Introduction
The relationship between poorer course of bipolar disorder and greater neurocognitive impairment has been previously reported (Martinez-Aran et al. Reference Martinez-Aran, Vieta, Torrent, Sanchez-Moreno, Goikolea, Salamero, Malhi, Gonzalez-Pinto, Daban, Alvarez-Grandi, Fountoulakis, Kaprinis, Tabares-Seisdedos and Ayuso-Mateos2007; Lopez-Jaramillo et al. Reference Lopez-Jaramillo, Lopera-Vasquez, Gallo, Ospina-Duque, Bell, Torrent, Martinez-Aran and Vieta2010), as well as the impact of neurocognitive deficits on psychosocial functioning (Tabares-Seisdedos et al. Reference Tabares-Seisdedos, Balanza-Martinez, Sanchez-Moreno, Martinez-Aran, Salazar-Fraile, Selva-Vera, Rubio, Mata, Gomez-Beneyto and Vieta2008; Bonnin et al. Reference Bonnin, Martinez-Aran, Torrent, Pacchiarotti, Rosa, Franco, Murru, Sanchez-Moreno and Vieta2010; Depp et al. Reference Depp, Mausbach, Harmell, Savla, Bowie, Harvey and Patterson2012; Torrent et al. Reference Torrent, Martinez-Aran, del Mar, Reinares, Daban, Sole, Rosa, Tabares-Seisdedos, Popovic, Salamero and Vieta2012). As a consequence of these findings, the interest in the development of therapeutic strategies that contribute to restoring or mitigating the impact of sustained neurocognitive and functional impairment has grown in the last few years (Fuentes-Dura et al. Reference Fuentes-Dura, Balanza-Martinez, Ruiz-Ruiz, Martinez-Aran, Giron, Sole, Sanchez-Moreno, Gomez-Beneyto, Vieta and Tabares-Seisdedos2012; Vieta, Reference Vieta2014). However, only two publications have addressed this issue so far in well-defined samples of bipolar patients. An open study, using a programme called Cognitive Rehabilitation (Deckersbach et al. Reference Deckersbach, Nierenberg, Kessler, Lund, Ametrano, Sachs, Rauch and Dougherty2010) and a randomized controlled trial, implementing a programme named Functional Remediation (Torrent et al. Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013). Both studies reported the positive impact of the interventions on functional outcomes but the improvement of neurocognition was mild (Deckersbach et al. Reference Deckersbach, Nierenberg, Kessler, Lund, Ametrano, Sachs, Rauch and Dougherty2010) or non-significant at 6-month follow-up (Torrent et al. Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013). Specifically, Deckersbach et al. (Reference Deckersbach, Nierenberg, Kessler, Lund, Ametrano, Sachs, Rauch and Dougherty2010) reported an improvement in residual depressive symptoms, occupational and psychosocial functioning. Moreover, changes in executive functioning accounted, in part, for the improvements in occupational functioning. In a multicentre trial carried out by Torrent et al. (Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013) no significant effect of treatment group was found on neurocognitive variables, despite significant improvements in psychosocial functioning. The authors hypothesized that this result was most likely due to the study design which did not require a defined cognitive impairment but functional impairment at study entry. Hence, a number of patients showing functional but not cognitive impairment at enrolment may have caused ceiling effects on cognitive measures. Due to an important heterogeneity in the severity of neurocognitive impairment in bipolar disorder (Altshuler et al. Reference Altshuler, Ventura, van Gorp, Green, Theberge and Mintz2004; Martino et al. Reference Martino, Strejilevich, Scapola, Igoa, Marengo, Ais and Perinot2008, Reference Martino, Strejilevich, Marengo, Ibáñez, Scápola and Igoa2014), we decided to conduct a subanalysis of the original multicentre randomized controlled trial by Torrent et al. (Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013), to analyse the response to functional remediation in terms of neurocognitive performance. This subsample consisted of patients who performed below 2 standard deviations from the mean of normative data in any neurocognitive test. The main objective was to ascertain whether neurocognitive enhancement was a potential ingredient of functional remediation, given the high density of neurocognitive tasks involved in the therapy. We hypothesized that the group of neurocognitively impaired patients receiving functional remediation would improve not only their functional outcome but also their neurocognitive functioning, as compared with patients who were randomized to psychoeducation or treatment as usual (TAU).
