www.elsevier.com/locate/yebeh

Epilepsy & Behavior 5 (2004) 976–980

Determinants of quality of life in epilepsyq

David W. Loringa,*, Kimford J. Meadora, Gregory P. Leeb

a Department of Neurology, University of Florida, Gainesville, FL, USAb Department of Neurology, Medical College of Georgia, GA, USA

Received 25 June 2004; revised 25 August 2004; accepted 27 August 2004

Abstract

Although depression is associated with diminished quality of life (QOL) in epilepsy patients, the relative contributions of epilep-sy-specific concerns, as well as clinical and cognitive variables of QOL, have not been simultaneously investigated. A comprehensiveneuropsychological test battery including the Beck Depression Inventory (BDI), Epilepsy Foundation of America�s (EFA) ConcernsIndex, MMPI-2, QOLIE-89, WAIS-III, and Selective Reminding was administered to 115 epilepsy surgery candidates with normalFull Scale IQs. Linear regression analyses were performed to identify significant predictor combinations of QOLIE-89 total score.Regression analysis demonstrated that depressive symptomatology, whether reflected by the BDI (R2 = 0.45) or Depression scale ofthe MMPI-2 (R2 = 0.36), was a robust individual QOL predictor. Seizure Worry from the EFA Concerns Index was nearly as effec-tive as the BDI in predicting QOLIE-89 (R2 = 0.42). When the BDI and EFA Concerns Index were combined into the same regres-sion, both factors continued to contribute significantly to the QOLIE-89 total score, with both variables accounting for 61% of thevariance. Although patients who developed their seizures at an older age had poorer QOL and patients with higher educationallevels reported higher QOL, neither factor was related to QOL after accounting for the effects of psychological variables and epi-lepsy-related concerns. Although quality of life has multiple determinants, symptoms of depression and seizure worry are the mostimportant factors affecting QOL in patients with intractable epilepsy.� 2004 Elsevier Inc. All rights reserved.

Keywords: Depression; Quality of life; EFA Concerns Index; Epilepsy

1. Introduction

Over the past decade, there has been increased inter-est in how epilepsy and its treatment affect the quality oflife (QOL) of epilepsy patients. Disease-specific QOLinfluences contrast with general life satisfaction factorsthat independently affect QOL. In epilepsy, QOL mea-sures have been used not only to characterize disease

1525-5050/$ - see front matter � 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.yebeh.2004.08.019

q Presented at the 2004 Meeting of the American Academy ofNeurology, San Francisco, CA.

* Corresponding author. Department of Neurology, University ofFlorida, McKnight Brain Institute, PO Box 100236, Gainesville,Fl 32610. Fax: +1 352-392-6893.

E-mail address: david.loring@neurology.ufl.edu (D.W. Loring).

burden, but also to reflect the effects of AED treatment[1–3], seizure severity [4], and epilepsy surgery [5–7].

A common measure of QOL scales is the Quality ofLife in Epilepsy—89 (QOLIE-89) [8]. Although multiplecontent scores can be derived, the most common sum-mary measure is the overall QOLIE-89 score. The use-fulness of the summary score is evident in the multipleshort forms of the QOLIE that contain many fewerquestions, but generate an overall score with fewer con-tent scales (i.e., QOLIE-31, QOLIE-10) [9,10].

A parallel development to study the experience of epi-lepsy on everyday activities is reflected in the EpilepsyFoundation of America (EFA) Concerns Index [11].The EFA Concerns Index is a 20-item scale on whichpatients rate specific epilepsy concerns (e.g., being a bur-den or worry to family), and was derived from epilepsy

Table 1Patient scores on QOLIE-89 and depression measures, and patientfrequencies by depression severity

QOLIE-89 Total (mean, SD) 41.9 (11.0)

Beck Depression Inventory (mean, SD) 13.6 (9.1)Normal (0–9) N = 41 (35.7%)Mild–moderate (10–18) N = 46 (40.0%)Moderate–severe (19–29) N = 20 (17.4%)Severe (30+) N = 8 (7.0%)

MMPI-2 Depression (Scale 2) (mean, SD) 68.3 (14.9)Normal (<64) N = 55 (47.8%)Mild (65–74) N = 24 (20.9%)Moderate (75–84) N = 19 (16.5%)Severe (84+) N = 17 (14.8%)

D.W. Loring et al. / Epilepsy & Behavior 5 (2004) 976–980 977

patients who listed their concerns over living with recur-rent seizures in order of importance [12]. The EFA Con-cerns Index has been successful in demonstrating thebeneficial effects of anterior temporal lobectomy [11].

