Journal of Exposure Science and Environmental Epidemiology (2008) 18, 360–368
r 2008 Nature Publishing Group All rights reserved 1559-0631/08/$30.00
www.nature.com/jes
Creatinine corrections for estimating children’s and adult’s pesticide intake
doses in equilibrium with urinary pesticide and creatinine concentrations
DAVID T. MAGEa, RUTH H. ALLENb AND ANURADHA KODALIa
a
Danya International Inc., 8737 Colesville Road, Suite 1100, Silver Spring, Maryland 20910, USA
United States Environmental Protection Agency, Office of Prevention, Pesticides and Toxic Substances, 1200 Pennsylvania Avenue NW 7509P,
Washington, District of Columbia 20460, USA
b
A urine contaminant concentration per se has uncertain meaning for human health because of dilution by hydration. However, the estimation of the
health-related daily intake dose of pollutant (mg/kg/day) that equilibrates with a spot urinary concentration of a pesticide residue or metabolite, or other
analyte, can be made using creatinine-corrected toxicant levels (mg analyte/mg creatinine) multiplied by an estimate of the subjects’ expected creatinine
excretion rates (mg creatinine/kg/day). The objective was to develop a set of equations predicting a person’s expected daily creatinine excretion (mg/kg) as
a function of age, gender, race and morphometry, from birth to old age. We review the creatinine excretion literature where infants, children and adults
provided 24 h total urine samples for creatinine analysis. Equations are developed for infants (r3 years), children (3–18 years) and adults (Z18 years)
that match at 3 and 18 years. A series of equations that estimate daily creatinine excretion (mg/day) are developed that are piecewise continuous from
birth through infancy through adolescence and through adulthood for males and females, and Black and White races. Complicating factors such as diet,
health status and obesity are discussed. We propose that these equations, with caveat, can now be used with measured urine concentrations to consistently
estimate the corresponding equilibrium intake doses of toxicants at ages from birth to 92 years for the healthy non-obese. We recommend that this system
of equations be considered for future development and reporting of applied doses in mg/kg/day of pollutants and toxicants that are measured in urine
samples, as in the National Health and Nutrition Examination Survey.
Journal of Exposure Science and Environmental Epidemiology (2008) 18, 360–368; doi:10.1038/sj.jes.7500614; published online 19 September 2007
Keywords: creatine, creatinine, NHANES, pesticides, children, adults, obesity.
Introduction
The Centers for Disease Control and Prevention (CDC)
National Center for Health Statistics (NCHS) conducts the
National Health and Nutrition Examination Study
(NHANES), in a probability sample of the non-institutionalized civilian population of the United States on a
continuing basis. As part of the NHANES, blood and spot
urine samples are collected from a random subset of the
NHANES probability sample and are analyzed for creatinine
and toxicants, including pesticide metabolites and residues,
described together as pesticide analytes (Barr et al., 2005).
The US Environmental Protection Agency (USEPA, 2003)
recognizes the value of the NHANES data and encourages
their use as indicators of exposure to environmental
pollutants. These values are reported in units of mg/dl blood
serum, mg/l urine and mg/g creatinine in the urine.
Interpretation of these values for pesticides is made difficult
because, by themselves a concentration in blood or urine may
1. Address all correspondence to: Dr. David T. Mage, 18 W. Periwinkle
Ln, Newark, DE 19711-6212, USA.
Tel: þ 1 302 235 7156. Fax: þ 1 302 338 8833.
E-mail: [email protected]
Received 3 April 2007; accepted 25 May 2007; published online 19
September 2007
not be directly related to a health effect. Alone, urinary
contaminant data are of uncertain value because of the highly
variable dilution caused by wide fluctuations of fluid intake.
The USEPA establishes reference doses (RfD) for
pesticides and other chemicals representing the daily intake
in mg/kg body weight over a 70-year lifetime that is not
expected to be associated with an adverse health effect. These
values are estimated from human and animal studies, and
include several safety factors that account for interspecies
differences and intraspecies variability. Thus, there is interest
in relating urine or blood concentrations of pesticides in the
population to exposures of pesticide residues and to how
these exposures relate to the RfD. Replicable children’s
creatinine correction methods extend the work previously
completed for adults (Mage et al., 2004) and the usefulness
of NHANES human biomonitoring data on pesticides for
regulatory risk analyses and indicator development, by
reduction of uncertainty in single spot urine sampling
reported as concentration alone.
There are two possible independent approaches to relate
the NHANES blood serum and urinary analyte measurements to an RfD for a parent pesticide:
1. By physiologic-based pharmaco-kinetic modeling. One
can estimate the equilibrium or ‘‘biological equivalent’’ blood
serum contaminant concentration (mg/dl), which would result
Mage et al.
Creatinine excretion rates of children and adults
from a constant daily dietary intake at the RfD level
(Timchalk et al., 2002; Hays et al., 2007). With a
physiologic-based pharmaco-kinetic (PBPK) submodel of
renal clearance and an assumption of the daily fluid intake
(l/day), estimate the average daily urinary contaminant
concentration, mg/l, which would be excreted in equilibrium
with the calculated blood serum contaminant concentration.
Then the measured serum value (mg/dl) or urine value (mg/l)
could be compared to those reference values corresponding to
the RfD (Rigas et al., 2001).
2. By creatinine correction. One can convert the urinary
analyte per mg of creatinine into an equilibrium daily
excretion of the pesticide analyte that corresponds to a daily
intake (mg/kg/day) for comparison with the RfD level of the
parent pesticide (Boeniger et al., 1993; Mage et al., 2004).
