Epilepsy Research 66 (2005) 173–183

Serum thyroid hormone balance and lipid profile

in patients with epilepsy
Sherifa A. Hamed a,∗ , Enas A. Hamed b , Mahmoud R. Kandil a ,
Hala K. El-Shereef c , Moustafa M. Abdellah d , Hanan Omar e
aDepartment of Neurology, Assiut University Hospital, Assiut, Egypt
b Department of Physiology, Assiut University Hospital, Assiut, Egypt
c Department of Internal Medicine, Assiut University Hospital, Assiut, Egypt
d Department of Pharmacology, Assiut University Hospital, Assiut, Egypt
e Department of Clinical pathology, Assiut University Hospital, Assiut, Egypt

Received 2 March 2005; received in revised form 29 July 2005; accepted 10 August 2005

Abstract

Purpose : Patients with epilepsy may exhibit changes in thyroid hormone balance, lipids and lipoproteins concentrations. The
suggestion that lipid abnormalities are associated with subclinical thyroid dysfunction remains controversial. The aim of this
study was to analyze whether thyroid dysfunction encountered in patients with epilepsy would also be associated with abnormal
lipid profile.
Methods : Eighty-eight patients with epilepsy and 30 control subjects were included in the study. A fasting blood sample for
thyroid hormones, lipid profile and GGT determination was obtained.
Results : The serum levels of FT3 was elevated in 10.2% of patients, FT4 was low in 28.4%, TSH was high in 4.6% and low in
2.3%. 13.6% of patients had high TC, 17.1% had high LDL-c, 60.2% had marked reduction of HDL-c levels (P < 0.0001) and only
2.3% had high TG levels. Abnormalities were predominated in CBZ-treated patients. 27.3% patients with abnormal hormones
had abnormal lipid profile. Significant association was identified between the serum TC, LDL-c, TG, GGT and EIAEDs and
between the duration of illness and TG (r = −0.411; P = 0.017), and FT4 (r = −0.412; P = 0.018). HDL was higher in women
than men (r = 0.416; P < 0.002). However, changes in HDL-c levels associated neither with duration of illness, type or serum
levels of AEDs nor with age or degree of control on AEDs.
Conclusions : Our results support that (1) altered lipid metabolism might be associated but not solely influenced by thyroid
hormones and (2) enzyme induction is not the main or only reason for altered thyroid function or HDL-c among patients with

Abbreviations: AEDs, antiepileptic drugs; CBZ, carbamazepine; EIAEDs, enzyme inducer AEDs; FT3 , free triiodothyronine; FT4 , free thy-
roxine; GGT, gamma glutamyl transferase; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol; OCBZ,
oxcarbamazepine; PHT, phenytoin; TC, total cholesterol; TG, high triglyceride; TRH, thyroid hormone releasing hormone; TSH, thyroid stim-
ulating hormone; VPA, valproate
∗ Corresponding author. Present address: Neuroscience Department, Saudi German Hospital — Aseer, P.O. Box 2553, Khamis Mushayt,

Saudi Arabia. Tel.: +966 7 2355000x6567; fax: +966 7 2354500.

E-mail address: hamed sherifa@yahoo.com (S.A. Hamed).