Epilepsy is relatively unique among chronic neuro-logic diseases in its potential influence on QOL. Epilepsyoften begins at a young age and may hinder social andcognitive development. In addition, epilepsy is episodic,occurs unpredictably, and typically involves loss of con-sciousness, leading to driving and employment restric-tions. Importantly, epilepsy is associated with highrates of psychiatric comorbidity [13–16], and amongpsychiatric inpatients, the association between diseaseand QOL is greatest for patients with depression [17].The effect of depressive symptomatology on QOL in epi-lepsy is strong, in many reports accounting for approx-imately half of the variance when compared with formalQOL measures [16,18,19]. After accounting for the ef-fects of depression, other potentially relevant clinicalfactors are either unrelated to QOLIE-89 or accountfor very small amounts of the QOLIE-89 total score,suggesting that dysphoria and stress-related complaintsovershadow the effects of other variables. In the presentstudy, we describe the relative contributions of specificepilepsy concerns to two commonly employed measuresof depression to determine their relative contributions toQOL. In addition, we explore the role of clinical factorsand cognition in QOL.

2. Methods

2.1. Subjects

All patients had medically refractory epilepsy andwere undergoing evaluation for epilepsy surgery at theMedical College of Georgia (MCG); 115 patients wereretrospectively identified from the MCG Epilepsy Sur-gery database with Full Scale IQ levels in the normalrange (i.e., FSIQ P 70). Patients with FSIQ < 70 wereexcluded (n = 11) to avoid the potential confound ofpoor reading and comprehension on our results.

Seventy-three patients subsequently underwent resec-tive surgery. Reasons for not having surgery includedpoorly localized seizure onset, bilateral independent sei-zures, generalized seizure onset, and change of interestin persuing a surgical option. This sample averaged34.2 (SD = 11.5) years of age and had a mean age ofhabitual seizure onset of 19.7 (SD = 14.1) years andaverage education of 12.4 (SD = 2.3) years. The meanFSIQ was 87.0 (SD = 12.2).

2.2. Tests

In addition to cognitive assessment including theWAIS-III [20] and Verbal Selective Reminding Test

[21], patients were administered the QOLIE-89 [8], BeckDepression Inventory [22], MMPI-2 [23], and EFA Con-cerns Index [11]. Patients with WAIS-III Full Scale IQsbelow 70 were excluded. Mean QOLIE-89 total scores,and BDI and MMPI-2 depression levels are summarizedin Table 1. In addition, the frequency of patients at dif-ferent levels of depression severity is included in thetable.

2.3. Analysis

Stepwise linear regression was employed to evaluatethe contribution of the above variables to QOLIE-89.Because of data redundancies (i.e., collinearity), we per-formed regression analyses on data subsets prior to theomnibus regression, which included all potential predic-tor variables. Although number of AEDs has been stud-ied in previous QOL reports [16], this variable was notpart of our database and could not be reliably estab-lished retrospectively. In each case, the summary QO-LIE-89 measure was used as the dependent variable.

For this first regression, we used variables includingage, education, age of habitual seizure onset, Full ScaleIQ, and Verbal Selective Reminding Memory Perfor-mance (Continuous Long-Term Retrieval and DelayedRecognition) as potential predictors. Additional inde-pendent analysis employed clinical MMPI-2 scales withand without the BDI, and the final analysis regressedEFA concern against QOLIE-89 Overall Score.

3. Results

For all analyses, the dependent variable was totalQOLIE-89 score. From the clinical and cognitive vari-ables, age at habitual seizure onset value accounted for14% of the QOLIE-89 total score variance—older pa-tients developing seizures had lower QOLIE-89 scoresthan those developing epilepsy at a younger age. Yearsof education accounted for an additional 11% of the var-iance, with higher education associated with a higher

Fig. 2. Scatterplot of QOLIE-89 total score and EFA Seizure Worry.

