Subjective perception of cognition is related to mood and not performance

Results Across all three experiments, significant correlations were more frequent ( χ 2 = 259, P ⩽ 0.000) for mood versus subjective cognitive perception (59%) compared with subjective versus objective cognition (2%) and mood versus objective cognitive performance (2%). Conclusions Subjective perception of cognitive effects is related more to mood than objective performance. Clinicians should be aware of this relationship when assessing patients’ cognitive complaints. Keywords Antiepileptic drugs Quality of life Parkinson’s disease Neuropsychological assessment Depression Cognition 1 Introduction Practitioners commonly rely on patient self-report of perceived difficulties with memory, attention, concentration, and language when assessing how much a disease state or its treatment might be affecting day-to-day cognitive functioning. Similarly, self-accounts of cognitive performance are often sought from participants in clinical research trials to monitor the safety, efficacy, and side effects of the therapy or drug being studied. Therefore, the extent to which subjective perceptions of cognitive functioning correspond to objective measures of cognitive performance is crucial in determining whether or not self-reports should be considered valid appraisals of neuropsychological status on which diagnostic, treatment, research, or regulatory decisions should be based. As subjective memory complaints may serve as an indicator of more general cognitive decline [1] , the relationship between the subjective perception of memory impairment and objective measures of memory function has been studied extensively, especially in the elderly. In a number of cross-sectional studies of patients without dementia, subjective memory complaints were more highly correlated with measures of mood, especially depression and anxiety, than with objective measures of cognitive performance [2–7] . Memory is also negatively impacted by epilepsy [8] , and patients with epilepsy are apt to report more significant memory disturbances than individuals without epilepsy [9] . Not only are self-reported memory problems by patients with epilepsy weakly correlated with standard memory and neuropsychological measures [10,11] , but patients with subjective memory complaints are more likely to be depressed or anxious than those who report no memory difficulties [10,12–14] . Using the Quality of Life in Epilepsy Inventory (QOLIE-89) to measure subjective perceptions of domains known to affect daily living as well as the Profile of Mood States (POMS), Perrine et al. [15] found that mood significantly correlated with self-reported QOLIE-89 cognition sub-scales for memory, attention/concentration, and language, as well as the overall QOLIE-89 score. They also reported smaller, but significant correlations between the objective neuropsychological tests and the three-self reported QOLIE-89 cognition subscale measures, suggesting that self-reports of memory impairment by persons with epilepsy may be a function of both mood and objective memory functioning. However, the degree to which neuropsychological measures used in these studies reflect the types of memory problems experienced in day-to-day living has been called into question [12,16] . Elixhauser et al. [17] sought to address this issue by using the Rivermead Behavioural Memory Test (RBMT), a measure explicitly designed to assess everyday memory function, and found similar associations between everyday behavioral memory performance, subjective perception of memory difficulties (QOLIE-89), and self-reported mood (POMS) in patients with epilepsy. Similar to Perrine et al. [15] , Elixhauser et al. [17] found a strong association between mood and perceived cognitive function, but a smaller, yet significant, correlation between objective memory performance and perceived cognitive function, regardless of whether neuropsychological tests or measures of everyday performance were used. Investigations into the validity of self-reports of cognitive function in patients with epilepsy have employed subjects whose seizures were controlled by antiepileptic drugs (AEDs). A cross-sectional study by Uijl et al. [18] attempted to tease out the contribution of AEDs