Triclosan in plasma and milk from Swedish nursing mothers
and their exposure via personal care products
Mats Allmyra,b,⁎, Margaretha Adolfsson-Ericib,
Michael S. McLachlanb, Gunilla Sandborgh-Englunda
aInstitute of Odontology, Karolinska Institutet, PO Box 4064, SE-141 04 Huddinge, Sweden
bDepartment of Applied Environmental Science, Stockholm University, SE-106 91 Stockholm, Sweden
Received 11 April 2006; received in revised form 2 August 2006; accepted 8 August 2006
Available online 26 September 2006
The bactericide triclosan is commonly used in e.g. plastics, textiles and health care products. In vitro studies on rat and human
biological systems indicate that triclosan might exert adverse effects in humans. Triclosan has previously been found in human
plasma and milk, but neither the primary source of human exposure nor the efficiency of triclosan transfer to human milk is known.
In this study, plasma and milk were sampled from 36 mothers and analyzed for triclosan. Scrutinization of the women's personal
care products revealed that nine of the mothers used toothpaste, deodorant or soap containing triclosan. Triclosan and/or its
metabolites were omnipresent in the analyzed plasma and milk. The concentrations were higher in both plasma and milk from the
mothers who used personal care products containing triclosan than in the mothers who did not. This demonstrated that personal
care products containing triclosan were the dominant, but not the only, source of systemic exposure to triclosan. The concentrations
were significantly higher in plasma than in milk, indicating that infant exposure to triclosan via breast milk is much less than the
dose in the mother.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Irgasan; Antibacterial; Human; Infant
Triclosan (2,4,4′-trichloro-2′-hydroxydiphenyl ether,
CAS 3380-34-5) (Fig. 1) is a lipophilic and phenolic
compound (log KOW=4.76, pKa=7.9) (Syracuse Re-
search Corporation, 2006; Merck Index 12:3, 2000). Due
to its antibacterial properties, triclosan has found wide-
spread use in a variety of consumer products includ-
ing toothpastes, deodorants, soaps, polymers and fibers
(Bhargava and Leonard, 1996; Adolfsson-Erici et al.,
2002). An assessment of liquid soaps available on the
market in the USA showed that 76% of 395 soaps of
different brands contained triclosan (Perencevich et al.,
singular human source of exposure to triclosan; an
estimated 20–25% of all toothpaste sold in Sweden con-
tains triclosan, accounting for ∼ 2 tons of use per year
(Edwardsson et al., 2005). The European Commission
care products, and a similar amount in skin care products.
Science of the Total Environment 372 (2006) 87–93
⁎Corresponding author. Department of Applied Environmental
Science, Stockholm University, SE-106 91 Stockholm, Sweden.
Tel.: +46 8 674 77 19.
E-mail address: email@example.com (M. Allmyr).
0048-9697/$ - see front matter © 2006 Elsevier B.V. All rights reserved.
Triclosan is not acutely toxic to mammals but in vitro
studies indicate that at low concentrations triclosan may
disturb metabolic systems and hormone homeostasis
(Hanioka et al., 1996; Schuur et al., 1998; Wang et al.,
2004; Jacobs et al., 2005). Considering the potential
adverse effects of triclosan, the magnitude of systemic
exposure via the routine use of triclosan containing
products in the general population should be clarified.
the mouth and intestinal tract when administered via
dental care products (Lin, 2000; Sandborgh-Englund
et al., 2006). In addition, in vitro studies have shown that
triclosan is absorbed through human skin, although it is
subject to substantial metabolism during absorption
(Moss et al., 2000). Sandborgh-Englund et al. (2006)
showed a rapid uptake and elimination of triclosan in
human plasma after a single oral dose. The observed
terminal half-life in plasma was 21 h and triclosan con-
centrations returned to baseline within 8 days after the
experimental dose. Chronic accumulation of triclosan in
humans has not been shown; however, Bagley and Lin
(2000) showed that the continuous dose interval associ-
ated with normal use of toothpaste for 12 weeks resulted
in the blood. In addition, Mustafa et al. (2003)
demonstrated that triclosan was taken up by and dis-
tributed to the cytoplasm and nucleus of human gingival
fibroblast cells in vitro, suggesting that a certain amount
of triclosan may be retained in human cells.
Triclosan has been identified in a pooled plasma sam-
ple from Swedish men (Hovander et al., 2002). Further-
more, Sandborgh-Englund et al. (2006) showed that
triclosan was present in varying concentrations in plasma
from 10 individuals, of which five were exposed and five
small study population, there was no apparent difference
and control group, and other sources of exposure than
personal care products were hypothesized.