Creatinine is a natural metabolic product of creatine that
occurs in muscle tissue and an approximately constant
amount of creatinine is produced daily with additional
excretion of dietary creatinine and metabolized dietary
creatine. It is a molecule of moderate size (MW ¼ 113.12)
that can be filtered and secreted by renal processes. It is
mainly eliminated from the blood by glomerular filtration
and to a lesser extent by tubular secretion. This leads to
possible nonlinearities if other larger ligand molecules (e.g.,
drugs, such as penicillin) compete with creatinine for the
same chaperone proteins that accompany the creatinine when
it is secreted through the renal tube walls, and a transport
maximum is reached.
One could also use expected 24-h urine volume to scale up
a spot urine sample concentration value. However, ‘‘within
an age group, the daily volume of urine probably increases
with body size’’ (ICRP, 1975), and the daily variability of
fluid intake and seasonality of mean fluid intake makes these
values much more uncertain than creatinine excretion (CE)
rates. The ICRP (1975) (Figure 71) reports the range of
lower to upper limit of urine excretion values for adult males
as 0.5–3.0 l/day. Comparing the 5–95 percentile ranges in
male children (9–13 years), the daily urine volume variability
is 12–43.2 ml/kg (Mattsson and Lindstrom, 1995), which is
twice as large as the daily CE variability of 11.3–27.7 mg/kg
(Remer et al., 2002).
The latter creatinine approach is simpler than the PBPK
approach, and it has been utilized by Mage et al. (2004) for
male and female adults (18oageo92 years) with the daily
adult CE relationships of Cockcroft and Gault (1976). They
compared the urinary pesticide analyte concentrations in
NHANES III to RfD for the appropriate parent pesticide.
Blount et al. (2006) also applied their equations to NHANES
2001–2002 adult urinary perchlorate data.
In this study, we extend this technique to children in the
age range 0–3 and 3–18 years using the tabulated CE values
of Viteri and Alvarado (1970) and Remer et al. (2002),
respectively, on White (non-Hispanic) children. NHANES
recently reported data by race as non-Hispanic White, nonJournal of Exposure Science and Environmental Epidemiology (2008) 18(4)
Hispanic Black and other races combined (e.g., Pacific
Islander, Eskimo and Asian), and by all races with Hispanic
ethnicity. A racial correction factor is introduced to account
for the higher lean-body mass (LBM) and CE rate of nonHispanic Black adults and older non-Hispanic Black children
as they approach adulthood, but no factor, if needed, is
available for other NHANES ethnicity groupings; for
example, Mexican-American and other Hispanic adults and
children. Ethnicity, like race, is self-reported and alone it is
not expected to have a major influence on musculature. In the
following sections, ‘‘Black’’ will refer to non-Hispanic Black
subjects and ‘‘White’’ will refer to non-Hispanic White
subjects as defined within NHANES. Cachexia (wasting
diseases) and obesity are important modifying factors in CE
estimation for all races and corrections for their effects are
discussed.
We address the necessary boundary condition that must be
met for any analysis that compares children and adult
exposures and doses. Because of sampling error in the
happenstance choice of subjects and unavoidable experimental
error in the analysis of creatinine and collection of 24-h total
urine volumes, these child and adult equations do not mesh at
3 or 18 years, respectively. They must be made to coincide at
these borders for consistency. This allows such a person to
have their CE predicted the same way at 3 years whether from
below as an infant or from above as a young child, and at 18
years from below as an adolescent and from above as an adult.
We therefore sought to adjust these equations to create a set of
equations F appropriately caveated for cachexia and obesity
F that might follow a child from birth and infancy through
childhood and into adulthood, that continuously closely
approximate the individual and independent data sets that
cover this entire range.
A subsequent paper will apply these creatinine corrections
to convert NHANES pesticide analyte concentrations into
pesticide intake doses in children and adults. The NHANES
data are quality assured by CDC and NCHS and after
careful review are made available as a public release data file.
This file lists, for each subject sampled for pesticide and
toxicant-related analytes, their race, age, gender, morphology
(e.g., height, weight and waist circumference) and urinary
creatinine concentration, along with a probability-based
sample weight that can be used to extend their data to the
general population of the US children and adults from which
they are drawn.
Materials and methods
The ICRP (1975) (Figure 72) summarizes the average data
from birth to 100 years on excretion of creatinine (mg/day)
as a function of gender. We use more detailed and recent CE
data, which also report height and weight to allow individual
subject variations to be modeled. Our equations, described
361
Mage et al.
Creatinine excretion rates of children and adults
Daily White child creatinine excretion, mg/cm
below, predict median values similar to those reported by
ICRP (1975) (e.g., 1.9 g creatinine/day for a 20-year-old
male of unspecified race, height and weight).
Viteri and Alvarado (1970) report expected child CE rates
from birth (35.5 mg/day at a length of 50 cm) to an age of 8.5
years (616.6 mg/day at a height of 129 cm) as derived from
the data of Daniels and Hejinian (1929) on healthy White
children. We use these data between birth and 36 months of
age (96.2 cm height) for males and females, as shown in
Figure 1.
Remer et al. (2002) measured daily CE (mg/day) and
morphometry (height, weight and LBM) of White German
children, 3–18 years old. Appropriate quality control
procedures were applied and if any suspicion arose that a
urine collection was not complete, those data were discarded.
Smoothed CE data were tabulated and reported for both
genders as a function of height from 90 to 186 cm (male) and
from 90 to 172 cm (female) in 2 cm steps, as shown in
Figure 1. All children consumed presumably omnivorous
self-selected diets so no control was placed on the amount of
red meat intake that contributes to a child’s daily CE.
The relationships developed based on these data can be
applied to the NHANES pesticide analyte data for White
children to report the expected pesticide dose that would
equilibrate with those analyte values. Koch et al. (2006) have
also used these same CE data to estimate average daily intake
of di-n-butylphthalate and butylbenzylphthalate in German
children based on their urine analyses.
