Outcome of cognitive performances in bipolar euthymic patients after a depressive episode: a longitudinal naturalistic study

Background Cognitive functions have been investigated across depressed, manic, hypomanic, mixed and euthymic episodes of bipolar disorder, but the stability or the progression of cognitive impairment is still under research. Objective The purpose of the present study was to assess the outcome of cognitive functions in bipolar patients following a depressive episode, after a 6-month period in the absence of mood symptoms. Method 63 bipolar patients were tested with a battery of neurocognitive tests both at baseline (during an acute depressive episode) and after 6 months of euthymia. The cognitive domains assessed included memory, attention, verbal fluency, processing speed and executive functions. Cognitive performances were compared with those of a control group (40 healthy control subjects), both in depression and in euthymia. Results Patients scored worse than control subjects in several cognitive domains, both in depression and euthymia. The most impaired cognitive functions were executive functions and verbal memory. Between the two moments of assessment bipolar patients obtained a significant improvement in memory, verbal fluency, attention and information processing speed. Psychomotor speed showed no difference between depression and euthymia. Conclusions Bipolar patients showed impairment in several cognitive domains during depression. A certain degree of impairment remained even after the remission of the affective episode in relationship with the executive functions. Between depression and euthymia, bipolar patients showed important cognitive improvements.

Data from the literature show impairment in the majority of cognitive areas (attention, memory, verbal fluency, set shifting) for patients with bipolar disorder. Among these cognitive domains, attention, verbal learning, declarative memory, and executive functions were proposed as trait markers for bipolar disorder [12][13][14][15].
Attention deficits were highlighted in several studies; Torres et al. [9] concluded that patients with affective episodes show performance difficulties in areas related to attention which concern visual and motor processing speed, accuracy and reaction time at the tasks of maintaining attention that require identifying targets. Sustained attention or vigilance is impaired in patients with bipolar disorder regardless of whether they are evaluated during episodes of mania or depression, as shown in the study of Najit et al. [16]. Another study [17] demonstrated impairment in attention and response inhibition (prolonged reaction times and decreased discriminability) and also an inability to delay reward within bipolar disorder.
Memory deficits for patients with bipolar disorder consist of verbal memory, working memory and visual memory impairments [11]. Significant impairment in verbal recall and recognition in bipolar depression was outlined by Bearden et al. in 2006 [18]. The study of Altshuler et al. [19] demonstrated lower CVLT (California verbal learning test) recall scores for bipolar patients with a poorer functioning as well as a relationship between the impairment in verbal declarative memory and poor role functioning.
Dixon et al. [20] observed a certain pattern of cognitive dysfunctions in manic, depressed and euthymic bipolar patients in terms of strategic thinking, inhibitory control and response initiation regardless of the affective episode. In a recent review, executive functions, alongside with verbal memory were proposed as potential cognitive endophenotypes for bipolar disorder [21].
The progression of cognitive impairment has been reported by a number of studies [22][23][24], whereas others stress that cognitive impairment is in a strong relationship with acute illness episodes [25,26].
Methods 63 bipolar patients, both genders, with ages between 18 and 60 were included into the study. All patients had a minimum level of 8 years of education. The patients underwent clinical and neurocognitive evaluations twice: first time during an acute depressive episode; the second time after 6 months of euthymia. The inclusion criteria for depressive patients were DSM IV-TR and ICD-10 diagnosis of Bipolar Disorder and Hamilton Depression Rating Scale (HAM-D) >8. For euthymia, inclusion criteria consisted of HAM-D scores <7 and Young Mania Rating Scale (YMRS) scores <6. Exclusion criteria were chronic alcoholism or any other substance dependence, dementia, mental retardation, history of head trauma, or any current medical condition which could interfere with the level of cognitive performances. Patients were compared to 40 healthy control subjects matched by demographic characteristics and who met the same exclusion criteria. The study was approved by the University of Medicine and Pharmacy Iuliu Hatieganu Ethics Committee and all patients signed an informed consent before being admitted in the study.

Demographic and clinical data
Demographic data (age, gender, education level) and clinical information (number of previous affective episodes, types of previous episodes, history of psychotic symptoms) were collected through the clinical interview and completed with data from medical records. First clinical assessment was performed upon admission to the hospital for a depressive episode. The diagnosis was set according to DSM IV-TR and ICD-10 diagnosis criteria for bipolar disorder and major depressive episode; an additional inclusion criterion was HAM-D score >8. Remission was defined by the absence of any affective symptoms for 6 months and scores on HAM-D <7 and YMRS <6. Depressive patients were recruited while being hospitalized in acute emergency wards; therefore, their episodes and symptoms were more severe than those of an outpatient population. From baseline to follow-up, all bipolar patients were treated with a combination of 2 or 3 medications. Thus, 14 patients were treated with a combination of antipsychotics and mood stabilizers; 26 patients were treated with a combination of antipsychotics and antidepressants; 23 patients underwent treatment with a combination of antipsychotics, mood stabilizers and antidepressants.

Neurocognitive assessment
Cognitive functions were assessed using the following test batteries: Brief Assessment of Cognition in Schizophrenia (BACS) [27], Trail Making Test A and B (TMT A, TMT B) [28] and Wisconsin Card Sorting Test (WCST)-128 Card Version [29]. Subtests of each battery were used to assess specific cognitive domains as follows:

Demographic and clinical characteristics of the patients and control group
Bipolar patients and healthy control group were matched for all demographic items (gender, age, level of education). Demographic and clinical information is summarized in Table 1.

Comparison of cognitive functions between depressed patients and control group
The assessment of differences between cognitive performances of depressed bipolar patients and healthy controls was performed using Independent Samples t test. The equality of variances of the two independent groups was analyzed through Levene test. Results are presented in Table 2.

