Cognitive-behavioral screening in elderly patients with new-onset epilepsy before treatment
Cognitive comorbidity at epilepsy onset reflects disease severity and provides a baseline estimate of reserve capacities with regard to the effects of epilepsy and its treatment. Given the high incidence of epilepsy at an older age, this study analyzed objective and subjective cognition as well as quality of life in elderly patients with new-onset focal epilepsy before initiation of anti-epileptic treatment.
Materials and methods
A total of 257 untreated patients (60–95 years of age) with new-onset epilepsy underwent objective assessment of executive function (EpiTrack) and performed subjective ratings of cognition (Portland Neurotoxicity Scale) and quality of life (QoL; QOLIE-31).
According to age-corrected norms, 58% of patients (N = 257) demonstrated deficits in executive function; major determinants were cerebrovascular etiology, neurological comorbidity, and higher body mass index. Subjective ratings indicated deficits in up to 27% of patients. Self-perceived deficits were associated with neurological, cardiovascular, and/or psychiatric comorbidity, whereas poorer QoL was related to neurological comorbidity and female gender. Objectively assessed executive functions correlated with subjective social functioning, energy, motor function, and vigilance.
We found a relatively high QoL, a low rate of subjective impairment, but a high incidence of objective executive deficits in untreated elderly patients with new-onset epilepsy. Neurological status and body mass index, rather than seizure frequency or severity, were risk factors for cognitive impairment. Given the relevance of cognition in the course of epilepsy and its treatment, routine screening before treatment initiation is highly recommended.
The cognitive status of patients with new-onset epilepsy has recently been identified as a neuropsychological research topic of major relevance . Early identification of cognitive problems would allow timely intervention before the emergence of negative psychosocial consequences. As far as studies in untreated adult patients with newly diagnosed epilepsy have been conducted, the results revealed deficits primarily in attention, executive functions, and memory [2-8]. The patients' mean age in these studies was in the range of 30–47 years. Considering that the incidence of epilepsy varies across the life span, with peaks during both maturation and senescence of the brain, data on the cognitive status of elderly patients with new-onset epilepsy would be highly appreciated. In elderly patients with epilepsy, comorbidities accumulate, and cerebrovascular etiologies (e.g., stroke) represent one of the leading causes of epilepsy [9, 10].
In this study, we analyzed executive function, subjective complaints, and quality of life (QoL) in a large sample of elderly patients with new-onset focal epilepsy before the initiation of anti-epileptic treatment.
Data were obtained from patients participating in an investigator-initiated, randomized, double-blind, active comparator, parallel-group, clinical trial comparing initial treatment with controlled-release carbamazepine (CBZ-CR), lamotrigine (LTG), or levetiracetam (LEV) in patients aged ≥60 years with new-onset focal epilepsy (NCT00438451, EudraCT 2005-003324-19). Briefly, patients were included if they were older than 60 years and either had two spontaneous epileptic seizures with one occurring in the last 6 months, or one seizure plus epileptiform activity on EEG or an enduring relevant lesion on imaging. Patients were not to be treated before with anti-epileptic drugs, except for transient emergency treatment after the first seizure. Major exclusion criteria were seizures as the consequence of acute brain damage, dementia, renal failure, and drug abuse. Overall, 361 patients took part in the trial.
This trial was conducted in accordance with the Declaration of Helsinki, with good clinical practice guidelines and with Austrian, German, and Swiss law, as applicable. National and local independent ethics committees, where appropriate, approved the protocol which included cognitive testing. Written informed consent was obtained from all patients before admission into the trial.
The original study design comprised assessments of cognition and quality of life before treatment initiation (baseline) and at two follow-ups. For the aim of this study, only the data sets from the evaluation at baseline were considered. Originally, 303 of the 361 recruited patients had undergone cognitive assessment. Due to incomplete test data in 46 of these patients, a total of 257 patients remained for the final analysis (Table 1).
Table 1. Patient demographics and clinical characteristics
|Mean (SD), range||71.5 (7.2), 60–95|
|≥70 years, n (%)||138/257 (54%)|
|Body mass index (kg/m2)||[N = 254]|
|Mean (SD), range||26.8 (3.9), 18–38|
|Type of epilepsy, n (%)|
|Focal symptomatic||225/257 (87.5%)|
|Focal cryptogenic||32/257 (12.5%)|
|Imaging, n (%)|
|Pathological CT||151/201 (75.1%)|
|Pathological MRI||134/159 (84.3%)|
|Pathological CT and/or MRI||225/257 (87.5%)|
|Etiology, n (%)|
|Localization, n (%)|
|Lateralization, n (%)|
|Seizure types, n (%)|
|Simple partial||35/257 (13.6%)|
|Complex partial||209/257 (81.3%)|
|Generalized tonic–clonic||139/257 (54.1%)|
|EEG, n (%)|
|Comorbidities, n (%)|
|Seizure frequency (life time seizures)||[N = 251]|
|Mean (SD), range||4.2 (9.8), 1–96|
The diagnostic package comprised an objective measure of executive function, subjective ratings of cognition and somatomotor function, and a QoL inventory.
