Sunteți pe pagina 1din 7

Pediatric Neurology 52 (2015) 153e159

Contents lists available at ScienceDirect

Pediatric Neurology
journal homepage: www.elsevier.com/locate/pnu

Original Article

Longitudinal Change of Vitamin D Status in Children With


Epilepsy on Antiepileptic Drugs: Prevalence and Risk Factors
Yun-Jin Lee MD, PhD a,1, Kyung Mi Park MD a,1, Young Mi Kim MD, PhD b,
Gyu Min Yeon MD c, Sang Ook Nam MD, PhD a, *
a
Department of Pediatrics, Pusan National University Childrens Hospital, Pusan National University School of Medicine, Yangsan,
Korea
b
Department of Pediatrics, Pusan National University Hospital, Busan, Korea
c
Department of Pediatrics, Kosin University Gospel Hospital, Kosin University, Busan, Korea

abstract
BACKGROUND: Our aim was to evaluate the prevalence and risk factors of vitamin D deciency and the changes of
vitamin D level among children with epilepsy on antiepileptic drugs. METHODS: The levels of serum 25-hydroxy
vitamin D were measured at the start of antiepileptic drugs and at 6- to 12-month intervals in children with
epilepsy taking antiepileptic drugs in Pusan National University Childrens Hospital. Vitamin D deciency was
dened as 25-hydroxy vitamin D levels <20 ng/mL and insufciency between 21 and 29 ng/mL. RESULTS: A total of
143 children (103 boys and 40 girls) with the mean age of 7.4  5.4 years were included. The mean follow-up
duration was 1.8  0.8 years. At the start of antiepileptic drugs and the last follow-up, vitamin D deciency or
insufciency was recognized in 56.6% (81 of 143) and 79.0% (113 of 143), respectively (P < 0.01). The mean value of
initial 25-hydroxy vitamin D was 31.1  14.7 ng/mL, which was signicantly decreased to 20.2  14.9 ng/mL
(P < 0.01) in the last follow-up. Polytherapy (16.0  13.6 ng/mL), longer duration of 2 years (23.5  9.1 ng/mL),
tube feeding (18.2  14.5 ng/mL), and overweight with body mass index of eighty-fth percentile or greater
(17.0  12.1 ng/mL) had a signicant negative effect for the longitudinal change of 25-hydroxy vitamin D. Age,
etiologies, seizure outcomes, and type of antiepileptic drugs (enzyme-inducing versus nonenzyme-inducing anti-
epileptic drugs) did not affect the longitudinal decrease of 25-hydroxy vitamin D. CONCLUSIONS: A high proportion of
these children on antiepileptic drugs had hypovitaminosis D and a signicant decrease between the initial and the
last follow-up. Polytherapy and longer duration of antiepileptic drugs, tube feeding, and overweight were inde-
pendently associated with longitudinally signicant decrease of 25-hydroxy vitamin D.
Keywords: vitamin D, hypovitaminosis D, child, epilepsy, antiepileptic drug
Pediatr Neurol 2015; 52: 153-159
2015 Elsevier Inc. All rights reserved.

Introduction autoimmune diseases, and cancer.1 Causes of vitamin D


deciency are inadequate exposure to sunlight, body mass
There has been increasing concern recently on the index (BMI) of >30 kg/m2, underlying disease (e.g., fat
impact of vitamin D on human health. Besides rickets and malabsorption syndromes, nephrotic syndrome), and
decreased bone mineral density, vitamin D has been related medications including antiepileptic drugs (AEDs).2 AED
to various disorders, e.g., diabetes mellitus, stroke, medication, particularly long-term therapy or polytherapy,
is recognized to be associated with hypovitaminosis D.3-5
Article History: Adults with epilepsy are known to be at considerably
Received July 27, 2014; Accepted in nal form October 8, 2014 increased risk for bone fractures compared with the general
* Communications should be addressed to: Dr. Nam; Department of population, and increasing age and duration of AED treat-
Pediatrics; Pusan National University Childrens Hospital; Pusan Na- ment are signicant risk factors.6-8
tional University School of Medicine; 20 Geumo-ro, Meulgeum-eup;
In 1979, Offerman et al.9 documented that 72% of 83
Yangsan-si, Gyeongsangnam-do 626-770, Korea.
E-mail address: neuroped@naver.com children aged 10 to 16 years with epilepsy and 50% of 16
1
Yun-Jin Lee and Kyung Mi Park contributed equally to this work and are control patients exhibited 25-hydroxy vitamin D
the rst coauthors. (25OHD) <15 ng/mL. Since then, several cross-sectional

0887-8994/$ - see front matter 2015 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.pediatrneurol.2014.10.008
154 Y.-J. Lee et al. / Pediatric Neurology 52 (2015) 153e159

studies evaluated the prevalence of hypovitaminosis D in TABLE 1.