978 D.W. Loring et al. / Epilepsy & Behavior 5 (2004) 976–980

QOLIE-89 score, yielding a two-variable total QOLIE-89 prediction of 25%.

This regression analysis including the BDI andMMPI-2 clinical scales revealed three significant predic-tors: the BDI accounted for 45% of the variance in QO-LIE-89 total score (see Fig. 1), the MMPI-2 Scale 3(Hysteria) accounted for 10% of the variance, andMMPI-2 Scale 2 (Depression) accounted for an addi-tional 2% of the variance (total explained vari-ance = 57%). In all cases, higher-scale elevations wereassociated with lower QOLIE-89 scores. Because ofthe considerable overlap between the BDI and MMPI-2 Depression scale (Scale 2) in identifying depressionin epilepsy patients [24], MMPI-2 scores were submittedto an independent regression analysis without the BDI.A similar pattern was present, with MMPI-2 Depressionthe strongest predictor and accounting for 36% of thevariance, followed by Scale 3 (7%), Scale 7 (4%), andScale 5 (2%) (total explained variance = 49%). Althoughslightly different patterns emerged from the two regres-sion analyses, both results indicate that depression is amore significant psychological factor associated withQOL than any remaining psychological content area as-sessed by the MMPI-2.

The regression analysis examining the EFA con-cerns yielded significant results for Seizure Worry(EFA14, R2 = 0.42) (see Fig. 2), Cognitive Difficulty(EFA18, incremental R2 = 13), Nervousness/Depres-sion (EFA19, incremental R2 = 0.04), Family Burden(EFA9, incremental R2 = 0.02), and Having to TakeMedication (EFA3, incremental R2 = 0.02).

The omnibus regression including all variables as po-tential QOLIE-89 predictors entered BDI withR2 = 0.45 first into the equation. This was followed bySeizure Worry (EFA14, incremental R2 = 0.16), whichyielded a two-test total R2 of 0.61. All remaining statis-

Fig. 1. Scatterplot of QOLIE-89 total score and Beck DepressionInventory scale.

tically significant effects provided incremental predic-tions of 6% or less. Cognitive Difficulty (EFA18) hadan incremental R2 of 0.06, MMPI-2 Scale 3 an incremen-tal R2 of 0.03, and MMPI-2 Scale 9 an incremental R2 of0.02. Nervousness/Depression (EFA19), Education, andAge each independently contributed an additional 1% tothe regression prediction.

4. Discussion

This report demonstrates that depression symptom-atology and seizure worry are the two most importantfactors associated with QOL in patients with medicallyrefractory epilepsy. Although other statistically signifi-cant predictors were identified, their contributions toQOLIE-89 after accounting for the BDI and seizureworry were 6% or less. Age at habitual seizure onsetand years of education were both related to QOLIE-89total; however, the magnitude of their contributionwas relatively small and did improve the regressionprediction beyond depression and seizure worry. Norelationship between QOLIE-89 and cognitive perfor-mance was detected.

A strong relationship between measures of depressionand mood and QOL in epilepsy has previously been de-scribed. For example, the mood factor derived from fac-tor analysis of the Profile of Mood States scoresaccounted for 47% of the variance in QOLIE-89 in anonsurgical epilepsy sample [19]. This mood factorwas derived from within the patient sample, however,and was not based on an independent measure.Although not explicitly addressing depression, psycho-logical distress, loneliness, adjustment and coping, andstigma perception contributed significantly to QOL inepilepsy patients derived from outpatient settings [25].Significant association between the QOLIE-89 and

D.W. Loring et al. / Epilepsy & Behavior 5 (2004) 976–980 979

either the Symptom Checklist-90-R Depression Scale[26] or Center for Epidemiologic Studies DepressionScale [18] has been noted. The studies, however, didnot employ established clinical measures of depressionsymptomatology.