EpiTrack® (second edition with recently extended and revised norms) is a screening tool for tracking adverse cognitive effects of anti-epileptic medication . The test includes six subtests: response inhibition, visuo-motor speed, mental flexibility, visuo-motor planning, verbal fluency, and working memory. Based on the subtest results, an age-corrected total score is calculated. Age-corrected norms from 689 healthy individuals (age range 16–87 years) are available . Higher scores reflect better performance with a maximum score of 49 points. A score in the range of 29–31 points indicates mild impairment, and the cutoff for significant impairment is ≤28 points.
A recent study demonstrated the usefulness of the EpiTrack for cognitive screening of people with newly diagnosed epilepsy . Given its sensitivity with regard to the overall drug load, that is, the number of concurrent anti-epileptic drugs (AEDs) [12, 13], it is particularly useful for monitoring the cognitive effects of pharmacological treatments [14-16].
Subjective ratings of cognitive and somatomotor function
The Portland Neurotoxicity Scale (PNS) is a brief patient-based survey of cognitive and somatomotor complaints commonly encountered with the use of AEDs . The scale assesses subjective impairment via 15 nine-tiered (1 = no problem, 9 = severe problem) ratings: vision, energy level, memory, walking, interest, coordination, tremor, concentration, speech, forgetfulness, sleepiness, moodiness, alertness, attention span, and motivation. Higher scores indicate more severe problems.
Quality of life
The widely used Quality of Life in Epilepsy (QOLIE)-31 questionnaire  is a short form of the QOLIE-89 with 31 items addressing health status and seven domains (seizure worry, overall quality of life, emotional well-being, energy–fatigue, cognitive functioning, medication effects, social functioning). The questionnaire has been translated and validated for use in Germany . We applied the standard scoring form; that is, the subscale answers were transformed into scores ranging from 0 to 100, and the total score was calculated by weighting according to predefined values and adding up the individual subscores. Higher scores reflect better QoL.
We applied descriptive statistics to characterize the cognitive status and QoL at epilepsy onset and stepwise regression analysis [combining forward (P < 0.05) and backward selection (P < 0.10)] to identify factors explaining variance in behavioral measures. Patients' results were compared to published test norms in order to determine the presence/absence of cognitive or behavioral problems. Values of > 1 and ≤ 2 standard deviations below the mean of the respective normative sample indicated mild impairment or below average performance, whereas values of > 2 standard deviations below the mean of the normative sample were categorized as marked impairment or well below average. Correlation analysis was conducted to determine potential relations between behavioral measures. Analyses of variance were used to compare group means.
Analysis of the age-corrected EpiTrack data revealed mild impairment of executive function in 15.2% and marked impairment in 43.2% of the patients, whereas 35.4% were classified as unimpaired and 6.2% as above average (Table 2). When considering demographic (age, sex) and clinical variables (etiology, seizure types, frequency of life-time seizures, pathological EEG, comorbidities) within a stepwise regression analysis, worse performance in executive function was associated with infarction (β = −0.28, t = 4.3, P < 0.001) or vascular etiology (β = −0.14, t = 2.3, P = 0.022), neurological comorbidity (β = −0.19, t = 2.9, P = 0.003), and higher body mass index (BMI) (β = −0.12, t = 2.0, P = 0.044) (final model F = 12.1, P < 0.001; R2 = 0.17). In the subgroup of patients with available information on the localization, extratemporal lobe epilepsies (N = 46) scored significantly worse than temporal lobe epilepsies (N = 66) (29.9 vs 27.6, F = 7.5, P = 0.007). The hemispheric lateralization of the lesion/epilepsy was not relevant.