Demographic Prole of the Patients
pediatric epilepsy. Among children taking AEDs, non-
ambulatory children,10 polytherapy,5,10,11 and those with Variable n (%)
poorly nutritional status12 were reported to have a high Total 143 (100.0)
prevalence of vitamin D deciency or insufciency or Sex
signicantly lower 25OHD levels. However, there are few Boy 103 (72.0)
studies that evaluate the longitudinal change of 25OHD Girl 40 (28.0)
Mean age  SD (yr) 7.4  5.4 (1.0-18.5)
levels among children with epilepsy on AEDs in accor-
Brain MRI
dance to diverse risk factors. No lesion 99 (69.2)
The aims of this study, therefore, were (1) to evaluate the Lesion 44 (30.8)
prevalence of vitamin D deciency or insufciency among AEDs
children with epilepsy on AED treatment (greater than Monotherapy 77 (53.8)
1 year); (2) to examine the longitudinal change of 25OHD Non-EIAEDs 58 (40.6)
levels during AED treatment; and (3) to evaluate the po- EIAEDs 19 (13.3)
Polytherapy 66 (46.2)
tential risk factors for hypovitaminosis D in a tertiary pe-
Duration of AEDs
diatric neurology clinic. 1-2 yr 91 (63.6)
>2 yr 52 (36.4)
Patients and Methods Mean duration (yr) 1.8  0.8 (1.0-4.9)
Seizure outcomes
Seizure free 76 (53.1)
Patients
No seizure free 67 (46.9)
Nonambulatory 42 (29.4)
We undertook a retrospective longitudinal cohort study of children
Tube feeding 20 (14.0)
on AEDs between January 2011 and March 2013 at a pediatric neurology
Ketogenic diet 16 (11.2)
clinic afliated with Pusan National University Childrens Hospital in
BMI (percentile for age)
South Korea. The inclusion criteria of the study subjects were as follows:
<85th 93 (65.0)
(1) the range of patients age was between 1 and 18 years; (2) patients
85th 50 (35.0)
were monitored for at least 1 year after the start of AED treatment;
Mean BMI (kg/m2) 18.2  4.9 (8.2-36.5)
and (3) their levels of serum 25OHD were measured before the start
of AED treatment and at 6 to 12 months interval thereafter. The exclu- Abbreviations:
sion criteria were as follows: (1) presence of any condition known AEDs Antiepileptic drugs
BMI Body mass index
to affect the bone metabolism (e.g., hepatic, renal, endocrine, and
EIAEDs Enzyme-inducing AEDs
metabolic disease) and other chronic diseases (e.g., cancer, gastrointes- MRI Magnetic resonance imaging
tinal disorders, liver insufciency, diabetes mellitus); (2) intake of SD Standard deviation
vitamin D and calcium supplements within the last 6 months; and (3)
have taken drugs that may potentially affect the bone metabolism (e.g.,
corticosteroids).
negative if no known etiology had been identied and brain MRI was
normal. MRI positive was dened if there was a brain lesion present, such
Methods
as previous acquired brain injuries or developmental brain malforma-
tions. BMI of greater than eighty-fth percentile for their age was
The levels of serum 25OHD of all patients were measured before dened as being overweight.
the start of AED treatment (25OHD-1; baseline 25OHD) and every 6 to
12 months interval thereafter (25OHD-2; 25OHD measured at the last
follow-up visit) by isotope-dilution liquid chromatography-tandem
mass spectrometry in a licensed laboratory. Recent reviews have Standard protocol approvals, registrations, and patient consents
dened vitamin D deciency as 25OHD <20 ng/mL and vitamin D
insufciency as between 21 and 29 ng/mL.13-15 These denitions were Ethics permission for this study was granted (number: 05-2014-066)
used in our study. The difference or change between 25OHD-1 and by the Institutional Review Board of Pusan National University Yangsan
25OHD-2 (D-25OHD) was investigated as the value of 25OHD-2 minus Hospital, and fully informed written consent was obtained from each
25OHD-1. We divided the patient population into two groups by the participant.
value of D-25OHD: (1) patients with unnoticeable decreased value (D-
25OHD  10 ng/mL, group A) and (2) patients with signicant
decreased value (D-25OHD < 10 ng/mL, group B). Decline of more
than 10 ng/mL from the baseline to the last follow-up was suggested Statistical analysis
as the longitudinally signicant decreased value.
Potential risk factors for hypovitaminosis D were evaluated by Statistical analyses were performed with SPSS 15.0 software using
comprehensively reviewing the patients medical records, which raw scores. For the initial and/or last 25OHD (25OHD-1 versus
included sex, age at the start of AED treatment, ambulatory status, 25OHD-2) comparison in all patients, a paired t test was used to
feeding status (oral versus tubal), ketogenic diet, BMI (kg/m2), underly- evaluate the signicant changes between the corresponding values
ing etiologies, seizure outcomes (seizure free versus no seizure free), (25OHD-2 minus 25OHD-1). Two-tailed chi-square and Student t tests
number of regular AEDs over the last 6 months, and duration of AEDs were used to evaluate the signicant differences in dependent, cate-
(<2 years versus 2 years). Patients with monotherapy were classied as gorical, and continuous variables between groups. Multivariate logis-
taking enzyme-inducing or nonenzyme-inducing AEDs. Enzyme- tic regression was used to identify the variables that were
inducing AEDs included phenytoin, phenobarbital, and carbamazepine. independently associated with longitudinally signicant decrease of
Nonenzyme-inducing AEDs included vigabatrin, lamotrigine, levetir- the 25OHD. Sex, age, ambulatory status, feeding status, ketogenic diet,
acetam, topiramate, clobazam, valproate, and ethosuximide. Seizure-free BMI, underlying etiologies, seizure outcomes, and number, types, and
outcome was dened as no seizure for at least 1 year until the last follow- duration of AEDs were all entered as independent variables. Statistical
up. Etiologies were dened as magnetic resonance imaging (MRI) signicance was determined as a two-tailed P value of <0.05.
Y.-J. Lee et al. / Pediatric Neurology 52 (2015) 153e159 155