in the number of subjective complaints that patients with epilepsy conveyed to their physician. Sixty-seven percent of the 173 patients with well-controlled epilepsy reported moderate to severe complaints, most often regarding cognitive dysfunction. The two factors most closely associated with the number and severity of these complaints were high scores on Psychoneuroticism (Symptom Check List-90) and AED polytherapy. These results support the influence of mood on the subjective perception of cognitive impairment, but Uijl et al. [18] did not validate those complaints by objective neuropsychological testing. In fact, there are no studies in the literature that have examined the direct effect of drug treatment on subjective perception of cognitive function and its relationship to both mood and objective performance despite the fact that this is the primary method of assessing cognitive side effects of drugs in clinical settings and in clinical trials for regulatory approval of drugs. We previously published two double-blind studies that compared the effects of two AEDs on neuropsychological function in healthy volunteers [19] and in patients with epilepsy [20] . In this article, we reanalyzed those data to investigate the relationship of subjective perception, mood, and objective cognitive performance as a function of effects of treatment with AEDs, that is, lamotrigine (LTG) and topiramate (TPM). We hypothesized that changes in perception of cognition across drug/nondrug conditions would be more related to changes in mood than to changes in objective cognitive performance. As the number of studies that have examined the validity of subjective perception of cognitive performance in patients with neurological diseases remains limited, we also examined the robustness of the relationship between subjective perception of cognition and mood in a new cross-sectional study, which is the first to examine this association in patients with Parkinson’s disease. 2 Materials and methods All studies were conducted under the principles of the Helsinki Accords and were approved by the local ethics committees. 2.1 Experiment 1: Healthy volunteer AED cognition study Methodology and prior results have been described in detail elsewhere [19] and are outlined below. 2.1.1 Study design and prior results This was a randomized, double-blind, repeated-measures, crossover investigation. As previously reported, subjects had significantly better performance on 80% of the neuropsychological variables for LTG compared with TPM; there were no variables that exhibited superior performance for TPM [19] . 2.1.2 Subjects Forty-seven healthy paid volunteers (19 men, 28 women; mean age = 37 years) without a history of neurological or psychiatric diseases including drug abuse completed this study. 2.1.3 Procedure Subjects were treated with each AED in monotherapy for a 12-week period (7 weeks of dose escalation followed by 4 weeks of maintenance, and then a 1-week taper off of AED) followed by a 4-week washout period off AED. Target dose was 300 mg/day for both drugs. Subjects underwent neuropsychological testing on four occasions: two AED conditions and two nondrug conditions. The mean (range) blood level for LTG on the day of neuropsychological testing was 4.7 μg/mL (1.6–7.0) and for TPM was 9.3 μg/mL (2.8–15.8). The mean LTG dose was 298 mg/day and the mean TPM dose was 300 mg/day. 2.1.4 Neuropsychological tests The test battery to assess AED neuropsychological effects consisted of 17 tests. 2.1.4.1 Cognitive performance measures 1. Attention/vigilance: Continuous Performance Task, Digit Cancellation Test, Visual Serial Addition Test. 2. Memory: MCG Paragraph Memory, Selective Reminding Test. 3. Language: Controlled Oral Word Association Test (COWA), Boston Naming Test, Semantic Fluency Test. 4. Cognitive and motor speed: Lafayette Grooved Pegboard, Choice Reaction Time. 5. Other cognitive tests: Stroop Color–Word Test, Symbol Digit Modalities Test (SDMT). 