The estimated CE rate from 18 years through adulthood
has been reported (Mage et al. 2004). They used a derivation
based on the creatinine clearance equations of Cockcroft and
Gault (1976), who measured 24-h urine collections for 249
male and 18 female adults in the age range from 18 to 92
years. These authors did not mention the race of their
subjects, but because the study was carried out in the early
1970s at a Veterans Hospital in Montreal, Canada, we
assume that they were predominantly of White race. These
267 subjects were medical ward patients who gave two 24-h
12
Girls
Boys
10
8
6
Results
CE for Ages 0 (Birth) to 36 Months
CE rates are reported for well-nourished White children
(Viteri and Alvarado, 1970; Viteri et al., 1971) that range
from 35.5 mg/day for a 50-cm length at birth (0 month) to
250 mg/day for a 36-month-old child of 96.2 cm height. We
note that there appears to be a rapid increase in CE rate
between birth and 3 months and a lower rate of increase from
3 to 36 months. We therefore fit these ranges with two Eqs.
(1) and (2), constrained to match each other at 3 months:
CE mg=day ¼ 36 þ 10:33 ðageÞ; ageo3 months
ð1Þ
CE mg=day
¼ 67 þ 5:47 ðage 3Þ; 3oageo36 months
ð2Þ
These equations apply equally to male and female, both
Black and White, because in infancy there is virtually no
difference in musculature development between the races and
genders (ICRP, 1975).
CE for Ages 3–18 Years
Smoothed 24-h CE values of White male and female
children, aged 3–18 years, as a function of height are
reported by Remer et al. (2002) as shown in Figure 1. We
analyzed this daily CE (mg/day) as a function of height (cm)
in three ways: CE vs. Ht; log(CE/Ht) vs. Ht; CE/Ht vs. Ht.
For males, the relation CE/Ht vs. Ht gives the best fit
(maximum correlation) with slope and intercept shown in
Table 1. For females the relation log(CE/Ht) vs. Ht has the
optimal fit shown in Table 1 over the complete range of
heights from 90 to 172 cm.
There is a discontinuity in the rate of change of male CE at
168 cm. Therefore, two separate models, Eqs. (3a) and (3b),
are used below and above that height, respectively, and
Eq. (4) applies to females.
Child White male CE mg=day
4
¼ Ht ½6:265 þ 0:0564 ðHt 168Þ Hto168cm
2
¼ Ht ½6:265 þ 0:2550ðHt 168Þ Ht4168cm
0
0
20
40
60
80
100
120
Child height, cm
140
160
180
200
Figure 1. White children’s creatinine excretion (mg/cm/day) vs. height
(cm). Data of Viteri and Alvarado (1970), birth to 3 years; data of
Remer et al. (2002), 3–18 years.
362
urine collections. Subjects rejected from the analysis had
either unstable serum creatinine, duplicate CE values that
differed by more than 20% or a CEo10 mg/kg/day. The
latter exclusion controls for the presence of muscle wasting
diseases such as Duchenne muscular dystrophy (Franciotta
et al., 2003) and Cushing’s syndrome (Hatton et al., 1989),
as they would significantly reduce the patient’s CE rate.
Child White female CE mg=day
¼ 2:045 Ht exp ½0:01552 ðHt 90Þ
ð3aÞ
ð3bÞ
ð4Þ
From knowledge of the children’s heights (in cm) the daily
expected CE (in mg/day) can be computed. Equations (3b)
Journal of Exposure Science and Environmental Epidemiology (2008) 18(4)
Mage et al.
Creatinine excretion rates of children and adults
Table 1. Models for predicting White children’s daily creatinine excretion (CE) rate (mg/day) as a function of height (cm).
Gender
Male
Male
Female
Bothb
Bothc
Height range (cm)
90–168
168–186
90–172
90–186
50–129
Prediction equation of CE (mg/day)
a
CE ¼ Ht (6.265+0.0564 (Ht–168))
CE ¼ Ht (6.265+0.2550 (Ht–168))a
CE ¼ 2.045 exp[0.01552 (Ht–90)]
CE ¼ 1.709 exp[0.02349 (Ht–90)]
CE ¼ (252/Ht)–8.01+0.0817 Ht
Correlation coefficient
r ¼ 0.9944a
r ¼ 0.9999a
r ¼ 0.9963
r ¼ 0.9327
Not reported
Source: Remer et al. (2002).
a
The male datum point at 168 cm was weighted by multiple entries to match both segments.
b
Remer et al. (2002).
c
Institute for Algorithmic Medicine (2006).
and (4b) can be extrapolated to male heights over 186 cm and
female heights over 172 cm, respectively. The median height
and weight for the US boys and girls as a function of age
from age 2 through 20 years are reported by Kuczmarski
et al. (2000). With these data the continuous change of CE
with age can be estimated for children of median stature,
and combined at age 18 years and above with their adult
estimates.
It should be noted here that these expected values are
ecologic and not specific to any particular child that Remer
et al. (2002) sampled. They are our model of their best
estimates of the average daily CE of White children of a given
gender and height in the age range of 3–18 years. These
equations give results that are similar to a previously reported
ecologic equation (Institute of Algorithmic Medicine, 2006,
based on the data of Viteri and Alvarado (1970)) for
both male and female daily CE (mg/day) ¼ (252/
Ht)8.01 þ 0.0817 Ht, with height in cm, age from birth to
8.5 years. This similarity supports the use of creatinine
correction as being more stable than a volume correction
(estimated daily volume of urine to scale up a single void
sample) that may lead to a higher estimate (Koch et al., 2006).
CE for Ages 18–92 Years
Mage et al. (2004) used Eqs. (5) and (6) for CE of White
adult males and females, respectively, from the models of
Cockcroft and Gault (1976) for their non-obese subjects
(Wt, kg; Ht, cm).