Infants represent a sensitive subgroup of the popula-
tion for which the exposure to contaminants is of central
importance in risk assessment. Triclosan has previously
been found in breast milk (Adolfsson-Erici et al., 2002),
which indicates a route of exposure of the infant to
triclosan via the food. Hence, there is a need to study to
the nursing mother.
In the present study, we investigated if the use of
personalcareproductscontainingtriclosan was the domi-
nant source of systemic exposure to triclosan in a popula-
tion of Swedish nursing mothers. We also studied the
transfer of triclosan from plasma to breast milk in the
2. Materials and methods
The study was approved by the Regional Ethical
Review Board, Huddinge University Hospital, Sweden
(dnr: 395/03). Informed consent by the mothers was
compulsory for their participation in the study.
Thirty-six mothers were recruited at a childcare center
in the City of Stockholm between October 2003 and May
2004. Inclusion criteria were that the mother was a first
time mother, healthy and that at least half the infant food
intake was breastmilk. Milk and plasmawere sampledon
two occasions, approximately 6 and 12 weeks after
delivery. In order to avoid any direct contamination from
personal care products, the mothers were given instruc-
tions to wash the breast with only water and dry it with a
triclosan free paper tissue provided by us prior to sam-
pling. The milk sample was expressed by the mother,
preferably before the first breastfeeding of the day, col-
lected in a polypropylene container (Sarstedt™, Nüm-
brecht, Germany), stored in a refrigerator and brought to
the clinic the same day. There their blood was sampled in
10 ml sterile heparin tubes (Vacutainer™, Becton-
Dickinson, Meylan Cedex, France). The whole blood
was centrifuged at 900×g for 5 min and the plasma was
transferred to polypropylene tubes (Sarstedt™, Nüm-
brecht, Germany). The samples were stored at −20 °C
until analysis. Utensils used for sampling and storing the
samples were extracted and analyzed for triclosan content
before use. The mothers' use of a breast milk pump was
documented to control for potential contamination bias.
2.2. Monitoring of personal care products containing
The mothers were instructed to collect their personal
care products (here defined as any chemical product
used for personal care, e.g. toothpaste, mouth rinse,
soap, shampoo, deodorant, cosmetic, moisturizer, etc.)
and bring them to the clinic on the first sampling occa-
sion. The product labels were scrutinized for triclosan
Fig. 1. Molecular structure of triclosan.
88M. Allmyr et al. / Science of the Total Environment 372 (2006) 87–93
content. In Sweden, all ingredients in personal care
products have to be labeled, as required in the Cosmetics
Directive within the European Community. The mothers
who used products containing triclosan were denoted as
“exposed”, while those who did not were denoted as
The mother's were asked to judge their knowledge
about triclosan on a scale ranging from “none” to “very
2.3. Chemical analysis
The analytical method for the analyzed plasma and
milk samples, including the sources and quantities of the
reagents used, is described elsewhere (Allmyr et al.,
2006). Briefly,13C-labeled triclosan was added to 3 g
milk or 5 g plasma as a surrogate standard. The sample
was submitted to acid hydrolysis using sulphuric acid to
release any triclosan bound in conjugates. The samples
were then extracted using hexane/acetone and cleaned
up using sulphuric acid. Triclosan was converted to its
pentafluorobenzoyl ester, and the extract was analyzed
using gas chromatography/mass spectrometry/electron
capture negative ionization (GC/ECNI/MS).
The triclosan concentration in the samples was quan-
tified with an internal standard method. The limit of
in the blanks. The area ratio of triclosan to13C-labeled
triclosan (Atriclosan/A13C triclosan) in the blanks was con-
sistent, and hence 4×(Atriclosan/A13C triclosan) in the blank
was defined as the LOQ. This corresponded to a triclosan
amount of 0.06 ng, or 0.018 ng/g milk and 0.009 ng/g
plasma. Concentrations below the LOQ were also quan-
tified and these data were used in pair-wise statistics, as
this was considered superior to excluding the data. All
samples were corrected for the blanks. Triclosan concen-
trations were determined as the sum of unchanged and
conjugated triclosan (ng/g). Since triclosan has been
shown to be rapidly eliminated from the body (Sand-
borgh-Englund et al., 2006), the time elapsed since a
given exposure was expected to be the major factor af-
fecting the triclosan concentration, not properties of the
body fluids such as lipid content. Therefore, the concen-
trations were calculated on a fresh weight basis.
