Analytical Biochemistry 384 (2009) 2733

Contents lists available at ScienceDirect

Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio

Measurement of phenol and p-cresol in urine and feces using vacuum

microdistillation and high-performance liquid chromatography
Roger A. King, Bruce L. May, Debbie A. Davies, Anthony R. Bird *
Food Futures National Research Flagship, CSIRO Human Nutrition, Adelaide, SA 5000, Australia

a r t i c l e i n f o a b s t r a c t

Article history: In this article, we describe a simple, sensitive, accurate, and repeatable method for the measurement of
Received 28 April 2008 phenol and p-cresol (4-methylphenol) in human urine and feces. We examined a number of parameters
Available online 26 September 2008 to identify an optimal extraction protocol. Purication of sample extracts was achieved by low-temper-
ature vacuum microdistillation. Separation was achieved in approximately 15 min by high-performance
Keywords: liquid chromatography (HPLC) with quantication by uorescence at 284/310 nm. Limits of detection for
Cresols phenol were 2 ng/ml for urine and 20 ng/g for feces, and those for p-cresol were 10 ng/ml for urine and
Feces
100 ng/g for feces. For comparison, approximate mean values for urine are 3 lg/ml for phenol and 30 lg/
HPLC
Human
ml for p-cresol, and those for feces are 1 lg/g for phenol and 50 lg/g for p-cresol. An experienced analyst
Phenol can process 60 samples each day using this method.
Urine Crown Copyright 2008 Published by Elsevier Inc. All rights reserved.
Vacuum microdistillation

Measurement of phenols and cresols in feces and urine is of
interest for two main reasons. First, these substances serve as
markers for monitoring environmental exposure to aromatic
hydrocarbons such as benzene and toluene [1,2]. Second, phenol
and p-cresol (4-methylphenol) are putative biomarkers of large bo-
wel health both in humans [35] and in animal models [6]. Diets
that promote the generation of these potentially toxic metabolites
are associated with enhanced rates of colonic DNA damage and
greater risk of large bowel cancer [6,7]. Although environmental
exposure can lead to the generation of a large number of phenols
and cresols [1,8], only phenol and p-cresol, which arise from bacte-
rial action on dietary aromatic amino acids [9], are present in nor-
mal samples.
A number of high-performance liquid chromatography (HPLC)1
and gas chromatography (GC) methods for the measurement of phe-
nols and cresols in urine have been published [1,814]. In contrast to
urine, we are aware of only two methods for feces [9,11]. In urine,
phenols and cresols are present largely as glucuronide and sulfate
conjugates [15,16], and before chromatography these must be
hydrolyzed using acid [2,913] or enzymatic [1,8,13,15] methods.
Hydrolysis is not required for fecal samples [9] because, as is the case
for bile acids [17], endogenous bacterial hydrolases generate the free
forms of phenols and cresols. Initial extracts have most commonly
been obtained by simple extraction from acidied samples using a
* Corresponding author. Fax: +61 88303 8899.
E-mail address: tony.bird@csiro.au (A.R. Bird).
1
Abbreviations used: HPLC, high-performance liquid chromatography; GC, gas
chromatography; UV, ultraviolet.
variety of solvents, including dichloromethane [1], isopropyl ether
[2,14,15], t-butyl ether [8], hexane [10], and diethyl ether [11]. Other
methods have used distillation [9,13] and solid phase extraction
[12]. Most methods [1,1115,18,19], but not all methods [2,9], use
an internal standard, and these have included 4-ethylphenol
[14,19], 4-chlorophenol [1,11], 2,6-dimethylphenol [13,15], and o-
cresol [18]. For HPLC methods, analytes are detected and quantied
with ultraviolet (UV) [1,2,11], uorescence [9,14,15], electrochemi-
cal [13], or chemiluminescence [12] methods.
Measurement of phenol and p-cresol in urine and feces presents
a challenge because phenol is usually present at concentrations at
least one to two orders of magnitude less than p-cresol and be-
tween-individual differences in concentrations can also be of a
similar magnitude. In the current study, and as expanded on below,
we examined a number of elements of the assay of phenols and
cresols in urine and feces with a view to optimizing and validating
their measurement.
With regard to choice of internal standards, we have exam-
ined a number of substances to identify the best candidate. For
sample cleanup, distillation is inherently superior to solvent
extraction because it removes all nonvolatile substances and
provides a purer extract [9,13]. This is particularly important
for fecal samples. However, previously reported distillation
methods for phenols and cresols have involved collection of
up to 50 ml of distillate [9,13,20] and process only one sample
at a time, thereby limiting throughput. For the measurement of
short-chain fatty acids in feces, our group has used a simple
multisample low-temperature vacuum microdistillation appara-
tus [21] based on a single-sample apparatus described by others

