What is the Contribution of Executive Dysfunction to the Cognitive Profile of ? A Well- Controlled Direct Comparison Study

Juan Lopera-Vasquez1 Vaughan Bell1 Carlos López-Jaramillo2

Abstract

Background: Large scale neuropsychological studies of patients with bipolar disorder have reported verbal memory and executive function deficits that persist during remission. A recent analysis by Thompson et al. (2009) indicated that verbal memory deficits could be entirely explained by the statistical variance attributed to primary executive function de- ficits. This study tests the hypothesis that verbal memory deficits in bipolar patients are largely the result of executive difficulties by direct comparison of verbal neuropsychological tests primarily differing in their executive load as well as examining potential interactions with medication status. Methods: 33 Bipolar I patients not taking medication, 40 Bipolar I patients taking medication, and 28 healthy controls were compared on measures of IQ, verbal fluency, category fluency, verbal recall, and category prompted recall. Results: After controlling for IQ, performance on tasks that involved additional executive involvement was significantly worse. Medication had a small but reliable effect on cognitive performance. Conclusions: The results provide support to the hypothesis that the most significant source of cognitive impairment in bipolar disorder stems from executive impairment and that verbal memory deficits may arise as a result of this, rather than from primary impairment to core verbal memory mechanisms.

Key words: Bipolar disorders, memory, neuropsychology, executive function.

Título: ¿Cuál es la contribución de la disfunción ejecutiva al perfil cognitivo del trastorno bipolar? Un estudio comparativo bien controlado

Resumen

Introducción: Los estudios neuropsicológicos a gran escala de pacientes con trastorno bipolar han reportado déficits en la memoria verbal y en la función ejecutiva que persisten durante la remisión. Un análisis reciente realizado por Thompson et al. indicó que los déficits en la

1 Research Group on Psychiatric Disorders, Department of Psychiatry, School of Medicine, University of Antioquia, Medellín, Colombia. Chairman Department of Psychiatry School of Medicine, University of Antioquia, Medellín, Colombia. 2 MD Psychiatrist. Coordinator of the Psychiatry Research Group (GIPSI) and Chair of the Department of Psychiatry at the Medicine School of the University of Antioquia. Medellín, Colombia.

64 S Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 What is the Contribution of Executive Dysfunction to the Cognitive Profile of Bipolar Disorder?... memoria verbal podían explicarse entera- and executive function being the mente por la varianza estadística atribuida a most commonly encountered areas los déficits en la función ejecutiva primaria. of cognitive impairment. The fact Este estudio demuestra la hipótesis de que los déficits en la memoria verbal de pacientes that these deficits are present when bipolares son mayormente el resultado de las patients are asymptomatic and, al- dificultades ejecutivas mediante una compa- beit less consistently, in non-affected ración directa de pruebas neuropsicológicas family members (Balanzá-Martínez verbales, que difieren principalmente en su et al., 2008) indicates that reduced carga ejecutiva, como también mediante la examinación las interacciones potenciales cognitive performance in key areas con el estado de la medicación. Métodos: could be a vulnerability marker or Se compararon 33 pacientes con trastorno endophenotype for the disorder (Ja- bipolar I que no estaban tomando medica- mrozinski, 2010). mentos con 28 controles saludables respecto Nevertheless, the issue of the a sus medidas de coeficiente intelectual (CI), fluidez verbal, fluidez de categorías, memoria neuropsychological impairments in verbal y memoria alentada por categorías. bipolar disorder goes beyond the Resultados: Después de realizar controles basic science and has clear clinical relacionados con el CI, el desempeño que implications as cognitive difficulties requería un mayor involucramiento ejecutivo have been reported to be one of era significativamente peor. Los medicamen- tos tenían un efecto pequeño pero confiable strongest predictors of functional sobre el desempeño cognitivo. Conclusiones: impairment in bipolar patients (on a Los resultados soportan la hipótesis de que par with mood symptoms; Sanchez- la fuente más significativa de trastornos cog- Moreno et al., 2009a) and remain nitivos en el trastorno bipolar es el trastorno a strong and significant predictor ejecutivo y que pueden surgir déficits en la memoria verbal como resultado de este, y no of functional outcome even when de un trastorno primario de los mecanismos mood symptoms, hospitalisations, centrales de la memoria verbal. suicide attempts and demographic variables are controlled for (Wingo Palabras clave: Trastorno bipolar, memoria, et al., 2009a). Similarly, a study neuropsicología, función ejecutiva by Gualtieri et al. (2008) found that bipolar has the highest rate (30%) of severe cognitive impair- Introduction ment among the mood disorders, defined as cognitive performance Meta-analyses of neuropsycholo- two standard deviations below the gical studies have reported neurocog- population mean in at least two nitive deficits in patients with bipolar cognitive domains. disorder that persist when euthymic But despite the fact that cog- and during periods of remission nitive difficulties are clearly asso- (Robinson et al., 2006; Torres et al., ciated with bipolar disorder, the 2007; Bora et al., 2009; Kurtz and core neurocognitive impairments Gerraty, 2009) with verbal memory identified as possible risk factors

Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 65 S Lopera-Vasquez J., Bell V., Lopez-Jaramillo C. for the condition have recently been Nevertheless, better controlled questioned. Although verbal memory studies are starting to appear that and have been include euthymic patients without associated with the condition, a significant differences in active symp- recent analysis by Thompson et al. toms that examine the impact of (2009) has suggested that the verbal substance misuse (Sanchez-Moreno memory difficulties could be entirely et al., 2009b) or evaluate the impact explained by the executive difficulties of psychiatric medication (Goswani rather than representing a distinct et al., 2009). One of the most cleanly domain of cognitive impairment. controlled studies in this regard is Furthermore, recent meta-analyses from Lopéz-Jaramillo et al. (2010a, of structural imaging studies have 2010b) which includes groups of been provided no evidence of changes euthymic bipolar patients both with in areas of the temporal cortices that and without medication, who addi- would be typically associated with tionally do not differ on the clinical primary verbal memory deficits (Ar- and demographic variables that have none et al., 2009; Vita et al., 2009) been identified as possible confoun- while reductions in the volume of ding variables. pre-frontal cortex, more associated Additionally, these participants with executive function, have been completed an extensive neurop- associated with bipolar disorder sychological battery that allows a (Arnone et al., 2009). direct comparison between verbal One of the obstacles to clearly memory tasks that are known to characterising the cognitive impact differ primarily in their reliance on of bipolar is the significant variability the executive system owing to the seen across studies in terms of pa- fact that category prompted recall is tient selection. In a recent review of known to require less spontaneous neurocognitive findings, Jamrozinski use of executive function-based (2010) notes that “This heterogeneity organisational recall strategies in results has been a challenge to (Shimamura, 1995). These eva- reviews and meta-analytic studies, luations include letter and category and the discrepancies cannot be re- fluency, with the frontal-executive solved as long as data on confounding having been shown to be much more variables are incompletely reported strongly involved in letter fluency in original studies” and highlights than category fluency (Baldo et al., how differences in diagnoses, ac- 2006; Birn et al., 2010); and free tive symptoms, medication status, recall and category prompted recall, history of substance abuse, patient which similarly have been shown demographics and neuropsychologi- to differ with free recall showing cal assessment methods contribute much stronger frontal-executive towards ambiguous results. involvement than category cued