2.5. Quality control
The method repeatability test was assessed by re-
peatedly analyzing triclosan in three different milk sam-
ples. The replicates were not analyzed together; rather
they were analyzed on different occasions with freezing
and thawing of the sample in between. Hence, the
method repeatability test also included certain aspects of
the sample handling procedures. The sample with higher
concentration was among the highest concentrations in
milk in this study and the lower concentration was near
the LOQ for milk.
statistical software program version 11.0.0 (SPSS Inc.,
Chicago, Illinois). Differences and correlations were
tested with non-parametric tests: the Mann–Whitney U-
test, the Wilcoxon signed rank test and the Spearman
rank correlation. The impact of age, body mass index of
the mother, days postpartum at sampling, the number of
personal care products brought to the reception and the
use of a breast milk pump on triclosan concentrations
was evaluated with multiple linear regression. The level
of significance was 0.05.
3.1. Monitoring of products and grouping the study
Nine mothers used personal care products labeled
as containing triclosan and were denoted as exposed.
Among these, seven used toothpaste, one used soap and
one used deodorant containing triclosan. Twenty-six
mothers were denoted as controls, based on the lack of
labeling of triclosan on their personal care products. In
addition, one mother used toothpaste purchased in
Turkey, with no labeling of ingredients. However, when
sold in Sweden, the same brand and sort contains triclo-
san. Since there was considerable uncertainty concerning
her exposure, she was excluded from the statistical
None of the exposed mothers and 19 of the controls
had any knowledge about triclosan; eight of the controls
had some knowledge.
3.2. Quality control
The coefficient of variation was 6% for the high
concentration (mean=0.84 ng/g, n=7), 1% for the
intermediate concentration (mean=0.31 ng/g, n=3) and
5% for the low concentration (mean=0.020 ng/g, n=3).
The instrumental precision was evaluated by analyzing
two samples three times each in a sequence of 184
89M. Allmyr et al. / Science of the Total Environment 372 (2006) 87–93
injections on the GC/ECNI/MS. In both cases, the in-
strumental coefficient of variation was 1%.
3.3. Triclosan in milk and plasma
In the following, triclosan concentrations refer to the
present in plasma or milk. The results from the analyzed
plasma and milk samples are summarized in Table 1.
Figs. 2 and 3 depict the results from the analyzed plasma
and milk samples from the first sampling occasion. One
individual could not sample milk at the second sampling
occasion. She was not included in the statistical com-
parison of the two sampling occasions.
in all analyzed plasma samples. Triclosan was found in
detectable amounts in all analyzed milk samples. How-
ever, on the first sampling occasion, 12 milk samples
the LOQ, while on the second sampling occasion the
number of milk samples with concentrations below the
LOQ was 17 (n=25).
Triclosan concentrations in plasma and milk were not
influenced by the age or the body mass index of the
mother, days postpartum at sampling or the number of
personal care products brought to the reception. The use
of a breast milk pump to express milk did not influence
the triclosan concentration.
On both sampling occasions, the median triclosan
concentration in plasma and milk was significantly
higher in the exposed than in the control group (Mann–
Whitney U, pb0.001) (Table 1) (Figs. 2 and 3). There
was no significant difference in triclosan concentration
between the first and second sampling occasion in either
plasma or milk from the two groups (Wilcoxon signed
rank test: control plasma, p=0.17; exposed plasma,
p=0.59; control milk, p=0.70; exposed milk, p=0.14).
Triclosan concentrations (ng/g fresh weight) in maternal blood plasma and milk from controls and exposed mothers at the first and second sampling
Occasion 1 Occasion 2
Plasma MilkPlasma Milk
Cont.Exp. Cont.Exp. Cont. Exp. Cont.Exp.
Abbreviations: Cont., controls; Exp., exposed.
Fig. 2. Triclosan concentrations (ng/g fresh weight) in maternal blood
plasma on the first sampling occasion. Open circle=control, filled
circle=exposed, dashed line=LOQ.
Fig. 3. Triclosan concentrations (ng/g) in milk on the first sampling
90 M. Allmyr et al. / Science of the Total Environment 372 (2006) 87–93
The triclosan concentration ratios of paired milk and
plasma samples (M/P) were calculated to obtain a mea-
sure of how much triclosan was transferred from plasma
to milk. The M/P range was 0.01–2 (M/PN1 in just two
cases) and was, in both the exposed and control groups,
negatively correlated with the triclosan concentration in
occasion: exposed, rs=−0.85, p=0.004 and control, rs=
−0.58, p=0.002; second sampling occasion: exposed,
rs=−0.77, p=0.016 and control, rs=−0.69, pb0.001).