2.1.4.2 Behavioral measures Profile of Mood States (POMS—adjective checklist to assess mood; subscales include Tension/Anxiety, Depression, Anger/Hostility, Vigor, Fatigue, and Confusion/Bewilderment), and subjective cognitive function rating scales (i.e., attention, language, and memory) from the Quality of Life in Epilepsy-89 (QOLIE-89). The A-B Neurotoxicity Scale, Adverse Events Profile, Side Effects and Life Satisfaction (SEALS), and SF-12 (generic 12-item quality of life scale) were also administered but were not included in the present analysis. Thus, a total of 14 tests with 33 variables were used in the present analysis including three subjective perception of cognition, seven mood, and 23 objective cognitive performance measures for each of the two AED conditions. 2.2 Experiment 2: Patients with epilepsy AED cognition study Study methodology and prior results have been described in detail elsewhere [20] and are outlined below. 2.2.1 Study design and prior results This was a randomized, double-blind, repeated measures, parallel-group investigation. As previously reported, cognitive performance at the end of the maintenance phase was better for LTG than TPM when using a combined summary score reflecting performance on all neuropsychological measures (415 vs 315, P < 0.001) [20] . 2.2.2 Subjects One hundred ninety-two patients (116 men, 76 women) aged 18 years or older (mean age = 40 ± 13) with a diagnosis of partial epilepsy who were currently receiving carbamazepine or phenytoin as monotherapy or combined with one additional AED that did not change the enzyme-inducing status of carbamazepine or phenytoin completed the study. The mean (±SD) baseline depression score (Center for Epidemiological Studies Depression Scale) was 17 (12). 2.2.3 Procedures In this randomized, double-blind, parallel-group design, randomization was stratified by age (<50 years, >50 years) so that comparable numbers of patients in different age groups received each treatment, thereby minimizing the influence of age-related changes in cognition. A screening phase of up to 2 weeks was followed by an 8-week dose-escalation phase during which LTG and TPM were introduced and titrated to target doses of 500 mg/day for LTG and 300 mg/day for TPM, on which patients were maintained for 8 weeks. Mean (±SD) daily doses were 493.6 mg (33.9 mg) for LTG and 299.3 mg (4.7 mg) for TPM. Mean (SD) trough serum concentrations were 4.8 (3.9) μg/mL for LTG and 4.9 (2.5) μg/mL for TPM at end of maintenance. 2.2.4 Neuropsychological tests The following neuropsychological measures were administered at screening, end of dose-escalation phase, and end of maintenance phase: COWA, Stroop Color–Word Test, Digit Cancellation, Lafayette Grooved Pegboard, Rey Auditory Verbal Learning Test (RAVLT), and SDMT, Center for Epidemiological Studies Depression Scale (CES-D), POMS, and the cognitive scale from the Quality of Life in Epilepsy-31 (QOLIE-31, single subjective cognitive measure). The Adverse Events Profile was also given but was not included in the present analyses. A total of 15 variables were analyzed: one subjective perception of cognition, eight mood, and six objective cognitive performance measures for each of the two AED groups. 2.3 Experiment 3: Parkinson’s disease cognition study 2.3.1 Study design This was a cross-sectional analysis with assessment of neuropsychological function. 2.3.2 Subjects Fifty-eight patients (19 female, 39 male) with idiopathic Parkinson’s disease with a mean age of 68.3 ± 8.5 years and disease duration of 6.88 ± 5.04 years served as subjects. None of the subjects had clinical depression; all subjects were on stable dopaminergic therapy. 2.3.3 Neuropsychological tests The battery of tests included: 1. General cognitive function: Mini Mental State Examination (MMSE) [21] and the Mattis Dementia Rating Scale Total Scaled Score (DRS Total Score) [22] . 2. Frontal function: COWA, Semantic Animal Naming Fluency, Stroop Task interference, Trail Making Test Part B [23] . 3. Memory: Hopkins Verbal Learning Test—Delayed Recall [24] . 4. Depression: Geriatric Depression Scale (GDS) [25] . 5. Perceived cognition/memory dysfunction: subscale items of the PDQ-39 [26] . There were 11 total variables: three subjective perception, one mood, and seven objective cognitive performance. 