Method
Participants
Participants in the original trial fulfilled the following inclusion criteria (Torrent et al. Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013): (1) diagnosed with bipolar I or II disorder (BD-I or BD-II) according to the Diagnostic and Statistical Manual of Mental Disorders (fourth edition, text revision; American Psychiatric Association, 2000); (2) assessed during euthymia defined as Young Mania Rating Scale (YMRS) score ⩽6 (Young et al. Reference Young, Biggs, Ziegler and Meyer1978; Colom et al. Reference Colom, Vieta, Martinez-Aran, Garcia-Garcia, Reinares, Torrent, Goikolea, Banus and Salamero2002), and Hamilton Depression Rating Scale-17 (HAMD) score ⩽8 (Hamilton, Reference Hamilton1960; Ramos-Brieva & Cordero-Villafáfila, Reference Ramos-Brieva and Cordero-Villafáfila1986); (3) suffering from at least a moderate level of functional impairment measured by means of the Functioning Assessment Short Test (FAST) scale ⩾ 18 (Rosa et al. Reference Rosa, Sanchez-Moreno, Martinez-Aran, Salamero, Torrent, Reinares, Comes, Colom, Van, Ayuso-Mateos, Kapczinski and Vieta2007), including a score of four points or more in the cognitive functioning domain of the FAST.
Exclusion criteria were: (1) an intelligence quotient (IQ) < 85 or any medical condition that could affect neuropsychological performance (such as neurological diseases); (2) any co-morbid psychiatric condition, including substance abuse or dependence within the past 3 months; (3) electroconvulsive therapy within the past year; or (4) participation in any structured psychological intervention, such as psychoeducation or cognitive remediation, within the past 2 years.
As mentioned in the original study, all patients provided written informed consent. Ethical approval for the study was granted by the Ethics Committee at every hospital involved in the study.
For the specific purpose of this study, we applied the criteria of Martino et al. (Reference Martino, Strejilevich, Scapola, Igoa, Marengo, Ais and Perinot2008) to assess neurocognitive impairment in the original sample (n = 239), resulting in 188 neurocognitively impaired patients.
Further details of this procedure are explained in the data analysis section.
Interventions
Patients in the original trial (Torrent et al. Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013) were assigned in a 1:1:1 ratio to receive 21 weeks of functional remediation, psychoeducation, or TAU. Randomization was stratified by age, sex and education level.
Functional remediation
The functional remediation programme consisted of 21 weekly sessions, each lasting 90 min. This intervention addresses neurocognitive issues such as attention, memory and executive functions, but it focuses even more on enhancing functioning in daily routine. The content of the intervention is based on ecological tasks to be performed in two settings, in the clinic as well as at home. Patients were trained with exercises for memory, attention, problem solving and reasoning and organization in order to improve their functional outcome. Most of the techniques were based on paper-and-pencil tasks and group activities. For detailed information on the rationale of this intervention, see Martínez-Arán et al. (Reference Martínez-Arán, Torrent, Solé, Bonnín, Rosa, Sánchez-Moreno and Vieta2011), Bonnin et al. (Reference Bonnin, Torrent, Vieta and Martínez-Arán2014) and Vieta et al. (Reference Vieta, Torrent and Martinez-Aran2014).
Psychoeducation
The psychoeducation also consisted of 21 weekly sessions of 90 min each, aimed at preventing recurrences of bipolar illness by improving four main issues: illness awareness, treatment adherence, early detection of prodromal symptoms of relapse, and lifestyle regularity (Colom et al. Reference Colom, Vieta, Martinez-Aran, Reinares, Goikolea, Benabarre, Torrent, Comes, Corbella, Parramon and Corominas2003; Colom & Vieta, Reference Colom and Vieta2006).
TAU
In the TAU group, participants received prescribed pharmacological treatment without any adjunctive psychological therapy.
Clinical, neuropsychological and functional assessment
Clinical, neurocognitive and functional variables were collected for all patients both at baseline and at 6 months follow-up (see Torrent et al. Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013).
Clinical variables were gathered through a clinical interview and from clinical history records. In addition to this, the HAMD and YMRS were administered to assess affective symptoms.