Adult White male CE mg=day
¼ 1:93 ½140 Age ðyearÞ Wt1:5 Ht0:5 =1000
ð5Þ
Adult White female CE mg=day
¼ 1:64½140 Age ðyearÞ Wt1:5 Ht0:5 =1000
ð6Þ
Adjustment of Equations (1)–(6) for Continuity
Given the three data sets we used were determined
independently, there is a necessary boundary condition that
they must satisfy for consistency. At age 18 years, the upper
Journal of Exposure Science and Environmental Epidemiology (2008) 18(4)
asymptote of the CE equations (3a), (3b) and (4) for the
children must match CE equations (5) and (6) for the adult at
their lower asymptote. We have investigated this boundary
condition as follows:
Eqs. (3a), (3b) and (5) for these males match identically at
age 18 years and a height of 168 cm for a weight of 49.3 kg,
which is below the normal US range for that height. The
median height and weight for the 18-year-old US male are
176 cm and 67 kg, respectively (Kuczmarski et al., 2000).
Under these conditions the child Eq (3b) predicts 1462 mg/day
and the adult Eq. (5) predicts 1713 mg/day.
We suggest adjusting the adult’s equation down and the
children’s equation up so that they both predict the same
average value at 176 cm and 67 kg, of (1462 þ 1713)/2
¼ 1587 mg/day.
The child’s equation estimated is scaled up by a factor of
1587/1462 ¼ 1.085. The adult’s equation estimate is scaled
down by a factor of 1587/1713 ¼ 0.926.
We also evaluate female Eqs. (4) and (6) for an 18-yearold child and adult, respectively, at the US female median
height of 164 cm and median weight of 56 kg. The adult
estimate is 1074 mg/day and the child estimate is 1058 mg/day,
with an average of 1066 mg/day. Reducing the adult female
estimate by a factor of 1066/1074 ¼ 0.993 and increasing the
child estimate by 1066/1058 ¼ 1.008, we obtain the factors
shown in the "smoothed" set of CE equations for children
and adults in Table 2.
We then also need to adjust the child equations of Viteri
and Alvarado (1970) and Remer et al. (2002) in Table 2 so
that they both predict the identical CE at an age of 3 years
for males and females. Equation (2) predicts at 36 months a
CE of 247 mg/day. The Remer et al. (2002) Eqs. (3a), (3b)
and (4) in Table 2 for males and females at age 3 years, at the
median US heights of 95 and 94 cm, respectively, predict CE
values of 221 and 204.5 mg/day, respectively.
The Remer et al. (2002) equations are based on more
modern creatinine analysis techniques (2002 vs. 1929). The
Folin creatinine procedure used by Daniels and Hejinian
(1929) did not correct for a positive bias from non-creatinine
chromogens in urine. Thus, we propose to adjust the Viteri
and Alvarado (1970) equations for male and female by
363
Mage et al.
Creatinine excretion rates of children and adults
Table 2. Proposed sets of creatinine excretion equations for healthy, non-obese. Black and White males and females, from birth through adulthood
(92 years).
Gender
Number of subjects (n)
Age (years)
Height (cm)
Male
Male
Male
Male
Male
Male
Male
Female
Female
Female
Female
Female
9
NA
203a
0–0.25
0.25–3
3–14
3–14
15–18
15–18
18–92
0–0.25
0.25–3
3–14
15–18
18–92
50–60
60–96
90–168
168–186
90–168
168–186
Unlimited
50–60
60–96
90–172
90–172
Unlimited
22a
249
0b
NA
211a
17a
18
Matched equations in mg creatinine/day, showing race correction
0.895 [36+124 (age)]
0.895 [67+65.6 (age–0.25)]
1.085 Ht [6.265+0.0564 (Ht–168)]
1.085 Ht [6.265+0.2550 (Ht–168)]
1.085 Ht [6.265+0.0564 (Ht–168)] [1+0.045 (age–14) B]
1.085 Ht [6.265+0.2550 (Ht–168)] [1+0.045 (age–14) B]
0.926 1.93 [140–age] (Wt1.5 Ht0.5 ) (1+0.18 B)/1000
0.828 [36+124 (age)]
0.828 [67+65.6 (age–0.25)]
1.008 Ht 2.045 exp[0.01552 (Ht–90)]
1.008 Ht 2.045 exp[0.01552 (Ht–90)] [1+0.045 (age–14) B]
0.993 1.64 [140–age] (Wt1.5 Ht0.5) (1+0.18 B)/1000
a
Estimated from 28 male and 22 female of subjects age 14–18 years. Height breakout for male ages is not available.
Creatinine values in male subjects (Daniels and Hejinian, 1929) are assumed identical to the values in female subjects by Viteri and Alvarado (1970).
B ¼ 1 if Black, B ¼ 0 if not Black. Modified to match at 3 and 18 years at median heights and weights for the US children (see Eqs. (13) and (14) for
corrections when child’s weight is not the median weight for age).
White male creatinine excretion, mg/day
b
1800
Model
Data
1600
1400
1200
1000
800
600
400
200
0
0
10
20
30
40
50
60
Age, years
70
80
90
100
Figure 2. Continuous model of creatinine excretion rate of a White
male compared to the data segments. Child: assumed 50 cm at birth
and 95 cm at age 3 years. Adult: assumed 176 cm and 67 kg from 18 to
92 years.
of 92 years, the upper age limit of the Cockcroft and Gault
(1976) sampled population. No correction is made for the
expected decrease in weight and height that occurs at older
ages (ICRP, 1975).
The CE increases monotonically until full growth is
attained, at which point there is a gradual decrease of
musculature with aging, leading to a decline through the term
(140 F age (years)). ICRP (1975) reports full body growth
at age 20 years, but Cockcroft and Gault (1976) grouped
their data as ages 18–29 years and the Remer et al. (2002)
data did not extend to 20 years, so we are unable to model
accurately this maximum region. In Figure 2, the unconnected infant, child and adult data segments are shown for
comparison.
