum)
395 Secondary Direct Additives
Chemical Contaminants (natural or man-made)
Mercury in shark
Shark consumption
8.9 x 106 lbs shark
-------------------- = 3.4 lbs shark/eater/year
2.6 x 106 eaters
= 1554 g/eater/year or 4 g/eater/day.
Intake of methylmercury
Exposure to Methylmercury from Shark,
µg/person/day
(Eaters-Only)
Consumption Percentile
Methylmercury Intake mean 3.8 90th 8.4 95th 12.3 99th 25.0
100th 142 Special Cases
SUMMARY
* Office of Premarket Approval, Center for Food Safety and
Applied Nutrition (HFS-200), Food and Drug Administration,
200 C St., SW., Washington, DC 20204 (See updated contact information)
Food Ingredients and Packaging
|
Guidance for Submitting Petitions
Foods Home
|
FDA Home
|
Search/Subject Index
|
Disclaimers & Privacy Policy
|
Accessibility/Help