0003-2697/$ - see front matter Crown Copyright 2008 Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.ab.2008.09.034
28 Phenol and p-cresol in urine and feces / R.A. King et al. / Anal. Biochem. 384 (2009) 2733
[22]. In the current article, we examined its application for the
measurement of phenols and cresols.
With regard to detection methods, especially in feces, the low-
est concentrations of phenol are close to or below the limits of
detection of UV methods [11]; therefore, uorescence detection
methods are preferable because they are much more sensitive.
They have the additional advantage that the detector sensitivity
can be changed by at least an order of magnitude during an HPLC
run to accommodate on-scale the (low) phenol concentrations as
well as the (high) p-cresol concentrations.
In this article, we describe a sensitive repeatable method for the
measurement of phenol and p-cresol in urine and feces using o-
cresol as internal standard, a multisample vacuum microdistilla-
tion apparatus for sample cleanup and HPLC with uorescence
detection for quantication. The method is capable of a throughput
of approximately 60 samples per day, which is comparable to sol-
vent extraction methods that produce less clean extracts.
Materials and methods
Reagents
Phenol, o-cresol (3-methylphenol), p-cresol (4-methylphenol),
indole, 2,6-dimethylphenol, and skatole (3-methylindole) (all
P98% pure) were obtained from SigmaAldrich (Castle Hill, New
South Wales, Australia), and 4-ethylphenol (>98% pure) was ob-
tained from Merck (Kilsyth, Victoria, Australia). Acetonitrile was
HPLC grade (Merck). All other reagents were analytical grade. Mil-
liQ water was used throughout.
Collection of urine and feces
Collection of feces and urine samples was approved by the hu-
man research ethics committee of CSIRO Human Nutrition (Ade-
laide, South Australia, Australia). Urine from six volunteers was
collected into a suitable container without preservative and stored
in aliquots at 20 C in capped plastic tubes. Feces from six volun-
teers were voided into a large plastic bag placed over the toilet
seat. The sample from each volunteer was made homogeneous
by hand kneading in the plastic bag, and subsamples of approxi-
mately 70 g were transferred to plastic vials and stored at
20 C. Pools of samples were subsequently made as appropriate.
Preparation of standards and other analytes
Stock 1-mg/ml solutions of all analytes were prepared in MilliQ
water and stored for up to 1 month at 4 C. Working solutions were
prepared by making appropriate dilutions in MilliQ water or, for
some tests, NaPO4 (0.1 M, pH 5.5) or HCl (0.1, 1, or 4 M).
Choice of internal standard and effect of extractant volume on
recoveries for feces
In other studies we have conducted, the concentration of phe-
nol in approximately 60% of fecal samples was less than 1 lg/g.
This is close to or below the limit of detection of UV methods
[11]. In an attempt to increase the concentrations of analytes
in extracts from feces in the current study, we examined the ef-
fect of reducing the volume of extractant on recoveries of phenol
and p-cresol as well as the internal standard candidates o-cresol
and 4-ethylphenol. Duplicate 0.5-g aliquots of a fecal sample
were extracted with 1.5, 4.5, and 10 ml of water or the same
volumes of a solution containing 10, 20, 20, and 20 lg/ml phe-
nol, p-cresol, o-cresol, and 4-ethylphenol, respectively. Concen-
trations were calculated by comparison to external standards,
and recoveries were calculated by subtracting the nonspiked va-
lue from the spiked value.
Effect of extractant composition and hydrolysis on extraction efciency
from fecal samples
The two published methods for the measurement of phenol and
p-cresol in feces used 0.1 M phosphate buffer (pH 5.5) for extrac-
tion [9,11]. One appeared not to hydrolyze the extract [9], whereas
the other hydrolyzed with HCl for 60 min at 100 C. Therefore, we
examined the effect of extractant composition and acid hydrolysis
on the efciency of extraction from feces as follows. Duplicate 0.5-
g fecal sample aliquots were extracted by thorough mixing with
10 ml of water, 10 ml of 0.1 M NaPO4 buffer (pH 5.5), 10 ml of
0.1 M HCl, 10 ml of 1 M HCl, or 10 ml of 4 M HCl containing
10 lg/ml o-cresol internal standard. Extracts were centrifuged,
and 150 ll was distilled and applied to the HPLC. We examined
the effect of hydrolysis as follows. After removal of the aliquot
for distillation as described above, the samples that had been ex-
tracted with 1 M HCl at room temperature were heated for
60 min at 100 C. Extracts from these samples were also centri-
fuged and distilled as described above, and the values were com-