Results signicantly lower values of D-25OHD than those on oral


feeding and normal weight, respectively.
Vitamin D status and the change from the baseline to the last follow-
up in total patients
Factors associated with longitudinally signicant decrease of 25OHD
A total of 143 children were included in this study. The
patients age ranged from 1 to 18.6 years, with a mean of Sixty-four patients were assigned to group A (D-
7.4 years (5.4 years). One hundred three (72.0%) were 25OHD  10 ng/mL), and seventy-nine were in group B
boys. The mean follow-up duration was 1.8  0.8 years. The (D-25OHD < 10 ng/mL) (Table 2). The mean values of D-
demographic proles of the total patients are presented in 25OHD were 1.3  10.5 ng/mL in group A
Table 1. Before the start of AEDs for all patients, their mean and 20.8  8.6 ng/mL in group B. Group B showed
value of 25OHD-1 as the baseline 25OHD was 31.1  14.7 ng/ frequently polytherapy (2.3  1.6 vs 1.7  1.2; P 0.03) and
mL. The mean value of 25OHD-2 of the last follow-up was longer duration (2.2  0.9 vs 1.4  0.2 years; P < 0.01) of
20.2  14.9 ng/mL, which showed a longitudinally signi- AEDs than those in group A. Nonambulatory state (39.2% vs
cant decline (D-25OHD 10.9  14.5 ng/mL; P < 0.01; Fig 17.2%; P 0.01), children on tube feeding (21.5% vs 4.7%;
1A). At the last follow-up, the distribution of the levels of P < 0.01), and ongoing ketogenic diet (19.0% vs 1.6%;
25OHD was tilted to the low end of the spectrum. The P < 0.01) were signicantly frequent in group B compared
vitamin D deciency or insufciency status was 56.6% (81 of with those in group A. Overweight children with elevated
143) in 25OHD-1 and 79.0% (113 of 143) in 25OHD-2 BMI (eighty-fth percentile or greater for age) were
(P < 0.01; Fig 1B). observed more often in group B than in group A (49.4% vs
17.2%; P < 0.01). Age, etiologies (MRI negative versus posi-
tive), seizure outcomes, and type of AEDs (enzyme inducing
Factors for aggravation of hypovitaminosis D in pediatric epilepsy on versus nonenzyme inducing) had no signicant difference
AEDs between the two groups.
In the multivariate model, polytherapy (odds ratio [OR],
We investigated the potential risk factors affecting the 5.07; 95% condence interval [CI], 0.06 to 0.52), longer dura-
change of 25OHD from the baseline to the last follow-up tion of 2 years (OR, 8.38; 95% CI, 2.78 to 25.26), tube feeding
during AED treatment (Fig 2). Polytherapy and longer (OR, 6.08; 95% CI, 5.22 to 12.51), and elevated BMI of eighty-
duration of 2 years showed a higher negative effect for the fth percentile or greater (OR, 8.03; 95% CI, 1.13 to 14.37)
change of 25OHD than monotherapy (16.0  13.6 were independently associated with longitudinally signi-
vs 7.5  13.9 ng/mL; P < 0.01) and shorter duration of cant decrease of 25OHD (Table 2). The remaining factors did
AEDs (23.5  9.1 vs 5.7  11.9 ng/mL; P < 0.01). Non- not contribute signicantly to the multivariate models.
ambulatory children had a much severe decrease between Two patients (one boy of 7 years old with seventy-eighth
25OHD-1 and 25OHD-2 than ambulatory children BMI, one girl of 11 years old with eighty-seventh BMI) had a
(16.2  13.7 vs 8.3  14.1 ng/mL; P < 0.01). Children on history of a fracture during the treatment. Both of them had
tube feeding (18.2  14.5 vs 9.6  14.1 ng/mL; P 0.01) taken the AEDs of three or more during the 2.8- and
and overweight children with BMI of eighty-fth percentile 2.2-year periods and underwent on tube feeding with
or greater (17.0  12.1 vs 6.5  13.9 ng/mL; P < 0.01) had nonambulatory state. A boy had a hip fracture (25OHD,

