International Journal of Hygiene and Environmental Health 214 (2011) 339– 347
Contents lists available at ScienceDirect
International Journal of Hygiene and
Environmental Hea
j o ur nal homepage: www.elsev ie
Mini-Revi
Are pot isp
Tinne Ge
a Toxicological in 1, 2
b Analytical an
a r t i c l
Article history:
Received 21 Ja
Received in re
Accepted 5 Ap
Keywords:
Bisphenol-A
BPA
Human exposu
BPA applicatio
Exposure asse
us ap
ans
ugh
een e
Sever
is po
be e
es fo
ation
affect
all so
re to
General information
Bisphenol-A (BPA) is a high production volume chemical; its
production volume amounted to 3.8 million tons in 2006 (Plastics
Europe, 201
tant plastic
epoxy resin
million ton
in consume
in the envir
2009; Vand
large numb
wastewater
Belfroid et a
Health e
(Bondesson
are not fully
controversi
dictory. At t
listing BPA
the Europea
which wou
intake of 50
However, t
advice had
∗ Correspon
E-mail add
exposure to the chemical should be minimized and this led to a
ban of PC baby bottles.
Review objective
1438-4639/$ –
doi:10.1016/j.0). Almost 65% of the BPA is polymerized to the resis-
polycarbonate (PC) and almost 30% is polymerized to
s (Plastics Europe, 2010), which still leaves some 0.2
s for additional applications. Due to its widespread use
r goods and commodities, BPA is ubiquitously present
onment (Crain et al., 2007) and in humans (Becker et al.,
enberg et al., 2007, 2010). BPA has been detected in a
er of environmental compartments (e.g. surface- and
, sediments and biota) (Arditsoglou and Voutsa, 2010;
l., 2002; Navarro-Ortega and Barceló, 2010).
ffects related to BPA have been extensively investigated
et al., 2009; Braun et al., 2009; Melzer et al., 2010), they
understood yet and their interpretation is sometimes
al. Also governmental decisions are sometimes contra-
he same time, Canada was the first country in the world
as a toxic substance (Environment Canada, 2010), and
n Food Safety Authority could not identify new studies
ld call for a revision of the current BPA tolerable daily
g/kg bw/day (European Food Safety Authority, 2010).
he Health and Consumer Commissioner said the EFSA
thrown up “areas of uncertainty” which meant infant
ding author. Tel.: +32 32652704; fax: +32 32652722.
ress: tinne.geens@ua.ac.be (T. Geens).
Though humans in the industrialized world are daily exposed
to BPA (Calafat et al., 2005, 2008); the totality of sources for
human exposure is still insufficiently clarified. While exposures
through polycarbonate bottles and epoxy resin coated food and
beverage cans have been extensively discussed in the literature
(Carwile et al., 2009; Geens et al., 2010; Mercea, 2009), the currently
observed exposure cannot be explained by dietary exposure alone.
The importance of the non-food and non-oral exposure route has
also gained research interest, while some studies have already con-
firmed a contribution of alternative exposure pathways, e.g. dermal
absorption (see further) (Biedermann et al., 2010; Braun et al., 2011;
Zalko et al., 2011). This review aims to provide a comprehensive
overview of the current BPA applications those where it is used
as a monomer and those where it is modified or polymerized to
form other end products. The review also summarizes the extent
of human exposure assessed through various pathways and indi-
cates possible (overlooked) sources, which can be a risk for human
exposure and which should be investigated more thoroughly.
Exposure assessment of BPA
Exposure assessment by biomonitoring
Recently, several biomonitoring studies on BPA concentrations
in urine were conducted worldwide (e.g. US, Germany, Belgium,
see front matter © 2011 Elsevier GmbH. All rights reserved.
ijheh.2011.04.005ew
ential sources for human exposure to b
ensa,∗, Leo Goeyensb, Adrian Covacia
Centre, Department of Pharmaceutical Sciences, University of Antwerp, Universiteitsple
d Environmental Chemistry, University of Brussels, Pleinlaan 2, 1050 Brussels, Belgium
e i n f o
nuary 2011
vised form 30 March 2011
ril 2011
re
ns
ssment
a b s t r a c t
This review summarizes the numero
human exposure. The exposure to hum
polycarbonate bottles, as well as thro
there seems to be a discrepancy betw
based on food/drink concentrations.
sources. Although the main use of BPA
used as an additive, from which it may
evidence for the contribution of sourc
used as an additive. Polymeric applic
of BPA present, as well as the factors
from these sources. It is clear that not
the necessity to study also the exposulth
r .de / i jheh
henol-A overlooked?