pared with corresponding samples that had not been hydrolyzed.
Extraction of fecal samples: Final method
Based on the results of the studies of the effects of extractant
composition and volume on recoveries, the following method
was adopted. Duplicate 0.45- to 0.55-g fecal samples were weighed
into 15-ml plastic centrifuge tubes, and 10 ml of internal standard
(10 lg/ml o-cresol in water) was added. Samples were mixed thor-
oughly by vortex and hand and then were centrifuged for 15 min at
2000g and 10 C. An aliquot of 150 ll was taken from each tube
into a 5-ml Quickt ask for distillation.
Extraction of urine samples: Final method
Because measurement of phenol and p-cresol in urine has been
well studied, we did not examine all elements of this assay. How-
ever, we did conrm that 60 min at 100 C led to complete hydro-
lysis of glucuronide and sulfate conjugates (data not shown). The
nal method we employed was as follows. An aliquot of frozen ur-
ine was thawed, mixed, and then centrifuged for 15 min at 1000g
and 10 C, and duplicate 1-ml subaliquots of supernatant were
transferred to 15  60-mm at-bottomed glass vials with Teon-
lined caps and containing 100 lg of internal standard (100 ll of
o-cresol, 1 mg/ml in water). MilliQ water (400 ll) and 500 ll of
4 M HCl were added, and the vials were capped, mixed, and heated
for 60 min at 100 C in a dry block heater. Tubes were cooled to
room temperature, and 150 ll was taken from each vial into a 5-
ml Quickt ask for distillation.
Recovery of analytes from urine and feces
Recoveries of spikes of phenol and p-cresol added to a urine
sample with low endogenous levels of these metabolites were
determined at two concentrations of analytes as follows. To obtain
nonspiked concentrations, duplicate 1-ml aliquots of urine were
extracted using the method for urine described above except that
100 ll of water replaced 100 ll of internal standard solution.
Duplicate urine samples for each spike concentration also were ex-
tracted using the method for urine described above except that
100 ll of appropriate spike solution mix containing phenol, p-cre-
sol, and o-cresol replaced 100 ll of internal standard solution. Con-
centrations of phenol, p-cresol, and o-cresol in the 100 ll added
were 20, 100, and 100 lg/ml for the low-level spike and 100,
 Phenol and p-cresol in urine and feces / R.A. King et al. / Anal. Biochem. 384 (2009) 2733
1000, and 1000 lg/ml for the high-level spike, respectively. Ex-
pressed per milliliter urine, these concentrations correspond to
2 lg/ml phenol, 10 lg/ml p-cresol, and 10 lg/ml o-cresol for the
low-level spike and 10 lg/ml phenol, 100 lg/ml p-cresol, and
100 lg/ml o-cresol for the high-level spike. An aliquot of 150 ll
was distilled, the distillate was diluted with 2.0 ml of water, and
30 ll was applied to the HPLC. Recoveries were calculated by sub-
tracting the nonspiked value from the spiked value and expressing
this value as a percentage of the assayed concentration of the
respective undistilled spike solution.
For feces, recoveries of spikes of phenol and p-cresol from a
sample with low endogenous levels were determined at three con-
centrations of analytes as follows. Nonspiked concentrations were
determined by extracting duplicate 0.5-g feces samples with 10 ml
of water. Duplicate 0.5-g aliquots for each spike concentration also
were extracted with 10 ml of solution containing phenol, p-cresol,
and o-cresol. The respective concentrations for phenol, p-cresol,
and o-cresol were 0.5, 1, and 1 lg/ml; 2.5, 5, and 5 lg/ml; 10, 20,
and 20 lg/ml, and 100, 100, and 100 lg/ml. An aliquot of 150 ll
of each extract was distilled using the method described above,
and 30 ll of distillate was applied to the HPLC. Recoveries were
calculated by subtracting the nonspiked value from the spiked va-
lue as described for urine.
Precision
Aliquots of urine (1 ml) containing low- or high-endogenous
levels of analytes were transferred to small screw-capped plastic
tubes and stored at 20 C. Feces containing low or high levels of
analytes were mixed thoroughly, and 0.5-g samples were trans-
ferred to 15-ml screw-capped plastic tubes and stored at 20 C.
Duplicate aliquots of urine and feces were then processed on 5
separate days using the nal method as described. Intra- and
interassay coefcients of variation were calculated according to
Rodbard [23].
Sample extract stability
To determine stability of distilled extracts, duplicate aliquots of
a typical sample of urine (1 ml) were hydrolyzed using the method
as described. Similarly, duplicate 0.5-g aliquots of feces were ex-
tracted using the described method for feces. Multiple aliquots of