FIGURE 1.
(A) The longitudinal change of the mean of 25OHD levels in children with epilepsy taking AEDs showed a signicant decline from 31.1  14.7 ng/mL to
20.2  14.9 ng/mL during the mean 1.8-year follow-up period (10.9  14.5 ng/mL; P < 0.01). (B) The distribution of the levels of 25OHD was tilted to the
low end of the spectrum. The rate of the hypovitaminosis D status of <20 ng/mL signicantly increased from 20.3% (29 of 143) to 61.5% (88 of 143) between
the baseline and the last follow-up (P < 0.01). 25OHD, 25-hydroxyvitamin D; AEDs, antiepileptic drugs.
156 Y.-J. Lee et al. / Pediatric Neurology 52 (2015) 153e159

(A) Polytherapy Monotherapy (B) > 2 yr < 2 yr


25OHD 25OHD
(ng/mL) (ng/mL)
38.1 (p<0.01) 38.5 (p<0.01)

26.1 28.8
22 23.1
18.6
15

Baseline Last follow-up Baseline Last follow-up

(C) No seizure-free Seizure-free (D) Non-ambulatory Ambulatory

35.2 (p=0.10) 35.7


(p<0.01)
27.4 28.7
22.2 20.4
18.4 19.6

Baseline Last follow-up Baseline Last follow-up

(E) Tube feeding Oral feeding (F) BMI >85% BMI <85%

36.8 (p<0.01)
(p=0.01) 32.9
30 30.1
23.6
20.4
18.7 15.9

Baseline Last follow-up Baseline Last follow-up


FIGURE 2.
Comparison of the longitudinally decreasing change of the mean 25OHD levels from the baseline to the last follow-up by the potential risk factors: (A)
monotherapy (n 77; 7.5  13.9 ng/mL) versus polytherapy (n 66; 16.0  13.6 ng/mL; P < 0.01), (B) duration of AEDs of <2 years
(n 91; 5.7  11.9 ng/mL) versus 2 years (n 52; 23.5  9.1 ng/mL; P < 0.01), (C) seizure-free (n 76; 9.0  15.9 ng/mL) versus no seizure-free
outcomes (n 67; 13.0  12.5 ng/mL; P 0.10), (D) ambulatory (n 101; 8.3  14.1 ng/mL) versus nonambulatory children (n 42; 16.2  13.7 ng/mL;
P < 0.01), (E) oral feeding (n 123; 9.6  14.1 ng/mL) versus tube feeding (n 20; 18.2  14.6 ng/mL; P 0.01), (F) BMI of less than eight-fth percentile
for age (n 93; 6.5  13.9 ng/mL) versus eight-fth percentile or greater for age (n 50; 17.0  12.1 ng/mL; P < 0.01). 25OHD, 25-hydroxyvitamin D;
AEDs, antiepileptic drugs; BMI, body mass index.

