Neurotoxicology and Teratology 34 (2012) 592597

Contents lists available at SciVerse ScienceDirect

Neurotoxicology and Teratology

journal homepage: www.elsevier.com/locate/neutera

Neonatal exposure to Thimerosal from vaccines and child development in the rst
3 years of life
Dorota Mrozek-Budzyn , Renata Majewska, Agnieszka Kieltyka, Malgorzata Augustyniak
Epidemiology and Preventive Medicine, Jagiellonian University Medical College, Krakow, Poland

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

Article history: Background: Despite the common use of Thimerosal as a preservative in childhood vaccines since the 1930s,
Received 22 November 2011 there are not many studies on ethylmercury toxicokinetics and toxicodynamics in infants. The knowledge of
Received in revised form 3 October 2012 ethylmercury's potential adverse effects is derived mostly from parallel methylmercury research or from an-
Accepted 5 October 2012 imal and theoretical models.
Available online 13 October 2012
Aim of the study: This study was designed to examine the relationship between neonatal exposure to
Thimerosal-containing vaccine (TCV) and child development.
Keywords:
Thimerosal-containing vaccine
Material and methods: The study sample consisted of 196 infants born between January 2001 and March 2003
Neonatal exposure to mothers attending ambulatory prenatal clinics in the rst and second trimesters of pregnancy in Krakow.
Child neurodevelopment Vaccination history (date and the type of the vaccine) was extracted from physicians' records. Child develop-
ment was assessed using the Bayley Scales of Infant Development (BSID-II) measured in one-year intervals
over 3 years. General Linear Model (GLM) and Generalized Estimating Equation (GEE) models adjusted for
potential confounders were used to assess the association.
Results: An adverse effect of neonatal TCV exposure was observed for the psychomotor development index
(PDI) only in the 12th and 24th months of life ( = 6.44, p b 0.001 and = 5.89, p b 0.001). No signicant
effect of neonatal TCV exposure was found in the 36th month. The overall decit in the PDI attributable to
neonatal TCV exposure measured over the course of the three-year follow-up (GEE) was signicantly higher
in TCV group ( = 4.42, p = 0.001). MDI scores did not show the adverse association with neonatal TCV
exposure.
2012 Published by Elsevier Inc.

1. Introduction
The discussion on the safety of Thimerosal-containing vaccines
(TCVs) started in 1999 (Joint Statement issued by the American
Academy of Pediatrics and the U.S. Public Health Service in July 1999).
It was suspected that ethylmercury contained in the vaccines had a
harmful effect on children's neurodevelopment. In fact, the toxicity of
low dose human exposure to ethylmercury has not been assessed suf-
ciently, although it is assumed to have a similar effect as exposure to
methylmercury (Pichichero et al., 2000, 2002), an organic compound
with demonstrated harmful consequences (Grandjean et al., 1997,
2010; Jedrychowski et al., 2006; Lederman et al., 2008; Oken et al.,
2008). A joint statement issued by the American Academy of Pediatrics
and the U.S. Public Health Service in July 1999 called for measures to re-
move Thimerosal from vaccines for infants. Since 2004, no vaccines
recommended and routinely used in the United States and the Europe-
an Union (EU) to protect preschool children have contained Thimerosal
Corresponding author at: Epidemiology and Preventive Medicine, Jagiellonian University
Medical College, Kopernika 7A, 31-034 Krakw, Poland. Tel.: +48 12 423 10 03; fax: +48 12
422 87 95.
E-mail address: dorota.mrozek-budzyn@uj.edu.pl (D. Mrozek-Budzyn).
(CDC, 2004; EMEA, 2001). Nevertheless, most countries continue to use
TCVs in their childhood immunization schedules. The World Health
Organization's (WHO) position is based on the high effectiveness of
vaccines in protecting children against infectious diseases and not on
specic studies addressing ethylmercury in population studies (WHO,
2000, 2006). It is difcult to argue with the WHO's strategy, since
there is a scarcity of evidence for the harmful inuence of ethylmercury
in population studies. Reports on the adverse health effects of
ethylmercury derive mainly from animal experiments (Hornig et al.,
2004; Hewitson et al., 2010) and in vitro tests in human cell-cultures
(Mutkus et al., 2005; Yel et al., 2005; Herdman et al., 2006). These ex-
perimental studies have shown consistent toxicity in neural cells caused
by Thimerosal at concentrations relevant to vaccines. Furthermore, in
animal studies, behavior changes in groups exposed to ethylmercury
were observed (Hornig et al., 2004; Hewitson et al., 2010). The applica-
bility of these studies to humans is unknown, but the consistency of
their results suggests biological consequences in the neurodevelopment
of susceptible infants.
The controversies as to the harmful effects of childhood TCVs and the
strategies to abandon TCV vaccinations in developed countries create
confusion among parents and health workers in some regions where
TCVs are still administered to children. In Poland, newborns and infants