150 ll of each initial extract of urine and feces were immediately
distilled, the distillates were pooled, and an aliquot was applied
to the HPLC in duplicate to obtain baseline values. The remainder
of the pool was aliquoted and assayed in duplicate after 1 and 4
weeks storage at 20 C.
Separation of possible interfering substances
For substances to potentially interfere with the current assay,
they must codistill and coelute with phenol, p-cresol, or o-cresol
and uoresce at 284/310 nm. Solutions of indole and 3-methylin-
dole (skatole), 4-ethylphenol, and 2,6-dimethylphenol, which are
distillable substances that may be present in feces, were chromato-
graphed, and their retention times were compared with those of
phenol, p-cresol, and o-cresol.
Distillation procedure
The apparatus used for low-temperature vacuum microdistilla-
tion of samples is shown in Fig. 1. It was constructed to our spec-
ications by a commercial glass blower (Labglass, Stafford,
Queensland, Australia). The unit was connected to a vacuum pump
capable of drawing a vacuum of less than 1 Torr in the apparatus.
29
Briey, the distillation procedure involves shell-freezing an ali-
quot of the sample extract in a small glass Quickt ask (FR5/1S)
attached to the distillation apparatus, applying vacuum, rapidly
applying heat, and collecting the distillate into a cooled small glass
Quickt collection ask. Using our apparatus, up to six samples can
be distilled simultaneously, but it should be possible to construct
an apparatus capable of distilling eight samples. The method is de-
scribed in detail as follows (refer to Fig. 1 to identify apparatus part
names):
1. Lightly grease both Quickt ttings of each Y-piece. Petro-
leum jelly is preferable to silicone grease because it is easily
removed from glassware with hot detergent.
2. Attach empty Quickt asks (the collection asks) to the
Y-pieces on one side of the apparatus.
3. Close all taps on the Y-pieces.
4. Pipette the desired volume of sample (usually 150 ll) into
small round-bottom Quickt asks (the sample asks). If
required, the volume distilled can be increased up to 500 ll.
5. For feces, one-fth the sample volume of 1 M H3PO4 (i.e.,
30 ll for 150-ll sample) is added to facilitate distillation
by protonating the analytes. Because urine samples are
already acidied with HCl to hydrolyze conjugates, these
do not require the addition of H3PO4.
6. Turn the tap that is nearest to the vacuum pump on the
apparatus backbone (the inlet tap) to the open position,
and turn the tap that is farthest from the pump (the vent
tap) to the closed position.
7. Turn on the vacuum pump. At this stage, only the apparatus
backbone is evacuated.
8. Shell-freeze the rst sample by swirling the ask in ethanol
at 80 C. We use ethanol kept at 80 C in a freezer, but an
ethanoldry ice mix could also be used.
9. Apply the rst sample ask to the rst Y-piece on the oppo-
site side of the backbone to its corresponding collection
ask. Immediately reimmerse the ask in 80 C ethanol
so that the sample remains frozen, and then immediately
open the Y-piece tap to apply vacuum to both the sample
and collection asks for that sample. We have found that
small (150200 ml) Styrofoam cups are ideal containers for
80 C ethanol.
10. Repeat steps 8 and 9 for the remaining samples.
11. Turn the taps on the Y-pieces to the off position. This isolates
each sample askcollection ask pair from the other asks,
isolates all asks and Y-pieces from the main backbone, and
maintains a vacuum in the asks. Therefore, cross-contami-
nation between samples is not possible. We also distilled
water from clean asks immediately after processing sam-
ples with high levels of phenol and p-cresol and were unable
to detect phenol or p-cresol, demonstrating that there was
no carryover of sample.
12. Lift the entire apparatus and position it so that the collection
asks are now immersed in 80 C ethanol. Ensure that the
coolant does not cover more than about the bottom half of
each ask. If the ask is immersed too deeply, part of the
sample may condense on the end of the (cold) Y-piece.
13. Immediately place a small insulated container under each
sample ask (150- to 200-ml Styrofoam drink cups as used
for 80 C ethanol are again ideal). Add water at 50 to
60 C until it covers approximately the bottom one-third of
each ask, leave 20 s, and then cover the ask completely
with water. This sequential process prevents bumping of
the sample that may otherwise occur if the water were
added in a single operation.
14. Isolate the system from the vacuum pump by turning the
inlet tap to the closed position and then turn off the pump.
30 Phenol and p-cresol in urine and feces / R.A. King et al. / Anal. Biochem. 384 (2009) 2733

