Seizure 34 (2016) 4853

Contents lists available at ScienceDirect

Seizure
journal homepage: www.elsevier.com/locate/yseiz

Interictal fatigue and its predictors in epilepsy patients:

A case-control study
Oh-Young Kwon a, Sung-Pa Park b,*
a
Department Neurology and Institute of Health Science, Gyeongsang National University School of Medicine, Jinju, Republic of Korea
b
Department of Neurology, School of Medicine, Kyungpook National University, Daegu, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Article history: Purpose: Fatigue impairs the quality of life (QOL) of epilepsy patients, but few studies have investigated
Received 8 October 2015 this issue and no systematic analysis of the predictors of fatigue in epilepsy patients has been performed.
Received in revised form 5 December 2015 Thus, we investigated the degree and predictors of fatigue in epilepsy patients.
Accepted 8 December 2015
Methods: We enrolled 270 consecutive adult patients with epilepsy and categorized them into three
subgroups: uncontrolled epilepsy (UCE), well-controlled epilepsy (WCE), and poorly controlled epilepsy
Keywords: (PCE). All subjects were asked to complete the Korean versions of the Fatigue Severity Scale (K-FSS), the
Epilepsy
Neurological Disorders Depression Inventory for Epilepsy (K-NDDI-E), the Generalized Anxiety Disorder-
Fatigue
7 (K-GAD-7) scale, and the short forms of the Patient-Reported Outcomes Measurement Information
Sleep
Depression System Sleep-Related Impairment (PROMIS-SRI) and Sleep Disturbance (PROMIS-SD) scales. Addition-
Seizures ally, 200 normal control subjects who completed the K-FSS, K-NDDI-E, and K-GAD-7 measures were
included. The K-FSS scores of the epilepsy subgroups and the control group were compared, and stepwise
multiple regression analysis was performed to identify predictors of high scores on the K-FSS among
epilepsy patients.
Results: The K-FSS, K-NDDI-E, and K-GAD-7 scores were higher in the epilepsy patients than in the
controls. The K-FSS scores of the UCE subgroup, but not of the PCE and WCE subgroups, were higher than
those of the control group. K-FSS scores of epilepsy patients were predicted by PROMIS-SRI and K-NDDI-
E scores.
Conclusions: Fatigue was more severe in epilepsy patients than in healthy controls without epilepsy,
especially when seizures were not controlled. Sleep-related impairments and depression aggravated
fatigue in epilepsy patients.
2015 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.

1. Introduction Moreover, fatigue in epilepsy patients may precipitate their

seizures [14,15] and, for this reason, a better understanding of
Fatigue has been dened as the experience of extreme and fatigue in epilepsy patients is crucial to effectively manage the
persistent tiredness, weakness, or exhaustion that can be mental, course of the disease and the treatment regimen. However, only a
physical, or both [1]. Fatigue is associated with and may be few studies have compared the degree of fatigue in epilepsy
aggravated by neurological disorders such as multiple sclerosis, patients with that of the general population and, even studies that
Parkinsons disease, and stroke [26]; similarly, fatigue is also a have assessed the propensity for fatigue in epilepsy patients, have
common complaint of epilepsy patients [712]. The prevalence of used relatively small numbers of patients and controls [1].
fatigue ranges from 35.0% to 66.7% in epilepsy patients [912] but Likewise, studies investigating the predictors of fatigue in
only occurs in 1025% of the general population [7,8,13]. The epilepsy patients are also relatively rare. Although several studies
fatigue experienced by epilepsy patients is more severe than that have shown that fatigue in epilepsy patients is related to sleep
of healthy volunteers and the degree of fatigue in these patients is quality, depression, and anxiety [1,12,16,17], the variables
comparable to that of patients with multiple sclerosis [1]. associated with epilepsy, including seizure types, seizure freedom,
and factors related to antiepileptic drugs (AEDs), have not been
correlated with fatigue [1,1618]. Of these studies, one [1]
* Correspondence ing author. Tel.: +82 53 420 5769; fax: +82 53 422 4265. observed a tendency for increased fatigue based on the number
E-mail address: [email protected] (S.-P. Park). of AEDs or the number of seizures, but without statistical

http://dx.doi.org/10.1016/j.seizure.2015.12.003
1059-1311/ 2015 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.
 O.-Y. Kwon, S.-P. Park / Seizure 34 (2016) 4853 49