2.4 Data analyses For Experiments 1 and 2, change scores between the nondrug condition and each drug state were calculated for each participant across the objective cognitive, subjective perception of cognitive function, and mood measures. Pearson correlations for the change scores between the objective cognitive, subjective cognitive, and mood measures were calculated, resulting in 502 correlations for Experiment 1 and 124 correlations for Experiment 2. For Experiment 3, a total of 31 Pearson correlations were used to assess the relationship of perceived cognitive functioning to objective cognitive performance and mood. We chose a P value <0.01 for all studies as a compromise between type I and II errors to provide some control for the multiple comparisons, but also to permit examination of the pattern of correlations across the three sets of experiments. As a further control for type I error, the pattern of correlations across all studies combined were compared using χ 2 analysis. 3 Results Correlation results for the combined experiments and for each individual experiment are summarized in Table 1 . The occurrence of significant correlations across all three experiments differed for the comparisons of subjective cognitive perception, mood, and objective cognitive performance ( χ 2 = 259, P ⩽ 0.000). Comparisons of mood with subjective cognitive perception were much more likely to exhibit significant correlations. The pattern of correlations was similar for each of the three experimental groups. 3.1 Experiment 1: Healthy volunteer AED cognition study 3.1.1 Relationship between perception of cognition and mood 3.1.1.1 LTG treatment The subjective measure of memory (QOLIE–Memory [nondrug average minus drug]) showed a significant (P < 0.01) negative correlation with the total score on the POMS ( r = −0.424), as well as with four of six subscales of the POMS: Depression ( r = −0.424), Anger ( r = −0.367), Fatigue ( r = −.411), Confusion ( r = −0.524). QOLIE–Attention was negatively correlated ( P < 0.01) with the total POMS score ( r = −0.691), as well as with five of six subscales: Tension ( r = −0.425), Depression ( r = −0.663), Anger ( r = −0.470), Fatigue ( r = −0.548), Confusion ( r = −0.733). QOLIE–Language was negatively correlated ( P < 0.01) with the total POMS score ( r = −0.416), as well as with three of six subscales: Anger ( r = −0.417), Fatigue ( r = −0.488), and Confusion ( r = −0.633). 3.1.1.2 TPM treatment There was no demonstrable relationship between QOLIE–Memory and any mood measures. QOLIE–Attention was negatively correlated ( P < 0.01) with POMS–Fatigue ( r = −0.519) and positively correlated with POMS–Vigor ( r = 0.416). QOLIE–Language showed a negative correlation with total POMS average score ( r = −0.444), as well as with Fatigue ( r = −0.408), and Confusion ( r = −0.482). 3.1.2 Relationship between cognitive performance and mood 3.1.2.1 LTG treatment There were significant positive correlations ( P < 0.01) between Choice Reaction Time–Initiation and total POMS average score ( r = 0.391) as well as with Anger ( r = 0.434), Confusion ( r = 0.500), and Fatigue ( r = 0.440). 3.1.2.2 TPM treatment The Animal Naming Fluency Test was negatively associated ( P < 0.01) with the total POMS score ( r = −0.417) and Confusion ( r = −0.461). Boston Naming Test was significantly associated with POMS–Vigor ( r = −0.373). Grooved Pegboard was positively associated with POMS–Anger ( r = 0.385). 3.1.3 Relationship between cognitive performance and perception of cognition 3.1.3.1 LTG treatment Both QOLIE–Language ( r = −0.456, P = 0.002) and QOLIE–Attention ( r = −0.384, P = 0.009) were associated with only a single measure of cognitive functioning, Choice Reaction Time–Initiation. QOLIE–Memory showed no association with cognitive performance. 3.1.3.2 TPM treatment No significant correlations were reported between any of the three measures of perception of cognition and those of cognitive performance. 3.2 Experiment 2: Patients with epilepsy AED cognition study 3.2.1 Relationship between perception of cognition and mood 3.2.1.1 LTG treatment Bivariate correlations on change scores across all points revealed significant correlations ( P < 0.01 with r values ranging from −0.316 to 0.626) for the QOLIE-31 cognitive subscale to all eight measures from the CES-D and POMS. 3.2.1.2 TPM treatment Five of eight measures (i.e., CES-D and POMS–Confusion, Fatigue, Tension, and Total) were significantly correlated with the QOLIE-31 cognitive subscale. 