Neurocognitive performance was evaluated through a comprehensive neuropsychological battery. The same battery was administered at baseline and at 6-month follow-up. It was composed of six domains: (1) estimated IQ, which was evaluated with the Wechsler Adult Intelligence Scale – III (WAIS-III) vocabulary subtest (Wechsler, Reference Wechsler1997a ); (2) the processing speed index, which consists of two subtests of the WAIS-III, the digit–symbol coding and symbol search (Wechsler, Reference Wechsler1997a ); (3) executive function, which was tested by set shifting, verbal fluency, planning, and response inhibition using the Computerized Wisconsin Card Sorting Test (Heaton, Reference Heaton1981), the Stroop Color–Word Interference Test (Golden, Reference Golden1978), the phonemic (F-A-S) and categorical (animal naming) components of the Controlled Oral Word Association Test (Benton & Hamsher, Reference Benton and Hamsher1976), the Trail Making Test, part B (Reitan, Reference Reitan1958), and the Rey–Osterrieth Complex Figure (Rey, Reference Rey1997); (4) visual memory and verbal learning/memory, which were assessed with the Rey–Osterrieth Complex Figure for visual memory and the California Verbal Learning Test (CVLT) (Delis et al. Reference Delis, Kramer, Kaplan and Ober1987) and the Logical Memory Scale (Wechsler Memory Scale-III; WMS-III) (Wechsler, Reference Wechsler1997b ) for learning/memory; (5) the working memory index, which was tested with three subtests of the WAIS-III (Wechsler, Reference Wechsler1997a ): arithmetic, digits forward and backward, and letter–number sequencing; and (6) attention, which was tested with the Trail Making Test, part A (Reitan, Reference Reitan1958), administered together with the Continuous Performance Test-II, version 5 (Conners, Reference Conners2000), to measure sustained attention.
Finally, psychosocial functioning was assessed by means of the FAST scale (Rosa et al. Reference Rosa, Sanchez-Moreno, Martinez-Aran, Salamero, Torrent, Reinares, Comes, Colom, Van, Ayuso-Mateos, Kapczinski and Vieta2007).
Data analysis
Data were analysed using SPSS v.18 for Windows (USA). First, the proportion of patients with neurocognitive impairment was calculated following the criteria established by Martino et al. (Reference Martino, Strejilevich, Scapola, Igoa, Marengo, Ais and Perinot2008). According to their criteria, a domain was compromised when performance in any test of that domain was below 2 standard deviations or more from the mean of normative data of each test.
In our sample, patients’ z scores on each measure were calculated according to data from a healthy control group (n = 30) that was comparable in terms of age (mean = 40.6, s.d. = 13.1, F 3,262 = 0.08, p = 0.77), estimated IQ (mean = 109.6, s.d. = 8.0, F 3,261 = 2.7, p = 0.04) and years of education (mean = 13.7, s.d. = 3.8, F 3,260 = 0.51, p = 0.47) with the remaining groups: functional remediation (n = 77); psychoeducation (n = 82); and TAU (n = 80). Even though significant differences were found in the estimated IQ, when a Tukey post-hoc test was performed, no significant differences were found between groups. Note that these comparisons were run with the total sample, before excluding the patients without neurocognitive impairment.
After excluding the patients without neurocognitive impairment, we proceeded with the second part of the analysis. Clinical and sociodemographic variables were analysed using one-way analysis of variance (ANOVA) for continuous variables or the χ2 test for categorical variables. Finally, repeated-measures ANOVAs were conducted in order to assess the impact of the treatment arms (functional remediation, psychoeducation and TAU) on participants’ scores in the different neuropsychological variables and psychosocial functioning at baseline and at follow-up. Repeated-measures analysis only took into account complete cases; hence missing cases due to being lost to follow-up were not analysed. Raw scores were used to compare the three groups since they did not differ in terms of confounding variables.
For the neuropsychological assessment, the following neuropsychological variables were analysed in order to cover the main cognitive domains that have been reported to be affected in bipolar disorder (verbal memory, attention and executive functions) (Bourne et al. Reference Bourne, Aydemir, Balanza-Martinez, Bora, Brissos, Cavanagh, Clark, Cubukcuoglu, Dias, Dittmann, Ferrier, Fleck, Frangou, Gallagher, Jones, Kieseppa, Martinez-Aran, Melle, Moore, Mur, Pfennig, Raust, Senturk, Simonsen, Smith, Bio, Soeiro-de-Souza, Stoddart, Sundet, Szoke, Thompson, Torrent, Zalla, Craddock, Andreassen, Leboyer, Vieta, Bauer, Worhunsky, Tzagarakis, Rogers, Geddes and Goodwin2013) and on neurocognitive domains we work on in the functional remediation programme: short free recall, short cued recall, delayed free recall, delayed cued recall, Trail Making Test part A and B, Wisconsin Card Sorting Test perseverative errors, Stroop interference, categorical and phonemic fluencies. Between- and within-effect sizes were calculated for each neurocognitive variable that was examined. For psychosocial functioning, only global scores measured by means of the FAST were analysed conducting another repeated-measure analysis.