Inlet tap Apparatus backbone

Vent tap

To vacuum pump To atmosphere

Y-piece tap

Sample flask
Y-piece Sample flask
Collection flask
Collection flask

Apparatus backbone

Y-piece tap

Rubber connector
Y-piece

Sample flask Collection flask

Fig. 1. Vacuum distillation apparatus. Top left: Side elevation photograph. The manifold consists of four main parts: the glass backbone, glass Y-pieces, Quickt sample asks,
and Quickt collection asks. Access of vacuum to the apparatus is controlled by the inlet and vent taps. Access of vacuum to the sample and collection asks is controlled by
the six Y-piece taps. Top right: End elevation photograph taken at a slight angle above bench level. Bottom: End elevation drawing at the level of the dashed line in the top left
picture.

15. Leave the samples to distill for approximately 5 min. Care-
fully observe each sample to ensure that the process is com-
plete (i.e., no sample remains). During this stage, each
sample distills from its sample ask and is condensed and
frozen in its corresponding collection ask.
16. Lift the entire apparatus and position it so that the sample
and collection asks are removed from the heating and cool-
ing baths.
17. Release the vacuum from the apparatus by opening the vent
tap and the Y-piece taps.
18. Allow the frozen distilled samples to thaw.
19. If necessary, dilute with water and take an aliquot for HPLC.
An experienced operator can extract and distill approxi-
mately 60 samples in 1 day. We uniformly recover 150 to
160 ll of distillate from each sample. Distillates from urine
are normally diluted with 2 ml of water. Note, however, that
provided that sufcient volume is collected for HPLC, the
volume of the distillate is not important because an internal
standard is included.
HPLC
HPLC was conducted using a Shimadzu LC-10AD instrument
with the sample compartment set to 7 C, an RF-10AXL uores-
cence detector set at excitation 284 nm and emission 310 nm
[24], an SIL-10A autoinjector, an LC-10AT pump, and a DGU-12A
degasser. The injection volume was 30 ll, and the ow rate was
1 ml/min. The instrument was controlled by Shimadzu LabSolu-
tions software (version 1.21). Separation was achieved with a
250  4.6-mm column with 5 lm C18 packing (cat. no.
R0086200C5, Varian, Melbourne, Victoria, Australia). The mobile
phase consisted of water/acetonitrile (7:3) adjusted to pH 3.2 with
glacial acetic acid. Because the concentrations of p-cresol are al-
most always approximately an order of magnitude or more higher
than those of phenol, the software was programmed to decrease
the sensitivity of the uorescence detector 30-fold by changing
from medium to low 10 min after sample injection to ensure that
all analyte peaks were on-scale.
Results and discussion
In this article, we have described a simple, highly sensitive,
accurate, and repeatable HPLC method for the measurement of
phenol and p-cresol in human urine and feces using multisample
vacuum microdistillation for sample cleanup. Because of its high
sensitivity, the method should also be applicable to samples from
rats where concentrations of these metabolites tend to be lower
than those in humans.
Methods for phenol and cresol measurement that use distilla-
tion for sample cleanup have been published [9,13], but they use
large sample volumes and distill only one sample at a time, thereby
limiting throughput. Our method has a throughput of approxi-
mately 60 samples per day, similar to methods that use solvent
extraction cleanup. However, it has the major advantage that the
low-temperature vacuum microdistillation apparatus produces
very clean samples for HPLC, ensuring simple chromatograms,
clear separation of analytes, and maximum column life.
 Phenol and p-cresol in urine and feces / R.A. King et al. / Anal. Biochem. 384 (2009) 2733 31

The use of o-cresol as internal standard overcomes possible er-

A 300
rors that could arise with 4-ethylphenol [14,19] because the latter
phenol
may be present in feces as a product of bacterial metabolism of
250
genistein [25], an isoavone found in soyfoods. Another advantage
of using o-cresol is that it elutes soon after p-cresol, thereby min-

Detector response (mV)

imizing run times. 200
The levels of phenol in human feces are close to the limits of p-cresol
detection using UV even when samples are extracted with solvent 150
and concentrated by evaporation [11]. Our method using uores-
cent detection is many orders of magnitude more sensitive than
100
UV and is easily capable of measuring the lowest concentrations o-cresol
of phenol in fecal samples.
Typical chromatograms for a standard and urine and fecal sam- 50
ples are shown in Fig. 2. They illustrate the presence of few other
peaks, clear separation of analytes, and high sensitivity of the as- 0
say. Standard curves were linear in the ranges used: 0 to 1 lg/ml 0 5 10 15 20
for phenol and 0 to 10 lg/ml for p-cresol and o-cresol. B 120

o-cresol
Choice of internal standard and effect of extractant volume on 100
recoveries from feces

Detector response (mV)