3.2.2 Relationship between cognitive performance and mood 3.2.2.1 LTG treatment There were no significant correlations between any of the objective measures and those of mood. 3.2.2.2 TPM treatment Only Digit Cancellation showed a significant association with a mood measure (POMS–Depression: r = 0.349, P < 0.005). 3.2.3 Relationship between perception of cognition and cognitive performance 3.2.3.1 LTG treatment No objective measures were correlated to the QOLIE-31 cognitive subscale. 3.2.3.2 TPM treatment COWA was the only objective measure of cognitive performance that was significantly associated with the QOLIE-31 cognitive subscales ( r = 0.304, P < 0.007). 3.3 Experiment 3: Parkinson’s disease cognition study 3.3.1 Relationship between perception of cognition and mood All three measures of perceived cognition had moderate positive association with the GDS ( P < 0.001). 3.3.2 Relationship between cognitive performance and mood There were no significant correlations between any of the objective measures of cognition and the GDS. 3.3.3 Relationship of perception of cognition and cognitive performance There was no significant association between any of the measures of perceived cognition and measures thought to assess frontal lobe functioning (i.e., COWA, Stroop, and Trail Making Test Part B). None of the measures of perception of cognition were correlated with the memory measure (i.e., Hopkins Verbal Learning Test—Delayed). When the MMSE and Mattis DRS were used as measures of general cognitive function, the perceived cognition item of the PDQ-39 was significantly associated with MMSE ( r = −0.364, P = 0.005), but none of the items of perceived cognition/memory dysfunction were significantly correlated with the DRS. 4 Discussion We have demonstrated in two double-blind, randomized, repeated-measures design studies of healthy volunteers (crossover) and patients with epilepsy (parallel group) that the subjective perception of AED cognitive effects is more related to mood than objective neuropsychological performance. Similar to prior cross-sectional studies in patients with epilepsy [15,17] , we have also demonstrated the same phenomenon in patients with Parkinson’s disease. Self-report of cognitive impairment can be affected by some diseases (e.g., anosognosia). However, the pattern of a strong relationship of mood to subjective perception of cognition in contrast to our other comparisons was also seen in healthy volunteers and in a variety of disease states including those without a primary effect on the brain. The pattern of significant correlations is unlikely due to chance given that the large majority of significant correlations occurred for mood versus subjective cognitive perception, which was statistically greater than other contrasts. Further, the findings in our present studies are consistent with a number of prior studies from disparate patient populations including epilepsy [15,17] , psychogenic nonepileptic seizures [27] , coronary artery bypass surgery [28] , Cushing’s disease [29] , cancer [30] , multiple sclerosis [31,32] , and psychiatric disorders [33,34] . The observation of a similar pattern across independent samples, different clinical conditions, different neuropsychological measures, and different experimental designs in our studies, as well as in others, replicates the finding that mood is a primary contributor to subjective cognitive complaints, and illustrates the robustness of this effect. Researchers and clinicians must consider with skepticism the subjective cognitive ratings of healthy subjects or patients with a wide range of medical conditions, as there is now sufficient evidence for the position that these reports are frequently a function of mood rather than actual cognitive functioning. One limitation of this study is the relatively large number of correlations performed. Because no consensus exists regarding type I error control, the choice between minimizing type I errors versus type II errors should be based on context-dependent pragmatic considerations where informed personal judgment plays a vital role [35] . We consider our approach using a conservative value of 0.01 to indicate statistical significance to represent an appropriate balance. Although it is possible that some of the significant correlations result from chance associations, the