Discussion
factors of 221/247 ¼ 0.895 and 204.5/247 ¼ 0.828, respectively, from birth to 36 months. Therefore, the formulae in
Table 2 are written showing each correction factor for the
reader to be able to readily follow our continuity-smoothing
procedures. Note that in Remer et al. (2002), their youngest
female children at heights 90–96 cm had higher smoothed CE
rates than males of similar height. Although this difference is
not expected and may be caused by dietary differences and
small sample size (n ¼ 3 male, n ¼ 13 female), we use these
smoothed data as published.
Figure 2 shows the application of the CE equations in
Table 2 for White males at the median weight for age and
height. The CE, (in mg/kg/day), increases from birth to age
18 years with the median US adult height of 167 cm and
weight of 76 kg (non-obese body mass index (BMI) ¼ 27.3),
which is assumed constant throughout adulthood to an age
364
These above results are ecologic in nature and apply to an
average healthy and non-obese White person of the given
age, gender, height and weight. Ascites (serous fluid retention
in the abdominal cavity), cachexia (muscle wasting diseases)
and anorexia, may cause CE to be significantly lower than
the predictions based on presumably healthy subjects. In
addition, there are important corrections that can be made,
such as for Black race and deviation from normal
morphology such as morbid obesity, that are discussed in
the following sections.
Increased CE Rates of Higher LBM Black Populations
Young Black adults have on the order of 20% greater LBM
than a White adult of similar age, gender and height and
weight, and have a correspondingly higher CE rate (James
et al., 1988; Goldwasser et al., 1997; Levey et al., 1999;
Journal of Exposure Science and Environmental Epidemiology (2008) 18(4)
Mage et al.
Creatinine excretion rates of children and adults
person is trying to increase muscularity. Seasonal variations
in physical exercise and dietary intake of red meat containing
creatine and its metabolite creatinine do not appear to create
a seasonal variation in CE rates (Robertson et al., 1975).
However, daily variations in exercise and red meat intake can
lead to variable CE rates from day-to-day for the same
person with a coefficient of variation (s/m) of 13% (Greenblatt et al., 1976). Diurnal fluctuations of order 5%–10%
from the 24-h mean value in CE have been noted and related
to meal patterns where red meat was ingested (Pasternack
and Kuhlbäck (1971). Subjects following a vegetarian diet
are identified in NHANES, and their CE rate will be lower
than for an omnivorous subject with the same age, gender,
race and morphology. Consequently, the CE data reported
by Remer et al. (2002) for children and Cockcroft and Gault
(1976) for adults may have had wider confidence intervals
than one would associate with a study where dietary red meat
intake per kilogram body mass was a constant.
National Kidney Foundation, 2006). As aging progresses
there may be a decline in the excess LBM of Black adults
relative to White adults of similar gender, age and
morphometry. However, in the absence of data on this effect
of aging, we assume that the excess of Black CE compared to
White CE continues throughout adulthood.
There is an indication of a similar increased CE for Black
children compared to White children, both at age 10 years.
‘‘However, after adjusting for body weight, differences in
CE were nonsignificant’’ (Pratt et al., 1993). This lack of an
excess Black CE rate is expected in prepubescent children and
it increases to the adult CE value of order 20% excess as their
adolescent musculature fully develops with approaching
adulthood. Thus, the above CE Eqs. (3a), (3b) and (4)
may not apply similarly to both Black and White children
after they reach puberty.
Remer et al. (2002) report that their sample of German
White males reach their maximum rate of growth at a mean
age of B16.5 years, consistent with the US data of
Kuczmarski et al. (2000), confirming that the period from
15 to 18 years is one of rapid increase in musculature. We
therefore introduce a multiplicative correction factor of
[1 þ 0.045 (age14) B], where B ¼ 1 if Black and B ¼ 0 if
non-Black, for ages 15–18 so that at age 18 years the
correction factor is 1.18 as reported for Black adults (Levey
et al., 1999; Palmer, 2004). There is an indication that
Mexican-American adults may also have a higher CE rate
than White adults in the NHANES III Study (Mattix et al.,
2002), but because these results are not standardized by age,
gender, height and weight, they are not used here to justify a
Hispanic correction factor for the White CE equations at this
time.
In a following paper, these equations for Black and nonBlack (White and Hispanic) children are applied to the
NHANES 2001–2002 children’s pesticide data. In addition,
we compare the race-corrected adult results to the urinary
perchlorate data, analyzed by Blount et al. (2007) for adults
20 years and older using the equations (Mage et al., 2004)
without the corrections for Black race.
Effects of Sample Error and Creatinine Measurement
Error
Table 3 shows the wide person-to-person variability of
measured CE of White male children, perhaps due to the
sampling error associated with a small number of children
sampled per cell and differences in amounts of red meat in
recent meals of the subjects prior to and during their 24-h
urine collection. These data do not constitute random
probability samples of all White children in their categories
of age, gender, weight and height, so there may also have
been some selection bias.
It is expected that taller and heavier children will usually
excrete more creatinine per day but that is not observed in
this small sample (see following discussion on the effect of
obesity). However, Remer et al. (2002) have smoothed their
data (see Table 3, column 6) so that their final best estimates
shown are consistent with the expectation that metabolic
creatinine production increases with body mass, and the heat
loss from the body surface area (BSA) increases with height
and weight. They found their daily CE rate based on BSA
correlated with age (r ¼ 0.73) better than when based on
LBM (r ¼ 0.49) or body weight (r ¼ 0.45). However, there is
controversy whether creatinine clearance should be normalized to BSA or LBM (Hallynck et al., 1981), and LBM is
discussed in the following section on obesity.