66 S Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 What is the Contribution of Executive Dysfunction to the Cognitive Profile of Bipolar Disorder?... recall (Wheeler et al., 1995; Tacon- trist who performed the Diagnostic nat et al., 1995). Interview for Genetic Studies (DIGS), This allows a test of Thompson which has been confirmed as a valid et al.’s (2009) suggestion that verbal and reliable diagnostic measure in a memory deficits in bipolar are largely Spanish-speaking population (Palacio accounted for by executive difficul- et al., 2004; Roca et al., 2007). This ties – although while Thompson evaluation was used to confirm that et al. examined the hypothesis by patients fulfilled the DSM-IV criteria statistically controlling for variance for this disorder. The research proto- among tests, the current data set col used was approved by the Ethics allows this hypothesis to be tested Committee, and all participants read, by directly comparing performance understood, and signed an informed on tests that differ primarily in their consent form before being assessed. executive involvement. On this basis, we hypothesise that bipolar patients Patients will show greater impairment in letter fluency than category fluency, and All BD-I patients from a local free recall compared to category cued mood disorders clinic (University recall, indicating a relatively greater of Antioquia Mood Disorder Clinic, impairment in executive function. University Hospital San Vicente de Paúl, Medellín, Colombia) and who Method fulfilled the inclusion criteria were invited to participate in the study. Participants All patients reported being euthymic for at least six months, confirmed by We evaluated a sample of 101 clinical record review, and the pre- participants for the study comprising sence of residual affective symptoms 33 patients diagnosed with Bipolar was controlled for, as they needed to Disorder I not taking medication, score < 8 on the Zung Self-Rated De- 40 patients diagnosed with Bipolar pression Scale (ZSDS) and < 6 on the Disorder I that were prescribed medi- Young Mania Rating Scale (YMRS). cation, and 28 healthy controls. Those Furthermore, patients were exclu- patients who did not take medication ded if individuals had consumed did so through their own choice des- illicit substances or benzodiazepines pite recommendations to the contrary during the four weeks prior to the by their doctor and understanding assessment, or had other psychia- the risks of not doing so. This group tric or neurological disorders, such agreed to have regular monitoring of as epilepsy, mental retardation (IQ their symptoms and this was noted in < 70), or any treatment with elec- the consent form they signed. All par- troconvulsive therapy. Exclusion ticipants were evaluated by a psychia- criteria were confirmed by a review

Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 67 S Lopera-Vasquez J., Bell V., Lopez-Jaramillo C. of the case notes. All participants version of this assessment where had to be between 18 and 60 years participants are asked to name as of age and have had 5 to 16 years of many words as they can think of schooling, restricting the potential in one minute. Three trials of the influence of low illiteracy levels on test were run, using the prompt neuropsychological performance. letters ‘F’, ‘A’ and ‘S’. Categorical Recruited patients were divided into verbal fluency was assessed by the two groups, depending on whether standard version of this test where they were currently prescribed me- participants are asked to name as dication or not. Out of the patient many examples of each category as group prescribed medication, 20 possible, using the prompts ‘ani- were prescribed lithium, 11 were mals’ and ‘fruits’. As discussed in the prescribed valproate, one was addi- Introduction, performance —mea- tionally prescribed an antipsychotic sured by correct responses— on and one was additionally prescribed these phonological and categorical an antidepressant at time of testing. verbal fluency has been shown to differentially involve the prefrontal Controls and temporal cortices (e.g. Baldo et al., 2009; Birn et al., 2010). Healthy controls conformed to Free recall was assessed by the the same exclusion criteria used for use of first free recall condition of the patient participants and who were the Spanish version of the California also assessed with the DIGS to con- Verbal Learning Test (CVLT; Jaco- firm the absence of . bs et al., 1997) where participants are asked to recall as many of the Materials words as they can in any order they wish from a list of 16 read out by Neuropsychological Assessment the assessor. Categorical recall was assessed by the use of the first ca- All participants completed an tegory prompted recall trials of the extensive neuropsychological ba- semantic memory with associative ttery which is fully described in increment test (SMAI; Pineda and López-Jamarillo et al. (2010b). With Ardila, 1991) where participants regards to this particular study, are given four category headings to results from five neuropsychological assist with recall of the items. tests are included in the analysis. All participants were assessed with Analysis the Spanish version of the Wechsler Adult Intelligent Scale 3rd edition. Potential demographic differen- Phonological verbal fluency was ces between groups were assessed assessed by the use of the standard with one-way between groups analy-