overall pattern of findings cannot be easily attributed to chance or spurious correlations alone, as evidenced by our χ 2 analysis. Thus, the pattern more likely reflects a dissociation between objective and subjective measures of cognition in their relationship to mood. We do not assert that there is no relationship between subjective and objective measures of cognitive abilities. Several studies including our own have shown a significant correlation between objective cognitive test performance and perceived cognitive function, though the correlations are not nearly as frequent or strong as those between mood and subjective reports [15,17,36] . For example, Elixhauser et al. found that both mood and objective memory performance were independent predictors of perceived memory functioning, but the explained variance for mood and memory performance combined ( R 2 = 0.58) is only slightly higher than the predictive value for the mood scales alone ( R 2 = 0.56) [17] . We report few significant correlations between mood and cognitive performance in the present study. Depression can produce a reversible cognitive impairment called pseudodementia [37] . The cognitive performance of both adult and elderly patients with unipolar depression has also been reported to improve after treatment with antidepressants [38] , but the improvement can be modest or absent [39] . Moreover, nonelderly patients with unipolar or bipolar depression have been shown to have impaired cognition even in a euthymic state [40] . Therefore, the degree to which depression affects cognitive performance is not consistent across all groups studied. We found few significant correlations between mood and cognitive performance in the present study. In subjects who manifested a negative mood, the extent of their depressed state may not have been sufficient to affect their objective cognitive performance, but was sufficient to affect their perception of everyday functioning; none of the present studies was designed to test this hypothesis. Nevertheless, our results clearly underscore the necessity for clinicians and researchers to take into account mood, medical condition, and drug treatment when assessing the validity of subjective cognitive ratings. In the clinical setting, physicians monitor adverse cognitive side effects of medications by spontaneous patient reports or by asking patients. Clinical trials conducted for regulatory approval of drugs depend primarily on patient report to determine safety with regard to cognitive side effects. Thus, patient care and regulatory safety assessments rely on patients’ perception to determine adverse drug-induced cognitive side effects, even though their perception is much more related to mood than actual cognitive performance. Patient perception alone may not adequately assess for important cognitive side effects in both clinical practice and research investigations including regulatory studies. This concern is especially important in groups with increased susceptibility such as children and the elderly. Even a modest untoward effect on cognitive performance might be problematic in children over the course of neurodevelopment or in the elderly due to their increased vulnerability to drug-induced cognitive side effects. Problems with objective measurement of drug-induced decrements in cognitive performance include time, costs, sensitivity of the measurements, test–retest effects, availability of trained personnel for cognitive assessments, and confounding effects of various clinical factors. Within research trials, these effects can be addressed and should be considered a part of the regulatory process. In a clinical practice setting, there is presently no proven and pragmatic methodology to assess patients on an individual basis. Standardized, brief computerized cognitive batteries and electroencephalographic measures combined with mood and drug toxicity questionnaires offer hope [41] , but additional research is needed. In the interim, clinicians need to consider the impact of mood on their patients’ complaints of cognitive problems. They should assess mood and evaluate its contribution in comparison to the underlying disease state and the potential risk that the drug may adversely affect cognitive function. Possible