EPA/600/A-96/06Q
Incorporating Mass Balance Concepts In Total Exposure Studies
David T. Mage1
Maria Donner2
1U. S. Environmental Protection Agency
National Exposure Research Laboratory
Research Triangle Park, NC 27711
2Chemical Industry Institute of Toxicology
Research Triangle Park, NC 27709
ABSTRACT
Total exposure studies require the monitoring of personal exposures
to pollutants over all five routes of exposure: 1) inhaling air;
2) drinking water; 3) eating food; 4) uptake through the skin; 5) other
unique incidents such as thumb sucking and chewing or smoking tobacco.
To evaluate their potential effect on human health, the exposures via
these five routes can be added as either a total applied dose or a total
absorbed dose over the period of the study (e.g., mg/kg body
weight/day). Since the absorbed doses via routes 4 and 5 are not
directly measurable in a field study they are commonly estimated from
unvalidated models. However, the applied dose can be determined by the
Law of Conservation of Mass;
Input = Output	+	Accumulation
Applied Dose = Output + Change of Body Burden
The portion of the applied dose that the body absorbs must also obey the
Law of Conservation of Mass;
Absorbed Dose = Bodily Eliminations + Change of Body Burden
We present three survey designs from the WHO/UNEP Human Exposure
Assessment Locations (HEAL) program, the NCI/NIEHS/EPA Agricultural
Health Study {AHS) and the EPA National Human Exposure Assessment Study
(NHEXAS) and discuss their abilities to estimate an applied dose or an
absorbed dose of target subjects using the mass balance equation.
INTRODUCTION
A fundamental tenet of toxicology is that the body responds to the
absorbed dose, defined as "the amount of a substance penetrating across
the exchange boundaries into body fluids and tissues after exposure"1.
The systemic response to this absorbed dose is independent of whether it
was inhaled, ingested or absorbed through the skin. Exposure is defined
as "Contact of an organism with a chemical, physical or biological
agent. Exposure is quantified as the amount of the agent available at
the exchange boundaries of the organism and available for absorption"1,
which represents the applied dose (as if given to a human to determine a
dose-response relationship). Given equal applied doses of a pollutant
to identical twins, the twin with the greater absorbed dose will be at
the higher risk, so absorbed dose is the relevant quantity for
evaluating total exposure and its risk effects.
The WHO/UNEP and two U.S. EPA programs initiated a series of human
exposure assessment pilot studies that recognize the need to evaluate
the total absorbed dose of a pollutant. Although some of these studies
discuss total exposure2 and cumulative exposures3, these terms are not
defined in those papers.
Total exposure studies measure or estimate exposure over all the
individual routes by which a pollutant can enter the body. We define