0920-1211/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.eplepsyres.2005.08.004
174 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183
epilepsy. Hypothalamic/pituitary dysregulation by precisely mechanism caused by epilepsy itself or AEDs seems possible and
(3) it is important to recognize that patients with epilepsy are at great risk for atherosclerosis, hence monitoring and correction
of the culprit risks are mandatory.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Antiepileptic drugs; Thyroid hormones; Fasting lipid profile; Lipoproteins
1. Introduction
Epilepsy is a common chronic medical problem
and often requires long-term therapy, which may even
continues throughout life (Hauser et al., 1993). Many
previous studies reported altered thyroid function (par-
ticularly low FT4 ) among patients with epilepsy par-
ticularly during treatment with CBZ or other EIAEDs.
This has been also attributed to induction of the hep-
atic P-450 enzyme system and the consequent increase
in the metabolism of thyroid hormones (Connell et al.,
1984; Isojarvi et al., 1995). However, serum thyrotropin
concentrations do not change, and the clinical signif-
icance of low serum thyroid hormone concentrations
is unclear (Fichsel and Knpfle, 1978). The reports on
the effects of VPA on serum thyroid hormone concen-
trations were controversial (Fichsel and Knpfle, 1978;
Bentsen et al., 1983; Ericsson et al., 1985; Isojarvi et
al., 2001; Verrotti et al., 2001).
In addition, many previous studies demonstrated
that AEDs might alter the metabolism of lipids and
lipoproteins. The increased risk was noted among
patients receiving EIAEDs including; CBZ, PHT and
PB (Yoo and Hong, 1999; Yilmaz et al., 2001; Tumer
et al., 2002). The effect of AEDs on the serum level
of lipids and lipoproteins could be explained to a great
extent on the basis of the different hepatic microso-
mal enzyme biotransformation pathways of the AEDs
(Luoma et al., 1980, 1982). VPA, an enzyme inhibitor
AED, was reported to have little or no effect on fasting
lipid profile (Isojarvi et al., 2001).
The suggestion that lipid abnormalities in patients
with epilepsy are associated with subclinical thyroid
disturbance remains controversial. The aim of this
study was to determine whether or not clinically euthy-
roid patients with epilepsy with abnormal fasting lipid
profile could be characterized by common abnormal-
ities of thyroid hormones balance. For this purpose,
we studied the thyroid function and the fasting lipid
profile in a group of children and adult patients with
chronic epilepsy by measuring the serum levels of FT4 ,
FT3 , TSH, TC, TG, LDL-c and HDL-c. Untreated and
treated patients with epilepsy on long-term conven-
tional AEDs (CBZ, PHT and/or VPA) were included
for comparison. We correlated lipid abnormalities with
the subclinical thyroid dysfunction. In addition, we
correlated different biochemical parameters with age,
gender, duration of illness, type and serum levels of
AEDs utilized.
2. Subjects and methods
A total of 88 patients with epilepsy (children = 33
and adults = 55) and 30 healthy volunteers served as
controls, were included in this study. This study was
carried out at epilepsy out-patient clinic of Assiut
University Hospital, Assiut, Egypt. It was approved
by the regional Ethics Committee. Informed consent
was obtained from all the patients and control sub-
jects or their parents. All participants were subjected
to full medical and neurological history and exam-
ination. Seizures were analyzed determining seizure
type, duration of illness, age at onset, type and dura-
tion of AEDs therapy and the degree of response to
AEDs. Epilepsy type was classified according to the
recommendations of the International League Against
Epilepsy (Commission on Classification and Termi-
nology, 1989). Regarding the degree of control on
AEDs, patients were considered controlled on AEDs
treatment, when seizure free for >1 year, partially con-
trolled when seizure frequencies were occasional or
rare and uncontrolled when seizures were frequent or
very frequent. The frequency of seizures were divided
as follows: (a) very frequent: seizures occurring sev-
eral times a day or at intervals shorter than 7 days, (b)
frequent: seizures at intervals longer than 7 days but
shorter than 30 days, (c) occasional: seizures at inter-
vals longer than 30 days but shorter than 1 year and (d)
rare: seizures at intervals longer than 1 year. Patients
 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183
with diseases other than epilepsy or taking other regular
medication besides AEDs were excluded. AEDs were
described in recommended doses and according to the
well-known guidelines. All treated patients gave his-
tory of compliance to AEDs. The serum levels of AEDs
were determined in the therapeutic drug monitoring
(TDM) laboratory, Assiut University Hospital, Assiut
Egypt, using fluorescence polarization immunoassay
system of Abbott (EPIA) using TDxFLX appara-
tus (Abbott Lab, Wiesbaden, Germany), as described
before (Goma et al., 2004) and they were measured as
part of the investigation in batched assays.
Routine laboratory blood tests were done for all
patients, including complete blood count, random
blood sugar, kidney and liver function tests. After an
overnight fast and patients were seizure free for at least
72 h, as any postictal central hormonal dysfunction is
recognized to reverse within hours (Pritchard, 1980),
blood samples were drawn at (8:00–10:00 a.m.) for the
analysis of serum levels of AEDs, FT3 , FT4 , TSH, TC,
TG, HDL-c and LDL-c. Serum samples were left to
coagulate at room temperature, centrifuged and frozen
at −20 ◦ C until utilized. TC, TG, HDL-c, LDL-c and
GGT were measured by enzymatic colourimetric meth-
ods (Roche).
For all patient and control subjects, the serum
concentrations of FT4 , FT3 and TSH were measured
as described by the manufacturer by immunoenzy-
metric assay kits (IMMULITE reproductive hormone
assays). The kits were obtained from DPC or Diag-
nostic Products Corporation, Los Angeles, USA
(Diagnostic Product Corporation, 1987; Witherspoon
et al., 1988; Babson, 1991). Samples were assayed
in duplicate. For FT3 , the analytical sensitivity of the
assay was 1.0 pg/ml. Within run, the mean precision
values were 1.7 ± 0.24 (CV = 14.1%), 2.9 ± 0.32
(CV = 11.0%), 3.7 ± 0.31 (CV = 8.4%), 4.6 ± 0.34
(CV = 7.4%), 5.6 ± 0.34 (CV = 6.1%) and 10.2 ± 0.59
(CV = 5.8%). For FT4 , the analytical sensitivity of
the assay was 0.15 ng/ml. Within run, the mean
precision values were 0.65 ± 0.044 (CV = 6.8%),
1.33 ± 0.077 (CV = 5.8%), 2.37 ± 0.097 (CV = 4.1%),
4.31 ± 0.188 (CV = 4.4%) and 5.02 ± 0.175
(CV = 3.5%). For TSH, the analytical sensitivity
was 0.01 ␮IU/ml. The intraassay precisions were:
1.3 ± 0.10 (CV = 7.7%), 1.4 ± 0.077 (CV = 0.10%),
4.8 ± 0.20 (CV = 4.2%), 7.6 ± 0.55 (CV = 7.2%),
24.2 ± 1.18 (CV = 4.9%) and 25.7 ± 1.71 (CV = 6.7%).
175
The interassay precisions were: 0.015 ± 0.004
(CV = 27%), 0.035 ± 0.006 (CV = 17%), 099 ± 0.009
(CV = 9.1%), 0.395 ± 0.36 (CV = 9.1%), 1.28 ± 0.134
(CV = 10.5%), 3.93 ± 0.246 (CV = 6.3%) and
24.5 ± 1.26 (CV = 5.1%).
3. Statistical analysis
Data are expressed as means ± 2S.D. Statistical
comparison among different groups was evaluated
using ANOVA tests. The logistic t-test was used to
evaluate the difference in the descriptive data of the
patient and the control group. Adjustment for differ-
ences in baseline covariates and changes in variables
during study were performed by analysis of covariance
using general linear models. Moreover, Pearson’s and
Spearman’s correlation was performed between differ-
ent biochemical parameters. Chi-square was utilized
for comparison of binomial data. Calculations were
done with the statistical package SPSS for windows,
version 10.0. Statistical significance was defined as
P < 0.05.
4. Results
The study series comprised a total of 88 patients
with primary epilepsy (children = 33 and adults = 55).
Children group of patients aged 7–8 years (mean;
13.1 ± 4.2 years) with male to female ratio 1:2.3
(M/F = 10/23) and duration of illness ranged from 0.5 to
15 years (mean; 4.7 ± 3.7 years). Twenty-five (75.8%)
children had generalized tonic–clonic convulsions
(GTC), 24.2% had other types of epilepsy (absence = 3,
psychomotor = 1, myoclonic = 1, mixed = 1 and focal
motor = 1). Eighteen patients were CBZ-treated, nine
were VPA-treated and one was PHT-treated. Regard-
ing the degree of seizure control on AEDs: 13 patients
(39.4%) were seizure free for ≥1 year (controlled),
11 (33.3%) had less frequent and occasional or rare
seizures (partially controlled) and 5 (15.2%) had very
frequent seizures (uncontrolled).
Adult group of patients aged 19–55 years (mean;
28.2 ± 8.7 years) with male to female ratio 3.2
(M/F = 34/21) and duration of illness ranged from
0.5 to 27 years (mean; 12.15 ± 6.7 years). Forty-five
(81.8%) patients had GTC, 18.2% had other types of
176 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183

epilepsy (focal motor = 5, mixed = 2, myoclonic = 2 and
psychomotor = 1). Twenty-four patients were CBZ-
treated, 11 were VPA-treated, one was PHT-treated
and 11 patients were on polytherapy with more than
one AEDs. Fifteen patients (27.3%) were seizure free
for ≥1 year (controlled), 20 (36.4%) had less frequent
and occasional or rare seizures (partially controlled)
and 12 (21.8%) had very frequent seizures (uncon-
trolled). Untreated group of patients with epilepsy par-
ticipated in this study comprised a total of 13 patients
(children = 5 and adults = 8). All patients included
in this study had drug concentrations at the refer-
ence therapeutic level (CBZ is 4–10 ␮g/ml; VPA is
50–100 ␮g/ml; PHT is 10–20 mg/l). Demographic data
of the studied group are illustrated in Table 1.
All patients were clinically euthyroid. Compared
to controls, nine patients (10.2%) (children = 4 and
adults = 5) had mild elevation in the serum level of
FT3 (children: 3.08 ± 0.71 pg/ml; P < 0.01 and adults:
2.90 ± 0.81; P < 0.05). Only one patient had low FT3
level (1.5 pg/ml). Twenty-five patients (28.4%) (chil-
dren = 8 and adults = 17) had significant reduction
in the serum level FT4 (children: 0.99 ± 0.12 ng/dl;
P < 0.01 and adults: 0.98 ± 0.28 ng/dl; P < 0.01). Four
patients (4.6%) (children = 1 and adults = 2) had ele-
vated TSH level (4.5, 10.0 and 11.0 ␮IU). Three
patients (3.4%) (children = 1 and adults = 2) had
reduced TSH levels (0.01–0.37 ␮IU) (Tables 2 and 4).
Among, one PHT-treated child had reduced level of
FT4 (0.55 pg/ml) and elevated TSH (4.5 ␮IU), one
VPA-treated adult had marked reduction in the serum
level of FT3 (1.5 pg/ml) and level FT4 at the lower
border of normal 0.80 ng/dl. One CBZ-treated adult
had FT3 level at upper level of normal (4.1 pg/ml)
with mild elevation in TSH level (4.4 ␮IU). Changes
were predominated among the treated group of patients
with epilepsy compared to the untreated group, par-
ticularly the CBZ-treated (FT3 : 3.05 ± 0.68 for CBZ
versus 2.84 ± 0.77 for VPA, 2.64 ± 0.65 for poly-
therapy and 3.14 ± 0.70 for the untreated; P < 0.05)
(FT4 : 0.93 ± 0.24 for CBZ versus 1.14 ± 0.0.30 for
VPA, 0.95 ± 0.23 for polytherapy and 0.99 ± 0.19 for
the untreated; P < 0.05). Although remained within
the normal levels range, levels of TSH were lower