The trial is registered under Clinicaltrials.gov identification number: NCT 01370668.
Results
Rate of neurocognitive impairment
Considering the patients who fulfilled the criteria of Martino et al. (Reference Martino, Strejilevich, Scapola, Igoa, Marengo, Ais and Perinot2008), we found that from the original sample (n = 239) only 20% of patients (n = 51) were free of neurocognitive impairment, that is, they did not present any impaired cognitive domain and they were excluded from the present study in order to avoid ceiling effects. The remaining patients (n = 188) presented cognitive impairment and they were allocated to the groups as follows: 56 patients in the functional remediation group; 69 patients in the psychoeducation group; and 63 patients in the TAU group. Of note is that there is a broad range of neurocognitive impairment in this sample, highlighting the heterogeneity of neurocognitive impairment that can be observed in bipolar disorder. As shown in Fig. 1, rates of neurocognitive impairment were equally distributed in the three groups (χ2 = 2.45, p = 0.65).
Fig. 1. Rates of neurocognitive impairment. TAU, Treatment as usual.
Clinical and sociodemographic variables
As shown in Table 1, all three patient groups were comparable in terms of sociodemographic (age, years of education, and pre-morbid IQ estimation) and clinical variables (number and type of episodes, chronicity, age at onset, HAMD, YMRS and FAST scores).
Table 1. Clinical and sociodemographic characteristics
Data are given as mean (standard deviation) unless otherwise indicated.
TAU, Treatment as usual; IQ, intelligence quotient; HAMD, Hamilton Depression Rating Scale; YMRS, Young Mania Rating Scale; FAST, Functioning Assessment Short Test.
Attrition rates
During the intervention a total of 27 patients discontinued the study, distributed as follows: 14.3% (n = 8), 11.6% (n = 8) and 17.5% (n = 11) in the functional remediation, psychoeducation and TAU groups, respectively. These rates were not significantly different between groups (χ2 = 0.92, p = 0.63). The main reason for discontinuation was being lost to follow-up. Finally, extra analyses were performed between drop-outs and completers regarding their clinical and sociodemographic characteristics at baseline. No differences were found between both groups in terms of these variables indicating that the drop-out patients were not a source of bias for the present analyses.
Treatment effects on neurocognitive variables and functional outcome
Compared with participants receiving TAU or psychoeducation, those who attended functional remediation improved in delayed free recall on the CVLT at 6-month follow- up, as shown by the significant group x time interaction in the repeated-measures analysis (F 2,158 = 3.37, p = 0.04) (see Fig. 2 and Table 2).
Fig. 2. Changes in delayed free recall at 6-month follow-up. TAU, Treatment as usual; CVLT, California Verbal Learning Test.
Table 2. Neuropsychological outcomes at baseline and at 6-month follow-up
Data are given as mean (standard deviation).
TAU, Treatment as usual; WCST, Wisconsin Card Sorting Test; SCWT, Stroop Color and Word Test.
a Phonemic (F-A-S) and categorical (animal naming) components of the Controlled Oral Word Association Test.
Tukey post-hoc analyses revealed that functional remediation group was only superior when compared with TAU (p = 0.04) but not to psychoeducation (p = 0.10). The between-groups effect size was small (Cohen's d = 0.2 and 0.3 for functional remediation v. TAU and psychoeducation, respectively). However, the within-group effect size, the comparison of pre- and post-treatment, in the functional remediation was moderate (Cohen's d = 0.64), in psychoeducation the within-group effect size was small (Cohen's d = 0.26) and in the TAU group it was small too (Cohen's d = 0.39).
No statistically significant differences were found in the remaining comparisons of neurocognitive variables, although a trend was detected in the group × time interaction towards an improvement in favour of the functional remediation group in verbal learning (F 2,158 = 2.8, p = 0.064) (Table 2), showing a medium within-group effect size (Cohen's d = 0.6). The between-groups effect size at follow-up was small for functional remediation when compared with both the TAU (Cohen's d = 0.3) and psychoeducation (Cohen's d = 0.4) groups.
The group receiving functional remediation also improved significantly their functional outcome when compared with psychoeducation and TAU as shown by the group x time interaction (F 2,158 = 4.26, p = 0.016). These results do not differ from the original study (Torrent et al. Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013). See Fig. 3 for further details.