80
Levels of phenol in feces are often very low. In an attempt to
obtain extracts sufciently concentrated to allow UV detection,
we initially tested the effect of reducing extractant volume. 60
The internal standard method requires that the recovery of the
internal standard is similar to the recoveries of the analytes of phenol
40
interest; otherwise, measurement errors will occur. The effects p-cresol
of extractant volume on recoveries from feces are shown in Ta-
20
ble 1. There was a clear increase in recovery with increasing
extractant volume, especially for 4-ethylphenol. Importantly,
the recovery of 4-ethylphenol was less than the recoveries of 0
phenol and p-cresol for all volumes, but especially for 1.5 ml, 0 5 10 15 20
making it unsuitable as an internal standard. The recovery of C 250
o-cresol using 1.5 ml of extractant was also less than the recov- p-cresol
eries of phenol and p-cresol, but for 4.5 and 10 ml the recovery
200 o-cresol
was similar to the recoveries of these analytes. Attempts to im-
Detector response (mV)

prove recoveries for 1.5 ml of extractant by a number of modi-

cations were unsuccessful (data not shown). We tested 10 fecal phenol
150
samples and, in agreement with published data [9], found no
interfering peaks at the position of o-cresol. It has the advantage
that it elutes soon after p-cresol, thereby allowing run times to 100
be as short as possible. Therefore, we chose o-cresol as the inter-
nal standard and 10 ml as the extractant volume for the remain-
der of our studies. 50

Effect of extractant composition and hydrolysis on extraction efciency

from fecal samples
There were no obvious differences in the measured levels of
phenol and p-cresol between any of the extraction conditions
tested, so a simple extraction with water was adopted (data not
shown). There was no effect of treatment with 1 M HCl for
60 min at 100 C (data not shown); this demonstrates that, as is
the case with bile acids [17], the glucuronide and sulfate forms
of phenols and cresols that enter the large intestine in bile are
hydrolyzed to their free forms by the action of the resident bacte-
rial populations. Therefore, no acid hydrolysis of fecal samples is
required.
Recovery of analytes from urine and feces, precision of the method,
and sample extract stability
The recoveries of phenol, p-cresol, and o-cresol added at a
number of concentrations to urine and feces using our nal as-
say methods are summarized in Tables 2 and 3. For both urine
0
0 5 10 15 20
Elution time (min)
Fig. 2. HPLC chromatogram of standards (0.25 lg/ml phenol, 2.5 lg/ml p-cresol)
(A) and typical samples of urine (B) and feces (C). Concentrations were approxi-
mately 1 lg of phenol/ml and 10 lg of p-cresol/ml for urine, and they were
approximately 3 lg of phenol/g and 90 lg of p-cresol/g for feces. The arrow
indicates a 30-fold decrease in detector sensitivity at 10 min.
and feces, recoveries of analytes at all concentrations were close
to 100%.
Intra- and interassay coefcients of variation (for means of
duplicates) for measurement of phenol and p-cresol of urine and
feces that contained concentrations in the lower and higher phys-
iological ranges are summarized in Table 4. These were less than
5% except for the low-level feces sample, which was between 5%
and 10%.
Distillates of extracts of urine and feces showed no obvious loss
on storage for up to 4 weeks at 20 C (data not shown).
32 Phenol and p-cresol in urine and feces / R.A. King et al. / Anal. Biochem. 384 (2009) 2733

Table 1 sol is required, the limit of detection for p-cresol could be reduced
Effects of extractant volume on recoveries of phenol, p-cresol, o-cresol, and by at least 30-fold by using the higher sensitivity setting of the
4-ethylphenol from feces.
detector for the entire run.
Extractant volume (ml) Phenol p-Cresol o-Cresol 4-Ethylphenol
1.5 85, 86 71, 68 71, 73 50, 51 Separation of possible interfering substances
4.5 95, 93 92, 89 92, 87 75, 73
10.0 95, 99 94, 99 91, 95 82, 87 Urine and feces may contain substances that can interfere
Note. The values are percentage recoveries. Shown are each of duplicate net values with the measurement of phenol and p-cresol. However, for
after subtraction of amounts measured in nonspiked fecal samples for each a substance to interfere in our assay, it must distill under
extractant volume. The spike contained 10 lg/ml phenol, 20 lg/ml p-cresol, 20 lg/ the assay conditions, coelute with an analyte of interest, and
ml o-cresol, and 20 lg/ml 4-ethylphenol, as described in Materials and methods.
uoresce at 284/310 nm. Indole, 4-ethylphenol, 2,6-dimethyl-
phenol, and 3-methyindole (skatole) did distill and were de-
tected at 284/310 nm, but they eluted well after o-cresol at
Limits of detection approximately 25, 23, 24, and 50 min, respectively (data not
shown). Therefore, none of these substances would interfere
With our HPLC method, the sensitivity of the detector is de- with the assay. Our method did not separate m-cresol from
creased by a factor of 30 for p-cresol compared with phenol. Using p-cresol. However, because m-cresol is not present in urine
a peak height of ve times baseline noise as the criterion, at these and feces of persons not exposed to aromatic hydrocarbons
sensitivities, the limits of detection for phenol were determined to [1,8,9], we did not pursue modications reported by others
be 2 ng/ml for urine and 20 ng/g for feces, and for p-cresol they to achieve separation [1,2]. As discussed above, in agreement
were determined to be 10 ng/ml for urine and 100 ng/g for feces. with published data [9], we found no interfering peaks at
These limits are more than adequate for measurements in human the position of o-cresol that would invalidate its use as an
urine and feces. However, if measurement of lower levels of p-cre- internal standard.