signicance. Nonetheless, the relationships between fatigue and a was diagnosed according to the criteria of the International League
number of epilepsy-related factors remain unclear. Against Epilepsy (ILAE) for seizures and epileptic syndromes [19].
Thus, the present study aimed to determine the degree and Subjects younger than 19 years of age and older than 70 years of
predictors of fatigue in epilepsy patients by comparing their age and subjects with severe neurological, psychiatric, or other
characteristics with those of healthy control subjects. To accom- disorders that prevented them from understanding the ques-
plish this, a wide variety of epilepsy-related variables, including tionnaires and fully cooperating with the study were excluded
socioeconomic factors, seizure-related features, and depression, from the nal analyses. The present study initially included 320
anxiety, and sleep-related problems were assessed to determine epilepsy patients, but 50 were excluded for the following reasons:
the predictors of fatigue in epilepsy patients. refusal to complete the questionnaires (n = 24), severe neurologi-
cal or other disorders (n = 19), psychosis (n = 1), being older than
2. Methods 70 years of age (n = 4), less than 1 year of AED treatment (n = 1),
and lack of education (n = 1). Thus, 270 epilepsy patients were
The present cross-sectional study was approved by the included in the nal analyses of the present study.
Institutional Review Board of Kyungpook National University The demographic, socioeconomic, and clinical characteristics of
Hospital. Upon enrollment, all subjects provided informed consent the study subjects are summarized in Table 1. The epilepsy
and were asked to complete a battery of reliable and validated self- patients were classied into three subgroups based on the state of
report health questionnaires that included the Korean versions of their seizure control: uncontrolled epilepsy (UCE), well-controlled
the Fatigue Severity Scale (K-FSS), the Neurological Disorders epilepsy (WCE), and poorly controlled epilepsy (PCE). UCE was
Depression Inventory for Epilepsy (K-NDDI-E), the Generalized dened as an average of more than one seizure per month for 18
Anxiety Disorder-7 (K-GAD-7) scale, and the short forms of the months and no seizure-free periods longer than 3 months, which
Patient-Reported Outcomes Measurement Information System were the criteria used to determine drug-refractory epilepsy as a
(PROMISTM) Sleep-Related Impairment (PROMIS-SRI) and Sleep failure in previous adequate trials of two AEDs [20]. WCE was
Disturbance (PROMIS-SD) scales. dened as freedom from seizures during the preceding year, and
PCE was dened as an intermediate degree of seizure control that
2.1. Subjects did not meet the criteria for UCE or WCE. Of the 270 epilepsy
patients, 49 were classied with UCE, 78 with PCE, and 143 with
Epilepsy patients who had been treated with AEDs for at least 1 WCE. The seizure-control classication for each epilepsy patient
year and who had attended the epilepsy clinic at Kyungpook was determined based on information about seizure frequency
National University Hospital between July 1, 2014 and January 31, obtained from their medical records. Additionally, 200 age- and
2015 were consecutively enrolled in the present study. Epilepsy sex-matched healthy adult volunteers were enrolled in the study

Table 1
Characteristics and questionnaire scores of eligible study subjects.

Characteristics Mean  SD (range) or percentage (%) p valuea

Epilepsy patients (n = 270) Controls (n = 200)

Age, years 39.8  12.4 (1970) 40.3  12.3 (1970) 0.677

Gender, male 168 (62.2%) 125 (62.5%) 1.000
Education, years 12.9  2.8 (620) 14.8  2.4 (620) <0.001
Job, yes 130 (48.1%) 137 (68.5%) <0.001
Household income, 1 (million KRW per month) 209 (77.4%) 192 (96.0%) <0.001
Drivers license, yes 162 (60.0%) 178 (89.0%) <0.001
Married but no divorce or bereavement 123 (45.6%) 125 (62.5%) <0.001
Concurrent medical disease 80 (29.6%) 28 (14.0%) <0.001
Age at onset, years 25.1  13.1 (163)
Duration of epilepsy, years 14.7  11.0 (157)
Type of seizure, partial 212 (78.5%)
Epilepsy syndrome
Temporal lobe epilepsy 131 (48.5%)
Extra-temporal lobe epilepsy 81 (30.0%)
Generalized epilepsy 50 (18.5%)
Unknown 8 (3.0%)
MRI, abnormal 122 (45.2%)
Family history of epilepsy 21 (7.8%)
History of febrile convulsions 60 (22.2%)
Duration of AED intake, years 11.6  10.0 (154)
AED regimen, monotherapy 125 (46.3%)
AED load 1.3  0.9 (0.24.6)
Seizure control
Well-controlled epilepsy 143 (53.0%)
Partially controlled epilepsy 78 (28.9%)
Uncontrolled epilepsy 49 (18.1%)
Co-administration of psychiatric drug 38 (14.1%)
PROMIS-SD 48.6  10.4 (28.976.5)
PROMIS-SRI 48.3  9.9 (30.080)
K-QOLIE-10 overall score 75.5  20.0 (7.5100)
a
Independent t-test or Chi-square test used for analysis.
KRW: Korean won, MRI: magnetic resonance imaging, AED: antiepileptic drug, PROMIS-SD: short form of Patient-Reported Outcomes
Measurement Information System-Sleep Disturbance, PROMIS-SRI: short form of Patient-Reported Outcomes Measurement
Information System-Sleep-Related Impairment, K-QOLIE-10: Korean version of Quality of Life in Epilepsy Inventory-10, SD: standard
deviation, MRI: magnetic resonance imaging
50 O.-Y. Kwon, S.-P. Park / Seizure 34 (2016) 4853