12 ng/mL), and a girl experienced a femur fracture (25OHD, mean 10.9  14.5 ng/mL between the baseline and the last
9 ng/mL). Bone mineral density was not performed. Their 25OHD in our tertiary hospital-based population of pediatric
fractures have been naturally healed. epilepsy on long-term AEDs (mean duration, 1.8  0.8 years).
Polytherapy (two or more AEDs) and longer duration
Discussion (2 years) of AEDs, tube feeding, and elevated BMI of
eighty-fth percentile or greater were independently asso-
We discovered a high proportion of vitamin D deciency ciated with longitudinally signicant decrease of 25OHD
(61.5%; 88 of 143) or insufciency (17.5%; 25 of 143) before and after the AED treatment. Our study included a
status with a longitudinally signicant decrease of the large number of children with long-term follow-up to
Y.-J. Lee et al. / Pediatric Neurology 52 (2015) 153e159 157

TABLE 2.
Univariate and Multivariate Regression Analysis of the Factors Associated With Longitudinal Decrease Between the Baseline and the Last 25OHD

Factors Univariate Model Multivariate Model


Group A (n 64), n (%) Group B (n 79), n (%) P OR 95% CI
Sex
Boy 39 (61.0) 64 (81.0) 0.01* 0.30 0.07-1.25
Girl 25 (39.0) 15 (19.0)
Mean age  SD (yr) 7.0  5.3 (1.0-16) 8.0  5.1 (1.0-18.5) 0.23 1.07 0.27-4.29
Brain MRI
No lesion 47 (73.4) 52 (65.8) 0.37 3.51 0.75-16.45
Lesion 17 (26.6) 27 (34.2)
AEDs
Monotherapy 42 (65.6) 35 (44.3) 0.01* 5.07y 0.06-0.52
Non-EIAEDsz 35 (54.7) 23 (29.1) 0.07 0.75 0.25-2.57
EIAEDsx 7 (10.9) 12 (15.2)
Polytherapy 22 (34.4) 44 (55.7) 0.01*
Mean number 1.7  1.2 2.3  1.6 0.03* 7.02y 0.27-0.95
Duration of AEDs
1-2 yr 62 (96.9) 29 (36.7) <0.01* 8.38y 2.78-25.26
> 2 yr 2 (3.1) 50 (63.3)
y
Mean duration (yr) 1.4  0.2 2.2  0.9 <0.01* 9.15 5.27-33.17
Seizure outcomes
Seizure free 40 (62.5) 36 (45.6) 0.06 0.39 0.12-1.33
No seizure free 24 (37.5) 43 (54.4)
Nonambulatory 11 (17.2) 31 (39.2) 0.01* 3.33 0.36-30.71
Tube feeding 3 (4.7) 17 (21.5) <0.01* 6.08y 5.22-12.51
Ketogenic diet 1 (1.6) 15 (19.0) <0.01* 5.09 0.08-19.98
BMI (percentile for age)
<85th 53 (82.8) 40 (50.6) <0.01* 8.03y 1.13-14.37
85th 11 (17.2) 39 (49.4)
Abbreviations:
25OHD 25-hydroxy vitamin D
AEDs Antiepileptic drugs
BMI Body mass index
CI Condence interval
EIAEDs Enzyme-inducing AEDs
OR Odds ratio
MRI Magnetic resonance imaging
SD Standard deviation
Group A Difference between the baseline and the last 25OHD  10 ng/mL
Group B Difference between the baseline and the last 25OHD < 10 ng/mL
* P < 0.05.
y
Reects the increase in the OR of signicant longitudinal decrease of 25OHD.
z
Non-EIAEDs included vigabatrin, lamotrigine, levetiracetam, topiramate, clobazam, valproate, and ethosuximide.
x
EIAEDs included phenytoin, phenobarbital, and carbamazepine.