therapeutic actions might include treatment of mood or other contributing diseases or alteration of the drug regimen. Contributors All investigators participated in the preparation of the article after the initial draft was prepared by the corresponding author. Role of the funding source Data from prior studies funded by GlaxoSmithKline and the INFORM PD Database, National Parkinson Foundation of Excellence, were used. Conflict of interest statement J.M. Miller and R.P. Kustra are employees of GlaxoSmithKline and hold equity in the company. K.J. Meador and M.A. Werz have received grants in excess of $10,000 from GlaxoSmithKline separate from the present study grant. M.A. Werz, and M.R. Schoenberg have received honoraria from GlaxoSmithKline. Remaining authors report no conflict of interest. Acknowledgments The following people performed statistical analyses: Experiment 1 (healthy volunteers)—co-author V.J. Vahle while at Washington University, St. Louis; Experiment 2 (patients with epilepsy)—A.E. Hammer, Biomedical Data Sciences, GlaxoSmithKline; Experiment 3 (Parkinson’s disease)—co-author A. Roy while at the University of Florida. This work was also supported by the Center of Excellence grant from the National Parkinson Foundation and the UF INFORM-PD database. Though the actual research reported in this paper was not supported by NIH, I was being supported by an NIH grant (100% effort) via a Career Development Award (K01) during the time that I wrote the paper. References [1] S. Jungwirth P. Fischer S. Weissgram W. Kirchmeyr P. Bauer K.H. Tragl Subjective memory complaints and objective memory impairment in the Vienna-Transdanube aging community J Am Geriatr Soc 52 2004 263 628 [2] G.J. Larrabee H.S. Levin Memory self-ratings and objective test performance in a normal elderly sample J Clin Exp Neuropsychol 8 1986 275 284 [3] K.I. Bolla K.N. Lindgren C. Bonaccorsy M.L. Bleecker Memory complaints in older adults: fact or fiction? Arch Neurol 48 1991 61 64 [4] A.P.J. Barker R. Jones Memory complaints in attendees at a self-referral memory clinic: the role of cognitive factors, affective symptoms and personality Int J Geriatr Psychiatry 10 1995 777 781 [5] C. Derouesne L. Lacomblez S. Thibault M. LePoncin Memory complaints in young and elderly subjects Int J Geriatr Psychiatry 14 1999 291 301 [6] R. Stewart C. Russ M. Richards C. Brayne S. Lovestone A. Mann Depression, APOE genotype and subjective memory impairment: a cross-sectional study in an African-Caribbean population Psychol Med 31 2001 431 440 [7] T.S.C. Minett J.L. Dean M. Firbank P. English J.T. O’Brien Subjective memory complaints, white-matter lesions, depressive symptoms, and cognition in elderly patients Am J Geriatr Psychiatry 13 2005 665 671 [8] B.P. Hermann M. Seidenberg A.R. Woodard Memory and epilepsy: current perspectives Clin Nurs Pract Epilepsy 3 1996 4 8 [9] P.J. Thompson R. Corcoran Everyday memory failures in people with epilepsy Epilepsia 33 Suppl. 6 1992 S18 S20 [10] A.R. Giovagnoli S. Mascheroni G. Avanzini Self-reporting of everyday memory in patients with epilepsy: relation to neuropsychological, clinical, pathological and treatment factors Epilepsy Res 28 1997 119 128 [11] A. Piazzini M.P. Canevini G. Maggiori R. Canger The perception of memory failures in patients with epilepsy Eur J Neurol 8 2001 613 620 [12] R. Corcoran P. Thompson Epilepsy and poor memory: who complains and what do they mean? Br J Clin Psychol 32 Pt. 2 1993 199 208 [13] S.M. Sawrie R.C. Martin R. Kuzniecky Subjective versus objective memory change after temporal lobe epilepsy surgery Neurology 53 1999 1511 1517 [14] A. Au P. Leung A. Kwok P. Li C. Lui J. Chan Subjective memory and mood of Hong Kong Chinese adults with epilepsy Epilepsy Behav 9 2006 68 72 [15] K. Perrine B.P. Hermann K.J. Meador The relationship of neuropsychological functioning to quality of life in epilepsy Arch Neurol 52 1995 997 1003 [16] J. Vermeulen A.P. Aldenkamp W.C.J. Alpherts Memory complaints in epilepsy: correlations with cognitive performance and neuroticism Epilepsy Res 15 1993 157 170 [17] A. Elixhauser N.K. Leidy K. Meador E. Means M.K. Willian The relationship between memory performance, perceived cognitive function, and mood in patients with epilepsy Epilepsy Res 37 1999 13 24 [18] S.G. Uijl