There have been, however, several studies examiningtraditional measures of depression and mood on QOLscales. By use of validated German measures of depres-sion and epilepsy QOL, depression was the strongestpredictor of health-related quality of life in a sampleof patients with temporal lobe epilepsy who were beingevaluated for surgery [27]. BDI scores accounted for51% of QOLIE-31 total score variance, with no effectsnoted for age, seizure frequency, or seizure duration ina sample of epilepsy surgery candidates [16]. With theBDI for classification, 46% were classified as normal,24% had mild-moderate depression, 23% had moder-ate-severe depression, and 7% displayed severe depres-sion. These figures are comparable to our results inwhich BDI scores accounted for 45% of the QOLIE-89total score. Although our sample has slightly fewer pa-tients in the normal range and more with mild-moderatedepressive symptoms, the overall severity classificationsappear similar in both reports. We also note that noneof our patients were formally diagnosed using DSM-IV criteria. Although the BDI and MMPI-2 Depressionscales are commonly used to assess dysphoria and stress-related complaints, neither should be used outside of aformal psychiatric or psychological interview for diag-nostic purposes.

A significant correlation between the MMPI-2Depression scale and QOLIE-89 Language, Memory,and Attention/Concentration factors was described ina cohort of epilepsy surgery candidates, although a cor-relation with overall QOLIE-89 score was apparentlynot performed [28]. Because the MMPI-2 Depressionscale assesses somatic complaints that may be endorsedby epilepsy patients independent of depressed mood, itmay have decreased specificity for depression in thispopulation [24], and our results in which we observeda stronger relation between QOLIE-89 and the BDIrather than MMPI-2 Depression scale support this view.

We observed a relationship between QOLIE-89 andage of habitual seizure onset, with older age associatedwith poorer QOL. A smaller but still significant relation-ship (3% of variance) was noted with age at first unpro-voked seizure [16], although this measure is slightlydifferent than the age of habitual seizure onset we em-ployed. We also observed that education levels modu-lated QOL, an effect that had previously beendescribed in a nonsurgical sample [29]. Although thesefactors did not statistically contribute to QOLIE-89 pre-diction after accounting for the effects of depression andseizure worry in our sample, these factors statistically re-lated to QOL and, in the absence of depressive symp-tomatology, will contribute to QOL.

This study demonstrates that depressive symptom-atology is a significant factor affecting QOL in epilepsy,but also that QOL reflects factors other than mood. Sei-zure worry appears to be the greatest epilepsy concernaffecting QOL, accounting for 42% of the total QO-LIE-89 variance by itself, which is only slightly less thanthe contribution of the BDI by itself (45%). In addition,there is a unique contribution of seizure worry (16%)after accounting for BDI effects. Although a strong rela-tionship exists between depression and seizure worry(r = 0.43), there are large independent contributions ofeach factor related to QOL. Thus, a large portion ofan epilepsy patient�s QOL depends on mood and thepresence of depressive symptomatology and, fromextension from seizure worry concerns, being seizure-free.

References

[1] Birbeck GL, Hays RD, Cui X, Vickrey BG. Seizure reduction andquality of life improvements in people with epilepsy. Epilepsia2002;43:535–8.

[2] Gilliam FG, Fessler AJ, Baker G, Vahle V, Carter J, Attarian H.Systematic screening allows reduction of adverse antiepilepticdrug effects: a randomized trial. Neurology 2004;62:23–7.

[3] Sackellares JC, Kwong WJ, Vuong A, Hammer AE, Barrett PS.Lamotrigine monotherapy improves health-related quality of lifein epilepsy: a double-blind comparison with valproate. EpilepsyBehav 2002;3:376–82.

[4] Vickrey BG, Berg AT, Sperling MR, et al. Relationships betweenseizure severity and health-related quality of life in refractorylocalization-related epilepsy. Epilepsia 2000;41:760–4.

[5] McLachlan RS, Sadler M, Pillay N, et al. Quality of life aftervagus nerve stimulation for intractable epilepsy: is seizure controlthe only contributing factor?. Eur Neurol 2003;50:16–9.

[6] Spencer SS, Berg AT, Vickrey BG, et al. Initial outcomes in theMulticenter Study of Epilepsy Surgery. Neurology2003;61:1680–5.

[7] Lowe AJ, David E, Kilpatrick CJ, et al. Epilepsy surgery forpathologically proven hippocampal sclerosis provides long-termseizure control and improved quality of life. Epilepsia2004;45:237–42.

[8] Devinsky O, Vickrey BG, Cramer J, et al. Development of thequality of life in epilepsy inventory. Epilepsia 1995;36:1089–104.