Effects of Diet on CE
Creatinine measurements made on human subjects are
influenced by dietary red meat intake. Its parent creatine is
also a food additive sometimes used with exercise when a
Table 3. The creatinine excretion data of Remer et al. (2002) on White male children, showing wide experimental variability that is adjusted by
smoothing in accord with theory.
Sample size n
4
3
8
5
Mean height (cm)
172.7
176.7
182.4
186.4
Mean weight (std) (kg)
63.4
68.5
73.5
75.4
CE (std) (mg/day)
(3.6)
(8.5)
(8.6)
(9.6)
Journal of Exposure Science and Environmental Epidemiology (2008) 18(4)
1277
2058
1816
1594
(485)
(308)
(478)
(505)
Smoothed height (cm)
Smoothed CE (mg/day)
172
176
182
186
1244
1459
1800
2014
365
Mage et al.
Creatinine excretion rates of children and adults
The different offsets between Cockcroft and Gault (1976)
and Remer et al. (2002) for male and female can possibly be
due to a combination of a small female sample (Cockcroft
and Gault, n ¼ 18) and different methods of creatinine
analysis. Remer et al. (2002) used a kinetic Jaffé analysis by a
Beckmann-2 creatinine analyzer (Beckmann Instruments,
Fullerton, CA, USA) and Cockcroft and Gault (1976) used
the Jaffé analysis via the autoanalyzer method N-11B
(Technicon Instruments, Tarrytown, NY, USA) for urine
and blood serum creatinine analysis. The method N-11B
corrected the Jaffé analysis for non-creatinine chromogens by
their differential rates of color development (Fujiwara et al.,
1997) so that it may not have been the cause of the
overprediction relative to Remer et al. (2002) at the same age
and height.
Effect of Obesity on Creatinine Estimation
Obesity can be related to a (BMI ¼ Wt (kg)/(Ht(m))2)430.
This is not a strict definition because a subject, such as a body
builder, could have a high BMI and not be considered obese.
NHANES measures and reports waist circumference, and
measures of skinfold thicknesses at the triceps and subscapular, that can be combined with BMI to separate the
muscular from the obese. An excessive increase in weight at a
fixed height if caused only by ascites (abdominal fluid
retention) or by an increase in adipose tissue (non-LBM) has
no direct impact because serous fluids and adipose tissue
metabolism do not produce creatinine. Consequently, use of
the Cockcroft and Gault (1976) relations for obese adults
(male420% body fat and female430% body fat) tends to
significantly overestimate CE, and those authors stated ‘‘a
correction to lean or ideal body weight is advised when these
abnormalities [ascites and obesity] are pronounced.’’
Such correction for CE in the obese may be made using
estimates of their LBM. The LBM may be computed from
weight and weight while submerged in water, or by bioelectric
impedance analysis (Segal et al., 1988). It may also be
estimated by use of charts and tables that predict the ideal
weight-for-height (Stavig et al., 1984). James (1976)
proposed Eqs. (7) and (8) for predicting the obese-adult
LBM from male and female weight and BMI, respectively:
Male LBM ðkgÞ ¼ Wt ð1:10 0:0128 BMIÞ
ð7Þ
Female LBM ðkgÞ ¼ Wt ð1:07 0:0148 BMIÞÞ
ð8Þ
The ideal weight for height (LBM) could then be substituted
for measured body weight in the child and adult Eqs. (3–6)
when indications of obesity are reported.
Snider et al. (1995) report the least biased CE rates for the
obese (race not specified) can be estimated by Eqs. (9) and
(10) derived from those of Salazar and Corcoran (1988)
Male mg=kg day ¼ ð137 AgeÞ ð0:0805
þ 3:42=BMIÞ
366
Female mg=kg day ¼ ð146 AgeÞ ð0:081
þ 2:75=BMIÞ
ð10Þ
Equations (5) and (6) for adults, based on Cockcroft and
Gault (1976), adjust for the increase in BSA of the non-obese
due to changes in height and weight from a reference BSA of
1.73 m2. However, when the increase is due primarily to
adipose tissue, Eqs. (5) and (6) overpredict the obese
subjects’ CE rates, and we suggest that in studies with obese
subpopulations such LBM corrections are made.
We use the Cockcroft and Gault (1976) White adult
equations that include BSA as a variable. An increase in
body weight at fixed height increases the BSA resulting in
additional heat loss. This leads to a need for more metabolic
energy production that releases additional creatinine. After
6 months an infant’s renal function begins to be predicted by
BSA (USEPA, 2005) but we ignore this effect below 3 years
of age. Remer et al. (2002) analyzed their CE data
normalized to 1.73 m2 BSA but did not provide any
equations or tables with their results. We therefore introduce
a correction for the body weight and BSA. Mage et al. (2004)
assumed that the BSA correction to the CE per unit weight is
approximately proportional to the square root of the weight
(Mosteller, 1987); therefore, an 186-cm tall 85-kg male will
have only a 6% higher CE/kg relative to the median weight
of 75 kg. This holds provided that the 10-kg weight increase
consists of the same proportion of LBM as for the 75-kg
male of the same height and it is not all adipose tissue. The
increase in BSA would then account for the 6% increase in
daily CE per kg body weight. However, if the 10-kg weight
increase is all adipose tissue, then the Salazar and Corcoran
(1988) Eqs. (9) and (10) predict that the CE mg/kg/day will
decrease even though the BSA increases. Given the recent
increases in childhood obesity, we recommend applying this
correction effect for children’s CE (mg/kg/day) between ages
3 and 18 years when a high BMI is noted.