evaluate the vitamin D status (numbers of children on ranging from 4% to 75%.3-5,16,17 These inconsistent results
long-term AEDs participated in previous reports ranged could be inuenced by multiple factors, including seasonal
between 38 and 125 children).2-5,16,17 It can be considered as variation, diverse place of residence, different inclusion
the rst longitudinal study from the baseline to the last criteria for the duration of AED treatment, different exclu-
follow-up, and not as a cross-sectional study, for the sion criteria of children who may have underlying disorders
hypovitaminosis D in pediatric epilepsy on AEDs. We also affecting bone metabolism, and dissimilar cutoff values for
investigated the risk factors for the hypovitaminosis D the hypovitaminosis D (<10 or <20 ng/mL). We have tried
through a comparison between patients with longitudinally to reduce these confounding factors by measuring 25OHD
signicant decrease of 25OHD and with unnoticeable levels including all children with epilepsy on AEDs for at
change (D-25OHD < 10 vs 10 ng/mL). least 1 year, excluding the children with any intrinsic con-
Data from 12- to 19-year-old American adolescents from dition or disease of impaired bone health, and using a recent
the National Health and Nutrition Examination Survey denition for the vitamin D deciency (<20 ng/mL) and
revealed striking differences in the prevalence, depending insufciency (21 to 29 ng/mL). Our patients had a consid-
on the inconsistent cutoff values. When the denition of erable change of the prevalence of vitamin D deciency or
vitamin D deciency was changed from <11 to <20 ng/mL, insufciency between the baseline (56.6%) and the last
the prevalence increased from 2% to 14%. Vitamin D levels follow-up (79.0%) during the AED treatment. The preva-
are also affected by age, sex, geographic location, use of lence of hypovitaminosis D among healthy children in our
supplements, and BMI, which make it complicated to dene hospitals region has not been investigated until now.
the hypovitaminosis D and estimate the prevalence.18 Pre- Several studies commented on the effect of polytherapy
vious studies of vitamin D status in pediatric epilepsy have of AEDs on vitamin D status in pediatric epilepsy. A
reported a diverse prevalence of the hypovitaminosis D, German cross-sectional study4 reported that polytherapy
158 Y.-J. Lee et al. / Pediatric Neurology 52 (2015) 153e159