68 S Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 What is the Contribution of Executive Dysfunction to the Cognitive Profile of Bipolar Disorder?... ses of variance (ANOVA), for conti- .304) and no significant differences in nuous variables shared between all LSD post-hoc comparisons between groups, and t-tests for continuous individual groups. However, when clinical variables only relevant to comparing WAIS IQ between groups patients. Non-parametric Kruskal- there was a significant effect (F(2, 98) Wallis tests were used to compare the = 3.636, p = .03) with post-hoc LSD three groups on non-continuous de- tests showing that while there was mographic variables (such as marital no difference between the patients and occupational status). A one-way group (p = .708) there was a signi- between groups multivariate analy- ficant difference between controls sis of covariance (MANCOVA) was and unmedicated patients (p = .014) completed to examine differences in and controls and medicated patients neuropsychological test performance (p = .027). between healthy controls, patients Using a non-parametric Krus- without medication and patients kal-Wallis test to compare the three with medication. Four dependent groups on non-continuous variables variables were included – number there was no significant difference of correct items on: phonological between the three groups in terms of verbal fluency, categorical verbal marital status (c2 = .342, p = .806), fluency, CVLT free recall, semantic occupational status (c2 = .704, memory with associative increment p = .806) and laterality (c2 = 1.166, test category recall. WAIS IQ was p = .558). There was a trend to signi- included as a co-variable. Including ficance in a comparison of the three all variables in a single analysis has groups in terms of previous abuse the advantage of controlling for Type of drugs and alcohol (c2 = 5.508, I errors by reducing problems with p = .08) although the trend no multiple comparisons. longer remains when the alpha is corrected for multiple comparisons. Results When comparing the two pa- tient groups using an independent Means for demographic variables samples t-test, there was no signi- of the groups are displayed in Table 1. ficant difference between number

There were no significant differences of depressive episodes (t(63) = .985, between the three groups, compared p = .328), number of manic episo- using a one-way independent sam- des (t(63) = .915, p = .368), number ples ANOVA, in terms of age (F(2, 95) = of hypomanic episodes (t(56) = .838, 1.743, p = .181), years of schooling p = .406) and years since first onset

(F(2, 95) =.182, p = .834), score on the (t(65) = .274, p = .785). There was a Zung Self-Rating Scale trend to significance for number

(F(2, 73) =.344, p = .710) or Young Ma- of mixed episodes (t(63) = 1.773, nia Rating Scale (F(2, 93) = 1.206, p = p = .081) but the trend no longer

Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 69 S Lopera-Vasquez J., Bell V., Lopez-Jaramillo C.

Table 1. Descriptive characteristics of control and patient groups

Healthy Controls Patients No Meds Patients With Meds (N = 28) (N = 31) (N = 40)

Age 38.96 (8.63) 38.52 (8.14) 42.22 (10.03) Schooling 10.68 (3.93) 10.45 (5.03) 10.05 (3.71) WAIS IQ 104.71 (18.65) 94.0 (15.48) 95.47 (16.2) Zung 34.25 (3.61) 36.04 (8.84) 35.02 (7.89) Young .81 (1.331) 1.15 (1.44) .69 (.98) Depressive episodes - 2.55 (6.03) 1.35 (3.54) Manic Episodes - 2.10 (2.27) 2.74 (3.22) Hypomanic Episodes - 0 (0) .03 (.17) Mixed Episodes - .52 (.92) 1.65 (3.43) Years since onset - 14.79 (8.28) 15.53 (13.24) remains after correction for multiple After adjusting for WAIS IQ, the- comparisons. re was a significant effect of group

A one-way between groups MAN- on the combined variables (F(8, 188) COVA was completed to examine = 2.458, Wilk’s lambda = .82, p = differences in neuropsychological .015). Considering the results of the test performance between healthy dependent variables separately there controls, patients without medication was a significant effect on phonolo- and patients with medication. Four gical verbal fluency (F(2, 97) = 3.207, dependent variables were included p = .045) but no significant effect of