as control subjects. The comparisons of the characteristics of the Cronbachs a coefcient of the Korean version of the FSS is 0.935,
epilepsy patients and the controls subjects are summarized in and a total score of 3.22 or more is suggestive of suffering from
Table 1. fatigue [29].

2.2. Study design 2.3.5. The PROMISTM and subscales

To improve the quality of the assessments of patient-reported
The electronic medical records of the enrolled patients were outcomes, a cooperative group founded by scientists from several
reviewed to obtain information regarding patient-related variables US-based academic institutions and the National Institutes of
such as age, gender, concurrent medical diseases, and the co- Health was created. This group developed the PROMISTM item
administration of psychiatric drugs; epilepsy-related variables banks, which precisely and efciently measure patient-reported
such as family history of epilepsy, age at onset, duration of symptoms in individuals with various chronic diseases and
epilepsy, duration of AED medication use, AED regimens, drug load conditions [30]. More advanced sleep-associated questionnaires
of AEDs, etiology of epilepsy, seizure type, seizure freedom over the that assess two different aspects of sleep-related problems have
preceding year, and febrile convulsions; and psychosocial variables been developed from the PROMISTM: the PROMISTM-SRI and the
such as education, job, income, marriage, and possession of a PROMISTM-SD. The full version of the PROMISTM includes 16
drivers license. Drug load of AED was estimated as the sum of the PROMISTM-SRI items and 27 PROMISTM-SD items, whereas the
ratios of prescribed daily dose over dened daily dose for each short forms of the PROMISTM-SRI and PROMISTM-SD each has eight
AED in the regimen of individual patients [21]. The dened daily items. The convergent and discriminant validities of the short
dose means the assumed average daily maintenance dose of the forms of PROMISTM-SD and PROMISTM-SRI reect correlations with
AED when used for its main indication [22]. The K-NDDI-E, K-GAD- the full versions of PROMISTM-SD and F- PROMISTM-SRI, respec-
7, K-FSS, PROMIS-SRI, PROMIS-SD, and the Quality of Life (QOL) in tively [31]. The responders report the details of their sleep and
Epilepsy Inventory-10 (K-QOLIE-10) scale were completed by sleep-related disturbances over the last 7 days using a ve-point
patients. The control subjects completed the K-NDDI-E, K-GAD-7, scale [31]; thus, total scores on the short forms of the PROMISTM-
and K-FSS scales. SRI and PROMISTM-SD each range from 0 to 40. For the nal
interpretation of these questionnaires, the individual items are
2.3. Questionnaires summed and the corresponding t-scores are estimated from a
nonlinear transformation. The Korean version of the short forms of
2.3.1. The K-NDDI-E PROMISTM-SRI and PROMISTM-SD were used in the present study.
The K-NDDI-E is a quick, reliable, and validated screening tool
for major depressive disorder (MDD) in epilepsy patients [23]. This 2.4. Statistical analysis
is a six-item measure using a four-point scale (14) to evaluate the
degree to which an epilepsy patient has been bothered by The descriptive statistics, including counts, percentages, means,
depression-related problems over the previous 2 weeks. The total and standard deviations, are summarized in Table 1. Independent
scores range from 6 to 24, and higher scores indicate a more t-tests and chi-square tests were used to compare the K-FSS,
intense level of depression. A total score of 12 or more is suggestive K-NDDI-E, and K-GAD-7 scores of the epilepsy patients in all the
of MDD, and the Cronbachs a coefcient is 0.898 [23]. subgroups with those of the control subjects. A one-way analysis of
variance (ANOVA) was used to compare the scores among the
2.3.2. The K-GAD-7 study groups, and independent t-tests were used to compare each
The GAD-7 is a self-report questionnaire used for the rapid potential pair of groups. In order to obtain the most valid and
detection of generalized anxiety disorder [24]. This is seven-item accurate information possible, a p value <0.01 rather than <0.05
measure uses a four-point scale (03) to assess the degree to which was considered to indicate statistical signicance for all compar-
a subject has been bothered by anxiety-related problems over the isons.
previous 2 weeks. The total GAD-7 scores range from 0 to 21, and To determine the relationship between K-FSS scores and each
higher scores indicate a more intense level of anxiety. The present study variable, a Pearsons correlation coefcient analysis was used
study utilized the K-GAD-7, which can be downloaded from the to select the variables that were correlated with these scores; a p
Patient Health Questionnaire website [www.phqscreeners.com; value <0.05 was considered to indicate statistical signicance.
[25]. The K-GAD-7 has been validated, a total score of 7 or more in Using the selected variables, a stepwise multiple regression
the Korean version of the measure is suggestive of generalized analysis was performed to identify the best combination of
anxiety disorder, and the Cronbachs a coefcient is 0.924 [26]. predictors of a high K-FSS score; a p value <0.01 was considered to
indicate statistical signicance, and dummy variables were used as
2.3.3. The QOLIE-10 the categorical variables. Because QOL is the ultimate psychosocial
The QOLIE-10 is a 10-item self-administered questionnaire reection of the lives of epilepsy patients, the effect of K-QOLIE-10
specically designed to measure QOL in patients with PCE [27]. scores may have obscured the results of the stepwise regression.
This measure consists of subscales that address epilepsy effects, For this reason, K-QOLIE-10 scores were not included in the
mental health, and role functioning, and higher scores are stepwise regression analysis even though this score was signi-
indicative of a better QOL. The present study utilized the K- cantly correlated with the K-FSS score. A collinearity statistical
QOLIE-10. The Cronbachs a coefcient of the Korean version is analysis was also conducted as a redundancy check. SPSS software
0.843 for the epilepsy effects and role function subscales and 0.606 (version 21, IBM Inc.; SPSS Inc., Chicago, IL, USA) was used for all
for the mental health subscale [27]. statistical analyses.