C. Uiterwaal A.P. Aldenkamp A cross-sectional study of subjective complaints in patients with epilepsy who seem to be well-controlled with anti-epileptic drugs Seizure Eur J Epilepsy 15 2006 242 248 [19] K.J. Meador D.W. Loring V.J. Vahle Cognitive and behavioral effects of lamotrigine and topiramate in healthy volunteers Neurology 64 2005 2108 2114 [20] D. Blum K. Meador V. Biton Cognitive effects of lamotrigine compared with topiramate in patients with epilepsy Neurology 67 2006 400 406 [21] M.F. Folstein S.E. Folstein P.R. McHugh “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician J Psychiatr Res 12 1975 189 198 [22] R. Schmidt W. Freidl F. Fazekas The Mattis Dementia Rating Scale: normative data from 1, 001 healthy volunteers Neurology 44 1994 964 966 [23] K. Arbuthnott J. Frank Trail Making Test, Part B as a measure of executive control: validation using a set-switching paradigm J Clin Exp Neuropsychol 22 2000 518 528 [24] S.P. Woods J.C. Scott E. Conover T.D. Marcotte R.K. Heaton I. Grant Test–retest reliability of component process variables within the Hopkins Verbal Learning Test—Revised Assessment 12 2005 96 100 [25] W.J. Burke W.H. Roccaforte S.P. Wengel The short form of the Geriatric Depression Scale: a comparison with the 30-item form J Geriatr Psychiatry Neurol 4 1991 173 178 [26] D.M. Bushnell M.L. Martin Quality of life and Parkinson’s disease: translation and validation of the US Parkinson’s Disease Questionnaire (PDQ-39) Qual Life Res 8 1999 345 350 [27] J.D. Fargo B.K. Schefft J.P. Szaflarski Accuracy of self-reported neuropsychological functioning in individuals with epileptic or psychogenic nonepileptic seizures Epilepsy Behav 5 2004 143 150 [28] P. Khatri M. Babyak C. Clancy Perception of cognitive function in older adults following coronary artery bypass surgery Health Psychol 18 1999 301 306 [29] A. Heald C. Parr C. Gibson K. O’Driscoll H. Fowler A cross-sectional study to investigate long-term cognitive function in people with treated pituitary Cushing’ disease Exp Clin Endocrinol Diabetes 114 2006 490 497 [30] G. Kibiger K.L. Kirsh J.R. Wall S.D. Passik My mind is as clear as it used to be: a pilot study illustrating the difficulties of employing a single-item subjective screen to detect cognitive impairment in outpatients with cancer J Pain Symptom Manage 26 2003 705 715 [31] D.A. Carone R.H. Benedict F.E. Munschauer 3rd I. Fishman B. Weinstock-Guttman Interpreting patient/informant discrepancies of reported cognitive symptoms in MS J Int Neuropsychol Soc 11 2005 574 583 [32] L.S. Middleton D.R. Denney S.G. Lynch B. Parmenter The relationship between perceived and objective cognitive functioning in multiple sclerosis Arch Clin Neuropsychol 21 2006 487 494 [33] R. Halari R. Mehrotra T. Sharma V. Kumari Does self-perceived mood predict more variance in cognitive performance than clinician-rated symptoms in schizophrenia? Schizophr Bull 32 2006 751 757 [34] J. Sanjuan E.J. Aguilar J.M. Olivares Subjective perception of cognitive deficit in psychotic patients J Nerv Ment Dis 194 2006 58 60 [35] R. Hubbard M.J. Bayarri Confusion over measures of evidence ( P ’s) versus errors (alpha’s) in classical statistical testing Am Statist 57 2003 171 178 [36] J.I. Breier K.L. Fuchs B.L. Brookshire Quality of life perception in patients with intractable epilepsy or pseudoseizures Arch Neurol 55 1998 660 665 [37] L.G. Kiloh Pseudo-dementia Acta Psychiatr Scand 37 1961 336 351 [38] E.L. Constant S. Adam B. Gillain X. Seron R. Bruyer A. Seghers Effects of sertraline on depressive symptoms and attentional and executive functions in major depression Depress Anxiety 21 2005 78 89 [39] R.D. Nebes B.G. Pollock P.R. Houck Persistence of cognitive impairment in geriatric patients following antidepressant treatment: a randomized, double-blind clinical trial with nortriptyline and paroxetine J Psychiatr Res 37 2003 99 108 [40] M.N. Pavuluri D.B. Henry J.A. Carbray G.A. Sampson M.W. Naylor P.G. Janicak A one-year open-label trial of risperidone augmentation in lithium nonresponder youth with preschool-onset bipolar disorder J Child Adolesc Psychopharmacol 16 2006 336 350 [41] K.J. Meador A. Gevins D.W. Loring Neuropsychological and neurophysiological effects of carbamazepine and levetiracetam Neurology 69 2007 2076 2084