4.1. Triclosan in plasma and milk and the source of
The present study shows that the commonly used
bactericide triclosan and/or its metabolites are omni-
present in plasma and milk from nursing mothers. Tri-
closan was found, in a broad range of concentrations, in
all plasma and milk samples from both the control and
the exposed group (Figs. 2 and 3). The presence of
triclosan in the whole study population suggests that
there are sources of human exposure other than personal
care products. Indeed, triclosan is used in textile and
plastics (Adolfsson-Erici et al., 2002) and has been
found in fatty food (Remberger et al., 2002), but the
influence of these sources on triclosan levels in the
human body has not been studied. We note that improp-
er labeling or incomplete collection of personal care
products in the control group may have occurred, con-
sequently placing exposed mothers in the control group.
Hence, no conclusions can be drawn from our data about
the upper range of plasma and milk concentrations re-
sulting from exposure to these other sources.
Sandborgh-Englund et al. (2006) found triclosan in
plasma at 0.1–8.1 ng/ml in 10 subjects, of which five
were exposed and five not exposed to triclosan via
personal care products. In that study, there was no ap-
parent difference in plasma triclosan concentrations be-
tween the two groups. In the present study, there was an
even broader range of triclosan concentrations in plasma
(0.010–38 ng/g) and a large within group variability.
However, in contrast to the findings by Sandborgh-
Englund et al. (2006), the present study showed that the
triclosan concentrations in both plasma and milk were
clearly and significantly higher in the exposed group
than in the control group (Table 1) (Figs. 2 and 3). It is
noteworthy that the two individuals who were exposed
via soap and deodorant had the lowest triclosan
concentration in the exposed group on the first sampling
occasion; the seven who used toothpaste all had higher
levels. The results support the main hypothesis of the
study, namely that triclosan containing personal care
products are the dominant source of exposure to tri-
closan in the general population.
4.2. Milk-to-plasma ratio and infant exposure to
triclosan via breast milk
On an individual basis, the triclosan concentration in
milk was lower than in plasma (Fig. 4). It is well rec-
ognized that many acidic compounds have an M/Pb1,
which may partly be explained by the fact that breast
milk is slightly more acidic than plasma (Fleishaker,
2003). Triclosan is a phenol and a weak acid (pKa=7.9),
thus the differences in acid ionization equilibrium for
triclosan in plasma (pH=7.5) and milk (pH=6.5) will
favor the ionized form in plasma compared to milk.
Assuming that the protonated triclosan equilibrates
across the mammary membrane, the total concentration
of triclosan in milk will be lower in milk than in plasma.
Another factor may be the presence and behavior of
triclosan conjugates in the plasma and milk.
Sandborgh-Englund et al. (2006) reported triclosan
conjugates in plasma to be approximately 70% of the
total triclosan concentration after a swallowed dose of
4 mg triclosan. The conjugated fraction remained rela-
tively unchanged over a range of plasma concentrations.
If the hydrophilic and highly acidic triclosan sulfate and
glucuronide conjugates are not transported across the
mined in this work (free triclosan+conjugates) would be
Fig. 4. Triclosan concentration ratios of paired milk and plasma
samples (M/P) plotted against the corresponding plasma concentration
(ng/g fresh weight). Open circle=control, filled circle=exposed; first
91 M. Allmyr et al. / Science of the Total Environment 372 (2006) 87–93
higherinplasma.Takentogetherwiththe influence ofpH
Other factors, e.g. protein binding and partitioning to
lipids in either matrix, may also affect the relative con-
centrations in plasma and milk. For example, Guvenius
Meironyté et al. (2003) observed lower concentrations
of polychlorobiphenylols (OH-PCBs) in milk than in
plasma from nursing mothers, which they attributed to
the specific and potent affinity of the identified OH-
PCBs to the thyroid hormone (T4) transport protein
(transthyretin, TTR) in plasma.