is associated with signicantly lower 25OHD levels is After the follow-up period, we instituted vitamin D
monotherapy. Bergqvist et al.5 reported a decline of supplementation in the children with vitamin D deciency.
approximately 7 ng/mL in 25OHD for each additional AED The oral vitamin D preparation used was sunny D drops
in children with refractory epilepsy. Other reports have11 cholecalciferol (D3, 400 IU/drop; GMP laboratories of
documented that, while the type of AEDs exerted no ef- America), and children took one drop (400 IU/day) of sunny
fect on vitamin D status or bone mineral density, poly- D drops according to the maintenance dose in children.
therapy exerted a negative impact on bone mineral Most of them showed an increase of 25OHD, although these
density. Our children with polytherapy or longer duration data are preliminary.
(longer than 2 years) of AEDs also had greater decline of Our study has limitations. First, we lack comprehensive
25OHD than those with monotherapy or shorter duration. dietary assessments and home backgrounds in our patients.
Because many AEDs induce hepatic CYP450 metabolism, Others have highlighted that diet plays an important role in
they result in an increased metabolism of vitamin D, determining vitamin D levels, and this should not be over-
leading to declining 25OHD levels, increased parathyroid looked.3,5 Second, we do not routinely examine the bone
hormone levels, and abnormally enhanced bone turn- mineral density in children with epilepsy. Therefore, we
over.19-25 However, even nonenzyme-inducing medica- cannot comment much regarding the potential impact of
tions were associated with poor bone health.19-23 low vitamin D levels on bone mineral density in our pa-
Therefore, hepatic enzyme induction is just one compo- tients. Nevertheless, current evidence has shown the asso-
nent of multifaceted mechanisms by which AEDs affect the ciation of low vitamin D levels with low bone mineral
health of bones. density in children.30 Last, we did not evaluate the effects of
Baer et al.10 studied vitamin D levels in relation to specic drugs. We just compared the effect between
ambulation in a large sample of children and documented enzyme-inducing and nonenzyme-inducing AEDs in chil-
that the risk of the hypovitaminosis D among non- dren undergoing monotherapy. Enzyme-inducing AEDs
ambulatory children was about twice that of ambulatory were not associated with the risk compared with others.
children (P < 0.01), even after adjusting for confounders. Nearly half of our patients had taken with new AEDs. These
Other studies excluded nonambulatory patients because of new AEDs, which are not potent hepatic enzyme inducers,
concerns about confounding results. We found that non- may have a lesser effect on 25OHD levels than old enzyme-
ambulatory children of epilepsy had a longitudinally greater inducing drugs. However, others have documented that
decrease of 25OHD than ambulatory children; however, oxcarbazepine24,31,32 and lamotrigine33 may inuence on
ambulation was not an independent risk factor in multi- 25OHD and/or bone mineral density.
variate model. We found that there is a longitudinally signicant
One study5 of children with intractable epilepsy on a decrease in 25OHD levels during the AED treatment and a
ketogenic diet reported that, before the diet therapy, 4% and high prevalence of hypovitaminosis D in children with ep-
51% had vitamin D deciency (25OHD of <11 ng/mL) and ilepsy. Polytherapy and longer duration of AEDs, tube
insufciency (<32 ng/mL), respectively. The levels of feeding, and elevated BMI were signicant as independent
25OHD were declined by approximately 0.5 ng/mL/mo risk factors for the longitudinal decrease of 25OHD. A high
during a 15-month ketogenic diet. In our study, both tube proportion of these children on AEDs are at further risk for
feeding and ketogenic diet were associated with the sig- bone injury due to seizures, comorbid neuromuscular
nicant decrease of 25OHD in a univariate model; however, dysfunction, and long-term AED treatment that affect the
only tube feeding was an independent risk factor (OR, 6.08; health of bones. Augmented concern and advice regarding
95% CI, 5.22 to 12.51). Generally, there is no need for addi- vitamin D status and bone health among children with
tional nutritional supplements when taking a liquid for- epilepsy are very important. Further studies evaluating the
mula (vitamin D, 2.5 mg/200 mL) or a ketogenic milk efcacy of vitamin D supplementation in these children is
(vitamin D, 3.2 mg/180 mL) for the tube-fed children necessary.
because it contains all the vitamins and minerals that are
required for physical growth based on the Dietary Reference This study was supported by a 2013 research grant from Pusan National University
Intake for Koreans (vitamin D, 5 mg/day in 1 year of age, Yangsan Hospital.
10 mg/day in >1 year of age). Nonetheless, children on tube
feeding, who might be exposed to poor eating capacity or
recurrent gastrointestinal disturbance, could be explained
References
the signicant drop of vitamin D levels.
In the general population of children, obesity is a major 1. Holick MF. Vitamin D deciency. N Engl J Med. 2007;357:266-281.
identied risk factor for vitamin D insufciency.26-28 In a 2. Fong CY, Riney CJ. Vitamin D deciency among children with epi-
recent study, obesity has been identied as a common lepsy in South Queensland. J Child Neurol. 2014;29:368-373.
comorbidity in children with newly diagnosed epilepsy.29 3. Shellhaas RA, Barks AK, Joshi SM. Prevalence and risk factors for
vitamin D insufciency among children with epilepsy. Pediatr
Shellhaas et al.3 documented that 25% of 78 children Neurol. 2010;42:422-426.
with epilepsy on AEDs had vitamin D deciency 4. Nettekoven S, Strohle A, Trunz B, et al. Effects of antiepileptic drug
(<20 ng/mL), and girl and children with showed BMI were therapy on vitamin D status and biochemical markers of bone
at increased risk for low 25OHD. We showed that over- turnover in children with epilepsy. Eur J Pediatr. 2008;167:
1369-1377.
weight children with an elevated BMI (eighty-fth 5. Bergqvist AG, Schall JI, Stallings VA. Vitamin D status in children
percentile or greater for age) had signicantly lower with intractable epilepsy, and impact of the ketogenic diet. Epilepsia.
25OHD levels than those of normal weight. 2007;48:66-71.
Y.-J. Lee et al. / Pediatric Neurology 52 (2015) 153e159 159