– number of correct items on: pho- categorical verbal fluency (F(2, 97) = nological verbal fluency, categorical .971, p = .382). Similarly there was verbal fluency, CVLT free recall, a significant effect for CVLT free semantic memory with associative recall (F(2, 97) = 6.412, p = .002) but increment test category recall with the effect for semantic memory with means displayed in Table 2. Owing associative increment test category to significant differences between recall only just reached a trend for patients groups and controls on Full significance (F(2, 97) = 2.475, p = .089). Scale WAIS IQ, this was included Post-hoc LSD tests were comple- as a co-variable in the analysis to ted for the comparisons that showed control for the influence of differen- overall significance (phonological ces in general cognitive functioning. verbal fluency and CVLT free recall). Initial testing showed no violations of There was a significant difference in assumptions of normality, linearity, phonological verbal fluency between outliers, homogeneity of variance or controls and patients with medi- multi-collinearity. cation (p =.014) with the compari-

70 S Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 What is the Contribution of Executive Dysfunction to the Cognitive Profile of Bipolar Disorder?...

Table 2. Neuropsychological test results from control and patient groups

Healthy Controls Patients No Meds Patients With (N = 28) (N = 31) Meds (N = 40) Verbal Fluency – Pho- 39.64 (12.11) 31.09 (10.04) 30.08 (11.88) nological Verbal Fluency – Ca- 37.82 (7.22) 34.12 (7.68) 33.3 (7.10) tegories CVLT Free Recall 11.68 (1.70) 9.30 (2.48) 9.30 (2.72) SMAI Category Recall 13.64 (1.77) 12.33 (2.33) 11.88 (2.82) son between controls and patients prescribed medication to perform without medication showing a trend worse than the unmedicated group, to significance p ( = .085) and no indicating a small but detectable significant difference between pa- impact of medication of cognitive tient groups (p = .465). For CVLT performance, although this was far free recall there was a significant outweighed by the effect of bipolar difference between both controls disorder itself. and patient with medication (p = The pattern of results provides .001) and patients without medi- additional evidence that the most cation (p = .003) and no difference significant source of cognitive im- between patient groups (p = .873). pairment in bipolar is from execu- tive impairment and verbal memory Discussion deficits may arise as a result of this, rather than from primary impairment The aim of the study was to to core verbal memory mechanisms, examine whether neuropsychological in line with the results of Thompson tests of verbal memory that required et al. (2009). The pre-frontal cortex a greater degree of executive control is known to be specialised for the would be differentially impaired in use of active strategies for encoding patients with bipolar disorder. In line and recall in (Mos- with the hypothesis, we found that covitch, 1992) and it is notable that performance on the more executi- the tests used in the present study vely demanding tasks, phonological have been shown to preferentially fluency and free recall, was signifi- engage this area (Baldo et al., 2006; cantly lower in patients than controls Birn et al., 2010; Taconnat et al., – in contrast to the less executively 1995; Wheeler et al., 1995) providing demanding tasks, category fluency additional evidence for the primacy and category cued recall, in which of pre-frontal dysfunction in bipolar no differences between patients and affective disorder. controls were apparent. There was Although individual structural a slight trend for the patient group neuroimaging studies have reported

Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 71 S Lopera-Vasquez J., Bell V., Lopez-Jaramillo C. conflicting results, meta-analyses One particular issue has been of structural imaging studies in the cognitive impact of medication bipolar have most typically repor- which has received increasing atten- ted lateral ventricular enlargement tion in the literature with the results (Kempton et al., 2008; Arnone et are largely in line with the findings al., 2009), increased rates of white from this study: medication has matter hyperintensities (Kempton et been found to have a detectable but al., 2008), grey matter reductions in minor effect of (Goswami et the anterior cingulate and bilateral al., 2009). Owing to its widespread insula (Ellison-Wright and Bullmore, use in bipolar disorder, lithium has 2010) and an increased volume of been of particular interest (Foun- the with reductions toulakis et al., 2008) with a recent in whole and / or prefrontal meta-analyses finding a small but structure volume (Arnone et al., reliable overall impact on cognition 2009). It is notable that changes (Wingo et al., 2009b). With respect to the medial-temporal and hippo- to other medications, benzodiazepi- campal structures, that we would nes have a well established impact expect if verbal memory deficits re- on memory (Vgontzas et al., 1995), sults from impairment to primary preliminary evidence suggests that memory mechanisms, are absent, valproate and carbamazepine have while frontal changes that are more been associated with lower cognitive likely to impact on executive function function in comparison to patients are reliably reported. treated with lamotrigine (Daban et Indeed, similar patterns of me- al., 2006) while antipsychotics have mory difficulty can be seen in pa- been associated with reduced cogni- tients with focal damage to prefrontal tive performance in bipolar patients circuits, such as in degenerative di- (Dittmann et al., 2008; Jamrozins- sorder (Glosser et al., 2002) or acqui- ki et al., 2009; Arts et al., 2010) red damage (Shimamura although it must be noted that the et al., 1991), and in effect sizes are typically small and where dysexecutive problems stron- well-controlled studies are still lac- gly contribute to episodic memory king. Perhaps more pertinently, bi- impairment (Ragland et al., 2009). polar patients are commonly treated However, it is important to note the with a combination of medications considerable variability that has been and yet little is known about the reported between both patients and cognitive impact of polypharmacy studies in this area with inconsisten- in this group (Fountoulakis, 2010). cies in medication regimes, clinical However, there is a growing body characteristics and demographics of evidence that clinical characte- of the samples cited as contributory ristics, particularly the evolution of factors (e.g. Jamrozinski, 2010). bipolar may lead to an increasing

72 S Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 What is the Contribution of Executive Dysfunction to the Cognitive Profile of Bipolar Disorder?... neurocognitive impact over-time. A that focus on their dysexecutive meta-analysis specifically focused on problems. structural brain changes during first- However, it is important to note episode bipolar patients reported a that the data from this study, as reduction in white matter volume with many other in the area, is pu- that did not overlap with results rely cross-sectional and longitudi- reported from studies of chronic nal studies will be needed to fully patients, indicating ongoing chan- understand the relation between ges associated with the evolution of clinical evolution and cognitive im- bipolar disorder (Vita et al., 2009). pairment. This is particularly per- A review of studies on the cognitive tinent in bipolar disorder, owing to evolution of bipolar (Robinson and evidence suggesting that, in contrast Ferrier, 2006) found that cognitive to patients with schizophrenia for impairment was associated with example, bipolar patients show an number previous manic episodes, intact developmental cognitive tra- hospitalizations and length of illness. jectory but impairments that reflect As with many studies in this area, worsening symptoms (Lewandowski poorly controlled samples was cited et al., 2010). Furthermore, it is also as an issue, although a recent and not clear which cognitive impair- better controlled study (Lopéz-Jara- ments might fluctuate as the clinical millo et al., 2010a) has supported picture alters and which might be this conclusion with clinical groups relatively invariant. However, our re- clearly matched on demographic sults, along with those of Thompson variables including both medicated et al. (2009) highlight the importance and unmedicated patients. of looking at cognitive performance Such evidence has led to the as an interacting system, rather than development of the ‘staging mo- assuming the existence of individual components on the basis that they del’ of bipolar disorder (Berk et al., are represented by distinct nume- 2007; Vieta et al., 2010) that argues rical indices. for a clear clinical evolution of the condition over time, having strong implications for the importance of Bibliography early intervention (Berk et al., 2010). Arnone D, Cavanagh J, Gerber D, et al. Our results have implications both Magnetic resonance imaging studies for early intervention and for cogni- in bipolar disorder and schizophre- nia: meta-analysis. Br J Psychiatry. tive rehabilitation, indicating that 2009;195:194-201. prompt and effective treatment may Arts B, Jabben N, Krabbendam L, et al. A better preserve essential executive 2-year naturalistic study on cognitive functioning in bipolar disorder. Acta functions, while those patients with Psychiatr Scand. 2011;123:190-205. neurocognitive problems may be best Baldo JV, Schwartz S, Wilkins D, et al. Role treated by rehabilitation programmes of frontal versus temporal cortex in

Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 73 S Lopera-Vasquez J., Bell V., Lopez-Jaramillo C.