Fig. 3. Improvement of functional outcome. TAU, Treatment as usual; FAST, Functioning Assessment Short Test.
Changes in the FAST scale and changes in the CVLT free delayed recall were not significantly correlated (r = 0.10, p = 0.22). However, at the end of the study there was a significant association between higher FAST total score and lower free delayed score in the CVLT (r = −0.25, p = 0.002). Finally, changes in CVLT free delayed recall did not correlate with changes in HAMD total score (r = 0.17, p = 0.29) or with YMRS total score (r = 0.14, p = 0.39).
Discussion
The main result of this study was that participants selected for being neurocognitively impaired improved after the administration of functional remediation. Specifically, they presented better performance in verbal memory measures such as the delayed free recall measured by the CVLT only when compared with TAU, with a moderate effect size (Cohen's d = 0.6). Although not significant, a trend in improving verbal learning was in the same direction. In addition to this, patients in the functional remediation group also improved their psychosocial functioning, which was the primary outcome of the original trial. Overall, it seems that functional remediation, when applied to neurocognitively impaired patients, enhances not only functional outcome but also verbal memory, and, more specifically, learning and delayed free recall.
There may be different reasons to explain why functional remediation improves verbal memory particularly. First, this intervention involves training exercises related to memory and learning skills, such as association, categorization and mental imagery. This module represents almost one-third of the whole intervention and it is the most extensive part when compared with the remaining neurocognitive training sessions. Furthermore, of all the neuropsychological functions involved, verbal memory is perhaps the most sensitive function related to the enhancement of psychosocial functioning, especially when measured with the CVLT, since there is an international agreement (International Society for Bipolar Disorders-Battery for Assessment of Neurocognition; ISBD-BANC) on this test as the most suitable one to assess verbal learning and memory in bipolar disorder (Yatham et al. Reference Yatham, Torres, Malhi, Frangou, Glahn, Bearden, Burdick, Martinez-Aran, Dittmann, Goldberg, Ozerdem, Aydemir and Chengappa2010). These results introduce a clinically relevant topic in the field of the rehabilitation of psychiatric patients which is whether the level of neurocognitive impairment may influence the response to psychological interventions or not. According to our results, neurocognitively impaired bipolar patients not only improved in psychosocial functioning, but also in delayed free recall. The fact that most of these patients were chronic with a long duration of illness and multiple episodes suggests that functional remediation may be effective in later stages of the illness. This is at odds with the results of Deckersbach et al. (Reference Deckersbach, Nierenberg, Kessler, Lund, Ametrano, Sachs, Rauch and Dougherty2010), who found that patients with neurocognitive impairment benefit less from their cognitive rehabilitation. However, patients in that trial were mildly symptomatic, which can influence neuropsychological and functional outcome (Bonnin et al. Reference Bonnin, Sanchez-Moreno, Martinez-Aran, Sole, Reinares, Rosa, Goikolea, Benabarre, Ayuso-Mateos, Ferrer, Vieta and Torrent2012). Regarding the present results, it is worth mentioning that the changes in delayed free recall were not correlated with changes in functional outcome. This was an unexpected result, since it has been consistently found that verbal memory impairment is related to worse functional outcome (Martínez-Arán et al. Reference Martinez-Aran, Vieta, Torrent, Sanchez-Moreno, Goikolea, Salamero, Malhi, Gonzalez-Pinto, Daban, Alvarez-Grandi, Fountoulakis, Kaprinis, Tabares-Seisdedos and Ayuso-Mateos2007; Martino et al. Reference Martino, Marengo, Igoa, Scapola, Ais, Perinot and Strejilevich2009; Bonnin et al. Reference Bonnin, Martinez-Aran, Torrent, Pacchiarotti, Rosa, Franco, Murru, Sanchez-Moreno and Vieta2010). Even though these are preliminary results, we have different hypotheses that may explain this lack of relationship between the two variables: first, it might be attributed to the loss of statistical power due to small sample size in the functional remediation group. Second, these results may suggest that, at least, in patients with neurocognitive impairment, the improvement of verbal memory (delayed free recall) and the improvement of functional outcome are independent from each other; despite the neurocognitive changes being subtle, patients have acquired a broad range of strategies to cope with difficulties in daily life. Controversially, in a recent study of functional remediation, a correlation between neurocognitive changes and functional outcome was found at 12-month follow-up (Bonnín et al. Reference Bonnín, Torrent, Arango, Amann, Solé, González-Pinto, Crespo, Tabarés-Seisdedos, Reinares, Ayuso-Mateos, García Portilla, Ibáñez, Salamero, Vieta and Martínez-Aránin press). In light of this, it may be hypothesized that the 6-month follow-up is not long enough to detect significant correlations and that the consolidation of learned skills requires time. It may depend mainly on training and sustained practice of such strategies that become evident at 12-month follow-up. Finally, changes in verbal memory were not related either to changes in mood symptomatology, measured by means of the HAMD and the YMRS scores changes.