-------
five classes of route of exposure, as opposed to the traditional three
{inhalation, ingestion and dermal) , to highlight the measurement
difficulties of several important pathways. These five applied doses
must equal the change in body burden plus the portion of the dose
eliminated from the body. The five major routes of exposure are
1) inhaling air; 2) drinking water; 3) eating food; 4) dermal uptake;
5) miscellaneous uptake activities, such as thumb sucking and pica.
ROUTES OF EXPOSURE
Route 1: Exposure via the inhalation route is quantified by the volume-
weighted-average (VWA) concentration of the pollutant in the volume of
air inhaled (mg/m3), times the volume (m3) , which is equal to the
applied dose (mg). Active tobacco smoking, which contaminates inhaled
air is discussed with Route 5. The exhaled air may contain a fraction
of the inhaled pollutant that is not absorbed or deposited. In
toxicological studies on nocturnal rodents, exposures of unrestrained
animals often specify whether the animals are awake (lights off) or
asleep (lights on) to account for variations in the animals' alveolar
ventilation rate as a function of activity.
Route 2: Exposure from drinking water is quantified as the pollutant
concentration ^g/liter) times the volume of water drunk (liters).
Route 3: A food or beverage that is eaten (e.g., not pica or drinking
water) is a potential source of exposure to a pollutant. Computation of
an applied dose via the diet requires either an analysis of all items
eaten or a single analysis of a duplicate diet.
Route 4: Dermal exposure is quantified by the average concentration of a
pollutant on the skin (mg/m2) times the skin area of the body (m2). The
fraction (x) of the applied dose entering the body through the skin
represents an absorbed dose. An estimate of the absorbed dose requires
the area of skin contaminated, the time of contact, and the mass
transfer coefficient if available from the literature4.
Route 5; Singular incidents, such as touching food with dirty hands,
thumb-sucking, pica, and chewing or smoking tobacco can lead to an
ingestion of a pollutant, but it will be impossible to quantify them
under field conditions. For example, smoking one cigarette can deliver
from less than 0.5 mg to 27 mg of tar (applied dose)5. However,
depending on the frequency and depth of inhalations, different absorbed
doses can result.
TOTAL EXPOSURE
The exposures over these five routes can be estimated individually
as an applied dose1. The unit doses by these routes are not equivalent
because, for a dermal applied dose of 1 mg, only a small fraction
(x » 0) may be absorbed by the skin before the surface is cleansed, as
opposed to 1 mg ingested or inhaled, which may have a large fraction
(x ~ 1) absorbed. Therefore, the absorbed dose is the more appropriate
quantity to report because it is more proximal to the resultant risk to
the person. The difference between applied dose and absorbed dose is
the crux of the matter. The risks for individuals can vary widely
because of their different absorption fractions for each route.
Therefore, it makes no sense to add applied doses, so total exposure
should refer to the total absorbed dose where the risk of an effect is
independent of the route and portal of entry. To assign ordinal
rankings to individuals, as shown on a histogram of total exposure such
as Figure 1 from Sexton et al.3, the absorbed dose should be used as the
index. For example, let identical twins A and B be simultaneously

-------
exposed. Twin A receives 10 mg of a pollutant on the skin (10 mg applied
dose) and absorbs only 1 mg into the body (1 mg absorbed dose). Twin B
swallows 10 mg of the same pollutant (10 mg applied dose) and absorbs it
all (10 mg absorbed dose). Twins A and B have identical applied doses,
but twin B would have a 900% higher absorbed dose than A. In this
scenario, twin B would have the higher cancer risk, and in that sense
would have the higher total exposure. How then can any person's total
exposure, including pica, dermal contacts and active tobacco smoking be
assigned an ordinal rank for evaluating whether or not it is above a
reference dose that is considered to be safe? The answer is through the
application of the Law of Conservation of Mass for a fixed period of
observation;
Input - Output = Accumulation	(1)
When applied to a total exposure field study it can be written as;
Applied Dose » Output + Accumulation in the Body	(2)
where the symbol » is used to indicate that experimental errors and
unmeasured routes of input (e.g., smoking) or output{e.g., milk, menses)
and unmeasured changes in some body compartments (e.g., adipose tissue,
bone) are expected to result in an incomplete closure of the mass
balance. Equation 2 provides an experimental basis for estimating the
applied dose on the left hand side of the equation by measuring the
quantities on the right hand side. For example, Calabrese and Stanek6
report on a pica study in which five of 24 children with normal mental
capabilities and an average age of three years, displayed soil pica
behavior (> 1 g soil ingested/day) by using fecal silicon as a soil
tracer. An observation of infant pica activity cannot be quantified
without fecal measurement.
APPLICATION OF THE MASS BALANCE: FROM EXPOSURE TO THE APPLIED DOSE
The linkage between exposure (as a VWA concentration) and applied dose
is described for the inhalation route7: "If a worker who weighed 70 kg
breathed in 7 m3 of air containing 50 ppm (200 mg/m3) ethylene
dichloride daily, he would be exposed to 20 mg/kg each day".
This is a correct way to treat these data because the units of mg/kg-day
are additive over all five types of exposures. The use of body weight
to normalize an applied dose corrects for the dilution of the dose into
the body fluids. Consequently, descriptions of daily intakes, such as
that for methylmercury "with health effects occurring with a daily
dietary intake of 280-420 pg"8 should be normalized by kg body weight,
as the effects relate to the daily pg/kg dose, rather than the daily pg
dose9.
APPLICATION OF THE MASS BALANCE: FROM APPLIED DOSE TO THE ABSORBED DOSE.
The form of Equation 2 that estimates the applied dose can be
adapted to estimate the absorbed dose. With the same caveats a third
mass balance equation for a subject in a field study can be written as;
Absorbed Dose ~ Bodily Eliminations + Change in Body Burden (3)
where body burden is defined as the total mass of the pollutant in the
various compartments of the body, and the bodily eliminations are the
external excretions of the systemic mass that had been previously
absorbed. In this equation, bodily eliminations do not include the
portion of a dermally applied dose that is cleansed off the skin and is
not absorbed, the unabsorbed gases that are exhaled, or the fraction
(1 - x) of ingested material that is not digested and is excreted in the
feces. The pollutant in the fecal output should approximate the portion
of the food and inhaled particles that are not digested and absorbed