2.3.4. The K-FSS 3. Results

The FSS consists of nine items that assess fatigue on a scale from
0 to 7 [28]. After summing the scores of the nine items, the total 3.1. Characteristics of epilepsy patients and control subjects
score is divided by 9, yielding values from 0 to 7. The FSS is useful in
clinical practice because it has fewer items than other ques- The demographic, socioeconomic, clinical, and psychosocial
tionnaires that evaluate fatigue and it is easy to score. The characteristics of epilepsy patients and control subjects are
 O.-Y. Kwon, S.-P. Park / Seizure 34 (2016) 4853 51

summarized in Table 1. Although the experimental and control Table 3

Variables correlated with K-FSS scores in epilepsy patients.
subjects were matched by age and gender, there were a number of
signicant differences (p < 0.001) between the two sets of Variable p value (r)a
subjects. The epilepsy patients were more likely to have concurrent Job 0.008 (0.162)
medical diseases; less likely to be married, have a job, or have a Drivers license 0.015 (0.148)
drivers license; had fewer total years of total education; and had a Married but no divorce or bereavement 0.004 (0.175)
lower level of income. Age at onset 0.016 (0.147)
Household income 0.019 (0.143)
AED regimen, monotherapy 0.016 (0.147)
3.2. Comparisons between epilepsy subgroups and control subjects Seizure control <0.001 (0.211)
Co-administration of psychiatric drugs <0.001 (0.283)
The K-FSS, K-NDDI-E, and K-GAD-7 scores of the epilepsy PROMIS-SD <0.001 (0.519)
PROMIS-SRI <0.001 (0.643)
patients, the epilepsy subgroups, and the control subjects are
K-QOLIE-10 <0.001 (0.546)
shown in Table 2. There were signicant differences in the K-FSS, K-NDDI-E <0.001 (0.497)
K-NDDI-E, and K-GAD-7 scores between the epilepsy patients and K-GAD-7 <0.001 (0.498)
the controls (p < 0.01). The comparisons between each potential a
Pearsons correlation analysis was used for the analysis.
pair of the three epilepsy subgroups and the control group revealed K-FSS: Korean version of Fatigue Severity Scale, AED: antiepileptic drug, PROMIS-
that the K-FSS, K-NDDI-E, and K-GAD-7 scores were higher in the SRI: short form of Patient-Reported Outcomes Measurement Information System
UCE subgroup than in the control group and that the K-NDDI-E Sleep-Related Impairment, PROMIS-SD: short form of Patient-Reported Outcomes
Measurement Information System Sleep-Disturbance, K-QOLIE-10: Korean version
scores were higher in the PCE subgroup than in the control group
of Quality of Life in Epilepsy Inventory-10, K-NDDI-E: Korean version of
(p < 0.01 for all comparisons). The comparisons of each potential Neurological Disorders Depression Inventory for Epilepsy, K-GAD-7: Korean
pair among the three epilepsy subgroups revealed that the K-FSS, version of Generalized Anxiety Disorder-7
K-NDDI-E, and K-GAD-7 scores were higher in the UCE subgroup
than in the WCE subgroup (p < 0.01).