The reason for the observed high variability and the
negative correlation of triclosan M/P with plasma con-
centration is not known. The mothers in the present
study were instructed to sample milk preferably before
the first nursing of the day. Yet, the milk sample may
have been expressed at any time before the time for
plasma sampling and before, within or after a feeding,
whenever it suited the mother. Perhaps more impor-
tantly, the individual dose and time point for exposure to
triclosan is unknown; hence, the time between exposure
and the sampling of milk and plasma respectively may
have varied considerably. The above-mentioned incon-
sistencies are all likely to have affected the absolute and
relative concentrations of triclosan in plasma and milkin
the individual subjects, resulting in a variable M/P.
Two daily brushes with 1–2 g toothpaste containing
0.3% triclosan results in a theoretical maximal dose of
6–12 mg. Triclosan was found in breast milk at con-
centrations ranging from b0.018 to 0.95 ng/g, which are
comparable to those determined by Adolfsson-Erici et al.
(2002) (0.9, 1.6 and 2.0 ng/g milk, fresh weight basis by
personal communication). The triclosan concentrations
were significantly higher in the exposed group; thus, the
infants' exposure to triclosan was related to the mothers'
exposure. However, for an infant weighing 4 kg with an
estimated milk intake of 150 ml/kg/day, extrapolating to
the concentrations found in single milk samples in the
present study, the dailyintake oftriclosan wouldbe b11–
570 ng/day. Conclusively, the infant is exposed to a
considerably smaller dose of triclosan via the breast milk
compared to the dose in the mother. The direct contact
with products containing triclosan may be of more
4.3. Health perspectives on triclosan exposure
The potential long-term negative effects should be
weighed against the benefits of triclosan use. However,
the long-term effects of a chronic systemic exposure to
triclosan in humans are not fully understood. Triclosan
is not acutely toxic to mammals as shown by traditional
toxicity tests (Bhargava and Leonard, 1996). However,
Schuur et al. (1998) demonstrated that triclosan inhibit-
ed the iodothyronine hormone sulfotransferase activity
in rat liver cytosol in vitro. In addition, Wang et al.
(2004) showed that triclosan, besides being a substrate
for certain phase II metabolizing sulfotransferase and
glucuronosyltransferase enzymes, was able to inhibit the
sulfonation and glucuronidation of 3-hydroxy-benzo[a]
pyrene in human liver preparations in vitro. Triclo-
san has also been shown to be a potent inhibitor of
induced activity of the phase I metabolizing enzymes 7-
pentoxyresorufin O-depentylase and 7-pentoxyresorufin
O-deethylase in liver microsomes obtained from rats
treated with phenobarbital and 3-methylcholanthrene
respectively (Hanioka et al., 1996). More recently, an in
vitro study showed that triclosan was an activator of the
human pregnane X receptor, which regulates the genetic
transcription of, among other things, the broad speci-
ficity metabolic enzyme CYP3A4, an important medi-
ator for biotransformation of steroids, pharmaceuticals
and other xenobiotics (Jacobs et al., 2005). Whether the
above-mentioned triclosan–enzyme interactions are of
any clinical importance is unknown. Nevertheless, the
findings suggest that triclosan may exert effects on
biological systems in the sense that the biotransforma-
tion of other exogenous and endogenous compounds
may be inhibited or induced. The possible long-term
clinical adverse effects of a chronic systemic exposure
to triclosan in humans need to be scrutinized. Addition-
ally, since the conjugated triclosan might be biologically
inactive, the fraction of unchanged triclosan in humans
should be investigated with appropriate specific analyt-
ical methods and studied in controlled experiments with
normal use of triclosan containing products.
The beneficial effects of triclosan to general hygiene
and dental health are not clinically evident,and therefore
the routine-like use of triclosan in toothpaste and soaps
has been questioned (Edwardsson et al., 2005; Larson
et al., 2003; Perencevich et al., 2001). In Sweden, all
ingredients in personal care products, as defined here,
have to be declared on the products. Moreover, in the
case of toothpaste containing triclosan, there is an ad-
on indication by a dentist. Apparently, this initiative to
help the customer make an informed choice has not been
very effective, since none of the exposed mothers in the
present study were aware of their use of triclosan.
We thank the mothers and the staff at Eken BVC that
participated in the project. Ann-Marie Sjöö is thanked
92M. Allmyr et al. / Science of the Total Environment 372 (2006) 87–93
for laboratory assistance. The project was supported by Download full-text
the Swedish Research Council for Environment,
Agricultural Sciences and Spatial Planning, the Swedish
Patent Revenue Fund for Research in Preventive Den-
tistry, and grants from Karolinska Institutet.