-------
plus some systemic excretions, such as bile. For example, 10% of the
radio-labeled dermal dosage of atrazine absorbed by humans is excreted
in the feces within 24 hours10.
The breath can excrete gaseous compounds that had been absorbed in
the blood as well as unabsorbed inhalation quantities. The urine, breath
and feces remove the chemical and its metabolites from the body burden.
Note that for a zero dose, the urinary, fecal and breath output will
equal the decrease in the body burden (BB) during the measurement period
T. For* zero dose of a pollutant with a boay burden that behaves as a
first order linear system, the body burden decreases as Equation 4,
BB (0) - BB (T) = BB (0) [ 1 - exp"<0,693 r/c4) ]	(4)
where tH = half-life of the chemical in the body stores,
BB(0) = body burden at start of study at time t = 0,
BB(T) = body burden at end of study at time t = T.
Therefore, it is critical to measure the change in the body burden
by paired before-and-after measurements on accessible body fluids (e.g.,
blood), and in some cases to also collect breath samples and total urine
over the entire study period. The sampling strategy will vary with each
pollutant and its metabolites according to their major routes of
excretion (e.g. carbon monoxide, breath only; lead, urine only; benzene,
breath and urine).
THE GEMS/HEAL PROJECT
The WHO/UNEP GEMS/HEAL project applied the mass balance (Equation
2) to the study of human exposures to lead and cadmium. A pilot study
involved the one-week measurement of the inhalation route of exposure by
personal air sampling and a one-week collection of duplicate diets for
several subjects11"13. In addition, a total feces collection was made for
lead and cadmium analysis as a data -uality assurance measure to
evaluate whether the sum of the measured applied doses times their
expected unabsorbed fractions (1 - x) was approximately equal to the
measured fecal output. Vahter et al.13 assumed that during a short
study period (seven-day) the change in the body burden of lead and
cadmium is negligible, so the measured fecal excretions should be
approximately equal to the net fraction (1 - x) of the applied dose that
is not absorbed. As shown in Figure 2, the estimated unabsorbed intake
of cadmium was approximately equal to the measured fecal output in
Beijing and Stockholm15. However, in Zagreb, the fecal output of
cadmium was greater than the unabsorbed dose stimated from only the air
and diet exposure concentrations. A special audit analysis confirmed
the accuracy of all the measurements of cadmium, so the discrepancy was
considered to be greater than expected (see explanation for the usage of
» in equations 2). An investigation of possible cadmium exposures from
the miscellaneous routes (lipstick, tooth paste, tooth fillings,
smoking, etc.) identified no other likely source of cadmium exposures.
THE AGRICULTURAL HEALTH STUDY (AHS)
In the AHS14 a thorough exposure analysis was performed on several
pesticide applicators in a field pilot study utilizing measurements of
the air, water, food and beverage, and dermal routes of exposure
(unpublished data). Personal air sampler values and dermal measurements
(hand wipes and skin patches) showed high atrazine exposures, but no
atrazine was detectable in the food and water. Assuming that all
atrazine inhaled is absorbed and none is exhaled, we estimated that the
absorbed dose by the air route was approximately 1 yg/75 kg by use of
20 liter/min inhalation for handling-mixing-loading and 15 liter/min for