0892-0362/$ see front matter 2012 Published by Elsevier Inc.

http://dx.doi.org/10.1016/j.ntt.2012.10.001
 D. Mrozek-Budzyn et al. / Neurotoxicology and Teratology 34 (2012) 592597
are still immunized with TCVs according to the recommended schedule.
In the rst 24 h of life, newborns are vaccinated against hepatitis-B,
mostly with Thimerosal. The same applies to diphtheria, tetanus, and
pertussis (DTP) vaccines administered to infants. The costs of
Thimerosal-free vaccines in Poland are placed on parents, and therefore
they often choose to immunize their children with cost-free TCVs. Un-
equal access to Thimerosal-free vaccines has led to hot debates about
the potential harmful inuence of TCVs on children's health. The current
concern about TCV exposure is related not only to advanced health
problems, but also to minor developmental disorders of infants and pre-
school children.
This study was designed to examine the hypothesis that neonatal
exposure to TCVs may have a negative impact on psychomotor devel-
opment and cognitive functions in children.
2. Materials and methods
This is a prospective molecular epidemiologic study, combining
environmental monitoring and molecular approaches with compre-
hensive neurodevelopment assessments. In the analysis, we used
data from an earlier established Krakow birth cohort of children,
part of an ongoing, collaborative study with Columbia University in
New York, on the vulnerability of the fetus and child to environmen-
tal factors. The study has received the approval of the Jagiellonian
University Ethical Committee.
The enrollment (November 2000August 2003) included only
non-smoking women, aged 1835 years, with a singleton pregnancy,
without illicit drug use and HIV infection, free from chronic diseases,
such as diabetes or hypertension, and residing in Krakow for at least
one year prior to pregnancy. In the 2nd trimester of pregnancy, all
women completed a detailed standardized questionnaire on demo-
graphic data, household characteristics, lifestyle and health. The en-
rolled women gave birth to 505 children. Weight, length and head
circumference at birth and delivery course were recorded for all in-
fants. During the rst three years of life, the children were followed
every 3 months over 2 years and every 6 months in the 3rd year of
life. Mothers were asked to provide information on the infants' health
and household characteristics by trained interviewers, who carried
out detailed face-to-face standardized interviews.
2.1. Vaccination data
The data of the infants' vaccination history (date of vaccination
and type of vaccine) were extracted from physicians' records.
2.2. Biological samples and analysis
Maternal and cord blood were sampled at delivery and concentra-
tions of cotinine and heavy metals (mercury, lead) were examined.
Details on blood sample collection and analysis were presented in
earlier publications (Jedrychowski et al., 2006).
2.3. Test of maternal non-verbal intelligence-3rd edition (TONI-3)
The TONI-3 is a language-free measure of general intelligence
designed for use with individuals aged 585 years and considered to
be relatively free of cultural bias. We have included this instrument to
adjust for the maternal contribution to child cognitive development.
2.4. Infant neurodevelopment testing
The Bayley Scales of Infant Development, second edition (BSID-II),
were administered in the 12th, 24th and 36th months of life (within
4 weeks of the target age). The Psychomotor Scale assesses control of
gross and ne muscle groups (rolling, crawling, creeping, sitting,
standing, walking, running, and jumping). The Mental Scale includes
593
items that assess memory, habituation, problem solving, early num-
ber concepts, generalization, classication, vocalization, language,
and social skills (Bayley, 1993). Test scores are adjusted for the age
of the child to obtain the Psychomotor Development Index (PDI)
and the Mental Development Index (MDI). Test results are in one of
four categories: 1) accelerated performance (score > 115), 2) within
normal limits (score, 85 to 114), 3) mildly delayed performance
(score, 70 to 84), and 4) signicantly delayed (score b 69).
The BSID-II were conducted at the Department of Epidemiology
and Preventive Medicine by trained examiners who were unaware
of the child's exposure.
2.5. Confounders
Data on possible confounders such as maternal age (continuous
variable) and education (university vs. non-university), maternal
non-verbal intelligence (result of TONI-3 as a continuous variable),
older siblings (yes or no), exposure to passive tobacco smoking in
pregnancy (yes or no), child's gender, gestational age (>37 weeks
vs. 37 weeks), birth weight (continuous variable, unit 100 g), cord
blood mercury level, cord blood lead level, breastfeeding (child
breastfed at least 6 months) and exposure to ethylmercury in later
life (from the 29th day to the 6th month) were collected and included
in the analysis.
The 196 children were eligible for the analysis for whom vaccina-
tion history, BSID scores, maternal TONI-3 results, and Pb levels in
cord blood were available. Of these, 168 children had measurements
of Hg levels in cord blood.
2.6. Statistical analysis
In the descriptive analysis, the differences in the distribution of
mother and newborn parameters grouped by neonatal ethylmercury
vaccination status were tested using 2 (for nominal variables) and
MannWhitney (for continuous variables) tests.
The association between PDI and MDI scores and neonatal expo-
sure to TCV at particular years was studied using the General Linear
Model (GLM). The impact of neonatal ethylmercury exposure on
child development over the rst three years of life was assessed
using Generalized Estimating Equations (GEE). The logistic models
were used to assess the risk of clinically assessed developmental
delay (PDI and MDI b 85). Possible confounding variables with
p b 0.1 in univariate models were considered in the multivariate anal-
ysis. Additionally, two variables were added despite being insigni-
cant in the univariate model: TCV exposure in later life and cord
blood mercury concentration. They were included to adjust for the ef-
fect of exposure to mercury in the earlier and later life of a child.
Statistical analyses were performed with SPSS software version
15.0 (SPSS Inc., 2006).
3. Results
In the analyzed group, all children were vaccinated against hepatitis
B in the neonatal period, 143 (73.0%) of them with TCV. Fifty-three
(27%) children were vaccinated with Thimerosal-free Engerix B; 108
(55.1%) with Euvax B, and 35 (17.9%) with Hepavax-Gene; the latter
two containing 25 g Thimerosal per dose. One-hundred eighty-eight
(95.9%) children were vaccinated within one day after birth and 194
(99.0%) within one week.
Exposure to TCV up to the age of 6 months was connected with fur-
ther doses of hepatitis B and DTP vaccines administered according to
the Polish Immunization Schedule. The children were vaccinated with
Polish manufactured DTP vaccine with Thimerosal (25 g/dose) and
Thimerosal-free vaccines such as: Infanrix IPV+Hib, Infanrix Hexa,
Tripacel and Tetracoq.
594 D. Mrozek-Budzyn et al. / Neurotoxicology and Teratology 34 (2012) 592597