In summary, we propose the use of a multiplier equal to
the square root of the ratio of subject weight to the US
median weight for height and age as a correction for BSA
only in the non-obese. The median weights of the US male
and female children (Kuczmarski et al., 2000) may be fit by
cubic functions of age (Eqs. (11) and (12)) from 3 to 18
years:
Male Median Weight ðkgÞ
¼ 14 þ 1:433 ðAge 3Þ þ 0:22 ðAge 3Þ2
0:00533ðAge 3Þ3
ð11Þ
Female Median Weight ðkgÞ
¼ 14 þ 0:40 ðAge 3Þ þ 0:52 ðAge 3Þ2
ð9Þ
0:024 ðAge 3Þ3
ð12Þ
Journal of Exposure Science and Environmental Epidemiology (2008) 18(4)
Mage et al.
Creatinine excretion rates of children and adults
The BSA corrections to the CE /kg for the weight (kg) then
become
Male correction ¼ ½Weight=ð14 þ 1:433 ðAge 3Þ
instruments does not imply recommendation for their use.
The authors declare they have no conflicting financial
interests.
þ 0:22 ðAge 3Þ2
0:00533 ðAge 3Þ3 Þ0:5
ð13Þ
Female correction ¼ ½Weight=ð14 þ 0:40 ðAge 3Þ
þ 0:52 ðAge 3Þ2
0:024 ðAge 3Þ3 Þ0:5
ð14Þ
These corrections for weight variation from the median body
weight in the non-obese range of BMI can then be made to
the equations for ages 3–18 years, as shown in Table 2.
Conclusion
We have developed an internally consistent set of equations
that predict a healthy non-obese person’s daily CE (mg/day)
from birth up to age 92 years. These equations allow the
conversion of a spot urine sample value for mg of analyte/mg
creatinine to units of the RfD in mg/kg/day when the
subject’s age, gender, race, height and weight are known, and
BMI is not in the range of obesity. The equations meet the
necessary condition of piecewise continuity.
These equations are therefore presented as our best
estimates of CE at this point of time. Although they could
be improved upon as additional experimental data become
available on the effects of children’s weight, this effect is likely
to be within the experimental error for most of the
observations in the non-obese status. New data on the
effects of Black and Hispanic race on the CE equations
developed for White children could make a worthwhile
improvement. Corrections for obesity need to be made when
BMI exceeds 30 for males and females. In a following article,
we will apply these equations to the NHANES 2001–2002
pesticide analyte data for adults and children as an example
of their utility.
Acknowledgements
An anonymous reviewer gave excellent advice on medical
aspects that considerably improved this paper. This work is
supported by EPA Contract GS-10F-0484N to Danya
International Inc.
Disclaimer
This paper has been cleared for publication by the USEPA.
It represents the views of the authors but does not necessarily
represent Agency policy. Mention of chemicals or commercial
Journal of Exposure Science and Environmental Epidemiology (2008) 18(4)
References
Barr D.B., Wilder L.C., Caudill S.P., Gonzalez A.J., Needham L.L., and Pirkle
J.L. Urinary creatinine concentrations in the U.S. population: implications for
urinary biologic monitoring measurements. Environ Health Perspect 2005:
113: 192–200.
Blount B.C., Valentin-Blasini L., Osterloh J.D., Mauldin J.P., and Pirkle J.L.
Perchlorate exposure of the US population, 2001–2002. J Expo Sci Environ
Epidemiol 2007: 17: 400–407.
Boeniger M.F., Lowry L.K., and Rosenberg J. Interpretation of urine results used
to assess chemical exposure with emphasis on creatinine adjustments: a review.
Am Ind Hyg Assoc J 1993: 54: 615–627.
Cockcroft D.W., and Gault M.H. Prediction of creatinine clearance from serum
creatinine. Nephron 1976: 16: 31–41.
Daniels A.L., and Hejinian L.M. Growth in infants from the standpoint of
physical measurements and nitrogen metabolism. I. Creatinine. Am J Dis Child
1929: 37: 1128–1138.
Franciotta D., Zanardi M.C., Albertotti L., Orcesi S., Berardinelli A., Pichiecchio
A., Uggetti C., and Tagliabue A. Measurement of skeletal muscle mass in
Duchenne muscular dystrophy: use of 24-h creatinine excretion. Acta Diabetol
2003: 40(Suppl 1): S290–S292.
Fujiwara Y., Takahashi T., and Yamakido M. Re: Prediction of carboplatin
clearance from standard morphological and biological patient characteristics.
Response to correspondence. J Natl Cancer Inst 1997: 89: 969–970.
Goldwasser P., Aboul-Magd A., and Maru M. Race and creatinine excretion in
chronic renal insufficiency. Am J Kidney Dis 1997: 30: 16–22.
Greenblatt D.J., Ransil B.J., Harmatz J.S., Smith T.W., Duhme D.W., and
Koch-Weser J. Variability of 24-h urinary creatinine excretion by normal
subjects. J Clin Pharmacol 1976: 16: 321–328.
Hallynck T.H., Soep H.H., Thomis J.A., Boelaert J., Daneels R., and Dettli L.
Should clearance be normalised to body surface or to lean body mass? Br J
Clin Pharmacol 1981: 11: 523–526.
Hatton J., Parr M.D., and Blouin R.A. Estimation of creatinine clearance in
patients with Cushing’s syndrome. Ann Pharmacother 1989: 23: 974–977.
Hays S.M., Becker R., Leung H.W., Aylward L.L., and Pyatt D.W.
Biomonitoring equivalents: a screening approach for interpreting biomonitoring
results from a public health risk perspective. Regul Toxicol Pharmacol 2007:
47: 96–106.
Institute for Algorithmic Medicine. The medical algorithms project. Chapter 14,
Renal and Urinary System. Expected daily urine creatinine excretion in
children and the creatinine height index of Viteri and Alvarado. http://
www.medal.org/visitor/www/active/ch14/ch14.01/ch14.01.07.aspx,Accessed
November 11, 2006.
ICRP (International Commission on Radiological Protection). No. 23. Report of
the Task Group on Reference Man, Pergamon Press: Oxford, 1975.
