Elsevier

Journal of Psychiatric Research

Volume 84, January 2017, Pages 207-213
Journal of Psychiatric Research

Memory performance predicts recurrence of mania in bipolar disorder following psychotherapy: A preliminary study

https://doi.org/10.1016/j.jpsychires.2016.10.008Get rights and content

Highlights

  • We investigated the association between general intelligence, memory and relapse.

  • We compared euthymic BD patients receiving either CBT or Supportive Therapy (ST).

  • Verbal recall reduces recurrence of mania to a greater extent in ST than CBT.

  • Verbal recall may be helpful in unstructured, non-directive therapeutical settings.

Abstract

Objective

Cognitive complaints are common features of bipolar disorder (BD). Not much is, however, known about the potential moderator effects of these factors on the outcome of talking therapies. The goal of our study was to explore whether learning and memory abilities predict risk of recurrence of mood episodes or interact with a psychological intervention.

Method

We analyzed data collected as part of a clinical trial evaluating relapse rates following Cognitive Behavioral Therapy (CBT) and Supportive Therapy (ST) (Meyer and Hautzinger, 2012). We included cognitive (Auditive Verbal Learning Test, general intelligence - Leistungsprüfsystem) and clinical measures from 76 euthymic patients with BD randomly assigned to either 9 months of CBT or ST and followed up for 2 years.

Results

Survival analyses including treatment condition, AVLT measures, and general intelligence revealed that recurrence of mania was predicted by verbal free recall. The significant interaction between therapy condition and free recall indicated that while in CBT recurrence of mania was unrelated to free recall performance, in ST patients with a better free recall were more likely to remain euthymic, and those with a poorer free recall were less likely to remain mania-free.

Conclusions

These findings constitute first evidence that, when considering treatment outcome in BD, differences in verbal free recall might interact with the kind of psychotherapy provided. More research is needed to determine what other areas of cognitive functioning are related to outcome in psychological interventions.

Introduction

The integration of non-pharmacological and pharmacological interventions has been shown to be beneficial and improves the long-term therapeutic outcomes of mood disorder patients (Lobban et al., 2007, Miklowitz et al., 2007, Oestergaard and Møldrup, 2011, Pfennig et al., 2013). The majority of studies in this field had ‘prevention of relapse’ as the primary outcome measure (Castle et al., 2010, Colom et al., 2003, Oestergaard and Møldrup, 2011), but improvements in medication adherence (Colom and Lam, 2005, Pampallona et al., 2004), mood symptoms (Colom et al., 2009), quality of life, and well-being have also been reported (Zilcha-Mano et al., 2014).

While some studies concluded that interventions such as cognitive behavioral therapy (CBT) are highly effective in bipolar disorder (BD), other trials including heterogeneous populations during both acute and euthymic phases of BD did not show significant changes in mood or relapse (Driessen and Hollon, 2010, Parikh, 2008). However, Colom et al.’s study showed that combining medication and psychosocial intervention in a stabilized population led to a reduced number of relapses and improved medication adherence, as well as improved psychosocial functioning. Notably these effects lasted up to 5 years post-intervention (Colom et al., 2009).

The number of studies investigating the factors predicting the efficacy of psychological treatments is limited and usually focused on predictors such as age of onset or number of prior mood episodes (Lam et al., 2009, Reinares et al., 2014). One potential predictor or moderator of outcome could be cognitive abilities. However, only few studies have investigated this area in psychiatry and, to date, no published study has examined cognitive functioning as part of psychological treatments for bipolar disorder (BD). In a study of older adults with anxiety disorders, there was a positive association between general intelligence and improvement in anxiety symptoms in the supportive counseling condition but not CBT (Doubleday et al., 2002). Similar findings by D'Alcante et al. revealed that higher verbal intelligence scores and immediate verbal recall predicted a better treatment response to both CBT and fluoxetine in adults with OCD (D'Alcante et al., 2012). Furthermore, lower intelligence scores appeared to be predictors of poor treatment response in adults with depression (Fournier et al., 2009) and Post-traumatic Stress Disorder (Rizvi et al., 2009). Looking at psychosis, only a very small number of studies have investigated this issue with one study reporting that treatment response to CBT in patients with psychosis was not related to cognitive performance (Granholm et al., 2008).