Between the 29th day and the 6th month of life, 86.2% of children
were vaccinated with TCV. The children immunized with ethylmercury-
containing vaccines in the neonatal period were more frequently vacci-
nated with ethylmercury later in their life compared to those not immu-
nized with TCV (91.7% vs. 71.7%, pb 0.001) (Table 1).
All analyzed vaccines (Thimerosal-containing and Thimerosal-free)
contained aluminum compounds adjuvant.
No signicant differences were found for maternal characteristics
such as age, education and TONI-3 scores between the children ex-
posed and non-exposed to TCV in the neonatal period. Passive tobac-
co smoking was less common among mothers of the TCV vaccinated
group (32.9% vs. 47.2%, p = 0.065). As for child characteristics, only
the lead level in cord blood was signicantly lower in TCV exposed
group than in non-exposed (1.3 vs. 1.8 g/L, p b 0.001) (Table 1).
In the TCV-exposed children, the PDI scores at the 12th and 24th Fig. 1. Average development tests scores in ethylmercury exposed and non-exposed
group (with standard deviation). MDI Mental Development Index. PDI Psychomo-
month of life were statistically lower than in non-exposed (95.3 vs. tor Development Index.
101.6, p = 0.002 and 96.8 vs. 102.6, p b 0.001 respectively). That dif-
ference in PDI scores disappeared at the 36th month (103.6 vs.
When we also adjusted the model for the cord blood mercury level,
104.3, p = 0.808). MDI scores were signicantly different only in the
the effect of neonatal exposure to TCV was also signicant and on a
36th month and were higher in children exposed to TCVs (104.9 vs.
similar level ( = 5.78, p = 0.009 in 12th month and = 7.64,
101.3, p = 0.028) (Fig. 1).
p b 0.001 in 24th month) (Table 3, Model 2, Fig. 2).
Comparison of the subgroup of children for whom complete data on
MDI scores did not show the adverse association with TCV expo-
cord blood mercury levels were available to a subgroup for whom such
sure (Table 3). On the contrary, when we adjusted for the cord
information was not obtained showed signicant differences in Pb level
blood mercury level, the children exposed to TCV had signicantly
(1.3 vs. 2.3 g/L, p b 0.001). There were also signicant differences be-
higher MDI scores than those non-exposed in the 36th month of life.
tween the subgroups for the frequency of Thimerosal-free vaccines in
Girls had signicantly higher PDI and MDI scores than boys ( = 2.5,
both the neonatal (16.1% vs. 92.9%, p b 0.001) and post-neonatal period
p = 0.011 and = 3.4, p = 0.002 respectively). Higher MDI scores over
(10.7% vs. 32.1%, p = 0.002) (Table 2).
the rst three years of life were also associated with a lack of environ-
mental tobacco smoke (ETS) during pregnancy (= 2.8, p = 0.015).
3.1. GLM and GEE linear models

After adjusting for potential confounders (Table 3, Model 1), a sta- 3.2. Logistic models
tistically signicant adverse effect of TCV exposure on the PDI was
seen at the 12th and the 24th month of life ( = 6.44, p b 0.001 To assess the variables responsible for child developmental delay,
and = 5.89, p b 0.001), but not at the 36th month. The overall def- logistic regression models for particular years and a GEE model for an
icit in the PDI attributable to neonatal TCV exposure measured over
the course of the three-year follow-up (GEE) was signicantly higher Table 2
Differences between subjects with and without measurement of mercury in cord
in the TCV group ( = 4.42, p = 0.001) (Table 3, Model 1, Fig. 2).
blood.