6. Vestergaard P. Epilepsy, osteoporosis and fracture riskda meta- 21. Akin R, Okutan V, Sarici U, Altunbas A, Gokcay E. Evaluation of bone
analysis. Acta Neurol Scand. 2005;112:277-286. mineral density in children receiving antiepileptic drugs. Pediatr
7. Souverein PC, Webb DJ, Weil JG, Van Staa TP, Egberts AC. Use of Neurol. 1998;19:129-131.
antiepileptic drugs and risk of fractures: case-control study among 22. Erbayat Altay E, Serdaroglu A, Tumer L, Gucuyener K, Hasanoglu A.
patients with epilepsy. Neurology. 2006;66:1318-1324. Evaluation of bone mineral metabolism in children receiving car-
8. Sheth RD, Gidal BE, Hermann BP. Pathological fractures in epilepsy. bamazepine and valproic acid. J Pediatr Endocrinol Metab. 2000;13:
Epilepsy Behav. 2006;9:601-605. 933-939.
9. Offermann G, Pinto V, Kruse R. Antiepileptic drugs and vitamin D 23. Kafali G, Erselcan T, Tanzer F. Effect of antiepileptic drugs on bone
supplementation. Epilepsia. 1979;20:3-15. mineral density in children between ages 6 and 12 years. Clin
10. Baer MT, Kozlowski BW, Blyler EM, Trahms CM, Taylor ML, Pediatr (Phila). 1999;38:93-98.
Hogan MP. Vitamin D, calcium, and bone status in children with 24. Babayigit A, Dirik E, Bober E, Cakmakci H. Adverse effects of anti-
developmental delay in relation to anticonvulsant use and ambu- epileptic drugs on bone mineral density. Pediatr Neurol. 2006;35:
latory status. Am J Clin Nutr. 1997;65:1042-1051. 177-181.
11. El-Hajj Fuleihan G, Dib L, Yamout B, Sawaya R, Mikati MA. Pre- 25. Chou IJ, Lin KL, Wang HS, Wang CJ. Evaluation of bone mineral
dictors of bone density in ambulatory patients on antiepileptic density in children receiving carbamazepine or valproate mono-
drugs. Bone. 2008;43:149-155. therapy. Acta Paediatr Taiwan. 2007;48:317-322.
12. Volpe SL, Schall JI, Gallagher PR, Stallings VA, Bergqvist AG. Nutrient 26. Saintonge S, Bang H, Gerber LM. Implications of a new denition of
intake of children with intractable epilepsy compared with healthy vitamin D deciency in a multiracial us adolescent population: the
children. J Am Diet Assoc. 2007;107:1014-1018. national health and nutrition examination survey III. Pediatrics.
13. Hamed SA. Inuences of bone and mineral metabolism in epilepsy. 2009;123:797-803.
Expert Opin Drug Saf. 2011;10:265-280. 27. Wagner CL, Greer FR, American Academy of Pediatrics Section on
14. Dawson-Hughes B, Heaney RP, Holick MF, Lips P, Meunier PJ, Breastfeeding, American Academy of Pediatrics Committee on
Vieth R. Estimates of optimal vitamin D status. Osteoporos Int. 2005; Nutrition. Prevention of rickets and vitamin D deciency in infants,
16:713-716. children, and adolescents. Pediatrics. 2008;122:1142-1152.
15. Adams JS, Hewison M. Update in vitamin D. J Clin Endocrinol Metab. 28. Reis JP, von Muhlen D, Miller 3rd ER, Michos ED, Appel LJ. Vitamin D
2010;95:471-478. status and cardiometabolic risk factors in the United States
16. Nicolaidou P, Georgouli H, Kotsalis H, et al. Effects of anticonvulsant adolescent population. Pediatrics. 2009;124:e371-e379.
therapy on vitamin D status in children: prospective monitoring 29. Daniels ZS, Nick TG, Liu C, Cassedy A, Glauser TA. Obesity is a
study. J Child Neurol. 2006;21:205-209. common comorbidity for pediatric patients with untreated, newly
17. Rauchenzauner M, Griesmacher A, Tatarczyk T, et al. Chronic anti- diagnosed epilepsy. Neurology. 2009;73:658-664.
epileptic monotherapy, bone metabolism, and body composition in 30. Bischoff-Ferrari HA. Vitamin D and fracture prevention. Rheum Dis
non-institutionalized children. Dev Med Child Neurol. 2010;52: Clin North Am. 2012;38:107-113.
283-288. 31. Cansu A, Yesilkaya E, Serdaroglu A, et al. Evaluation of bone turn-
18. Yetley EA. Assessing the vitamin D status of the US population. Am J over in epileptic children using oxcarbazepine. Pediatr Neurol. 2008;
Clin Nutr. 2008;88:558S-564S. 39:266-271.
19. Sheth RD, Wesolowski CA, Jacob JC, et al. Effect of carbamazepine 32. Mintzer S, Boppana P, Toguri J, DeSantis A. Vitamin D levels and
and valproate on bone mineral density. J Pediatr. 1995;127: bone turnover in epilepsy patients taking carbamazepine or
256-262. oxcarbazepine. Epilepsia. 2006;47:510-515.
20. Tekgul H, Serdaroglu G, Huseyinov A, Gokben S. Bone mineral status 33. Guo CY, Ronen GM, Atkinson SA. Long-term valproate and lamo-
in pediatric outpatients on antiepileptic drug monotherapy. J Child trigine treatment may be a marker for reduced growth and bone
Neurol. 2006;21:411-414. mass in children with epilepsy. Epilepsia. 2001;42:1141-1147.

S-ar putea să vă placă și