The high rate of impaired bipolar patients fulfilling criteria for neurocognitive impairment may sound bad news for the field; however, this result is most probably due to different artifacts of this subanalysis. First, it may be related to the inclusion criteria in the original study by Torrent et al. (Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013), which required a high level of functional impairment and as a consequence of this, more severe patients were selected. Second, it may also be related to the cut-off we have used to establish neurocogntive impairment (Martino et al. Reference Martino, Strejilevich, Scapola, Igoa, Marengo, Ais and Perinot2008). There is a great variability in the literature using different cut-offs to establish neurocognitive impairment (van der Werf-Eldering et al. Reference van der Werf-Eldering, Burger, Holthausen, Aleman and Nolen2010; Hellvin et al. Reference Hellvin, Sundet, Simonsen, Aminoff, Lagerberg, Andreassen and Melle2012; Martino et al. Reference Martino, Strejilevich, Marengo, Ibáñez, Scápola and Igoa2014; Volkert et al. Reference Volkert, Kopf, Kazmaier, Glaser, Zierhut, Schiele, Kittel-Schneider and Reif2015). The application of one or other criteria will inevitably lead not only to different results but also to different rates of neurocognitive impairment (Altshuler et al. Reference Altshuler, Ventura, van Gorp, Green, Theberge and Mintz2004; Reichenberg et al. Reference Reichenberg, Harvey, Bowie, Mojtabai, Rabinowitz, Heaton and Bromet2009; Burdick et al. Reference Burdick, Russo, Frangou, Mahon, Braga, Shanahan and Malhotra2014). One of the limitations of the criteria used for the present study is that the criteria of Martino et al. (Reference Martino, Strejilevich, Scapola, Igoa, Marengo, Ais and Perinot2008), together with other previous studies (Hellvin et al. Reference Hellvin, Sundet, Simonsen, Aminoff, Lagerberg, Andreassen and Melle2012; Volkert et al. Reference Volkert, Kopf, Kazmaier, Glaser, Zierhut, Schiele, Kittel-Schneider and Reif2015), might be among the less restrictive ones in the literature, especially with respect to other works that evaluate neurocognitive impairment in healthy controls where the occurrence of a single impaired test score on a neuropsychological battery is not uncommon (Binder et al. Reference Binder, Iverson and Brooks2009; Iverson et al. Reference Iverson, Brooks, Langenecker and Young2011). However, using the criteria of Martino et al. (Reference Martino, Strejilevich, Scapola, Igoa, Marengo, Ais and Perinot2008) in the present study ensured that the 20% of patients excluded to perform this analysis were the ones with the most preserved neurocognitive performance of the sample and allowed us to include all the heterogeneity in the neurocognitive impairment that can be observed in bipolar disorder (Martino et al. Reference Martino, Strejilevich, Marengo, Ibáñez, Scápola and Igoa2014). Because of this, the present results should be interpreted with caution and they also highlight the need to achieve a consensus about the cut-off to establish neurocognitive impairment in bipolar disorder.