Table 1 Characteristics Hg measured Hg not p

Characteristics of the study subjects by neonatal ethylmercury exposure. (n = 168) measured
(n = 28)
Characteristics Non-exposed Exposed to p
N % N %
(n = 53) TCV (n = 143)
Maternal education
N % N %
Elementary/vocational school 15 8.9 8 14.3 ns
Maternal education High school 57 33.9 13 46.4
Elementary/vocational school 5 9.4 14 9.8 ns University 96 57.1 11 39.3
High school 22 41.5 48 33.6 Passive tobacco smoking in pregnancy 59 35.1 13 46.4 ns
University 26 49.1 81 56.6 Child's gender
Passive tobacco smoking in pregnancy 25 47.2 47 32.9 0.065 Boys 94 56.0 12 42.9 ns
Child's gender Girls 74 44.0 16 57.1
Boys 28 52.8 78 54.5 ns Gestational age b37 weeks 4 2.4 0 ns
Girls 25 47.2 65 45.5 Breastfeeding up to 6 months 144 85.7 20 71.4 0.058
Gestational age b37 weeks 0 4 2.8 ns Parity
Breastfeeding up to 6 months 41 77.4 123 86.0 ns 1 110 65.5 16 57.1 ns
Parity 2 58 34.5 12 42.9
1 29 54.7 97 67.8 ns Ethylmercury exposure to 28th day of life
2 24 45.3 46 32.2 0 27 16.1 26 92.9 b0.001
Ethylmercury exposure from 29th day6th month of life 25 g 141 83.9 2 7.1
0 15 28.3 12 8.4 b0.001 Ethylmercury exposure from 29th day6th month of life
2550 g 12 22.6 66 46.2 0 18 10.7 9 32.1 0.002
>50 g 26 49.1 65 45.5 2550 g 73 43.5 5 17.9
>50 g 77 45.8 14 50.0
Mean SD Mean SD
Mean SD Mean SD
Mother's age at 2nd trimester 27.7 3.9 27.8 3.3 ns
Maternal non-verbal intelligence (TONI-3) 107.0 17.9 110.6 17.3 ns Mother's age at 2nd trimester 27.8 3.3 27.3 4.0 ns
Birth weight [g] 3472.3 454.7 3399.7 447.7 ns Maternal non-verbal intelligence (TONI-3) 109.9 17.0 107.9 20.4 ns
Cord blood mercury level g/L 0.9 0.5 1.1 0.5 ns Birth weight [g] 3424.2 434.8 3389.6 537.8 ns
Cord blood lead level g/L 1.8 1.1 1.3 0.5 b0.001 Cord blood lead level [g/L] 1.3 0.5 2.3 1.2 b0.001
 D. Mrozek-Budzyn et al. / Neurotoxicology and Teratology 34 (2012) 592597 595