Table 2
Recoveries of phenol, p-cresol, and o-cresol from feces.

Concentration of added Phenol p-Cresol o-Cresol

analyte in feces (lg/g)
Measured total concentration % Recovery Measured total concentration % Recovery Measured total concentration % Recovery
of analyte (lg/g) of analyte (lg/g) of analyte (lg/g)
0 1.00 0.10 28.3 1.2 0.02 0.00
10 10.6 0.4 96 4
20 47.6 0.7 97 2 19.6 0.7 98 4
50 54.1 1.4 107 1
100 139 2 110 0 105 2 105 2
200 218 2 109 1
400 463 3 109 0 403 5 108 1

Note. Values are means standard deviations of duplicate extractions and were calculated using external standards.

Table 3
Recoveries of phenol, p-cresol, and o-cresol from urine.

Concentration of added analyte Phenol p-Cresol o-Cresol

in urine (lg/ml)
Measured total concentration % Recovery Measured total concentration % Recovery Measured total concentration % Recovery
of analyte (lg/ml) of analyte (lg/ml) of analyte (lg/ml)
0 0.85 0.10 8.04 0.26 0.00 0.00
2 2.78 0.04 102 1 nd nd nd nd
10 11.0 0.0 104 0 17.5 0.1 99 0 9.43 0.41 104 2
100 nd nd 109 0.4 103 0 99.6 0.5 103 0

Note. Values are means standard deviations of duplicate extractions and were calculated using external standards. nd, not determined.

Table 4
Coefcients of variation for measurement of phenol and p-cresol in urine and feces using the nal assay methods.

Phenol p-Cresol
Urine Feces Urine Feces
Low (lg/ml) High (lg/ml) Low (lg/g) High (lg/g) Low (lg/ml) High (lg/ml) Low (lg/g) High (lg/g)
Mean 0.91 13 0.51 2.9 10 85 26 88
Intraassay 1.5 2.1 6.3 2.6 1.2 2.5 1.1 1.8
Interassay 4.6 2.9 9.4 3.0 2.8 3.3 4.7 2.1