Table 1
Demographic data of the studied group of patients
Demographic data (N = 88) Children (N = 33) Adult (N = 55)
Patients’ age (years) 7–18 (13.1 ± 4.2) 19–55 (28.2 ± 8.7)
Male/female 10/23 34/21
Age at onset (years) 1–12 (4.3 ± 3.9) 4–50 (28.2 ± 5.8)
Duration of illness (years) 0.5–15 (4.7 ± 3.7) 0.5–27 (12.2 ± 8.1)
Type of epilepsy
Primary generalized; # (%) 30 (34.1) 49 (55.7)
Partial; # (%) 3 (3.4) 6 (6.8)
Treated/untreated; # (%) 28/5 47/8
AEDs utilized; # (%)
VPA 9 (10.2) 11 (12.5)
CBZ 18 (20.5) 24 (27.3)
PHT 1 (1.1) 1 (1.1)
Polytherapy 0 11 (12.5)
Drug dose (mg/d)
VPA 400–800 (605.6 ± 160.7) 400–1200 (736.4 ± 283.8)
CBZ 200–800 (572.2 ± 187.3) 600–1200 (787.5 ± 227.1)
PHT 300 600
Degree of control on AEDs; # (%)
Controlled (seizure free for >2 years) 13 (14.8) 15 (17.1)
Partially controlled (occasional and rare) 11 (12.5) 20 (22.7)
Uncontrolled (frequent and very frequent) 5 (5.7) 11 (12.5)

Data are expressed as # (%).

 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183 177

Table 2 Table 3
Serum thyroid hormone and thyrotropin concentrations in patient Abnormal serum fasting lipid profile in the studied group of patients
with epilepsy with epilepsy
Children Adult Children Adult
FT3 TC
Range (pg/ml) 1.90–4.50 1.5–4.9 Range (mg/dl) 64–213 73–376
Mean ± S.D. 3.08 ± 0.71** 2.90 ± 0.81* Mean ± S.D. 138.60 ± 32.28* 162.59 ± 55.63***
Normal Normal
Range 1.90–3.80 1.5–4.1 Range 64–198 73–200
Mean ± S.D. 2.91 ± 0.56 2.84 ± 0.63 Mean ± S.D. 134.10 ± 31.30 142.54 ± 34.38
FT4 TG
Range (ng/dl) 0.49–1.53 0.47–1.82 Range (mg/dl) 55–194 54–222
Mean ± S.D. 0.99 ± 0.12** 0.98 ± 0.28** Mean ± S.D. 116.1 ± 037.2 114.6 ± 43.8
Normal Normal
Range 0.82–1.73 0.81–2.00 Range 45–194 50–200
Mean ± S.D. 1.06 ± 0.29 0.39 ± 1.10 Mean ± S.D. 111.1 ± 42.7 113.1 ± 44.8
TSH HDL-c
Range (␮IU/ml) 0.39–4.5 0.49–11.00 Range (mg/dl) 10–64 5–63
Mean ± S.D. 2.11 ± 1.02 2.45 ± 1.68 Mean ± S.D. 30.33 ± 10.29*** 25.39 ± 6.06***
Normal Normal
Range 0.9–3.66 0.4–4 Range 35–64 35–63
Mean ± S.D. 2.02 ± 1.42 2.23 ± 1.77 Mean ± S.D. 43.60 ± 8.3 44.40 ± 8.64
∗ P < 0.05. LDL-c
∗∗ P < 0.01. Range (mg/dl) 32–167 38–329
Mean ± S.D. 89.79 ± 32.28*** 109.90 ± 50.56***
Normal
in the CBZ-treated groups of patients (monotherapy Range 32–124 38–125
and polytherapy) compared to others (1.61 ± 0.93 for Mean ± S.D. 81.31 ± 23.60 88.97 ± 25.06
CBZ versus 2.50 ± 2.88 for VPA, 2.11 ± 2.08 for * P < 0.05.
PHT, 1.00 ± 0.57 for polytherapy and 1.24 ± 0.48 for ∗∗ P < 0.01.
untreated group). We noted enlargement of the thy- ∗∗∗ P < 0.001.

roid gland and presence of thyroid cysts in 20% of

CBZ-treated patients (data not shown). No significant
difference was identified between adult and children
levels.
Compared to controls, 12 patients (13.6%)
(children = 2 and adults = 10) had higher level of
TC (children: 138.60 ± 32.28 mg/dl; P < 0.05 and
adults: 162.59 ± 55.63; P < 0.001). Only two adult
patients had elevated level of TG (222 mg/dl).
Fifty-three (60.2%) (children = 23 and adults = 30)
patients had marked reduction of HDL-c levels
(children: 30.33 ± 10.29 mg/dl; P < 0.001 and adults:
25.39 ± 6.06; P < 0.001). Fifteen patients (17.1%) had
high levels of LDL-c (children = 4 and adults = 11)
(children: 89.79 ± 32.28 mg/dl; P < 0.001 and adults:
109.90 ± 50.56; P < 0.001). Among were 10 (11.4%)
adult patients (VPA-treated = 7 and untreated = 3) had
reduced levels of both HDL-c and LDL-c (range:
17–30 mg/dl and 32–55 mg/dl, respectively). All 12
patients with total cholesterol levels had also elevated
levels of LDL-c. Only three CBZ-treated patients had
elevated levels of total cholesterol, LDL-c and low
levels of HDL-c (Tables 3 and 4). Changes were pre-
dominated among the treated group of patients with
epilepsy compared to the untreated group, particularly
the CBZ-treated (TC: 158.74 ± 55.75 for CBZ ver-
sus 141.47 ± 40.09 for VPA, 166.14 ± 50.03 for poly-
therapy, 157.67 ± 40.00 for PHT and 144.23 ± 43.03
for the untreated; P < 0.05) (TG: 102.93 ± 59.45 for
CBZ versus 89.95 ± 43.47 for VPA, 95.14 ± 19.99 for
polytherapy, 85.33 ± 26.54 for PHT and 82.31 ± 25.58
for the untreated). While elevated LDL-c was pre-
dominated among polytherapy and PHT-treated group:
114.57 ± 45.84 for polytherapy and 104.67 ± 39.43 for
PHT versus 95.47 ± 31.35 for VPA, 53.34 ± 14.12 for
178 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183