610 Antwerp, Belgium
plications of bisphenol-A (BPA) and the potential sources for
is believed to occur mainly through food contamination from
food and beverage cans coated with epoxy resins. However,
xposure assessments based on biomonitoring data and those
al recent studies indicated also the importance of non-food
lymerization to polycarbonate and epoxy resins, it can also be
asily released. Several studies have already provided scientific
r dermal BPA absorption, such as thermal paper where BPA is
s of BPA require further investigation regarding the amounts
ing its release and potential dermal or non-dermal exposure
urces of BPA have been identified. This overview emphasizes
these unexpected sources of BPA.
© 2011 Elsevier GmbH. All rights reserved.

340 T. Geens et al. / International Journal of Hygiene and Environmental Health 214 (2011) 339– 347
China, Korea) to assess the human exposure to BPA (Becker
et al., 2009; Calafat et al., 2005, 2008; He et al., 2009; Milieu en
Gezondheid, 2010; Teitelbaum et al., 2008; Völkel et al., 2008;
Yang et al., 2009). Biomonitoring data provide reliable information
about exposure, because it does not require an identification of all
exposure sources (Vandenberg et al., 2010). Both single spot urine
samples and 24 h urine samples have been used for this purpose;
however, some conflicting opinions about the correct sampling
method have been reported.
For single spot samples, the urinary concentrations are con-
verted into a daily exposure simply by multiplying the BPA
concentration with an average daily urine output or creatinine
excretion for a certain age group. Since BPA excretion is fast, single
spot urine samples merely provide information on the actual expo-
sure and not on chronic or long-term intake (Lakind and Naiman,
2008). Additionally, the time of urine collection after exposure to
a relevant source of BPA can have considerable influence on the
results. Individual high BPA concentrations should not be used to
estimate th
shortly (∼h
itoring stud
values and o
to day) are a
Dekant and
single spot
estimations
Next to s
collected. A
collected ur
lyte concen
they also fo
samples of
surements s
of a subject
ing the aver
(Arakawa e
A selecti
summarize
frequencies
as compare
Indirect exp
The exp
whereby an
most impor
investigated
food, and in
BPA exposu
packaging m
and, more importantly, in the lining of food and beverage cans made
of epoxy resins. A limited number of studies summarized in Table 2
estimated the exposure to BPA through canned food and other food
products in combination with a food consumption pattern (Geens
et al., 2010; Higuchi et al., 2004; Lim et al., 2009; Mariscal-Arcas
et al., 2009; Thomson and Grounds, 2005). The BPA concentrations
in non-canned food are lower compared to canned food (Geens
et al., 2010). Sajiki et al. (2007) found a range of <LOD-842 ng/g
BPA in 48 canned foods, while the concentrations of 23 similar food
items in plastic containers ranged from <LOD-14 ng/g and those
of food items in paper packages from <LOD-1 ng/g in 16 foods,
respectively. Geens et al. (2010) also found substantially higher
concentration in 21 canned foods (range: 0.2–169 ng/g) compared
to 16 of their non-canned counterparts (range: 0.1–1.3 ng/g) packed
in plastic, glass, paper and Tetra Pak. Polycarbonate drinking bot-
tles might also contribute significantly to adults’ BPA exposure
(Carwile et al., 2009), whereas for infants, polycarbonate baby bot-
tles are expected to be the major source of BPA (von Goetz et al.,
ause
tions
ant e
foun
m da
ir in
foun
rban
ura,
t ing
on to
, 200
en 53
t sam
of 0
ere w
003)
amo
l et a
rison
ches
htly
posu
obt
inclu
d low
nces
s dif
non
Table 1
Overview of m ng est
Country e
y)
US
US
Germany
China
Korea
Belgium (14
Germany
US (6–10 ye
N: sample size
a Calculated
b Median.