Note. Phenol and p-cresol were measured on ve occasions in samples of urine and feces containing low or high endogenous levels.
 Phenol and p-cresol in urine and feces / R.A. King et al. / Anal. Biochem. 384 (2009) 2733
Acknowledgment
We thank Corinna Bennett for excellent technical assistance.
References
[1] A. Brega, P. Prandini, C. Amaglio, E. Pafumi, Determination of phenol, m-, o-,
and p-cresol, p-aminophenol, and p-nitrophenol in urine by high-performance
liquid chromatography, J. Chromatogr. 535 (1990) 311316.
[2] M. Yoshikawa, Y. Taguchi, K. Arashidani, Y. Kodama, Determination of cresols
in urine by high-performance liquid chromatography, J. Chromatogr. 362
(1986) 425429.
[3] A. Birkett, J. Muir, J. Phillips, G. Jones, K. ODea, Resistant starch lowers fecal
concentrations of ammonia and phenols in humans, Am. J. Clin. Nutr. 63 (1996)
766772.
[4] V. De Preter, T. Coopmans, P. Rutgeerts, K. Verbeke, Inuence of long-term
administration of lactulose and Saccharomyces boulardii on the colonic
generation of phenolic compounds in healthy human subjects, J. Am. Coll.
Nutr. 25 (2006) 541549.
[5] J.G. Muir, E.G. Yeow, J. Keogh, C. Pizzey, A.R. Bird, K. Sharpe, K. ODea, F.A.
Macrae, Combining wheat bran with resistant starch has more benecial
effects on fecal indexes than does wheat bran alone, Am. J. Clin. Nutr. 79 (2004)
10201028.
[6] S. Toden, A.R. Bird, D.L. Topping, M.A. Conlon, High red meat diets induce greater
numbers of colonic DNA double-strand breaks than white meat in rats:
attenuation by high-amylose maize starch, Carcinogenesis 28 (2007) 23552362.
[7] R.K. Le Leu, G.P. Young, Fermentation of starch and protein in the colon:
Implications for genomic instability, Cancer Biol. Ther. 6 (2007) 259260.
[8] R.L. Dills, G.M. Bellamy, D.A. Kalman, Quantitation of o-, m-, and p-cresol and
deuterated analogs in human urine by gas chromatography with electron
capture detection, J. Chromatogr. B 703 (1997) 105113.
[9] K.E. Murray, R.F. Adams, Determination of simple phenols in faeces and urine by
high-performance liquid chromatography, J. Chromatogr. 431 (1988) 143149.
[10] M. Balikova, J. Kohlicek, Gas chromatography of simple phenols in biological
uids, J. Chromatogr. 497 (1989) 159167.
[11] A.M. Birkett, G.P. Jones, J.G. Muir, Simple high-performance liquid
chromatographic analysis of phenol and p-cresol in urine and feces, J.
Chromatogr. B 674 (1995) 187191.
[12] E. Orejuela, M. Silva, Sensitive determination of low molecular mass phenols
by liquid chromatography with chemiluminescence detection for the
determination of phenol and 4-methylphenol in urine, Analyst 127 (2002)
14331439.
33
[13] W.E. Schaltenbrand, S.P. Coburn, Determination of phenol and p-cresol in
urine, Clin. Chem. 31 (1985) 20422043.
[14] Y. Tsuruta, S. Kitai, S. Watanabe, H. Inoue, 2-Methoxy-4-(2-
phthalimidinyl)phenylsulfonyl chloride as a uorescent labeling reagent for
determination of phenols in high-performance liquid chromatography and
application for determination of urinary phenol and p-cresol, Anal. Biochem.
280 (2000) 3641.
[15] G. Lesaffer, R. De Smet, F.M. Belpaire, B. Van Vlem, M. Van Hulle, R. Cornelis, N.
Lameire, R. Vanholder, Urinary excretion of the uraemic toxin p-cresol in the
rat: contribution of glucuronidation to its metabolization, Nephrol. Dial.
Transplant. 18 (2003) 12991306.
[16] N. Ogata, N. Matsushima, T. Shibata, Pharmacokinetics of wood creosote:
glucuronic acid and sulfate conjugation of phenolic compounds, Pharmacology
51 (1995) 195204.
[17] A.K. Batta, G. Salen, K.R. Rapole, M. Batta, P. Batta, D. Alberts, D. Earnest, Highly
simplied method for gasliquid chromatographic quantitation of bile acids
and sterols in human stool, J. Lipid Res. 40 (1999) 11481154.
[18] A. Gostner, M. Blaut, V. Schaffer, G. Kozianowski, S. Theis, M. Klingeberg, Y.
Dombrowski, D. Martin, S. Ehrhardt, D. Taras, A. Schwiertz, B. Kleessen, H.
Luhrs, J. Schauber, D. Dorbath, T. Menzel, W. Scheppach, Effect of isomalt
consumption on faecal microora and colonic metabolism in healthy
volunteers, Br. J. Nutr. 95 (2006) 4050.
[19] T. Niwa, Phenol and p-cresol accumulated in uremic serum measured by HPLC
with uorescence detection, Clin. Chem. 39 (1993) 108111.
[20] L. Weber, Gas chromatographic determination of urinary phenol conjugates
after acid hydrolysis/extractive acetylation, J. Chromatogr. 574 (1992) 349
351.
[21] G.S. Patten, A.R. Bird, D.L. Topping, M.Y. Abeywardena, Effects of convenience
rice congee supplemented diets on guinea pig whole animal and gut growth,
caecal digesta SCFA, and in vitro ileal contractility, Asia Pac. J. Clin. Nutr. 13
(2004) 92100.
[22] J.E. Tyler, G.H. Dibdin, Gas chromatography of volatile fatty acids: method
involving separation from biological material by vacuum distillation, J.
Chromatogr. 105 (1975) 7177.
[23] D. Rodbard, Statistical quality control and routine data processing for
radioimmunoassays and immunoradiometric assays, Clin. Chem. 20 (1974)
12551270.
[24] R. De Smet, F. David, P. Sandra, J. Van Kaer, G. Lesaffer, A. Dhondt, N.
Lameire, R. Vanholder, A sensitive HPLC method for the quantication of
free and total p-cresol in patients with chronic renal failure, Clin. Chim. Acta
278 (1998) 121.
[25] R.A. King, Digestion, absorption, and metabolism of isoavones, in: G.S. Gilani,
J.J.B. Anderson (Eds.), Phytoestrogens and Health, AOCS Press, Champaign, IL,
2002, pp. 209234.