Table 3 4. Discussion
Association between neonatal ethylmercury exposure and BSID-II test scores up to
3 years of age (multivariate GLM and GEE models).
Vaccine clinical trials and surveillance of post-license TCV safety
Model 1 Model 2 conducted up to now have mostly concentrated on the short-term clini-
B (95%CI) p B (95%CI) p cal symptoms and diseases after immunization rather than ethylmercury
toxicity. In our study, we focused on the post-vaccination delayed side ef-
PDI
12th month 6.44 (10.0; 2.89) b0.001 5.78 (10.07; 1.50) 0.009 fects that may be clinically imperceptible. We found no relationship be-
24th month 5.89 (8.79; 2.99) b0.001 7.64 (11.21; 4.12) b0.001 tween neonatal TCV exposure and mental developmental delays in
36th month 0.98 (4.48; 2.53) ns 0.20 (4.01; 4.41) ns children up to 3 years of age. On the other hand, our results have
GEE modela 4.42 (6.94; 1.91) 0.001 4.28 (5.22; 0.04) 0.007 shown that ethylmercury is not completely harmless for the rst stage
MDI
of life and may be responsible for poorer outcomes of psychomotor de-
12th month 2.69 (0.53; 5.91) ns 3.77 (0.31; 7.85) 0.070 velopment in children. The lower psychomotor scores associated with
24th month 0.51 (4.42; 3.40) ns 1.17 (5.85; 3.51) ns TCV neonatal exposure were observed only during the rst 2 years of
36th month 2.10 (1.03; 5.24) ns 4.08 (0.15; 8.00) 0.042 life. That negative inuence disappeared by 36 months. Our results are
GEE modela 1.43 (1.04; 3.90) ns 1.43 (1.04; 3.90) ns
consistent with the observations of other authors (Andrews et al.,
Model 1: standardized to gender, exposure to Thimerosal from 28th day up to 6th 2004; Heron et al., 2004) who did not conrm the inuence of TCV on
month (number of doses), breastfeeding up to 6th month, mother's higher education, mental development, and one who reported its negative association
birth weight, Pb in cord blood (continuous), maternal TONI test and prenatal ETS.
Group: N= 196 (53 without Thimerosal, 143 with Thimerosal).
with the psychomotor development (Marques et al., 2009).
Model 2: additional standardization to Hg in cord blood. Group: N = 168 (27 without Simultaneously, the reports from some other studies indicated both
Thimerosal, 141 with Thimerosal). benecial and adverse effects of TCVs on children's neurodevelopment
a
Standardized additionally to age. (Drea, 2010). Inconsistent results of epidemiological studies may be
derived from the differences in the analyzed parameters, surveyed
populations, and control of different potential confounders. The greater
entire period were constructed. The risk of mild or signicant devel- the time elapsed between TCV exposure and neurodevelopment assess-
opmental delay both in PDI and MDI (scores b85) was not related ment, the greater the chances that the measured outcomes can be af-
to TCV exposure in any of the logistic models. The only variable asso- fected by confounding variables. Our observations derived from the
ciated with the analyzed risk over three years was maternal universi- early years of infant development compared to other studies, and we
ty education, which lowered the risk of MDI delay (OR = 0.28; p = found a harmful inuence of TCV only during the rst 2 years of life. A
0.009). variety of confounders at later stages of life may have an effect on the

Model 1: standarized to gender,

exposure to Thimerosal from 28th day
up to 6 months (number of thimerosal
doses), breastfeeding up to the 6th
month, mothers higher education,
maternal TONI scores, ETS, birth
weight, Pb in cord blood (as
continuous) Group: N=196 (53 without
Thimerosal, 143 with Thimerosal)

Model 2: additional standardization to

Hg in cord blood. Group: N=168 (27
without Thimerosal, 141 with
Thimerosal)

