In the light of several recent reports on long-term effectiveness of psychoeducation the Dutch Foundation for Bipolar Disorders (DFBD) decided to update the psychoeducation program. Most systematic reviews on psychoeducation in bipolar disorder date from 2008 or before and even newer ones do not include all relevant studies.

The aims of this review are;

1. 1) to determine the long-term effectiveness of addition of psychoeducation to treatment as usual in terms of symptoms, relapse, suicide, quality of life and other secondary outcomes (see below) in bipolar patients;

2. 2) to review the acceptability of psychoeducation to patients and key-persons, measured by numbers of withdrawal from this intervention;

3. 3) to determine long-term beneficial effects on key-persons if involved in psychoeducation.

A protocol-based systematic literature search and review.

We found three randomized controlled trials (RCTs) of group psychoeducation, one trial of individual psychoeducation, one trial of multifamily psychoeducation and three trials of caregiver grouppsychoeducation showing efficacy of psychoeducation in mood stability of the patient and/or better coping and less burden in caregivers. The trials show similarities but also differences in content and delivery of psychoeducation.

Psychoeducation is efficacious in bipolar disorder. Based on the extensive experience in bipolar psychoeducation in the Netherlands and in the light of the results of the RCTs we suggest a prototype of psychoeducation in a multifamily format with 12 sessions and additional booster sessions.

Previous work suggests that impairments in executive function and verbal memory in particular may persist in euthymic bipolar patients and serve as an indicator of genetic risk (endophenotype).

A systematic review of the literature was undertaken. Effects sizes were extracted from selected papers and pooled using meta-analytical techniques.

In bipolar patients, large effect sizes (d>0.8) were noted for executive functions (working memory, executive control, fluency) and verbal memory. Medium effect sizes (0.5<d<0.8) were reported for aspects of executive function (concept shifting, executive control), mental speed, visual memory, and sustained attention. Small effect sizes (d<0.5) were found for visuoperception. In first-degree relatives, effect sizes were small (d<0.5), but significantly different from healthy controls for executive function and verbal memory in particular.

Executive function and verbal memory are candidate bipolar endophenotypes given large deficits in these domains in bipolar patients and small, but intermediate, cognitive impairments in first-degree relatives.