3.3. Pearsons correlation coefcients for the relationship between K- two signicant variables, their effects were independent of each
FSS scores and each study variable other.

The study variables that were signicantly correlated with the 4. Discussion
K-FSS scores are listed in Table 3. The K-FSS scores were higher in
the following subgroups of epilepsy patients: those with no job, a The present study demonstrated that the fatigue experienced
low household income, no drivers license, who were unmarried, by epilepsy patients was more severe than that experienced by the
who had an early age of onset, who had no seizure freedom, who healthy control subjects. A previous study found that the severity
were undergoing a polytherapy, and who were undergoing the co- of fatigue is higher in epilepsy patients than in healthy controls
administration of psychiatric drugs. The K-FSS scores were also but the numbers of subjects in each group in that study were
higher in epilepsy patients with high PROMIS-SRI, PROMIS-SD, K- relatively small [1]. The present study included 270 epilepsy
NDDI-E, and K-GAD-7 scores and in epilepsy patients with low patients and 200 healthy control subjects, which is a greater
total K-QOLIE-10 scores. However, the duration of AED medication number of participants than included in other studies investigat-
use and drug load of AEDs were not associated with KFSS scores. ing fatigue in epilepsy patients [1,16,17]. This may render the
differences in fatigue observed in epilepsy patients and controls in
3.4. Predictors of the K-FSS score the present study more reliable and valid. Additionally, the
present study also compared the degree of fatigue of each epilepsy
The stepwise multiple regression analysis in the present study subgroup with age- and sex-matched control subjects. The K-FSS
produced a model with two variables that explained 44.9% of the scores of the UCE subgroup, but not of the PCE or WCE subgroups,
variance. The strongest predictors of K-FSS scores were PROMIS- were signicantly higher than those of the control group. The
SRI (b = 0.526, p < 0.01) and K-NDDI-E (b = 0.232, p < 0.01; present study also assessed numerous variables describing
Table 4) scores. The standardized b value revealed that the effect various characteristics of the subjects, including socioeconomic
of PROMIS-SRI scores on K-FSS scores were 2.27 times stronger and seizure-related variables, depression, anxiety, and sleep-
than that of K-NDDI-E scores. Because the redundancy check related problems, to identify the predictors of fatigue more
showed that the variance ination factors were less than 10 for the accurately. The results show that the fatigue in epilepsy patients

Table 2
Fatigue, depression, and anxiety in epilepsy patients compared with control subjects.

Mean  SD (range)

Total patients with epilepsy (n = 270) UCE (n = 49) PCE (n = 78) WCE (n = 143) Controls (n = 200)

K-FSS 3.1  1.5 (1.07.0)b 3.8  1.6 (1.16.7)b,** 3.1  1.4 (1.06.7) 2.9  1.4 (1.07.0) 2.7  1.3 (1.06.6)
K-NDDI-E 10.1  4.5 (624)b 12.5  5.7 (624)b,** 10.2  4.4 (624)b 9.2  3.8 (624) 8.7  2.9 (118)
K-GAD-7 4.7  5.4 (021)b 7.6  6.2 (021)b,** 5.3  5.9 (021)a,* 3.4  4.2 (021) 3.5  3.6 (019)
a
p < 0.05, Comparisons with controls.
b
p < 0.01, Comparisons with controls.
*
p < 0.05, Comparison with WCE.
**
p < 0.01, Comparison with WCE.
A one-way ANOVA was used for the comparisons of the scores of WCE, PCE, and UCE patients with those of the control subjects. Independent t-tests were performed to
compare the scores between the epilepsy patients and control subjects and between each potential pair of the three subgroups of epilepsy patients and control subjects. UCE:
uncontrolled epilepsy, PCE: poorly controlled epilepsy, WCE: well-controlled epilepsy, K-FSS: Korean version of Fatigue Severity Scale, K-NDDI-E: Korean version of
Neurological Disorders Depression Inventory for Epilepsy, K-GAD-7: Korean version of Generalized Anxiety Disorder-7, SD: standard deviation
52 O.-Y. Kwon, S.-P. Park / Seizure 34 (2016) 4853