c Calculatede average daily intake, since they result from exposures
ours) prior to the urine collection. However, in biomon-
ies with a large number of individuals, these extreme
ther influencing factors (such as variable diets from day
veraged out. Therefore, Lakind and Naiman (2008) and
Völkel (2008) report that BPA intake calculated from
urine samples of large cohorts are considered as good
of the average population exposure.
ingle spot urine samples, 24 h urine samples can also be
rakawa et al. (2004) reported that the analysis of 24 h
ine can overcome the within-day variation in the ana-
tration caused by the short half life of BPA. Nevertheless,
und a large variation in daily BPA intake in 24 h urine
consecutive days and concluded that also these mea-
hould not be used to estimate the long-term BPA intake
. However, 24 h urine samples can be used for estimat-
age exposure of a population in a cross-sectional study
t al., 2004).
on of studies with sufficiently high sample numbers is
d in Table 1. Both the calculated intakes and detection
in Asian countries (China and Korea) are slightly lower
d to European countries and the US.
osure assessment
osure can also be assessed by an indirect approach
exposure model is supplied with the BPA levels of the
tant sources (von Goetz et al., 2010). The predominantly
sources for BPA are food, water, dust and air. For adults,
particular canned food, is considered a major source for
re (von Goetz et al., 2010). BPA can be present in food
aterials, such as polycarbonate and PVC (see further)
2010).
Bec
centra
import
(2001)
air fro
door a
(2010)
from u
Kawam
Dus
tributi
Völkel
betwe
tic dus
intake
dust w
et al. (2
studies
(Völke
Compa
approa
Slig
rect ex
results
which
showe
differe
source
well as
ost important biomonitoring studies measuring BPA in urine with the correspondi
N Det. freq (%) GM (ng/mL) Average intak
(ng/kg bw/da
394 95 1.33 33.3a
2517 93 2.60 46.8
147 – 1.20b 30.0
922 50 0.87 21.8a
485 76 0.56 14.0a
–16 years) 198 99.5 2.22 55.5a
599 99 2.66 60
ars) 159 95 3.40 68.0c
; Det. freq: detection frequency; GM: geometric mean.
for a urinary output of 1500 mL and body weight of 60 kg.
for a urinary output of 600 mL and body weight of 30 kg. of its low vapor pressure and, therefore, its low con-
in air, inhalation of BPA from air is unlikely to be an
xposure source (Dekant and Völkel, 2008). Wilson et al.
d concentrations between <0.1 and 1.8 ng/m3 in indoor
ycare centers and between <0.1 and 2.5 ng/m3 in out-
North Carolina, US and recently Fu and Kawamura
d comparable concentrations (0.004–17.4 ng/g) in air
sites in India, China, Japan and New Zealand (Fu and
2010).
estion and inhalation are expected to have a minor con-
the overall exposure compared to food (Dekant and
8; Geens et al., 2009). Geens et al. (2009) found levels
5 and 9730 ng/g (median 1460 ng/g dust) in 18 domes-
ples from Belgium which resulted in a calculated BPA
.0004 g/kg bw/day for an adult. The levels of BPA in
ithin the same range as the results obtained by Rudel
and Völkel et al. (2008). The median values in the latter
unted to 555 ng/g dust and 820 ng/g dust, respectively
l., 2008).
of the biomonitoring and indirect exposure
lower intake estimations were observed for the indi-
re assessment through canned food compared to the
ained by biomonitoring data. The exposure models
ded microwave containers or prepared meals, clearly
er estimations than biomonitoring data (Fig. 1). These
suggest the existence and importance of exposure
ferent from canned food, which could include food as
-food sources.
imated daily intake.
High intake
(ng/kg bw/day)
Reference
129.5a (p. 95) Calafat et al. (2005)
274.2 (p. 95) Calafat et al. (2008) and Lakind and
Naiman (2008)
233 (max) Völkel et al. (2008)
131.5 (p. 75)a He et al. (2009)
504 (p. 95)a Yang et al. (2009)
174.5 (p. 90)a Milieu en Gezondheid (2010)
370 (p. 95) Becker et al. (2009)
532 (p. 95)c Teitelbaum et al. (2008)