James G.D., Sealey J.E., Alderman M., Ljungman S., Mueller F.B., Pecker M.S.,
and Laragh J.H. A longitudinal study of urinary creatinine and creatinine
clearance in normal subjects. Race, sex, and age differences. Am J Hypertens
1988: 1: 124–131.
James W.P.T. DHSCC/MRC Report on research on obesity, HMSO: London,
1976.
Koch H.M., Becker K., Wittassek M., Seiwert M., Angerer J., and KolossaGehring M. Di-n-butylphthalate and butylbenzylphthalate F urinary
metabolite levels and estimated daily intakes: pilot study for the German
Environmental Survey on children. J Expo Sci Environ Epidemiol 2006.
September 27; (E-pub ahead of print).
Kuczmarski R.J., Ogden C.L., Grummer-Strawn L.M., Flegal K.M., Guo S.S.,
Mei Z., Curtin L.R., Roche A.F., and Johnson C.L. CDC Growth Charts:
United States. Advance data from vital and health statistics; no. 314.
Hyattsville, Maryland. National Center for Health Statistics 2000.
Levey A.S., Bosch J.P., Lewis J.B., Greene T., Rogers N., and Roth D. A more
accurate method to estimate glomerular filtration rate from serum creatinine: a
new prediction equation. Modification of Diet in Renal Disease Study Group.
Ann Intern Med 1999: 130: 461–470. Comments in Ibid: 1999; 131: 629–630
and 2004;140: 934–935.
367
Mage et al.
Mage D.T., Allen R., Gondy G., Smith W., Barr D.B., and Needham L.L.
Estimating Pesticide Dose from Pesticide Exposure Data by Creatinine
Correction in the Third National Health and Nutrition Examination Survey
(NHANES-III). J Exposure Anal Environ Epidemiol 2004: 14: 457–465.
Mattix H.J., Hsu C.Y., Shaykevich S., and Curhan G. Use of the albumin/
creatinine ratio to detect microalbuminuria: implication of sex and race. J Am
Soc Nephrol 2002: 13: 1034–1039.
Mattsson S., and Lindstrom S. Diuresis and voiding pattern in healthy
schoolchildren. Br J Urol 1995: 76: 783–789.
Mosteller R.D. Simplified calculation of body-surface area. N Engl J Med 1987:
317: 1098.
National Kidney Foundation. http://www.kidney.org/professionals/KLS/
gfr.cfm#9,Accessed November 29, 2006.
Palmer B.P. Approach to the patient with renal disease. in ACP Medicine, Section
10, Chapter III, American College of Physicians, Northbrook, IL, 2004.
Pasternack A., and Kuhlbäck B. Diurnal variations of serum and urine creatine
and creatinine. Scand J Clin Lab Invest 1971: 27: 1–7.
Pratt J.H., Manatunga A.K., Bloem L.J., and Li W. Racial differences in
aldosterone excretion: a longitudinal study in children. J Clin Endocrinol
Metab 1993: 77: 1512–1515.
Remer T., Neubert A., and Maser-Gluth C. Anthropometry-based reference
values for 24-h urinary creatinine excretion during growth and their use in
endocrine and nutritional research. Am J Clin Nutr 2002: 75: 561–569.
Rigas M.L., Okino M.S., and Quackenboss J.J. Use of a pharmacokinetic model
to assess chlorpyrifos exposure and dose in children, based on urinary
biomarker measurements. Toxicol Sci 2001: 61: 374–381.
Robertson W.G., Peacock M., Marshall R.W., Speed R., and Nordin B.E.
Seasonal variations in the composition of urine in relation to calcium stoneformation. Clin Sci Mol Med 1975: 49: 597–602.
368
Creatinine excretion rates of children and adults
Salazar D.E., and Corcoran G.B. Predicting creatinine clearance and renal drug
clearance in obese patients from estimated fat-free body mass. Am J Med 1988:
84: 1053–1060.
Segal K.R., Van Loan M., Fitzgerald P.I., Hodgdon J.A., and Van
Itallie T.B. Lean body mass estimation by bioelectrical impedance
analysis: a four-site cross-validation study. Am J Clin Nutr 1988: 47:
7–14.
Snider R.D., Kruse J.A., Bander J.J., and Dunn G.H. Accuracy of estimated
creatinine clearance in obese patients with stable renal function in the intensive
care unit. Pharmacotherapy 1995: 15: 747–753.
Stavig G.R., Leonard A.R., Igra A., and Felten P. Indices of relative body weight
and ideal weight charts. J Chronic Dis 1984: 37: 255–262.
Timchalk C., Nolan R.J., Mendrala A.L., Dittenber D.A., Brzak K.A., and
Mattsson J.L. A physiologically based pharmacokinetic and pharmacodynamic (PBPK/PD) model for the organophosphate insecticide chlorpyrifos
in rats and humans. Toxicol Sci 2002: 66: 34–53.
USEPA. EPA Handbook for use of data from the National Health and Nutrition
Examination Surveys (NHANES): A goldmine of data of environmental health
analyses, EPA/600/R-02/044, National Center for Environmental Assessment,
Washington DC, March 2003.
USEPA. Guidance on Selecting Age Groups for Monitoring and Assessing
Childhood Exposure to Environmental Contaminants, EPA/630/P-03/003F,
National Center for Environmental Assessment, Washington DC, November
2005.
Viteri F.E., and Alvarado J. The creatinine height index: its use in the estimation of
the degree of protein depletion and repletion in protein calorie malnourished
children. Pediatrics 1970: 46: 696–706.
Viteri F.E., Alvarado J., and Alleyne G.A.O. Creatinine-height index: reply to
Drs Mendez and Buskirk. Am J Clin Nutr 1971: 24: 386–387.
Journal of Exposure Science and Environmental Epidemiology (2008) 18(4)