Fig. 2. Mean PDI scores during the rst three years of life in TCV exposed and non-exposed groups of children. Model 1: standardized to gender, exposure to Thimerosal from 28th
day up to 6 months (number of Thimerosal doses), breastfeeding up to the 6th month, mother's higher education, maternal TONI scores, ETS, birth weight, Pb in cord blood (as
continuous) Group: N = 196 (53 without Thimerosal, 143 with Thimerosal). Model 2: additional standardization to Hg in cord blood. Group: N= 168 (27 without Thimerosal,
141 with Thimerosal).
596 D. Mrozek-Budzyn et al. / Neurotoxicology and Teratology 34 (2012) 592597
association between TCV and psychomotor development and even
positively correlate with cognitive outcome at the age of 36 months.
Similar to our results, a transient neurodevelopmental sensitivity at
the 6th month of life related to the total mercury exposure (included
ethylmercury derived from TCV) was revealed in a group of breastfed
infants (Marques et al., 2009). In that study, observed postnatal
neurodevelopmental delays associated with TCV were overcome at
60 months (24 months later than in our cohort) and that effect was
positively correlated with the length of breastfeeding. Early infant nutri-
tion and growth are fundamental to neurodevelopment (Kramer,
2010). In our cohort, the breastfeeding rate at 6 months was signicant-
ly higher compared to children from other study populations (Kelly et
al., 2006). This could be a reason for overcoming the poorer psychomo-
tor development associated with TCV exposure at an earlier age (by the
age of 36 months), and for why there was no negative inuence on cog-
nitive outcomes. However, signicantly lower psychomotor outcomes
in early infancy that disappear with age cannot be attributed solely to
breastfeeding.
The development of children depends on multiple conditions that
have the potential to eliminate the harmful effect of neonatal
ethylmercury exposure (Drea, 2010). However, in the populations
with higher co-exposure to other neurotoxic substances, especially
when the ability to detoxify mercury has been compromised by addi-
tional circumstances related to mercury metabolism, even a subtle neg-
ative effect of TCV may indicate greater risk of developmental delays
(Custodio et al., 2005; Drea, 2009; Gundacker et al., 2007). It is recog-
nized that mercury may also increase the pathological inuence of
other neurotoxic metals. Most TCVs also contain aluminum adjuvants.
On the other hand, there is no evidence of harmful effects of these adju-
vants on a child's neurodevelopment, and neurotoxic effects were re-
vealed only in animal experiments (Petrik et al., 2007; Shaw and
Petrik, 2009). The binary mixture in TCVs has not yet been fully
addressed and we could not study it in our cohort because the vaccines
administered in our infants contained the same kind of aluminum adju-
vants. The animal studies undertaken to determine the maximum toler-
ated dose of Thimerosal revealed gender-selective toxicity in mice that
was much higher in males than in females (Branch, 2009). It has been
reported that estrogens are protective and testosterone exposure may
be a risk factor for neurotoxic effects of mercury (Ekstrand et al.,
2010). In our cohort, boys had lower psychomotor outcomes related
to TCV exposure compared to girls, but that difference was not statisti-
cally signicant. Further studies need to take into consideration a po-
tential gender selectivity of the effects of TCV.
In our study, we concentrated on newborns vaccinated against hep-
atitis B, mostly on the rst day of life. At this stage of life, newborns
greatly differ in birth weight, gestational age, maturity and many
other medical conditions that may have an impact on ethylmercury
toxicokinetics and toxicodynamics (Pichichero et al., 2008). The dose
of ethylmercury received with TCV (per kg of body mass) can be 4
times higher for newborns with very low birth weight compared to
those with normal weight. The rise of mercury concentrations after an
administration of TCV in an infant's blood seems to be faster and lasts
longer than in adults. Half of the blood samples in newborns vaccinated
with TCV against hepatitis B showed the mercury concentration above
safe limits of 5 ng/mL (12 h after exposure) (Pichichero et al., 2008).
According to Stajich et al. (2000), immunization with the hepatitis B
vaccine containing 12.5 g of ethylmercury resulted in an increase of
mean mercury blood level in preterm infants from a baseline level of
0.54 g/L to 7.36 g/L, whereas in term infants an increase from
0.04 g/L to 2.24 g/L was detected (23 days after exposure). Further-
more, the half-life of some drugs administered in the rst postnatal day
is signicantly associated with gestational age (Allegaert et al., 2004).
The same could be related to mercury compounds.
On the other hand, the short-term consequences (as measured by
the Gesell schedules at 6 months) of the time of TCV rst exposure (1
vs. 24 postnatal days) were not statistically signicant (Marques et
al., 2007). Since immaturity is not a contraindication to hepatitis-B
vaccination, nearly all newborns are vaccinated during the rst days
of life, independently of their birth weight and gestational age. In
the case of a subtle negative inuence of TCV on psychomotor devel-
opment of infants with an adequate birth weight, one can expect
more such effects in children with low or very low birth weight. In
our study, we had no possibility to address this issue because of the
insufcient number of such children in our cohort.