This kind of research is relevant because memory, executive functions and pre-morbid intelligence are essential for completing daily activities involving goal setting and planning (Lezak, 2004). Thus, it is possible that pre-existing interindividual differences in cognitive abilities affect the extent to which people benefit from different psychological interventions (Doubleday et al., 2002). However, the literature in this field is a) very heterogeneous and b) also still controversial as a number of studies have not found a consistent link between cognitive performance and response to therapy across mood disorders (Knekt et al., 2014, Voderholzer et al., 2013). Thus, further research is needed to test this link.

Results about a potential link between cognitive functioning and outcome of talking therapies are highly relevant to BD as impairment in cognitive functions have been observed across all phases of BD (Bora et al., 2009, Glahn et al., 2007, Quraishi and Frangou, 2002), and cognitive impairment is especially pronounced during the manic and depressive phases of the disorder (Robinson et al., 2006). Additionally, all of the studies we identified and that looked at whether cognitive performance is related to outcome of talking therapies have focused on acute patient populations. However, most studies evaluating psychological interventions for BD, so far, have focused on relapse prevention and recruited patients in remission or immediately after an acute episode. Therefore, it is not known whether neuropsychological functioning in remitted patients will predict or moderate the outcome of talking therapies.

In sum, studies focusing on the impact of neuropsychological performance on treatment response to talking therapies in general and to CBT specifically are lacking. Furthermore, it is unclear whether differences in cognitive performance in patients with BD are prospectively associated with the risk of relapse. The current paper focuses on unpublished neuropsychological data collected as part of a randomized controlled trial (RCT) (Meyer and Hautzinger, 2012) comparing CBT and Supportive Therapy (ST) for remitted patients with BD. In this RCT, CBT and ST were matched with regards to number and duration of sessions, and both conditions included psychoeducation and a mood diary. While CBT additionally included typical cognitive and behavioral strategies and techniques to prevent relapse or how to cope with symptoms (e.g Basco and Rush, 1996). ST had a client-centered focus and was less structured and less directive. While this RCT found no overall difference in relapse rates between CBT and ST over almost 3 years, a higher number of prior mood episodes and a lower number of attended therapy sessions were associated with a shorter time before relapse. These findings suggest that characteristics shared by both treatments may contribute to the outcome. However, this prompted us to explore post hoc other potential moderators, and in this case whether indicators of general intelligence, verbal learning and memory predict recurrence of mood episodes during follow-up and whether there is an indication that cognitive performance and treatment interact. Verbal learning and memory were originally chosen because prior research showed that patients with BD demonstrated deficits in this area (Deckersbach et al., 2004, van Gorp et al., 1998).

Section snippets

Participants

One-hundred-forty adults with BD were either referred by local hospitals, psychiatrists, or self-referred in response to advertisements. Nine of 107 participants who completed the baseline assessment voluntarily withdrew from the study (n = 9) and 22 were excluded because they did not have bipolar disorder (n = 20) and had current opiate or alcohol dependency (n = 2). The analyses for the current paper include the clinical and cognitive data from all 76 participants which were randomized into

Demographics

As reported in Meyer and Hautzinger's study (Meyer and Hautzinger, 2012), participants allocated to the CBT and ST therapy were comparable in terms of age, gender, and clinical status as measured by the lifetime number of mood episodes, age of onset of the disease, and severity of the depressive and manic symptoms (e.g. HAMD, SRMI). In terms of cognitive measures participants displayed a comparable performance on LPS, Trials 1–5, Free Recall and Recognition scores (p > 0.05) (Table 1). As

Discussion

To the authors’ knowledge this is the first study exploring whether indicators of general intelligence, verbal learning and memory predict recurrence of mood episodes in euthymic adults with BD who received either 9 months of CBT or ST. The most important finding of the post-hoc analyses of this RCT is that although there was no overall significant difference in the risk of recurrence between CBT and ST (Meyer and Hautzinger, 2012) and general intelligence did not predict outcome of

Conflicts of interest

Dr Bauer has no conflicts of interest.

Professor Martin Hautzinger has received honorarium from Lundbeck, Sevier, and Merk.

Professor Thomas D. Meyer has been a speaker for Pfizer and Lundbeck.

Contribution

IB analyzed the data and wrote the first draft of the manuscript. MH and TDM designed the study, wrote the protocol and collected the data. All authors contributed to the interpretation of the data, and have approved the final manuscript.