Similarly, the fact that 80% of the patients from the Torrent et al. (Reference Torrent, Bonnin, Martinez-Aran, Valle, Amann, Gonzalez-Pinto, Crespo, Ibanez, Garcia-Portilla, Tabares-Seisdedos, Arango, Colom, Sole, Pacchiarotti, Rosa, Ayuso-Mateos, Anaya, Fernandez, Landin-Romero, Alonso-Lana, Ortiz-Gil, Segura, Barbeito, Vega, Fernandez, Ugarte, Subira, Cerrillo, Custal, Menchon, Saiz-Ruiz, Rodao, Isella, Alegria, Al-Halabi, Bobes, Galvan, Saiz, Balanza-Martinez, Selva, Fuentes-Dura, Correa, Mayoral, Chiclana, Merchan-Naranjo, Rapado-Castro, Salamero and Vieta2013) study qualified for this subanalysis [compared with 62% in the Martino et al. (Reference Martino, Strejilevich, Marengo, Ibáñez, Scápola and Igoa2014) study, using the same criteria] may be attributable to different reasons. First, and as mentioned before, one of the inclusion criteria in the original study was to have a FAST score ⩾18, which means that patients were markedly impaired in their psychosocial functioning at baseline. In addition, patients had to have a score of at least 4 points in the cognitive domain of this scale; this indicated that patients presented neurocognitive complaints detected by the clinician, since this scale is not self-administered. Clinicians hereby took into account not only patients’ subjective complaints but also the information by relatives and clinical history. Second, patients in the present study were recruited from different specialized reference centres in Spain which usually treat difficult-to-treat and severe cases, hence this potential source of bias cannot be ruled out. Finally, another reason for the discrepancy of our results compared with those of Martino et al. (Reference Martino, Strejilevich, Marengo, Ibáñez, Scápola and Igoa2014) may be the different percentage of BD-I and BD-II patients in each sample. While in the study of Martino et al. (Reference Martino, Strejilevich, Marengo, Ibáñez, Scápola and Igoa2014), BD-I and BD-II were equally distributed (48% BD-I and 52% BD-II), in our sample, the vast majority of patients were diagnosed with BD-I (up to 76%), who usually show more severe neurocognitive deficits than BD-II patients (Sole et al. Reference Sole, Martinez-Aran, Torrent, Bonnin, Reinares, Popovic, Sanchez-Moreno and Vieta2011). All these differences taken together may help to explain the discrepancy of impaired patients in both studies, even following the same method as described by Martino et al. (Reference Martino, Strejilevich, Marengo, Ibáñez, Scápola and Igoa2014).
Methodological caveats of our study must be taken into account when translating results to clinical practice. First, the 6-month follow-up period is short and the results cannot be generalizable for long-term outcome. For the same reason, evident conclusions cannot be deduced whether the improvement remains stable or varies over time. Second, as this was an exploratory subanalysis, we have not conducted any statistic procedure to control for multiple comparisons when analysing neurocognition; instead we selected a number of variables based on previous literature and on neurocognitive domains we work on in the functional remediation programme. Despite all, caution is needed, since only one out of 11 comparisons (delayed recall) among the neurocognitive variables was found to be statistically significant, and a trend was detected for another variable (verbal learning).
Finally, as stated before, the majority of subjects included in this trial were chronic patients with multiple admissions and with marked functional impairment which might interfere with a generalization of less severe cases with bipolar disorder.
Despite the previous limitations, this study shows that functional remediation at 6-month follow-up is effective not only at improving psychosocial functioning, but also at enhancing delayed free recall in the CVLT in a sample of patients with neurocognitive impairment. It supports the notion that neurocognitive enhancement is an active ingredient of functional remediation, at least in this subsample of patients, and that verbal memory seems to be more amenable to treatment than other neurocognitive domains. If the present results are confirmed by further studies, these findings may have important consequences in treating chronic bipolar patients, bringing hope to the subset of patients with neurocognitive and functional decline, and for whom there is the utmost unmet need (Reinares et al. Reference Reinares, Sanchez-Moreno and Fountoulakis2014).
Acknowledgements
This study was supported by an ETES grant from the Spanish Ministry of Economy and Competitiveness: PI080180, PI08/90825, PI08/90327, PI08/90675, PI08/90224, PI08/90654, PI08/90189, PI08/90916, PI08/90416, PI08/90094, PI11/00637, PI12/00912, integrated in the Plan Nacional de I + D + I and cofinanced by el Fondo Europeo de Desarrollo Regional (FEDER), the CIBERSAM and the Comissionat per a Universitats i Recerca del DIUE de la Generalitat de Catalunya to the Bipolar Disorders Group (2014 SGR 398).
M.R. was supported by Beatriu de Pinós, Secretaria d'Universitats i Recerca del Departament d'Economia i Coneixement, de la Generalitat de Catalunya i del programa COFUND de les Accions Marie Curie del 7è Programa marc de recerca i desenvolupament tecnològic de la Unió Europea.
M.R.-C. was supported by a Sara Borrell Health Research Fellowship from the Institute of Health Carlos III, Spanish Ministry of Economy and Competitiveness, an Alicia Koplowitz Research Grant and an Alicia Koplowitz Grant for Short-Term Placements from the Alicia Koplowitz Foundation (Madrid, Spain).