The plausibility of the harmful effect of TCVs on child development
was a sufcient argument to remove Thimerosal from all infant vac-
cines in the USA and the EU more than 10 years ago. Knowing that
children are exposed to mercury from many environmental sources,
public health ofcials should take measures to lower the number of
such sources as much as possible (WHO, 2006). The replacement of
TCVs by formulations without Thimerosal could prevent an occur-
rence of negative side effects, such as the delay in psychomotor devel-
opment. Moreover, it would improve immunization coverage by
increasing social acceptance of vaccinations.
The main strength of our study is the active follow-up of pregnant
mothers and their children with the possibility to control many im-
portant confounders potentially affecting infant development, espe-
cially the prenatal mercury (cord blood level) and post-neonatal
TCVs exposure (calculated as a dose of ethylmercury up to 6 months).
The important strength of our study is the high reliability of data
about infants' exposure to ethylmercury from TCVs because these
were taken directly from physicians' records. The groups of infants
under comparison did not signicantly differ with respect to most
of the potential confounders. There were also no ethnic differences
among children in our cohort. In comparison to previous studies, we
were able to evaluate the potential association between TCV neonatal
exposure and more subtle neurodevelopmental effects in early child-
hood that are not diagnosed by physicians. Previous studies mostly
used clinically diagnosed symptoms or the outcomes of neuropsycho-
logical tests that have been performed only in older children, many
years after vaccinations. Only Heron et al. (2004) assessed the ne
motor skills and tics in infants below 3 years of life. However, they fo-
cused on DTP vaccinations (administered in the post-neonatal peri-
od). Assuming that there could be a long latency of TCVs' inuence
on neurodevelopment, the ongoing follow-up gives us the possibility
to analyze similar relationships in the future with respect to the de-
velopment of older children. This study is the rst report regarding
the association of neonatal TCV exposure and early infant develop-
ment. Furthermore, our study is the rst one conducted in Eastern
Europe where children are still exposed to TCVs as single-dose vials
(technologically not requiring a preservative).
The limitation of this analysis is the study population, which may
not be sufciently representative of the general infant population be-
cause the enrollment covered only pregnant, nonsmoking women be-
tween 18 and 35 years with singleton pregnancies who were free
from selected chronic diseases. On the other hand, these criteria gave
us the possibility to exclude infants who were at higher risk for
neurodevelopment disorders due to maternal conditions. In addition,
the power to detect signicant associations was rather high. However,
the results of our analysis require conrmation in additional studies.
The results of our study showed that TCVs should be replaced by
Thimerosal-free formulas in countries that can afford it economically,
especially if it concerns single-dose vials. It is important to remove
Thimerosal from hepatitis-B vaccine for newborns. Despite the lack
of clinical evidence for developmental delays associated with neona-
tal TCV exposure in the rst 3 years of life, the risk of subtle negative
inuences on psychomotor development is plausible.
Conict of interest statement
All authors declare that they participated in the design, execution,
and analysis of the paper by entitled Neonatal exposure to Thimerosal
 D. Mrozek-Budzyn et al. / Neurotoxicology and Teratology 34 (2012) 592597
from vaccines and child development in the rst 3 years of life, that
they have seen and approved the nal version and that it has neither
been published nor submitted elsewhere. We also declare that we
have no conict of interest.
Acknowledgments
The study received funding from an R01 grant entitled Vulnerabil-
ity of the Fetus/Infant to PAH, PM2.5 and ETS (R01ES010165-01
NIEHS) and from the Lundin Foundation, the John and Wendy Neu
Foundation, and the Gladys T. and Roland Harriman Foundation. Princi-
pal investigator: Prof. FP Perera; co-investigator: Prof. W Jedrychowski.
References
Allegaert K, Van der Marel CD, Debeer A, Pluim MA, Van Lingen RA, Vanhole C, et al.
Pharmacokinetics of single dose intravenous propacetamol in neonates: effect of
gestational age. Arch Dis Child Fetal Neonatal Ed 2004;89:F258.
Andrews N, Miller E, Grant A, Stowe J, Osborne V, Taylor B. Thimerosal exposure in in-
fants and developmental disorders: a retrospective cohort study in the United
Kingdom does not support a causal association. Pediatrics 2004;114(3):58491.
Bayley N. Manual for the Bayley Scales of Infant Development. 2nd ed. San Antonio:
The Psychological Corporation; 1993.
Branch DR. Gender-selective toxicity of Thimerosal. Exp Toxicol Pathol 2009;61:1336.
Centers for Disease Control, Prevention. Recommended childhood and adolescent im-
munization schedule United States. Jan-Jun [online]. Available from URL:
http://www.cdc.gov/nip/recs/child-schedule.pdf 2004. [Accessed 2004 Jan 12].
Custodio HM, Harari R, Gerhardsson L, Skerfving S, Broberg K. Genetic inuences on the
retention of inorganic mercury. Arch Environ Health 2005;60:1723.
Drea JG. Breastfeeding is an essential complement to vaccination. Acta Paediatr
2009;98:124450.
Drea JG. Making sense of epidemiological studies of young children exposed to Thi-