Table 4 present study showed that seizure control was not a signicant
Predictors of K-FSS scores in epilepsy patients according to a multiple stepwise
predictor of fatigue in epilepsy patients.
regression analysis.
No correlations between fatigue and epilepsy-related variables,
Variable Standardized p value Collinearity Adjusted including seizure type, seizure freedom, and factors associated
coefcients () (VIF) R2
with AED treatment, have been observed in previous studies [1,16
0.449 18]. Of these studies, one observed that the number of AEDs and
PROMIS-SRI 0.526 <0.001 1.340 the number of seizures tended to increase fatigue, but these
K-NDDI-E 0.232 <0.001 1.340
relationships did not exhibit statistical signicance [1]. To date,
K-FSS: Korean version of Fatigue Severity Score, PROMIS-SRI: short form of Patient- only one study has found a high frequency of seizures to be a
Reported Outcomes Measurement Information System Sleep-Related Impairment,
signicant predictor of fatigue in epilepsy patients, but the study
K-NDDI-E: Korean version of Neurological Disorders Depression Inventory for
Epilepsy, VIF: variance ination factor
did not include psychosocial factors or sleep-related problems in
the multiple regression analysis [12].
There is a vicious cycle between sleep-related problems and
seizure control. Seizures are likely a primary cause of sleep-
was primarily associated with sleep-related impairments and related problems, and sleep-related problems may aggravate
depression. seizure control states [3335]. A few studies have investigated
Two previous studies found that sleep-related problems are seizure precipitants in epilepsy patients, and these found that
crucial predictors of fatigue in epilepsy patients, but the particular fatigue and sleep deprivation tended to precipitate seizures
aspects of the sleep-related problems differ among studies. Neves [14,15]. Thus, it is impossible to ignore the association between
et al. [16] included sleep quality and daytime sleepiness as fatigue and seizure control in epilepsy patients. In the context of
candidate variables in their analysis and found that sleep quality, these ndings, our observation suggests important associations
but not daytime sleepiness, was a signicant predictor of fatigue in among fatigue, sleep-related problems, and seizure control. This
epilepsy patients. In contrast, Hamelin et al. [17] found that study found a partial relationship among these variables; the
daytime sleepiness was an important predictor of fatigue in stepwise regression analysis showed that K-FSS scores were
epilepsy patients, although this study did not include sleep quality determined by PROMIS scores and that K-FSS scores were higher
as a candidate predictor. The present study utilized advanced in UCE patients but not in WCE patients compared with control
forms of questionnaires to evaluate sleep quality and daytime subjects (Table 2). Furthermore, PROMIS-SRI scores were higher
sleepiness: the PROMIS-SD and PROMIS-SRI, respectively. PROMIS- in the UCE subgroup than in the WCE subgroup (Table 5). These
SRI scores were signicant predictors of fatigue in epilepsy ndings may be interpreted to mean that fatigue has an indirect
patients, but PROMIS-SD scores were not, which is concordant relationship with seizure control. Although seizure control per se
with the ndings of Hamelin et al. [17]. may not directly be involved with the experience of fatigue, the
Epilepsy and depression have a bidirectional relationship and, seizure control state may inuence the fatigue of epilepsy patients
as a corollary of this relationship, depression is often comorbid via indirect mechanisms that are mediated by sleep-related
with epilepsy. It has been found that 937% of epilepsy patients impairments.
suffer from depression [32], and several studies have shown that Patients with UCE are more likely to be given additional AEDs
fatigue in epilepsy patients is determined by the presence of than are those with WCE and patients who receive more AEDs may
depression [9,16]. Neves et al. [16] determined that depression and be associated with greater levels of fatigue than those given lesser
sleep quality are signicant predictors of fatigue in epilepsy amounts of AEDs [1]. This association was also evident in the
patients and that the effects of depression are stronger than those correlation analysis in the present study; in which the K-FSS
of sleep quality. In the present study, sleep-related impairments, scores were higher in the epilepsy patients who underwent
but not sleep quality, was a signicant predictor of fatigue, and this polytherapy than in those who underwent monotherapy. AED
effect was 2.27 times stronger than that of depression. loads were not correlated with K-FSS scores and, despite the
The present study compared the degrees of fatigue between relationship between K-FSS scores and drug number, K-FSS scores
each potential pair of the three epilepsy subgroups and the control were not predicted by drug number. This suggests that the drug
group and found that K-FSS scores were signicantly higher in the number may also have indirect effects on fatigue by inuencing
UCE subgroup than in the control group. However, the K-FSS scores sleep-related impairments, in a similar way to the effects of
of the PCE and WCE subgroups did not differ from that of the seizure control on fatigue.
control group. These ndings suggest that fatigue may be more As observed in this and previous studies [9,16,17], sleep-related
severe in epilepsy patients when their seizures are not controlled. problems and depression may intensify fatigue in epilepsy
On the other hand, the stepwise multiple regression analysis in the patients. Thus, screening for depression and sleep-related

Table 5
Differences in sleep-related problems among the epilepsy subgroups according to seizure control in epilepsy patients.