Table 4
Number of patients with epilepsy having abnormal serum thyroid hormone balance and fasting lipid profile in relation to different AEDs utilized
N = 88 VPA (n = 20) CBZ (n = 42) PHT (n = 2) Polytherapy (n = 11) Untreated (n = 13)
FT3 (n = 10, 11.4%) 2 (high), 1 (low) (5%) 7 (high) (16.7%) 0 0 0
FT4 (n = 25, 28.4%) 3 (low) (15%) 19 (low) (45.2%) 1 (low) (50%) 0 3 (low) (23.1%)
TSH (n = 7, 8%) 1 (low) (0.05%) 3 (high) 2 (low) (11.9%) 1 (high) (50%) 0 0
TC (high) (n = 12) 0 8 (19.1%) 0 1 (9.1%) 3 (23.1%)
TG (high) (n = 2) 2 (10%) 0 0 0 0
HDL-c (low) (n = 53) 16 (80%) 30 (71.4%) 1 (50%) 1 (9.1%) 5 (38.5%)
LDL-c (high) (n = 19) 1 (5%) 14 (33.3%) 1 (50%) 1 (9.1%) 2 (15.4%)

CBZ and 97.31 ± 37.44 for the untreated. However,
reduced HDL-c was predominated in all groups of
patients (33.60 ± 11.64 for CBZ, 28.05 ± 13.22 for
VPA, 36.00 ± 3.61 for PHT, 32.43 ± 12.29 for poly-
therapy and 30.54 ± 9.17 for untreated group).
Changes were also predominated among the
controlled treated group of patients with epilepsy
(Table 5). Twenty-four patients (27.3%) (children = 10
and adults = 14) with abnormal thyroid hormone levels
had in addition abnormal fasting lipid profile. Although
did not reach a significant level, inverse correlation was
identified between FT4 and TC (r = −0.113; P = 0.540),
TG (r = −0.195; P = 0.284), LDL-c (r = −0.048;
P = 0.793) and HDL-c (r = −0.018; P = 0.899). Signifi-
cant positive inverse correlation was identified between
HDL-c and TC (r = −0.391; P = 0.0001) while no
significant correlation was identified between HDL-c
and FT4 (r = −0.018; P = 0.899). Low levels of GGT,
which were used as markers of liver enzyme induction,
was identified in 12 (13.6%) patients among which 7
(58.3%) were VPA-treated and 3 (25%) on polytherapy
with VPA while high levels were identified among 19
(21.6%) patients among which 8 (43.4%) were CBZ-
treated and 4 (21.1%) on polytherapy with CBZ. No
correlation was noted between thyroid hormone con-
centrations and GGT levels, but significant correla-
tion was identified between the serum TC (r = −0.435;
P = 0.006), LDL-c (r = −0.611; P = 0.002) and TG
(r = −0.513; P = 0.006), GGT (r = −0.821; P = 0.001)
and EIAEDs (CBZ and PHT utilized as monotherapy
or part of polytherapy AED treatment). Significant pos-
itive inverse association was identified between the
duration of illness and TG (r = −0.411; P = 0.018),
while non-significant inverse association was existed
between the duration of illness and FT4 (r = −0.112;
P = 0.420). In addition, changes in the serum thy-
roid hormones and HDL-c levels associated neither
with duration of therapy, type of antiepileptic drugs
(AEDs) utilized or plasma AEDs levels nor age but
gender difference has been demonstrated with signif-
icant increase of HDL among women and not men
(r = 0.416; P < 0.002). Changes in other serum lipids
associated with neither duration of therapy or plasma
antiepileptic drug levels nor age or gender.
5. Discussion
In this study and consistent with many others, 28.4%
of patients with epilepsy had serum FT4 level below

Table 5
Patients with epilepsy with alteration of serum thyroid hormones and fasting lipid profile: Relationship degree of control on AEDs
Degree of control on AEDs Controlled (n = 28) Partially controlled (n = 31) Uncontrolled (n = 16) Untreated (n = 13)
FT3 (n = 10) 6 (21.4%) 1 (3.2%) 1 (6.3%) 2 (15.4%)
FT4 (n = 25) 10 (35.7%) 11 (35.5%) 2 (12.5%) 2 (15.4%)
TSH (n = 2) 4 (14.3%) 2 (6.5%) 1 (6.3%) 0
TC (n = 12) 6 (21.4%) 4 (12.9%) 2 (12.5%) 0
TG (n = 2) 0 0 0 0
HDL-c (n = 53) 15 (53.6%) 25 (80.7%) 6 (37.5%) 7 (53.9%)
LDL-c (n = 19) 8 (28.6%) 6 (19.4%) 3 (18.8%) 2 (15.4%)

Data are expressed as # (%).