Table 2. Executive function, subjective complaints, and quality of life
|Total score (age-corrected)||257||29.3 (6.5)|| |
|Above average||257||16 (6.2%)|| |
|Unimpaired||91 (35.4%)|| |
|Mild impairment||39 (15.2%)|| |
|Marked impairment||111 (43.2%)|| |
|Total score||237||31.8 (15.3)||−0.25||P < 0.001|
|Vision||248||2.1 (1.9)||−0.18||P < 0.01|
|Energy level||250||2.1 (1.6)||−0.19||P < 0.01|
|Walking||252||2.3 (2.0)||−0.25||P < 0.01|
|Interest||252||1.9 (1.6)||−0.13||P < 0.05|
|Coordination||250||1.9 (1.5)||−0.17||P < 0.01|
|Tremor||245||1.8 (1.4)||−0.22||P < 0.01|
|Concentration||247||2.1 (1.4)||−0.21||P < 0.01|
|Speech||246||2.0 (1.5)||−0.20||P < 0.01|
|Sleepiness||246||2.5 (1.7)||−0.21||P < 0.01|
|Alertness||246||1.9 (1.3)||−0.16||P < 0.05|
|Attention span||247||2.1 (1.4)||−0.14||P < 0.05|
|Motivation||246||1.9 (1.4)||−0.17||P < 0.01|
|Above average||237||24 (10.1%)|| |
|Unimpaired||149 (62.9%)|| |
|Mild impairment||23 (9.7%)|| |
|Marked impairment||41 (17.3%)|| |
|Overall score||233||73.6 (13.8)||0.22||P < 0.01|
|Seizure worry||250||71.6 (24.9)||0.11||P < 0.05|
|Overall quality of life||253||71.3 (17.9)||0.18||P < 0.01|
|Emotional well-being||251||72.4 (16.6)||0.19||P < 0.01|
|Energy–fatigue||252||63.5 (18.6)||0.19||P < 0.01|
|Cognitive functioning||251||72.9 (18.8)||0.15||P < 0.05|
|Medication effectsc||238||74.5 (26.5)||−0.03||ns|
|Social functioning||242||82.7 (20.7)||0.21||P < 0.01|
|Above average||233||100 (42.9%)|| || |
|Average||125 (53.6%)|| || |
|Below average||7 (3%)|| || |
|Well below average||1 (0.4%)|| || |
While impairment of executive functions was objectively identified in a high proportion of patients, there were relatively few subjective reports about cognitive problems as assessed by the PNS (Table 2). Nevertheless, the mean PNS total score of 31.8 ± 15.3 indicated increased (t = 2.76, P < 0.01) problems compared with ratings of 55 healthy controls (27.5 ± 8.9) reported by Salinsky & Storzbach . Overall, 10% of patients had mild and 17% had marked impairment (Table 2). The mean scores of the individual PNS items ranged from 1.7 to 2.7 of possible nine points (Table 2). Forgetfulness and memory deficits were the most frequently reported problems: 14.6–16.8% of patients indicated that it is at least often a problem (i.e., scores ≥5). Stepwise regression analysis (F = 9.6, P < 0.001; R2 = 0.11) identified the following determinants of a higher (i.e., worse) PNS total score: presence of neurologic (β = 0.21, t = 3.3, P = 0.001), cardiovascular (β = 0.18, t = 2.8, P = 0.006), and/or psychiatric comorbidity (β = 0.15, t = 2.4, P = 0.016). An additional subgroup analysis revealed no impact of the site or side on the PNS total score.
With a mean overall score 73.6 ± 13.8 (maximum 100), the reported health-related QoL was better than the norm based on 304 patients with epilepsy presented in the QOLIE-31 scoring manual (62.9 ± 16.3; t = 8.23, P < 0.001). Only 3.4% scored poorer than that group (Table 2). The subscale with the lowest ratings was energy–fatigue (63.5 ± 18.6), whereas social functioning received the highest ratings (82.7 ± 20.7) (Table 2). According to regression analysis (F = 7.4, P = 0.001; R² = 0.06), better QoL was associated with the absence of neurological comorbidity (β = −0.19, t = −3.0, P = 0.003) and male sex (β = −0.17, t = −0.17, P = 0.011). According to a separate analysis, site or side were not related to the QOLIE-31 overall score.
The total scores of the PNS and QOLIE-31 were highly correlated (r = −0.70, P < 0.001, N = 220) and also correlated with EpiTrack performance (PNS r = −0.25, P < 0.001, N = 237; QOLIE-31 r = 0.22, P = 0.001, N = 233). Regarding the subordinate ratings and subscales, an explorative correlation analysis (Table 2) revealed highest correlations with executive function for walking (r = −0.25, P < 0.001), tremor (r = −0.22, P = 0.001), concentration (r = −0.21, P = 0.001), and sleepiness (r = −0.21, P = 0.001) of the PNS and with social functioning (r = 0.21, P = 0.001), emotional well-being (r = 0.19, P = 0.001), and energy (r = 0.19, P = 0.001) of the QOLIE-31.
Given the high incidence of epilepsy at an older age, this study analyzed executive function, subjective complaints, and QoL in elderly patients with new-onset focal epilepsy before initiation of anti-epileptic treatment. To our knowledge, this is the first such study in patients in this age group (60–95 years) with a mean age of more than 70 years. Considering age-corrected normative data, we found a large proportion of patients (58%) with objective impairment in executive function even at this early stage of the disease (43% marked, 15% mild impairment). The prevalence of deficits is higher than that reported for younger patients with a mean age of 47 ± 19 years (49% overall; 19% marked, 30% mild) .