Mean  SD (range)

UCE (n = 49) PCE (n = 78) WCE (n = 143)

PROMIS-SDa 51.86  11.05 (28.9076.50)** 49.34  10.06 (28.9076.50) 47.01  10.01 (28.9076.50)
PROMIS-SRIb 53.39  10.27 (30.0076.90)** 49.22  9.56 (30.0073.30)* 46.07  9.33 (30.0080.00)

A one-way ANOVA was conducted for the comparisons of the scores among WCE, PCE, and UCE patients. Independent t-tests were
performed to compare the scores between each potential pair of the three subgroups. UCE: uncontrolled epilepsy, WCE: well-controlled
epilepsy, PCE: poorly controlled epilepsy, PROMIS-SD: short form of Patient-Reported Outcomes Measurement Information System-
Sleep Disturbance, PROMIS-SRI: short form of Patient-Reported Outcomes Measurement Information System Sleep-Related
Impairment, SD: standard deviation
a
p < 0.05, Comparison among all three subgroups.
b
p < 0.01, Comparison among all three subgroups.
*
p < 0.05, Comparison with WCE.
**
p < 0.01, Comparison with WCE.
 O.-Y. Kwon, S.-P. Park / Seizure 34 (2016) 4853 53

problems should be conducted in epilepsy patients to prevent [13] Pawlikowska T, Chalder T, Hirsch SR, Wallace P, Wright DJ, Wessely SC.
Population based study of fatigue and psychological distress. Br Med J
fatigue. The present results also indicate that fatigue in epilepsy 1994;308:7636.
patients may be more severe than fatigue in healthy people when [14] Ferlisi M, Shorvon S. Seizure precipitants (triggering factors) in patients with
the seizures of the patients are not controlled and, therefore, epilepsy. Epilepsy Behav 2014;33:1015.
[15] Balamurugan E, Aggarwal M, Lamba A, Dang N, Tripathi M. Perceived trigger
fatigue levels must be considered when seizure control is an issue. factors of seizures in persons with epilepsy. Seizure 2013;22:7437.
The fatigue of epilepsy patients may be reduced by improving sleep [16] Neves GS, Gomes Mda M. Fatigue in patients with epilepsy and its association
hygiene and treating depression and, more specically, the present with depression and sleep quality. Arq Neuropsiquiatr 2013;71:5336.
[17] Hamelin S, Kahane P, Vercueil L. Fatigue in epilepsy: a prospective inter-ictal
study showed that the fatigue of UCE patients may be ameliorated and post-ictal survey. Epilepsy Res 2010;91:15360.
by improving seizure control. Conversely, the reduction of fatigue [18] Malow BA, Bowes RJ, Lin X. Predictors of sleepiness in epilepsy patients. Sleep
may also aid in the control of seizures. Sleep-related problems, 1997;20:110510.
[19] Proposal for classication of epilepsies and epileptic syndromes. Commission
depression, and fatigue have been shown to impair QOL in epilepsy
on classication and terminology of the international league against epilepsy.
patients [9,33,36,37]; thus, it is possible that reducing fatigue and Epilepsia 1985; 26:26878.
depression and improving sleep hygiene and seizure control may [20] Berg AT, Shinnar S, Levy SR, Testa FM, Smith-Rapaport S, Beckerman B. Early
lead to enhanced QOL in epilepsy patients. development of intractable epilepsy in children: a prospective study. Neurol-
ogy 2001;56:144552.
[21] Deckers CL, Hekster YA, Keyser A, Meinardi H, Renier WO. Reappraisal of
Conic of interest statement polytherapy in epilepsy: a critical review of drug load and adverse effects.
Epilepsia 1997;38:5705.
[22] World Health Organization Collaborating Centre for Drug Statistics Method-
The authors report no nancial interests or potential conicts of ology. About the ATC/DDD system, hhttp://www.whocc.no/atcddd/i;
interest related to the present study. [accessed 31.01.14].
[23] Ko PW, Hwang J, Lim HW, Park SP. Reliability and validity of the Korean version
of the Neurological Disorders Depression Inventory for Epilepsy (K-NDDI-E).
Acknowledgments Epilepsy Behav 2012;25:53942.
[24] Spitzer RL, Kroenke K, Williams JB, Lowe B. A brief measure for assessing
generalized anxiety disorder: the GAD-7. Arch Intern Med 2006;166:10927.
The authors report no nancial interests or potential conicts of [25] Pzer. Patient Health Questionnaire (PHQ) screeners. Available at: hhttp://
interest related to the present study. www.phqscreeners.com/i; [accessed 10.1.13].