 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183
the reference control range. The risk was dominated
among CBZ-treated patients with epilepsy (76% or
19/25). However, TSH concentrations were not sig-
nificantly elevated in patients with low serum thyroid
hormone concentrations (Table 2) (Isojarvi et al., 1989,
1990, 2001). The altered thyroid function during CBZ
medication has been attributed to induction of the hep-
atic P-450 enzyme system and the consequent increase
in the metabolism of thyroid hormones (Connell et
al., 1984; Isojarvi et al., 1995). We noted enlarge-
ment of the thyroid gland and presence of thyroid cysts
20% of CBZ-treated patients (data not shown). This
is consistent with many studies. Enlargement of thy-
roid gland has been reported in patients taking either
CBZ or PHT for epilepsy, and it was suggested to be
compensatory due to low serum thyroid hormone con-
centrations (Hegedus et al., 1985). However, data have
demonstrated that the changes induced by these drugs
are transient (Strandjord, 1981; Verrotti et al., 1998,
2001).
In consistent with many studies, only 15% of our
VPA-treated patients demonstrated low FT4 levels
below reference values (Isojarvi et al., 2001). In ear-
lier reports, the effects of VPA on serum thyroid
hormone levels have been inconsistent and conflict-
ing (Liewendahl et al., 1978; Bentsen et al., 1983;
Ericsson et al., 1985; Eiris-Punal et al., 1999; Iso-
jarvi et al., 1990, 2001). Recently, Isojarvi et al.
(2001) and Verrotti et al. (2001) did not find any
alterations in thyroid hormone metabolism in VPA-
treated patients with epilepsy, and the TSH response to
thyroid-releasing hormone also was normal. Slightly
increased serum TSH levels and normal serum thyroid
hormone concentrations have previously been reported
in adults taking VPA for epilepsy (Isojarvi et al.,
1990).
Our results and that of others highly suggest the
presence of mechanisms other that enzyme induc-
tion that may contribute to thyroid hormone dysfunc-
tion among patients with epilepsy. In support: (1) we
reported normal levels of TSH in CBZ-treated patients
despite low FT4 levels. Isojarvi et al. (1990) reported
normal TSH responses to TRH in CBZ-treated patients
with epilepsy, and therefore pituitary interference by
the drug seems unlikely, (2) we reported no correla-
tion between serum thyroid hormone concentrations
and GGT levels, which were used as markers of liver
enzyme induction and this is consistent with others
179
(Isojarvi et al., 1994), (3) although, initial studies
demonstrated that oxcarbamazepine or OCBZ does
not appear to induce the hepatic P-450 enzyme sys-
tem (Larkin et al., 1991) and replacing CBZ with
OCBZ resulted in deinduction of liver enzyme lev-
els and short-term restoration of normal endocrine
function in men with epilepsy (Isojarvi et al., 1995),
however, Isojarvi et al. (2001) reported that both CBZ
and OCBZ (in high or low doses) induce a decrease
in serum T4 and FT4 concentrations, which did not
correlate with serum GGT levels. OCBZ is mainly
reduced instead of being oxidized as in CBZ and it
may induce the hepatic P-450 enzyme system when
given in high doses (Klosterskov Jensen et al., 1992).
The authors suggested that CBZ and OCBZ are most
likely decrease serum thyroid hormone concentra-
tions through a mechanism other than hepatic P-450
enzyme system induction and consequent accelerated
metabolism of thyroid hormones. Hypothalamic inter-
ference of regulation of thyroid hormone production
by the drugs seems possible, (4) Surks and DeFesi
(1996) showed that patients taking CBZ medication
had low FT4 and FT3 levels when serum samples
were analyzed by a commercial procedure from diluted
serum, but not when an ultrafiltration method from
undiluted serum was used. The authors concluded that
their results further support the view that hepatic P450
enzyme induction is not the main or the only reason
for decreased thyroid hormone concentrations during
CBZ treatment, (5) Isojarvi et al. (2001) demonstrated
that despite the increased levels of thyroid peroxi-
dase (TPO)-ab and/or thyroglobulin (TG)-ab in CBZ-
and OCBZ-treated patients, these changes were not
associated with altered serum thyroid hormone con-
centrations and their concentration were normal in
VPA-treated men, suggesting that immunologic mech-
anism seems unlikely and (6) Villa and Alexander,
1987 suggested that inhibit iodine uptake by the thy-
roid gland might be contributed as one of the mech-
anisms by which CBZ can induce thyroid dysfunc-
tion.
In this study, we reported significantly elevated TC,
TG and LDL-c levels in 13.6, 2.3 and 21.6% of our
studies patients, respectively. Many previous studies
demonstrated that AEDs are able to modify lipid pro-
file. However, little is known about the precise nature
and underlying mechanisms of such changes. High lev-
els of LDL-c and as well as TG have been reported
180 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183
in treated patients (Calandre et al., 1991; Isojarvi et
al., 1994; Eiris et al., 1995). High serum TC concen-
trations have been reported among patients receiving
CBZ, PHT and PB (Franzoni et al., 1992). In con-
trast, lower values of TC, HDL-c and LDL-c similar
to healthy controls have been found in patients with
epilepsy receiving VPA (Calandre et al., 1991; Fran-
zoni et al., 1992; Eiris et al., 1995). CBZ, PHT and
PB are principally metabolized in the hepatic P-450
system. This enzyme system also catalyzes the transfor-
mation of the cholesterol in biliary acids. Thus, chronic
treatment with these drugs might compete with choles-
terol in the utilization of those enzymes, and this leads
to reduction in the transformation of cholesterol in
bile acids with increased total serum cholesterol level.
Previous studies have suggested direct links running
from the serum anticonvulsant levels to the extent of
hepatic microsomal enzyme induction, and further to
the plasma HDL-c concentrations (Luoma et al., 1980,
1982). On the other hand, the main route of VPA trans-
formation in humans is represented by glucuronidation
(Rettenmeier et al., 1989).
Interestingly, in this study, we reported higher per-
centage of patients with epilepsy with significantly
low HDL-c (60.2%) (P < 0.001). For our knowledge,
this is the first report of low levels of HDL-c among
various groups of patients with epilepsy (treated and
untreated), with different ages and despite the type of
AEDs utilized or the degree of control on AEDs. In
contrast, previous studies reported increase in HDL in
CBZ-treated patients with epilepsy (Yoo and Hong,
1999; Yilmaz et al., 2001; Tumer et al., 2002). Pre-
vious studies reported that CBZ is associated with