[9] Cramer JA, Perrine K, Devinsky O, Meador K. A briefquestionnaire to screen for quality of life in epilepsy: theQOLIE-10. Epilepsia 1996;37:577–82.

[10] Cramer JA, Perrine K, Devinsky O, Bryant-Comstock L, MeadorK, Hermann B. Development and cross-cultural translations of a31-item quality of life in epilepsy inventory. Epilepsia1998;39:81–8.

[11] Gilliam F, Kuzniecky R, Meador K, et al. Patient-orientedoutcome assessment after temporal lobectomy for refractoryepilepsy. Neurology 1999;53:687–94.

[12] Gilliam F, Kuzniecky R, Faught E, Black L, Carpenter G,Schrodt R. Patient-validated content of epilepsy-specific quality-of-life measurement. Epilepsia 1997;38:233–6.

[13] Hermann BP, Seidenberg M, Bell B. Psychiatric comorbidity inchronic epilepsy: identification, consequences, and treatment ofmajor depression. Epilepsia 2000;41:S31–41.

[14] Kanner AM, Palac S. Depression in epilepsy: a common but oftenunrecognized comorbid malady. Epilepsy Behav 2000;1:37–51.

980 D.W. Loring et al. / Epilepsy & Behavior 5 (2004) 976–980

[15] Gilliam F, Hecimovic H, Sheline Y. Psychiatric comorbidity,health, and function in epilepsy. Epilepsy Behav 2003;4:26–30.

[16] Boylan LS, Flint LA, Labovitz DL, Jackson SC, Starner K,Devinsky O. Depression but not seizure frequency predictsquality of life in treatment-resistant epilepsy. Neurology2004;62:258–61.

[17] Russo J, Roy-Byrne P, Reeder D, et al. Longitudinal assessmentof quality of life in acute psychiatric in patients: reliability andvalidity. J Nerv Ment Dis 1997;185:166–75.

[18] Cramer JA, Blum D, Reed M, Fanning K. The influence ofcomorbid depression on quality of life for people with epilepsy.Epilepsy Behav 2003;4:515–21.

[19] Perrine K, Hermann BP, Meador KJ, et al. The relationship ofneuropsychological functioning to quality of life in epilepsy. ArchNeurol 1995;52:997–1003.

[20] Wechsler D. WAIS-III administration and scoring manual. SanAntonio, TX: Psychological Corp.; 1997.

[21] Buschke H, Fuld PA. Evaluation of storage, retention, andretrieval in disordered memory and learning. Neurology1974;11:1019–25.

[22] Beck AT. An inventory for measuring depression. Arch GenPsychiatry 1987;4:561–71.

[23] Butcher JN, Dahlstrom WG, Graham JR. Manual for therestandardized Minnesota Multiphasic Personality Inventory:MMPI-2. Minneapolis: Univ. of Minnesota Press; 1989.

[24] Karzmark P, Zeifert P, Barry J. Measurement of depression inepilepsy. Epilepsy Behav 2001;2:124–8.

[25] Suurmeijer TPBM, Reuvekamp MF, Aldenkamp BP. Socialfunctioning, psychological functioning, and quality of life inepilepsy. Epilepsia 2001;42:1160–8.

[26] Hermann BP, Seidenberg M, Bell B, Woodard A, Rutecki P,Sheth R. Comorbid psychiatric symptoms in temporal lobeepilepsy: association with chronicity of epilepsy and impact onquality of life. Epilepsy Behav 2000;1:184–90.

[27] Lehrner J, Kalchmayr R, Serles W, et al. Health-related quality oflife (HRQOL), activity of daily living (ADL) and depressive mooddisorder in temporal lobe epilepsy patients. Seizure 1999;8:88–92.

[28] Breier JI, Fuchs KL, Brookshire BL, et al. Quality of lifeperception in patients with intractable epilepsy or pseudoseizures.Arch Neurol 1998;55:660–5.

[29] Djibuti M, Shakarishvili R. Influence of clinical, demographic,and socioeconomic variables on quality of life in patients withepilepsy: findings from Georgian study. J Neurol NeurosurgPsychiatry 2003;74:570–3.