Adolfsson-Erici M, Pettersson M, Parkkonen J, Sturve J. Triclosan, a
commonlyused bactericide foundin humanmilk andin the aquatic
environment in Sweden. Chemosphere 2002;46:1485–9.
Allmyr M, McLachlan MS, Sandborgh-Englund G, Adolfsson-Erici
M. Determination of triclosan as its pentafluorobenzoyl ester in
human plasma and milk using electron capture negative ionization
mass spectrometry. Anal Chem 2006,___________________
Bagley DM, Lin YJ. Clinical evidence for the lack of triclosan
accumulation from daily use in dentifrices. Am J Dent 2000;13:
Bhargava HN, Leonard PA. Triclosan: applications and safety. Am J
Infect Control 1996;24:209–18.
Edwardsson S, Burman L, Adolfsson-Erici M, Backman N. Risker och
nytta med triklosan i tandkräm [Risks and benefits with triclosan
containing toothpastes]. Tandlakartidningen 2005;97:58–64.
Fleishaker JC. Models and methods for predicting drug transfer into
human milk. Adv Drug Deliv Rev 2003;55:643–52.
Guvenius Meironyté D, Aronsson A, Ekman-Ordeberg G, Bergman A,
NorénK. Human prenataland postnatal exposure to polybrominated
diphenyl ethers, polychlorinated biphenyls, polychlorobiphenylols,
Hanioka N, Omae E, Nishimura T, Jinno H, Onodera S, Yoda R, et al.
Interaction of 2,4,4′-trichloro-2′-hydroxydiphenylether with mi-
crosomal cytochrome P450-dependent monooxygenases in rat
liver. Chemosphere 1996;33:265–76.
Hovander L, Malmberg T, Athanasiadou M, Athanassiadis I, Rahm S,
Bergman A, et al. Identification of hydroxylated PCB metabolites
and other phenolic halogenated pollutants in human blood plasma.
Arch Environ Contam Toxicol 2002;42:105–17.
Jacobs MN, Nolan GT, Hood SR. Lignans, bacteriocides and
organochlorine compounds activate the human pregnane X
receptor (PXR). Toxicol Appl Pharmacol 2005;209:123–33.
Larson E, Aiello A, Lee LV, Della-Latta P, Gomez-Duarte C, Lin S.
Short-and long-term effects of handwashing with antimicrobial or
Lin YJ. Buccal absorption of triclosan following topical mouthrinse
application. Am J Dent 2000;13:215–7.
Merck Index 12:3, 2000.
Moss T, Howes D, Williams FM. Percutaneous penetration and dermal
metabolism of triclosan (2,4,4′-trichloro-2′-hydroxydiphenyl
ether). Food Chem Toxicol 2000;38:361–70.
Mustafa M, Wondimu B, Hultenby K, Yucel-Lindberg T, Modeer T.
Uptake, distribution and release of 14C-triclosan in human
gingival fibroblasts. J Pharm Sci 2003;92:1648–53.
of prevalent antibacterial soaps. Am J Infect Control 2001;28:281–3.
Remberger M, Sternbeck J, Strömberg K. Screening av triclosan och
vissa bromerade fenoliska ämnen i Sverige. [Screening of triclosan
and some brominated, phenolic compounds in Sweden]. IVL Rapp
Sandborgh-Englund G, Adolfsson-Erici M, Odham G, Ekstrand J.
Pharmacokinetics of triclosan following oral ingestion in humans.
J Toxicol Environ Health 2006;69:1861–73.
Schuur AG, Legger FF, van Meeteren ME, Moonen MJH, van
Leeuwen-Bol I, Bergman A, et al. In vitro inhibition of thyroid
hormone sulfation by hydroxylated metabolites of halogenated
aromatic hydrocarbons. Chem Res Toxicol 1998;11:1075–81.
Syracuse Research Corporation. PhysDrop Database Demo. http://
www.syrres.com/esc/physdemo.htm [accessed March 2006].
The European Commission. Opinion on triclosan resistance, adopted
on 27–28 June 2002. http://europa.eu.int/comm/food/fs/sc/ssc/
out269_en.pdf [accessed March 2006].
Wang LQ, Falany CN, James MO. Triclosan as a substrate and
inhibitor of 3′-phosphoadenosine 5′-phosphosulfate-sulfotransfer-
ase and UDP-glucuronosyl transferase in human liver fractions.
Drug Metab Dispos 2004;32:1162–9.
93M. Allmyr et al. / Science of the Total Environment 372 (2006) 87–93