increase in HDL2 lipoprotein cholesterol, which results
in an increased hepatic synthesis of apolipoprotein
A1, the major component of HDL-c (Luoma et al.,
1980, 1982), whereas PHT is specifically associated
with increase of TG (Pita-Calandre et al., 1998). High
levels of HDL-c have been reported in normal sub-
jects receiving PB (Eiris et al., 2000). However, in
contrast, few studies reported that CBZ therapy does
not affect serum lipids (TG, TC, HDL-c and LDL-c)
(Deda et al., 2003). However, Lower values of TC,
HDL-c and LDL-c similar to healthy controls have
been found in patients with epilepsy receiving VPA
(Franzoni et al., 1992; Eiris et al., 1995). Gender dif-
ference has also been demonstrated with significant
increase of HDL among women and not men and
this is consistent with previous studies. Sex differ-
ence in serum lipoprotein concentrations (i.e., higher
LDL cholesterol and TG but lower HDL cholesterol in
men) is independent of the dose of CBZ and plasma
concentrations and considered to be correlated with
the previously reported diminished rate of death from
coronary heart disease in patients with epilepsy as
HDL exerts an antiatherogenic effect (Sudhop et al.,
1999). Our finding of low HDL-c among patients with
epilepsy is highly supporting the presence of mech-
anisms other that enzyme induction that may con-
tribute to abnormal lipid profile among patients with
epilepsy.
In this study, inverse correlation was identified
between TG and HDL-c levels (although did not reach
a significant level). Previous studies have demonstrated
that endogenous TG metabolism influences HDL-c
level in patients with epilepsy undergoing anticonvul-
sant treatment. Patients with epilepsy with elevated
TG or a combination of elevated TG and LDL-c have
been found to have lower plasma HDL-c level than
normolipidemic patients with epilepsy (Luoma et al.,
1980).
From this study and that reported by others, we
suggested that thyroid hormone balance might be
associated with characterized pattern of abnormal
fasting lipid profile among patients with epilepsy
which is not solely explained by the enzyme induction
mechanism of AEDs. Whether the increase in lipids
and lipoproteins in the treated group of patients with
epilepsy is mediated through effects of epilepsy
per se or the effect of AEDs on thyroid hormones
still has to be clarified. In support: (1) a significant
number of patients with abnormal lipid profile was
found to have abnormal thyroid hormone levels; (2)
thyroid hormones are well known to be involved in
the regulation of lipoprotein metabolism, inducing
significant changes in the concentration, size and
composition of plasma HDL (Huesca-Gomez et
al., 2002; Bramswig et al., 2002). Many studies
suggested that abnormal levels of plasma HDL
commonly reflect altered metabolism of the major
HDL-apolipoprotein A–I (apo A–I) (Huesca-Gomez et
al., 2002). From our results, there seems to be a high
possibility that thyroid hormones may influence HDL
through apo A–I, (3) many studies have confirmed that
hypopituitarism is associated with deranged plasma
lipid levels with lower HDL cholesterol and higher TG
 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183
level (Rosen et al., 1993; Beshyah et al., 1995; Mon-
son et al., 2000); (4) Halle et al. (1999) suggested that in
men with low HDL-c levels had mild to moderately ele-
vated serum TG which strongly suggest the presence of
other metabolic cardiovascular risk factors and in par-
ticular of a more atherogenic LDL subfraction profile
of increased concentration of small, dense LDL parti-
cles that are depleted in surface lipids; (5) many studies
suggested that CBZ seems not to influence endogenous
cholesterol synthesis or intestinal absorption directly,
neither also relates to increased ApoB production nor
to decreased catabolism but it changes the conversion
cascade of intermediate density lipoprotein (IDL)
particles by most likely indirect effect through a
decrease in thyroid hormones (Strandjord et al., 1981;
Isojarvi et al., 2001; Verrotti et al., 2001; Bramswig
et al., 2002); (6) in general, a higher prevalence of
subclinical hypothyroidism in a population with hyper-
cholesterolaemia was determined when compared to a
population with normal cholesterol levels (Mishkel and
Crowther, 1977); (7) many studies demonstrated a
significant reduction in both total and LDL cholesterol
concentrations has been reported after administration
of thyroxine in a small group of hypercholesterolaemic
patients with basal TSH levels in the upper range of
normal values (Deschampheleire et al., 1999; Canturk
et al., 2003); (8) many experimental studies demon-
strated that in ponies, similarly as reported in rats and
pigs, HDL cholesterol ester can be transferred to LDL
despite the absence of plasma cholesteryl ester transfer
protein (CETP) activity and hepatic lipase (HL), both
are important proteins that modulate the metabolism
of HDL, and that the magnitude of this transfer is
related to the levels of HDL CE (cholesterol ester) as
induced by the amount of fat in the diet (Geelen et al.,
2001). And hence high amount of fat in diet increases
the conversion of HDL into LDL.
However and despite all the above evidences, altered
lipid metabolism might not be solely influence by thy-
roid hormones. It should be kept in mind that subclin-
ical hypothyroidism is a relatively common condition
with incidence between 3 and 7% in the general pop-
ulation (Spacilova et al., 2003). The prevalence of the
disease is approximately 20% (Weissel, 2003). Hence,
probably, such frequency rates for developing subclin-
ical hypothyroidism might be increased among the
patients with epilepsy group of population, which is
also common.
181
The above information alerts the neurologists
for the high atherosclerotic risk exposed by patients
with epilepsy (reviewed in Hamed and Nabeshima,
2005). Epidemiologic studies have shown that dys-
lipoproteinemia with low concentrations of HDL-c
and elevated serum TG and LDL-c is associated with a
particularly high incidence of coronary artery disease
(Caligiuri et al., 1999; Artenie et al., 2003; Jouko
et al., 2004). HDL-c is a strong inverse predictor of
atherosclerosis risk (Halle et al., 1999). The protective
effects of HDL-c have been attributed to its role
in reverse cholesterol transport and its effects on
endothelial cells. Several reports have suggested that
HDL has anti-oxidative actions (Caligiuri et al., 1999;
Halle et al., 1999). Recent data indicate that subclinical
hypothyroidism is also associated with enhanced risk
for arteriosclerotic cardiovascular disease, which
could be attributed to multiple mechanisms such as
hyperlipidemia, hypercoagulable state, direct effects