Memory
Verbal memory Verbal memory was evaluated with List learning-a subtest within BACS battery. Patients performed worse than controls (mean score 6.231; SD = 2.31958 vs mean score 8.2500; SD = 2.42709), obtaining lower total scores for the five consecutive trials of the cognitive test.
Working memory The mean score for working memory in patients was 15.75 (SD = 5.995) while healthy controls obtained a mean score of 18.10 (SD = 6.184). Even if controls scored higher, the difference between the two groups had a low statistical significance (p < 0.1).

Verbal fluency
For category instances test, the difference between the two groups was low (p = 0.026). For the controlled oral word association test, the mean for the control group was 17.45; SD = 7.423 and the mean for patients was 12.92; SD = 5.589 (p = 0.001).

Motor speed
The psychomotor speed was tested with the token motor test within BACS. Controls performed better with a mean score of 82.60 (SD = 10.862) while the mean score for patients was 75.89 (SD = 13.272).

Executive functions
Executive functions were assessed with several cognitive tests. For Tower of London test within BACS, the difference between patients and controls was significant for the threshold of 5 %. For WCST, we compared raw scores for the total number of correct trials, total errors, total preservative errors, conceptual levels and the number of categories completed. While for the number of total correct trials and number of conceptual levels the two groups performed similarly (p = 0.184 and p = 0.332), for the total number of errors and preservative errors, as well as for the number of categories completed, patients performed worse than controls (p < 0.001). For the B version of TMT, the differences between the two groups were also significant at the threshold of 1 % (p < 0.001) both for the time needed to complete the trial and for the total number of errors.

Comparison of mean scores of bipolar patients between depression and euthymia
We used paired samples t test to compare the scores obtained by bipolar patients between the two times of assessment, as presented in Table 3.
Results displayed in Table 3 show that bipolar patients obtained significant improvement of cognitive performances between the two moments of assessment (depression and euthymia) for the following domains: verbal memory (Pearson correlation coefficient = 0.933, p < 0.001), working memory (Pearson correlation coefficient = 0.854, p < 0.001), verbal fluency (category instances subtest-Pearson correlation coefficient = 0.763; p < 0.001; controlled oral association-Pearson correlation coefficient = 0.716; p < 0.05), attention (Pearson Table 1 Table 4 shows the differences of cognitive performances obtained by the patients after 6 months of euthymia and healthy controls. The results show similar performances in most cognitive areas. For executive functions assessed with TMT-B, the patients obtained significantly lower scores than controls (p < 0.001) and had a higher number of errors (p = 0.004). The other executive functions scores were only significant for a threshold of 5 % (p = 0.030 for Tower of London subtest = 0.013 for WCST-conceptual levels). Scores obtained by the control group for verbal fluency and motor speed were also higher than those of the patients (p = 0.012 for controlled oral association subtest and p = 0.017 for token motor test).

Discussions
Several studies have outlined cognitive impairment in different cognitive domains for bipolar patients. The present study is a longitudinal, naturalistic, 6 months follow-up study intended to asses the changes of cognitive performances in bipolar patients after a depressive episode. Cognitive functions evaluated during depression revealed impairment in several cognitive areas when the patients were compared to healthy controls. Verbal memory was the most affected function within BACS battery (p < 0.001). Patients group, when compared to the controls, recorded a smaller number of words recalled for each of the five verbal memory trials. Also, the patients group did not show a significant progression in the number of words remembered after each try, unlike the control group, which did. Our findings are consistent with those reported by other studies [15,18,30]. There were no important differences between the 2 groups for working memory. Differences between controls and patients for verbal fluency tests were reported to be significant for the threshold of 5 %. While verbal fluency was reported as persistently impaired in manic episodes [31], data regarding this cognitive domain in the depressive phase of bipolar disorder are limited. Data from previous studies regarding verbal fluency in bipolar disorder are inconsistent. While some research showed differences in verbal fluency tasks for bipolar patients [32,33], other found no discrepancy between patients with bipolar disorder and control groups [1]. Psychomotor speed was assessed with Token motor test. Depressive patients performed worse than controls (p < 0.05), placing, in total, less tokens and having more tokens placed incorrectly. Both tests used for assessing attention showed no differences between the depressed and control group. For executive functions tests, patient performed worse than controls in all measures of assessment. (p < 0.001 for TMT-B and WCST; p < 0.05 for Tower of London), displaying thus difficulties in set shifting, task switching, conceptual skills, planning and problem solving.
Repeated measures of cognitive functions after 6 months showed significant differences for almost all cognitive areas. These results may conclude that cognitive impairment could be episode related in bipolar disorder. Both verbal and working memory scores improved significantly for remitted patients, who obtained similar performances when compared with controls (p < 0.001). Attention and verbal fluency were also corrected for euthymic patients. Motor speed did not improve between the two assessments, suggesting that for this cognitive area, factors other than depression are involved. The study of Mora et al. [34] who recruited 28 bipolar patients and 26 controls showed the persistence of cognitive impairment in areas such as executive functions, inhibition processing speeds and verbal memory which remained affected

Conclusions
During a depressive episode bipolar patients show impairments in most cognitive areas when compared to a healthy control group. After 6 months free of symptoms, patients display similar results with controls for neurocognitive testing for almost all cognitive functions. Executive functions such as the ability to switch between tasks and cognitive flexibility remain impaired even during euthymia. The results of the present study do not support the progression of cognitive impairment, showing that for the bipolar patients, after 6 months free of symptoms, significant improvement is present in most cognitive domains. Nevertheless, all patients included into the study were under psychiatric medication, for the entire