[26] Seo JG, Cho YW, Lee SJ, Lee JJ, Kim JE, Moon HJ, et al. Validation of the
generalized anxiety disorder-7 in people with epilepsy: a MEPSY study.
References Epilepsy Behav 2014;35:5963.
[27] Lee SA, Yun SC, Lee BI. Reliability and Validity of the Korean Version of QOLIE-
[1] Hernandez-Ronquillo L, Moien-Afshari F, Knox K, Britz J, Tellez-Zenteno JF. 10 in Epilepsy. J Clin Neurol 2006;2:23845.
How to measure fatigue in epilepsy? The validation of three scales for clinical [28] Krupp LB, LaRocca NG, Muir-Nash J, Steinberg AD. The fatigue severity scale.
use. Epilepsy Res 2011;95:11929. Application to patients with multiple sclerosis and systemic lupus erythema-
[2] Vercoulen JH, Hommes OR, Swanink CM, Jongen PJ, Fennis JF, Galama JM, et al. tosus. Arch Neurol 1989;46:11213.
The measurement of fatigue in patients with multiple sclerosis. A multidi- [29] Chung K, Song C. Clinical usefulness of fatigue severity scale for patients
mensional comparison with patients with chronic fatigue syndrome and with fatigue, and anxiety or depression. Korean J Psychosom Med 2001;9:
healthy subjects. Arch Neurol 1996;53:6429. 16473.
[3] Kos D, Kerckhofs E, Nagels G, DHooghe MB, Ilsbroukx S. Origin of fatigue in [30] Cella D, Riley W, Stone A, Rothrock N, Reeve B, Yount S, et al. The Patient-
multiple sclerosis: review of the literature. Neurorehabil Neural Repair Reported Outcomes Measurement Information System (PROMIS) developed
2008;22:91100. and tested its rst wave of adult self-reported health outcome item banks:
[4] Abe K, Takanashi M, Yanagihara T. Fatigue in patients with Parkinsons disease. 20052008. J Clin Epidemiol 2010;63:117994.
Behav Neurol 2000;12:1036. [31] Yu L, Buysse DJ, Germain A, Moul DE, Stover A, Dodds NE, et al. Development of
[5] Lou JS, Kearns G, Oken B, Sexton G, Nutt J. Exacerbated physical fatigue and short forms from the PROMIS sleep disturbance and sleep-related impairment
mental fatigue in Parkinsons disease. Mov Disord 2001;16:1906. item banks. Behav Sleep Med 2011;10:624.
[6] Christensen D, Johnsen SP, Watt T, Harder I, Kirkevold M, Andersen G. Dimen- [32] Kwon OY, Park SP. Depression and anxiety in people with epilepsy. J Clin
sions of post-stroke fatigue: a two-year follow-up study. Cerebrovasc Dis Neurol 2014;10:17588.
2008;26:13441. [33] de Weerd A, de Haas S, Otte A, Trenite DK, van Erp G, Cohen A, et al. Subjective
[7] Chen MK. The epidemiology of self-perceived fatigue among adults. Prev Med sleep disturbance in patients with partial epilepsy: a questionnaire-based
1986;15:7481. study on prevalence and impact on quality of life. Epilepsia 2004;45:
[8] Hyyppa MT, Lindholm T, Lehtinen V, Puukka P. Self-perceived fatigue and 1397404.
cortisol secretion in a community sample. J Psychosom Res 1993;37:58994. [34] Allen RP. Article reviewed: reduction of rapid eye movement sleep by diurnal
[9] Ettinger AB, Weisbrot DM, Krupp LB, Coyle PK, Jandorf L, Devinsky O. Fatigue and nocturnal seizures in temporal lobe epilepsy. Sleep Med 2000;1:2478.
and depression in epilepsy. J Epilepsy 1998;11:1059. [35] Bazil CW, Castro LH, Walczak TS. Reduction of rapid eye movement sleep by
[10] Soyuer F, Erdogan F, Senol V, Arman F. The relationship between fatigue and diurnal and nocturnal seizures in temporal lobe epilepsy. Arch Neurol
depression, and event-related potentials in epileptics. Epilepsy Behav 2000;57:3638.
2006;8:5817. [36] Piperidou C, Karlovasitou A, Triantafyllou N, Terzoudi A, Constantinidis T,
[11] Senol V, Soyuer F, Arman F, Ozturk A. Inuence of fatigue, depression, Vadikolias K, et al. Inuence of sleep disturbance on quality of life of patients
and demographic, socioeconomic, and clinical variables on quality of life of with epilepsy. Seizure 2008;17:58894.
patients with epilepsy. Epilepsy Behav 2007;10:96104. [37] Park SP, Song HS, Hwang YH, Lee HW, Suh CK, Kwon SH. Differential effects of
[12] Kim K, Ha YS. Factors inuencing on fatigue in patients with epilepsy. J Korean seizure control and affective symptoms on quality of life in people with
Epilepsy Soc 2009;13:426. epilepsy. Epilepsy Behav 2010;18:4559.