on vascular smooth muscle or endothelial (Caparevic
et al., 2003). Hence, the risk of atherosclerosis
might increase among patients with epilepsy with
subclinical hypothyroidism which are more liable for
developing hypercholesterolaemia and dyslipidemia
as demonstrated above. Many prospective studies have
demonstrated that a multi-ingredient vitamin formula
(Vitamin B6, Vitamin B12, folate, Vitamin C, Vitamin
E and beta-carotene) with anti-oxidant properties has
measurable effects LDL-c oxidation indices (Earnest
et al., 2003).
6. Conclusions
From our results, we concluded that: (1) altered
lipid metabolism might be associated but not solely
influenced by thyroid hormones, (2) our findings
support the notion that serum thyroid and HDL-c
concentrations in patients taking EIAEDs are altered
by some mechanism other than liver enzyme induction.
Therefore, interference with hypothalamic/pituitary
regulation by epilepsy itself or unknown complex cen-
tral regulating mechanism of thyroid function by AEDs
seems possible, (3) our results suggest that the effect of
long-term AED therapy on lipid profile increases the
risk of atherosclerosis-related disease, hence, a need
for monitoring serum thyroid hormones and fasting
lipid profile particularly in patients receiving EIAEDs.
182 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183
References
Artenie, R., Ungureanu, D., Artenie, A., Botnariu, G., Anisie, E.,
2003. HDL-cholesterol—active or passive participant in the
pathogenesis of atherosclerosis. Rev. Med. Chir. Soc. Med. Nat.
Iasi 107 (2), 282–287.
Babson, A.L., 1991. The IMMULITE 2000 Automated Immunoas-
say System. J. Clin. Immunoassay 14, 83–88.
Bentsen, K.D., Gram, L., Veje, A., 1983. Serum thyroid hormones
and blood folic acid during monotherapy with carbamazepine or
valproate. Acta Neurol. Scand. 67, 235–241.
Beshyah, S.A., Henderson, A., Niththyananthan, R., Skinner, E.,
Anyaoku, V., Richmond, W., Sharp, P., Johnston, D.G., 1995.
The effects of short and long-term growth hormone replacement
therapy in hypopituitary adults on lipid metabolism and carbo-
hydrate tolerance. J. Clin. Endocrinol. Metab. 80, 356–363.
Bramswig, S., Kerksiek, A., Sudhop, T., Luers, C., von Bergmann,
K., Berthold, H.K., 2002. Carbamazepine increases atherogenic
lipoproteins: mechanism of action in male adults. Am. J. Physiol.
Heart Circ. Physiol. 282 (2), 704–716.
Calandre, E.P., Rodriquez-Lopez, C., Blazquez, A., Cano, D., 1991.
Serum lipids, lipoproteins and apolipoproteins A and B in epilep-
tic patients treated with valproic acid, carbamazepine or pheno-
barbital. Acta Neurol. Scand. 83 (4), 250–253.
Caligiuri, G., Levy, B., Pernow, J., Thoren, P., Hansson, G.K., 1999.
Myocardial infarction mediated by endothelin receptor signaling
in hypercholesterolemic mice. Proc. Natl. Acad. Sci. U.S.A. 96
(12), 6920–6924.
Canturk, Z., Cetinarslan, B., Tarkun, I., Canturk, N.Z., Ozden, M.,
2003. Lipid profile and lipoprotein (a) as a risk factor for car-
diovascular disease in women with subclinical hypothyroidism.
Endocr. Res. 29 (3), 307–316.
Caparevic, Z., Bojkovic, G., Stojanovic, D., Ilic, V., 2003. Dyslipi-
demia and subclinical hypothyroidism. Med. Pregl. 56 (5–6),
276–280 (Abstract).
Commission on Classification and Terminology of the International
League Against Epilepsy, 1989. Proposal for revised classifica-
tion of epilepsies and patients with epilepsy syndromes. Epilepsia
30, 3890–3899.
Connell, J.M., Rapeport, W.G., Gordon, S., Brodie, M.J., 1984.
Changes in circulating thyroid hormones during short-term hep-
atic enzyme induction with carbamazepine. Eur. J. Clin. Pharma-
col. 26 (4), 453–456.
Deda, G., Caksen, H., Icagasioglu, D., 2003. Effect of long-term
carbamazepine therapy on serum lipids, Vitamin B12 and folic
acid levels in children. J. Pediatr. Endocrinol. Metab. 16 (2),
193–196.
Deschampheleire, M., Luyckx, F.H., Scheen, A.J., 1999. Thyroid
disorders and dyslipidemias. Rev. Med. Liege 54 (9), 746–
750.
Diagnostic Product Corporation, 1987. Albumin-Related Artifacts
in Free Thyroid Hormone Assays. Los Angeles.
Earnest, C.P., Wood, K.A., Church, T.S., 2003. Complex multivita-
min supplementation improves homocysteine and resistance to
LDL-c oxidation. J. Am. Coll. Nutr. 22 (5), 400–407.
Eiris, J.M., Lojo, S., Del Rio, M.C., Novo, I., Bravo, M., Pavon,
P., Castro-Gago, M., 1995. Effects of long-term treatment with
antipatients with epilepsy drugs on serum lipid levels in children
with epilepsy. Neurology 45 (6), 1155–1157.
Eiris-Punal, J.M., Del Rio-Garma, M., Del Rio-Garma, C., Lojo-
Rocamonde, S., Novo-Rodriguez, I., Castro-Gago, M., 1999.
Long-term treatment of children with epilepsy with valproate or
carbamazepine may cause subclinical hypothyroidism. Epilepsia
40, 1761–1766.
Eiris, J., Novo-Rodriguez, M.I., Del Rio, M., Meseguer, P., Del
Rio, M.C., Castro-Gago, M., 2000. The effects on lipid and
apolipoprotein serum levels of long-term carbamazepine, val-
proic acid and phenobarbital therapy in children with epilepsy.
Epilepsy Res. 41 (1), 1–7.
Ericsson, U.B., Bjerre, I., Forsgren, M., Ivarsson, S.A., 1985. Thy-
roglobulin and thyroid hormones in patients on long-term treat-
ment with phenytoin, carbamazepine, and valproic acid. Epilep-
sia 26 (6), 594–596.
Fichsel, H., Knpfle, G., 1978. Effects of anticonvulsant drugs on
thyroid hormones in patients with epilepsy children. Epilepsia
19, 323–336.
Franzoni, E., Govoni, M., D’Addato, S., Gualandi, S., Sangiorgi,
Z., Descovich, G.C., Salvioli, G.P., 1992. Total cholesterol,
high-density lipoprotein cholesterol, and triglycerides in chil-
dren receiving antipatients with epilepsy drugs. Epilepsia 33 (5),
932–935.
Geelen, S.N., Lemmens, A.G., Terpstra, A.H., Wensing, T., Bey-
nen, A.C., 2001. High density lipoprotein cholesteryl ester
metabolism in the pony, an animal species without plasma
cholesteryl ester transfer protein activity: transfer of high den-
sity lipoprotein cholesteryl esters to lower density lipoproteins
and the effect of the amount of fat in the diet. Comp. Biochem.
Physiol. B Biochem. Mol. Biol. 130 (2), 145–154.
Goma, A., Abdel-Regal, S., Abdellah, M., Hamed, S., 2004. Moni-
toring antiepiletptic drug therapy in Upper Egypt: a retrospective
6-years experience in carbamazepine. J. Egypt Soc. Pharmacol.
Exp. Ther. 25 (2), 319–335.
Halle, M., Berg, A., Baumstark, M.W., Konig, D., Huonker, M.,
Joseph, K., 1999. Influence of mild to moderately elevated
triglycerides on low density lipoprotein subfraction concentra-
tion and composition in healthy men with low high density
lipoprotein cholesterol levels. Atherosclerosis 143 (1), 185–192.
Hamed, S.A., Nabeshima, T., 2005. The high atherosclerotic risk
among epileptics: the atheroprotective role of multivitamins. J.
Pharmacol. Sci. 98 (4), 340–353.