In the current sample, greater impairment of performance was found to be associated with cerebrovascular etiology, neurological comorbidity, and higher BMI. Furthermore, consistent with the essential role of the frontal lobes for executive functions, patients with extratemporal lobe epilepsy scored worse than those with temporal lobe epilepsy. Other demographic or clinical variables such as lifetime seizures, seizure type, or pathological EEG findings had no predictive value.
In contrast to the high prevalence of impaired executive function identified through objective testing, subjective performance ratings indicated low insight into cognitive impairments. The PNS revealed some self-perceived problems, but did not appear to be very sensitive considering the younger age of the normative sample. Subjective complaints were associated with neurological, psychiatric, and/or cardiovascular comorbidities. Results of the QOLIE-31 indicated that patients reported better QoL than the younger norm group with established epilepsies. However, similar to the PNS, the QOLIE-31 lacks age-dependent norms. This result may also reflect the difference between new-onset and chronic epilepsy. Presumably, epilepsy variables at an early stage of the disease do not have yet a relevant impact on QoL. Instead, better QoL was related to the absence of neurological comorbidity, and male sex. Social functioning received the highest ratings, while items on fatigue and energy received the lowest. Subjective ratings of social functioning, energy, motor function, and vigilance were related to the objective performance in the test on executive functions. EpiTrack performance correlated well with the neurological status of patients and self-perceived problems in everyday functioning and QoL. It therefore has ecological validity beyond its already known sensitivity with regard to cognitive effects of anti-epileptic pharmacotherapy.
The fact that only executive function was assessed objectively is a limitation of this study. Given the reported subjective memory complaints, future studies should include objective memory assessment in addition to executive functions [8, 14]. Other limitations comprise lack of information on global intellectual functioning or education, which might have explained some of the variance. Although dementia was an exclusion criterion, a screening for mild cognitive impairment could have been worthwhile in this age group. Positive aspects of the study are the large sample size and the fact that only new-onset and not just newly diagnosed patients were included.
In conclusion, the findings indicate a high prevalence of cognitive deficits in untreated elderly patients with new-onset epilepsy. The results also highlight that relying solely on patients' reports of cognitive or neurological symptoms or QoL would result in an underestimation of cognitive problems at this early stage of the disease. Underreporting of cognitive problems has already been observed in the younger sample of patients with new-onset epilepsies reported in 2012 . Accordingly, an early cognitive screening should be considered mandatory to provide a baseline for the valid evaluation of the course of the disease and the effects of pharmacological treatment . Facing an increasingly older population, subjective measures in epilepsy with norms considering age and education are needed urgently. Finally, the increasing incidence of epilepsy in the elderly calls for more studies focusing on this population with consideration of age-related etiologies, comorbidities, and requirements for assessment and treatment [10, 20, 21].
A special thank goes to Mrs. C Beierle (statistical programing), Mrs. H Hofmann (study nurse), and Mrs. A Laupert (clinical monitor) without whose efforts this study would not have been possible. In addition, we would like to thank Dr. Azita Tofighy, UCB Pharma, for the editorial assistance, specifically, for reviewing the manuscript for the English language.
Conflict of interest and sources of funding statement
The authors would like to acknowledge UCB Pharma for providing an educational grant to support this study. JAW has no conflict of interest. KJW has received consultation/advisory board/speaking honoraria from UCB Pharma, Pfizer, Eisai, GlaxoSmithKline, Jansen-Cilag, and Desitin. KJW has received research funding from UCB Pharma and has been employed by UCB Pharma since January 2012. GK has received advisory board and speaker honoraria from Desitin, Eisai, GlaxoSmithKline, Lundbeck, Pfizer, UCB Pharma, and ViroPharma. CR has no conflict of interest. ET has acted as a paid consultant to Eisai, Ever Neuropharma, Biogen Idec, Medtronics, Bial, and UCB and has received speaker honoraria from Bial, Eisai, GL Lannacher, GlaxoSmithKline, Boehringer, Viropharma, Actavis, and UCB Pharma. ET has received research funding from UCB Pharma, Biogen-Idec, Red Bull, Merck, the European Union, FWF Österreichischer Fond zur Wissenschaftsförderung, and Bundesministerium fur Wissenschaft und Forschung. CH has received speaker honoraria from Desitin, Eisai, UCB Pharma, Viamed, GlaxoSmithKline, Pfizer and license fees from Viamed, EISAI, UCB Pharma.