merosal in vaccines. Clin Chim Acta 2010;411(2122):15806.
Ekstrand J, Nielsen JB, Havarinasab S, Zalups RK, Sderkvist P. Hultman P Mercury
toxicokineticsdependency on strain and gender. Toxicol Appl Pharmacol
2010;243(3):28391.
Grandjean P, Weihe P, White RF, Debes F, Araki S, Yokoyama K, et al. Cognitive decit in
7-year-old children with prenatal exposure to methylmercury. Neurotoxicol
Teratol 1997;19(6):41728.
Grandjean P, Satoh H, Murata K, Eto K. Adverse effects of methylmercury: environmen-
tal health research implications. Environ Health Perspect 2010;118(8):113745.
Gundacker C, Komarnicki G, Jagiello P, Gencikova A, Dahmen N, Wittmann KJ, et al.
Glutathione-S-transferase polymorphism, metallothionein expression, and mercu-
ry levels among students in Austria. Sci Total Environ 2007;385:3747.
Herdman ML, Marcelo A, Huang Y, Niles RM, Dhar S, Kiningham KK. Thimerosal in-
duces apoptosis in a neuroblastoma model via the cJun N-terminal kinase path-
way. Toxicol Sci 2006;92(1):24653.
Heron J, Golding J, ALSPAC Study Team. Thimerosal exposure in infants and develop-
mental disorders: a prospective cohort study in the United Kingdom does not sup-
port a causal association. Pediatrics 2004;114(3):57783.
Hewitson L, Houser LA, Stott C, Sackett G, Tomko JL, Atwood D, et al. Delayed acquisi-
tion of neonatal reexes in newborn primates receiving a Thimerosal-containing
hepatitis B vaccine: inuence of gestational age and birth weight. Toxicol Environ
Health A 2010;73(19):1298313.
Hornig M, Chian D, Lipkin WI. Neurotoxic effects of postnatal Thimerosal are mouse
strain dependent. Mol Psychiatry 2004;9(9):83345.
597
Jedrychowski W, Jankowski J, Flak E, Skarupa A, Mroz E, Sochacka-Tatara E, et al. Effects
of prenatal exposure to mercury on cognitive and psychomotor function in
one-year-old infants: epidemiologic cohort study in Poland. Ann Epidemiol
2006;16(6):43947.
Joint Statement issued by the AAP and the PHS in July 1999. Established the goal of re-
moving the vaccine preservative Thimerosal as soon as possible from vaccines rou-
tinely recommended for infants. from http://www.vaccinesafety.edu/AAFP-AAP-
ACIP-thimerosal.htm. [accessed 2012 Jan 12].
Kelly YJ, Watt RG, Nazroo JY. Racial/ethnic differences in breastfeeding initiation and
continuation in the United Kingdom and comparison with ndings in the United
States. Pediatrics 2006;118(5):e142835. [Nov].
Kramer MS. Long-term behavioral consequences of infant feeding: the limits of obser-
vational studies. J Pediatr 2010;156:5234.
Lederman SA, Jones RL, Caldwell KL, Rauh V, Sheets SE, Tang D, et al. Relation between
cord blood mercury levels and early child development in a World Trade Center
cohort. Environ Health Perspect 2008;116(8):108591.
Marques RC, Drea JG, Manzatto AG, Bernardi JV, Bastos WR, Malm O. Time of perinatal
immunization, Thimerosal exposure and neurodevelopment at 6 months in
breastfed infants. Acta Paediatr 2007;96:8648.
Marques RC, Drea JG, Bernardi JV, Bastos WR, Malm O. Pre- and post-natal mercury
exposure, breastfeeding and neurodevelopment during the rst ve years. Cogn
Behav Neurol 2009;22:13441.
Mutkus L, Aschner JL, Syversen T, Shanker G, Sonnewald U, Aschner M. In vitro uptake
of glutamate in GLAST- and GLT-1-transfected mutant CHO-K1 cells is inhibited by
the ethylmercury-containing preservative Thimerosal. Biol Trace Elem Res
2005;105(13):7186.
Oken E, Radesky JS, Wright RO, Bellinger DC, Amarasiriwardena CJ, Kleinman KP, et al.
Maternal sh intake during pregnancy, blood mercury levels, and child cognition
at age 3 years in a US cohort. Am J Epidemiol 2008;167(10):117181.
Petrik MS, Wong MC, Tabata RC, Garry RF, Shaw CA. Aluminum adjuvant linked to Gulf
War illness induces motor neuron death in mice. Neuromolecular Med 2007;9(1):
83-100.
Pichichero ME, Gentile A, Giglio N, Umido V, Clarkson T, Cernichiari E, et al. Mercury
levels in newborns and infants after receipt of Thimerosal-containing vaccines. Pe-
diatrics 2000;121(2):e20814.
Pichichero ME, Cernichiari E, Lopreiato J, Treanor J. Mercury concentrations and metab-
olism in infants receiving vaccines containing Thimerosal: a descriptive study. Lan-
cet 2002;360(9347):173741.
Pichichero ME, Gentile A, Giglio N, Umido V, Clarkson T, Cernichiari E, et al. Mercury
levels in newborns and infants after receipt of Thimerosal-containing vaccines. Pe-
diatrics 2008;121(2):e20814.
Shaw CA, Petrik MS. Aluminum hydroxide injections lead to motor decits and motor
neuron degeneration. J Inorg Biochem 2009;103(11):155562.
SPSS Advanced Statistics 15.0. SPSS Inc.; 2006
Stajich GV, Lopez GP, Harry SW, Sexson WR. Iatrogenic exposure to mercury after hep-
atitis B vaccination in preterm infants. J Pediatr 2000;136(5):67981.
The European Agency for the Evaluation of Medicinal Products, Committee for Proprietary
Medicinal Products. Points to consider on the reduction, elimination or substitution of
Thimerosal in vaccines. Apr 26; EMEA/CPMB/BWP2517/00 [on-line]. Available from
URL: http://www.emea.eu.int/pdfs/human/bwp/251700en.pdf 2001. [Accessed 2004
Jan 12].
WHO. Thiomersal as a vaccine preservative. Wkly Epidemiol Rec 2000;75(2):9-16.
WHO. Global Advisory Committee on Vaccine Safety Statement on Thiomersal. Available
from http://www.who.int/vaccine_safety/topics/thiomersal/statement_jul2006/en/
2006. [Accessed 2012 Jan 12].
Yel L, Brown LE, Su K, Gollapudi S, Gupta S. Thimerosal induces neuronal cell apoptosis
by causing cytochrome c and apoptosis-inducing factor release from mitochondria.
Int J Mol Med 2005;16(6):9717.