Hauser, W.A., Annegers, J.F., Kurland, L.T., 1993. Incidence of
epilepsy and unprovoked seizures in Rochester, Minnesota:
1935–1984. Epilepsia 34, 453–468.
Hegedus, L., Hansen, J.M., Luhdorf, K., Perrild, H., Feldt-
Rasmussen, U., Kampmann, J.P., 1985. Increased frequency of
goitre in patients with epilepic patients on long-term phenytoin
or carbamazepine treatment. Clin. Endocrinol. 23 (4), 423–429.
Huesca-Gomez, C., Franco, M., Luc, G., Montano, L.F., Masso, F.,
Posadas-Romero, C., Perez-Mendez, O., 2002. Chronic hypothy-
roidism induces abnormal structure of high-density lipopro-
teins and impaired kinetics of apolipoprotein A–I in the rat.
Metabolism 51 (4), 443–450.
Isojarvi, J.I., Pakarinen, A.J., Rautio, A., Pelkonen, O., Myllyla,
V.V., 1994. Liver enzyme induction and serum lipid levels after
 S.A. Hamed et al. / Epilepsy Research 66 (2005) 173–183
replacement of carbamazepine with oxacarbazepine. Epilepsia
35, 1217–1220.
Isojarvi, J.I.T., Airaksinen, K.E.J., Mustonen, J.N., et al., 1995. Thy-
roid and myocardial function after replacing carbamazepine by
oxcarbazepine. Epilepsia 36, 810–816.
Isojarvi, J.I.T., Pakarinen, A.J., Myllyla, V.V., 1989. Thyroid function
in patients with epilepsy patients treated with carbamazepine.
Arch. Neurol. 46, 1175–1178.
Isojarvi, J.I.T., Pakarinen, A.J., Ylipalosaari, P.J., Myllyla, V.V.,
1990. Serum hormones in male patients with epilepsy patients
receiving anticonvulsant medication. Arch. Neurol. 47, 670–676.
Isojarvi, J.I., Turkka, J., Pakarinen, A.J., Kotila, M., Rattya, J., Myl-
lyla, V.V., 2001. Thyroid function in men taking carbamazepine,
oxcarbazepine, or valproate for epilepsy. Epilepsia 42, 930–934.
Jouko, I., Isojarvi, T., Turkka, J., Pakarinen, A.J., Kotila, M., Rut-
tyu, J., Kawashima, S., 2004. Malfunction of vascular con-
trol in lifestyle-related diseases: endothelial nitric oxide (NO)
synthase/NO system in atherosclerosis. J. Pharmacol. Sci. 4,
411–419.
Klosterskov Jensen, P., Saano, V., Haring, P., Svenstrup, B., Menge,
G.P., 1992. Possible interaction between oxcarbazepine and an
oral contraceptive. Epilepsia 33, 1149–1152.
Larkin, J.G., McKee, P.J., Forrest, G., Beastall, G.H., Park, B.K.,
Lowrie, J.I., Lloyd, P., Brodie, M.J., 1991. Lack of enzyme induc-
tion with oxcarbazepine (600 mg daily) in healthy subjects. Br.
J. Clin. Pharmacol. 31, 65–71.
Liewendahl, K., Majuri, H., Helenius, T., 1978. Thyroid function test
in patients on long-term treatment with various anticonvulsant
drugs. Clin. Endocrinol. 8, 185–191.
Luoma, P.V., Myllyla, V.V., Sotaniemi, E.A., Lehtinen, I.A., Hokka-
nen, E.J., 1980. Plasma high-density lipoprotein cholesterol in
patients with epilepsys treated with various anticonvulsants. Eur.
Neurol. 19 (1), 67–72.
Luoma, P.V., Myllyla, V.V., Hokkanen, E., 1982. Relationship
between plasma high-density lipoprotein cholesterol and anti-
convulsant levels in patients with epilepsys. J. Cardiovasc. Phar-
macol. 4 (6), 1024–1027.
Mishkel, M.A., Crowther, S.M., 1977. Hypothyroidism, an impor-
tant cause of reversible hyperlipidemia. Clin. Chim. Acta 74 (2),
139–151.
Monson, J.P., Abs, R., Bengtsson, B.A., Bennmarker, H., Feldt-
Rasmussen, U., Hernberg-Stahl, E., Thoren, M., Westberg, B.,
Wilton, P., Wuster, C., 2000. Growth hormone deficiency and
replacement in elderly hypopituitary adults. KIMS Study Group
and the KIMS International Board. Pharmacia and Upjohn Inter-
national Metabolic Database. Clin. Endocrinol. 53, 281–289.
Pita-Calandre, E., Rodriguez-Lopez, C.M., Cano, M.D., Pena-
Bernal, M., 1998. Serum lipids, lipoproteins, and apolipoproteins
in adult patients with epilepsys treated with carbamazepine, val-
proic acid, or phenytoin. Rev. Neurol. 27 (159), 785–789.
Pritchard, P.B., 1980. Hyposexuality: a complication of complex par-
tial epilepsy. Trans. Am. Neurol. Assoc. 105, 193–195.
183
Rettenmeier, A.W., Howald, W.N., Levy, R.H., Witek, D.J., Gordon,
W.P., Porubek, D.J., Baillie, T.A., 1989. Quantitative metabolic
profiling of valproic acid in humans using automated gas chro-
matographic/mass spectrometric techniques. Biomed. Environ.
Mass Spectrom. 18 (3), 192–199.
Rosen, T., Eden, S., Larson, G., Wilhelmsen, L., Bengtsson, B.A.,
1993. Cardiovascular risk factors in adult patients with growth
hormone deficiency. Acta Endocrin. 129, 195–200.
Spacilova, J., Limanova, Z., Aschermann, M., 2003. Subclinical
hypothyroidism with emphasis on the cardiovascular system.
Cas. Lek. Cesk. 142 (11), 643–647 (Abstract).
Strandjord, R.E., Aanderud, S., Myking, O.L., Johannessen, S.I.,
1981. Influence of carbamazepine on serum thyroxine and tri-
iodothyronine in patients with epilepsy. Acta Neurol. Scand. 63,
111–121.
Sudhop, T., Bauer, J., Elger, C.E., von Bergmann, K., 1999. Increased
high-density lipoprotein cholesterol in patients with epilepsy
treated with carbamazepine: a gender-related study. Epilepsia 40
(4), 480–484.
Surks, M.I., DeFesi, C.R., 1996. Normal serum free thyroid hor-
mone concentrations in patients treated with phenytoin or carba-
mazepine. JAMA 275, 149–158.
Tumer, L., Serdaroglu, A., Hasanoglu, A., Biberoglu, G., Aksoy, E.,
2002. Plasma homocysteine and lipoprotein (a) levels as risk fac-
tors for atherosclerotic vascular disease in patients with epilepsy
children taking anticonvulsants. Acta Paediatr. 91 (9), 923–
926.
Verrotti, A., Basciani, F., Domizio, S., Sabatino, G., Morgese, G.,
Chiarelli, F., 1998. Serum lipids and lipoproteins in patients
treated with antipatients with epilepsy drugs. Pediatr. Neurol.
19 (5), 364–367.
Verrotti, A., Basciani, F., Morresi, S., Morgese, G., Chiarelli, F., 2001.
Thyroid hormones in patients with epilepsy children receiving
carbamazepine and valproic acid. Pediatr. Neurol. 25, 43–46.
Villa, S.M., Alexander, N.M., 1987. Carbamazepine (Tegretol)
inhibits in vivo iodide uptake and hormone synthesis in rat thy-
roid glands. Endocr. Res. 13, 385–397.
Weissel, M., 2003. Possible consequences of subclinical hypothy-
roidism. Acta Med. Austriaca 30 (4), 93–97 (Abstract).
Witherspoon, L.R., el Shami, A.S., Shuler, S.E., Neely, H., Son-
nemake, R., Gilbert, S.S., Alyea, K., 1988. Chemically blocked
analog assays for free thyronines. I. The effect of chemical block-
ers on T4 analog and T4 binding by albumin and by thyroxin-
binding globulin. Clin. Chem. 34, 9–16.
Yilmaz, E., Dosan, Y., Gurgoze, M.K., Gungor, S., 2001. Serum
lipid changes during anticonvulsive treatment serum lipids in
patients with epilepsy children. Acta Neurol. Belg. 101 (4), 217–
220.
Yoo, J.H., Hong, S.B., 1999. A common mutation in the methylenete-
trahydrofolate reductase gene is a determinant of hyperhomocys-
teinemia in patients with epilepsy patients receiving anticonvul-
sants. Metabolism 48 (8), 1047–1051.