Role of funding source

This research was supported by a grant provided from the Deutsche Forschungsgemeinschaft (DFG ME 1681/6-1 to 6.3) and, in part, by NIMH grant R01 085667 and the Pat Rutherford, Jr. Endowed Chair in Psychiatry (UTHealth).

Acknowledgement

We are indebted to all the therapists and independent raters, especially Dr. Peter Peukert and Dr. Katja Salkow as well as research assistants for their enormous work, support and help.

References (49)

  • S. Quraishi et al.

    Neuropsychology of bipolar disorder: a review

    J. Affect. Disord.

    (2002)
  • M. Reinares et al.

    Psychosocial interventions in bipolar disorder: what, for whom, and when

    J. Affect. Disord.

    (2014)
  • S.L. Rizvi et al.

    Cognitive and affective predictors of treatment outcome in cognitive processing therapy and prolonged exposure for posttraumatic stress disorder

    Behav. Res. Ther.

    (2009)
  • L.J. Robinson et al.

    A meta-analysis of cognitive deficits in euthymic patients with bipolar disorder

    J. Affect. Disord.

    (2006)
  • U. Voderholzer et al.

    Cognitive functioning in medication-free obsessive-compulsive patients treated with cognitive-behavioural therapy

    J. Obsessive Compuls. Relat. Disord.

    (2013)
  • S. Zilcha-Mano et al.

    Changes in well-being and quality of life in a randomized trial comparing dynamic psychotherapy and pharmacotherapy for major depressive disorder

    J. Affect. Disord.

    (2014)
  • P.D. Allison

    Multiple Regression: A Primer

    (1999)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorders: DSM-IV

    (1994)
  • J.R. Ball et al.

    A randomized controlled trial of cognitive therapy for bipolar disorder: focus on long-term change

    J. Clin. Psychiatry

    (2006)
  • M. Basco et al.

    Cognitive–Behavioral Treatment of Manic Depressive Disorder

    (1996)
  • P. Bech et al.

    The use of rating scales exemplified by a comparison of the Hamilton and the Bech-Rafaelsen Melancholia scale

    Acta Psychiatr. Scand.

    (1980)
  • J.S. Beck

    Cognitive Therapy

    (1995)
  • D. Castle et al.

    Group-based psychosocial intervention for bipolar disorder: randomised controlled trial

    Br. J. Psychiatry

    (2010)
  • F. Colom et al.

    A perspective on the use of psychoeducation, cognitive-behavioral therapy and interpersonal therapy for bipolar patients

    Bipolar Disord.

    (2004)
  • Cited by (15)

    • Associations between emotional and non-emotional cognition and subsequent mood episodes in recently diagnosed patients with bipolar disorder: A 16-month follow-up study

      2023, Journal of Affective Disorders
      Citation Excerpt :

      Moreover, the follow-up period of 16-months was relatively short. Nevertheless, the short follow-up is likely not able explain why cognition generally did not predict depression as studies with longer follow-up times ranging 24 to 71 months have found no association between none-emotional cognitive performance and depressive relapse (Bauer et al., 2017; Sanchez-Morla et al., 2019; Valerio et al., 2020). It was a limitation that we assessed emotional cognition using only two tasks (i.e., the Facial Expression Recognition Test and the Social Scenarios Test); of which the latter lacks reliability and validity data.

    • Weekly telephone mood monitoring is associated with decreased suicidality and improved sleep quality in a clinical sample

      2022, Psychiatry Research
      Citation Excerpt :

      Nonetheless, the ASRM has high clinical utility (see Cerimele et al., 2019 for review). Further, while there is an ongoing debate regarding the reliability and validity of self-report measures, especially in terms of mania, Bauer and colleagues (2017) found that self-rated subsyndromal manic symptoms, but not clinician-rated, predicted manic relapses. The ASRM was used in studies in South African samples (Savitz et al., 2008; van der Watt et al., 2018c).

    • Associations between cognition and subsequent mood episodes in patients with bipolar disorder and their unaffected relatives: A systematic review

      2022, Journal of Affective Disorders
      Citation Excerpt :

      One study found lower verbal memory domain scores to be associated with more annual (hypo)manic episodes measured within 71 months (Valerio et al., 2020). The other study reported that low free recall on the AVLT predicted shorter time to (hypo)manic relapse during 24 months (Bauer et al., 2017). In contrast, a study of visual learning and memory (Boston Naming Test) found no correlation between the number of depressive episodes per year over 71 months (Valerio et al., 2020).

    View all citing articles on Scopus
    View full text