verbal fluency as revealed by voxel- patients do not differ. Acta Psychiatr based lesion symptom mapping. J Int Scand. 2009;120:456-63. Neuropsychol Soc. 2006;12:896-900. Gualtieri CT, Morgan DW. The frequency Balanzá-Martínez V, Rubio C, Selva-Vera G, of cognitive impairment in patients et al. Neurocognitive endophenotypes with , depression, and bipolar (endophenocognitypes) from studies disorder: an unaccounted source of of relatives of bipolar disorder subjects: variance in clinical trials. J Clini Psy- a systematic review. Neurosci Biobehav chiatry. 2008;69:1122-30. Rev. 2008;32:1426-38. Jamrozinski K. Do euthymic bipolar patients Berk M, Hallam K, Malhi GS, et al. Evidence have normal cognitive functioning? and implications for early intervention Curr Opin Psychiatry. 2010;23:255-60. in bipolar disorder. J Ment Health. Jamrozinski K, Gruber O, Kemmer C, et al. 2010;19:113-26. Neurocognitive functions in euthymic Berk M, Hallam KT, McGorry PD. The poten- bipolar patients. Acta Psychiatr Scand. tial utility of a staging model as a course 2009;119:365-74. specifier: a bipolar disorder perspecti- Kempton MJ, Geddes JR, Ettinger U, et al. ve. J Affect Disord. 2007;100:279-81. Meta-analysis, database, and meta- Birn RM, Kenworthy L, Case L, et al. Neural regression of 98 structural imaging systems supporting lexical search gui- studies in bipolar disorder. Arch Gen ded by letter and semantic category Psychiatry. 2008;65:1017-32. cues: a self-paced overt response fMRI Kurtz MM, Gerraty RT. A meta-analytic in- study of verbal fluency. Neuroimage. vestigation of neurocognitive deficits in bipolar illness: profile and effects 2010;49;1099-107. of clinical state. Neuropsychology. Bora E, Yucel M, Pantelis C. Cognitive en- 2009;23:551-62. dophenotypes of bipolar disorder: a Lewandowski KE, Cohen BM, Ongur D. meta-analysis of neuropsychological Evolution of neuropsychological dys- deficits in euthymic patients and their function during the course of schizo- first-degree relatives. J Affect Disord. phrenia and bipolar disorder. Psychol 2009;113:1-20. Med. 2011;41:225-41. Daban C, Martínez-Arán A, Torrent C, et López-Jaramillo C, Lopera-Vásquez J, al. Cognitive functioning in bipolar Gallo A, et al. Effects of recurrence on patients receiving lamotrigine: prelimi- the cognitive performance of patients nary results. J Clin Psychopharmacol. with Bipolar I disorder: implications for 2006;26:178-81. relapse prevention and treatment adhe- Dittmann S, Hennig-Fast K, Gerber S, et rence. Bipolar Disord. 2010;12:557-67. al. Cognitive functioning in euthymic López-Jaramillo C, Lopera-Vásquez J, Ospina- bipolar I and bipolar II patients. Bipolar Duque J, et al. Lithium treatment effects Disord. 2008;10:877-87. on the neuropsychological functioning Ellison-Wright I, Bullmore E. Anatomy of patients with bipolar I disorder. J Clin of bipolar disorder and schizophre- Psychiatry. 2010;71:1055-60. nia: a meta-analysis. Schizophr Res. Moscovitch M. A neuropsychological mo- 2010;117:1-12. del of memory and consciousness. Fountoulakis KN. An update of evidence- En: Squire LR, Butters N, eds. Neu- based treatment of bipolar depression: ropsychology of memory. New York: where do we stand? Curr Opin Psychia- Guilford; 1992. p. 5-22. try. 2010;23:19-24. Palacio CA, García J, Arbeláez MP, et al. Glosser G, Gallo JL, Clark CM, et al. Memory Validación de la entrevista diagnóstica encoding and retrieval in frontotempo- para estudios genéticos (DIGS) en ral and Alzheimer’s disease. Colombia. Biomédica. 2004;24:56-62. Neuropsychology. 2002;16:190-6. Ragland JD, Laird AR, Ranganath C, et al. Goswami U, Sharma A, Varma A, et al. The Prefrontal activation deficits during neurocognitive performance of drug- episodic memory in schizophrenia. Am free and medicated euthymic bipolar J Psychiatry. 2009;166:863-74.