-------
application. Blood serum levels of atrazine were 0.05 ppb pre- and 1.42
ppb post-application; thus, the accumulation in the blood could be a
measure of the change in the body burden. Assuming that 5% of the body
weight of a 75 kg applicator is blood serum, we estimated that the
uptake of atrazine in the blood serum during the handling-mixing-loading
and application stages was about 1.37 ppb * 3.75 kg blood serum « 5 jig
atrazine. The urinary excretion of the atrazine metabolite was not
detectable and the fecal excretion, of order 10% of the urinary
excretion, was also assumed to be negligible. If the human partition
coefficients for atrazine between blood and other tissues were known, we
could estimate the total accumulation that occurred in other body stores
such as adipose tissue. Because the inhalation dosage is only about
1 pg, the total accumulation would be almost all from dermal exposure.
THE NATIONAL HUMAN EXPOSURE ASSESSMENT SURVEY (NHEXAS) PILOT STUDIES
The NHEXAS has been designed by a U.S. federal interagency task
force and a team of outside cooperative investigators3. Two of its goals
are to conduct pilot studies towards measuring the national
distributions of total exposure and to identify the individuals with the
highest exposures to a mix of target pollutants, ranging from heavy
metals to volatile organic chemicals and pesticides3. We believe that
the individual with the highest total exposure to a substance during the
study period should be defined as the one who has the greatest absorbed
dose during the study period, rather than the greatest applied dose.
However, the estimate of the absorbed dose may require a material
balance to estimate the uptake of the pollutant into the systemic
circulation.
Several NHEXAS pilot studies are now in progress which are testing
procedures for measuring and estimating applied doses. However, these
pilot studies, in contrast to the GEMS/HEAL design, provide for no
analyses of total fecal collections that lead to an estimate of the
total applied dose of a nonvolatile pollutant and thus may not be able
to identify a case where an unmeasured route of exposure is important.
The NHEXAS pilot studies only measure blood levels at the end of the
exposure study period, and cannot provide an estimate of the
accumulation in the blood, or any other measure of change in the
internal dose during the study.
An exposure pilot study with each respondent providing a single
blood sample at the end cannot provide a measure of accumulation.
Unless the exposure measurements cover a period of order five or more
biological half-lives (T z 5 th in equation 4) of the pollutants in the
body the initial condition BB (0) cannot be neglected. For such a long
duration, the initial mass of pollutant in the body will decrease by a
factor of order (0.5)5 ~ 1/32, so the error in neglecting the initial
body burden will be less than 3% of the initial value. Because the
NHEXAS pilots cover multiple target pollutants with half-lives varying
between days and months, a one-week study period may not be sufficiently
long enough to correct for the absence of an initial blood sample at the
start of the study for most of the pollutants.
Therefore, the NHEXAS may need another pilot study to estimate the
total excretions of pollutants and the changes in the body stores of the
pollutant to demonstrate the ability to quantify the uptake of
pollutants via dermal routes of exposure and the ability to quantify the