F. C. would like to thank the support and funding of the Spanish Ministry of Health, Instituto de Salud Carlos III, CIBERSAM. F.C. is also funded by the Spanish Ministry of Economy and Competitiveness, Instituto Carlos III, through a ‘Miguel Servet’ contract (CP08/00140) and the Fondo de Investigación Sanitaria (FIS) (PI 12/00910).
A.M.-A.'s project is supported in part by a 2013 NARSAD, Independent Investigator Grant from the Brain & Behavior Research Foundation.
B.L.A. would like to thank the support and funding of the Spanish Ministry of Health, Instituto de Salud Carlos III, for a ‘Miguel Servet’ contract (CP06/00359) which has been extended in 2013 as a stabilization contract (CES 12/024).
Declaration of Interest
E.V. has served as consultant, advisor or speaker for the following companies: Alexza, Almirall, AstraZeneca, Bial, Bristol-Myers Squibb, Elan, Eli Lilly, Ferrer, Forest Research Institute, Gedeon Richter, Glaxo-Smith-Kline, Janssen-Cilag, Jazz, Johnson & Johnson, Lundbeck, Merck and Co. Inc., Novartis, Organon, Otsuka, Pfizer Inc., Roche, Sanofi-Aventis, Servier, Schering-Plough, Shire, Sunovion, Takeda, United Biosource Corporation and Wyeth.
A.M.-A. has served as speaker or advisor for the following companies: Bristol-Myers Squibb, Otsuka, Lundbeck and Pfizer.
C.A. has been a consultant to or has received honoraria or grants from Abott, AMGEN, AstraZeneca, Bristol-Meyers Squibb, Caja Navarra, CIBERSAM, Fundación Alicia Koplowitz, Instituto de Salud Carlos III, Janssen Cilag, Lundbeck, Merck, Ministerio de Ciencia e Innovación, Ministerio de Sanidad, Ministerio de Economía y Competitividad, Mutua Madrileña, Otsuka, Pfizer, Roche, Servier, Shire, Takeda and Schernig Plough.
V.B.-M. has received grants and served as consultant, advisor or continuing medical education (CME) speaker during the last 3 years for the following entities: Almirall, Angelini, AstraZeneca, Bristol-Myers-Squibb, Janssen; Juste, Lundbeck, Otsuka, and Fundación Alicia Koplowitz.
M.P.G.-P. has served as consultant and/or speaker for Janssen-Cilag, Lilly, Lundbeck, Otsuka, Pfizer, Servier, Roche, and Rovi, and has received grants from the Ministry of Economy and Competitiveness (Instituto de Salud Carlos III, FIS), European Comission (FP7) and CIBERSAM.
F.C. has served as advisor or speaker for the following companies (lifetime): Adamed, Astra Zeneca, Bristol-Myers, Eli Lilly, Glaxo-Smith-Kline, Lundbeck, MSD Merck, Otsuka, Pfizer Inc., Rovi, Sanofi-Aventis, Sanovel, Shire and Tecnifar. He has received copyright fees from: Cambridge University Press, Igaku-Shoin Ltd, Solal Ed., Ars Médica, Giovani Fioriti Ed., Medipage, La Esfera de Los Libros, Morales i Torres Ed., Panamericana, Mayo Ed. & Columna.
A. I. has served as speaker or advisor for the following companies: Adamed, Bristol-Myers Squibb, Ferrer, Janssen-Cilag, Lundbeck, Pfizer, Servier and Otsuka.
B.L.A. has served as a speaker for Bristol-Myers Squibb, Otsuka and Janssen-Cilag.
J.M.C. has served as advisor or speaker for the following entities: Almirall, AstraZeneca, Bristol-Myers Squibb, Janssen-Cilag, Ferrer, Lundbeck, Merck, Novartis, Otsuka, Pfizer, Sanofi-Aventis, Servier, the Spanish Ministry of Science and Innovation (CIBERSAM), the Ministry of Science (Carlos III Institute) and Wyeth.
A.G.-P. has received grants and served as consultant, advisor or CME speaker for the following entities: AstraZeneca, Bristol-Myers Squibb, Cephalon, Eli Lilly, Janssen-Cilag, Lundbeck, Merck, Otsuka, Pfizer, Sanofi-Aventis, Servier, the Spanish Ministry of Science and Innovation (CIBERSAM), the Ministry of Science (Carlos III Institute), the Basque Government, the Stanley Medical Research Institute and Wyeth.
The remaining authors have no conflicts of interest.