74 S Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 What is the Contribution of Executive Dysfunction to the Cognitive Profile of Bipolar Disorder?...

Robinson LJ, Thompson JM, Gallagher P, et cued-recall test: the role of executive al. A meta-analysis of cognitive deficits functions and fluid intelligence. Brain in euthymic patients with bipolar disor- and Cognition. 2007;64:1-6. der. J Affect Disord. 2006;93:105-15. Thompson JM, Gray JM, Crawford JR, et al. Robinson LJ, Ferrier IN. Evolution of cog- Differential deficit in executive control nitive impairment in bipolar disorder: in euthymic bipolar disorder. J Abnorm a systematic review of cross-sectional Psychol. 2009;118:146-60. evidence. Bipolar Disord. 2006;8:103- Torres IJ, Boudreau VG, Yatham LN. 16. Neuropsychological functioning in Roca M, Martin-Santos R, Saiz J, et al. euthymic bipolar disorder: a meta- Diagnostic Interview for Genetic Stu- analysis. Acta Psychiatr Scand Suppl. dies (DIGS): inter-rater and test-retest 2007;434:17-26. reliability and validity in a Spanish po- Vieta E, Reinares M, Rosa AR. Staging pulation. Eur Psychiatry. 2007;22:44-8. bipolar disorder. Neurotox Res. Sánchez-Moreno J, Martínez-Aran A, Colom 2011;19:279-85. F, et al. Neurocognitive dysfunctions Vita A, De Peri L, Sacchetti E. Gray matter, in euthymic bipolar patients with and white matter, brain, and intracranial without prior history of alcohol use. J volumes in first-episode bipolar di- Clin Psychiatry. 2009;70:1120-7. sorder: a meta-analysis of magnetic Sánchez-Moreno J, Martínez-Aran A, Ta- resonance imaging studies. Bipolar barés-Seisdedos R, et al. Functioning Disord. 2009;11:807-14. and disability in bipolar disorder: an Vgontzas AN, Kales A, Bixler EO. Benzo- extensive review. Psychother Psycho- diazepine side effects: role of pharma- cokinetics and pharmacodynamics. som. 2009;78:285-97. Pharmacology. 1995;51:205-23. Shimamura AP. Memory and the pre- Wingo AP, Harvey PD, Baldessarini RJ. frontal cortex. Ann N Y Acad Sci. Neurocognitive impairment in bipolar 1995;769:151-9. disorder patients: functional implica- Shimamura AP, Janowsky JS, Squire LR. tions. Bipolar Disord. 2009:11:113-25. What is the role of frontal lobe damage Wingo AP, Wingo TS, Harvey PD, et al. in memory disorders? En: Levin HS, Effects of lithium on cognitive perfor- Eisenberg HM, Benton AL, eds. Frontal mance: a meta-analysis. J Clin Psy- lobe function and dysfunction. New chiatry. 2009;70:1588-97. York: Oxford University Press; 1991. Wheeler MA, Stuss DT, Tulving E. Frontal p. 173-95. lobe damage produces episodic me- Taconnat L, Clarys D, Vanneste S, et al. mory impairment. J Int Neuropsychol Aging and strategic retrieval in a Soc. 1995;1:525-36.

Conflicts of interest: The authors have not conflicts of interest.

Recibido para evaluación: 17 de junio del 2011 Aceptado para publicación: 30 de junio del 2011

Corresponding author Carlos López-Jaramillo Departamento de Psiquiatría, Facultad de Medicina Universidad de Antioquia Calle 64 No. 51 D 38 Medellín, Colombia [email protected]

Rev. Colomb. Psiquiat., vol. 40, Suplemento 2011 75 S