-------
immeasurable uptake of pollutants by the incidents of exposure (e.g.,
cigarette smoking). The present study design being evaluated in a pilot
study may not be able to measure the exposure of the maximally exposed
individuals above the 90th percentile of the absorbed dose distribution
or the complete absorbed dose distribution for two of its targets,
benzene and cadmium, because the highest 10% of the benzene and cadmium
exposed individuals in the target population are likely to be smokers.
In addition, the NHEXAS pilots do not take complete duplicate diets
over the study periods2. Experience from GEMS/HEAL11"13 shows that at
least a one-week duplicate diet is necessary to get a stable average
dietary exposure to metals because of repetitive weekly patterns (e.g.,
eating fish on Friday). Figure 3 shows an example in which one day out
of seven had an exceedingly high cadmium and lead applied dose. Because
the NHEXAS pilots are only collecting duplicate diets for one-day2 or
four-days during a seven-day exposure study period15, another pilot
study may be necessary to evaluate the collection of a one-week diet.
A NHEXAS pilot study2 also plans to monitor personal air exposures for
only one out of seven days, and to estimate the occupational air
exposures from literature values for similar occupational exposures.
Therefore, a series of untestable assumptions (e.g., hand-to-mouth
uptake of lead from a measure of lead in house dust) will be necessary
to estimate applied doses of all the pollutants for those routes of
exposure in which measurements are not taken in a pilot study. Another
pilot study of the collection of jbe/ore-and-after blood values, a
collection of excretions of urine and total feces over the study, and
measured exposures over all seven days of the one-week study period may
be necessary to supplement the NHEXAS phase 1 pilot study to validate
the current estimates of total exposure. Therefore, we recommended that
any future NHEXAS work consider including another pilot study with full
mass balance components in a small cohort to provide data that would
show whether the objectives of the phase 1 pilot have been met.
SUMMARY AND CONCLUSIONS
Future total exposure survey designers should be conscious that a
choice to save funds by incomplete monitoring in a pilot study may not
be cost effective16. We recommend that all future studies of total
exposure be framed in the context of measurement of the absorbed dose
and that total exposure be computed by summing over all exposure routes
as mg/kg-day. All measurable major routes of exposure should be
measured and not estimated (e.g., diet and occupational exposures) in a
subsample of the population. Measurements of total accumulation by
before-and-after measurements of body fluids and total excretions, such
as .urine and feces, should be required in the subsample of a pilot study
to validate estimation procedures that may be adopted to reduce costs16.
The estimated contributions by unmeasurable routes of exposure (e.g.,
smoking, dermal, hand-to-mouth) to the applied and absorbed doses should
be estimated by a difference technique, using mass balance equations.
DISCLAIMER
This article has been subjected to reviews by the U.S. Environmental
Protection Agency. The views and opinions expressed are solely those of
the authors, and do not necessarily reflect the views or policies of the
Agency.

-------
TECHNICAL REPORT DATA

2.

4. TITLE AND SUBTITLE
Incorporating Mass Balance concepts in total
exposure studiesr
5.REPORT DATE
6.PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
David T. Mage, USEPA/NERL/EAB and Maria Conner, CUT
8.PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
USEPA/NERL/AERD/EAB
MD-56, RTP, NC 27711
10.PROGRAM ELEMENT NO.
Human Exposure
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
USEPA/NERL/AERD/EAB
MD-56, RTP, NC 27711
13.TYPE OF REPORT AND PERIOD COVERED
Proceedings of 1996 EPA/AWMA
Symposium on Measurements of
Toxic Substances
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Total exposure studies require the monitoring of personal exposures to pollutants
over all five routes of exposure: 1) Inhaling air; 2) Drinking water; 3) Eating
food; 4) uptake through the skin; 5) Other unique incidents, such as thumb sucking,
and chewing or smoking tobacco. To evaluate their potential effect on human health,
the exposures via these five routes can be added together as a total applied dose or
a total absorbed dose over the period of the study {e.g., mg/kg/day). The absorbed
doses via routes 4 and 5 are not directly measurable in a field study, and are
commonly estimated from unvalidated models. However, the applied dose can be
determined by the Law of Conservation of Mass, which states,
Applied Dose - Output from the Body + Change of Body Burden.
The portion of the applied dose that the body absorbs must also obey the Law of
Conservation of Mass;
Absorbed Dose = Bodily Eliminations + Change of Body Burden.
We present three exposure survey designs from the WHO/UNEP Human Exposure Assessment
Locations (HEAL) Programme, the NCI/NIEHS/EPA Agricultural Health Study (AHS), and
the EPA National Human Exposure Assessment Study (NHEXAS) and discuss their
abilities to estimate an applied dose or an absorbed dose of target subjects using a
mass balance approach.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/ OPEN ENDED
TERMS
c.COSATI



18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This
Report)
21.NO. OF PAGES
20. SECURITY CLASS (This
Page)
22. PRICE

-------