NICE guidelines advise to consider admission for patients with borderline personality disorder (BPD) for the management of crises involving significant risk to self or others. Furthermore, to consider structured psychological interventions of greater than three months’ duration and twice-weekly sessions according to patients’ needs and wishes.

We aimed to assess reasons for admission and access to psychological interventions in an acute inpatient BPD population.

Case notes of patients with a diagnosis of BPD (ICD-10 F60.3 and F60.31), discharged from four acute general adult wards in Sheffield during a period of twelve months were studied retrospectively, using a structured questionnaire based on BPD NICE guidance.

Of the 83 identified BPD patients, seventy-eight percent were female and 82% between 16–45 years old. Eleven patients had four or more admissions. Eighty percent reported suicidal ideation at admission, with 50% having acted on it (70% by overdose, 50% cutting, 10% hanging). Of this cohort, 58% reported they intended to die. Psychosocial factors at admission were identified in 59 cases, including relationship breakdown (47.5%), alcohol/drug use (30.5%) and accommodation issues (17%). Disturbed/aggressive behaviour was documented in 27.1% of these cases. Sixty-eight percent of patients had psychology input in the 5 years preadmission: 38% (21 patients) received structured therapy, whilst 62% received only one assessment or advise to teams.

Patients were mainly admitted for risk management. A high proportion received unstructured psychological interventions. Services offering structured psychological interventions should be supported, as hospitalisations only temporarily address BPD patients’ suicidality and psychosocial difficulties.

The authors have not supplied their declaration of competing interest.

In this study, we aim to assess the role of weight suppression (WS) in eating disorders, not only from a quantitative point of view but also assessing the speed of the weight loss by using a new parameter: the weight loss rate (WLR). We analysed the role of these two indexes in different eating disorders domains, considering both eating behaviours and outcome profiles.

The sample consisted of 414 patients, including 62 with AN binge purge subtype (ANBP), 146 with AN restrictive subtype (ANR) and 206 with bulimia nervosa (BN). Data about response to treatment were available for a subsample of 201 patients. A cross-sectional design was used for the clinical symptoms detected during the initial assessment and a longitudinal design was adopted for the response to treatment analysis.

No significant relationship emerged between both WS and WLR and variables collected at baseline assessment. We observed, on the contrary, a significant association between WS and weight gain at the end of treatment. High WLR predicted remission of binge eating and compensatory behavior in BN patients. We further analyzed our data to identify threshold values of both WS and WLR of clinical utility.

The role of WS and WLR as predictive factors in the outcome of eating disorders is of great interest and these initial results remark the usefulness of collecting these data during the initial assessment in order to plan a tailored therapeutic intervention.

The authors have not supplied their declaration of competing interest.

Adolescence is a period of increasing demands in independent social functioning where parenting style may have an impact on social anxiety. A context-dependent effect of oxytocin on human social behavior has been suggested, however research on the gene coding for oxytocin (OXT) has mostly been reported without considering contextual factors.

This study investigated interactions between parenting style and polymorphic variations in the OXT gene in association with social anxiety symptoms in a community sample of adolescents.

The study group consisted of 1359 adolescents. Two single nucleotide polymorphisms located near OXT, rs4813625 and rs2770378, were genotyped. Social anxiety and perceived parenting style were assessed by behavioral questionnaires.

Significant joint effects in line with the differential susceptibility framework were observed for rs4813625 with parenting style. The levels of social anxiety among C allele carriers were conditional on the level of supportive parenting style whereas homozygote G carriers’ levels of social anxiety were unaffected by supportive parenting style. The nature of the interactions between rs2770378 and parenting style was in line with the diatheses-stress model. However, associations of rs2770378 and parenting style with social anxiety became nonsignificant in nonlinear models.

The study provides preliminary evidence for a modifying effect of supportive parenting style on the relationship between rs4813625 and social anxiety symptoms in adolescents, independent of sex. The findings may be interpreted from the perspective of the social salience hypotheses of oxytocin, with rs4813625 affecting social anxiety levels along a perceived unsafe–safe social context dimension.

The authors have not supplied their declaration of competing interest.

The authors have not supplied their declaration of competing interest.

The authors have not supplied their declaration of competing interest.

Obsessive-compulsive disorder (OCD) is described as a chronic condition. However, relatively little is known about predictors affecting its long-term outcome.

To examine the contribution of clinical and interpersonal determinants in predicting remission status of 254 OCD patients at one-year follow-up (FU1).

We used the baseline and FU1 data of The Netherlands obsessive-compulsive disorder association (NOCDA) study. Clinical predictors were chronicity level, depressive and obsessive-compulsive symptom severity. Suspected interpersonal factors included attachment style, social support and expressed emotion. Remission status was defined using the Yale-Brown obsessive-compulsive symptom (Y-BOCS) scale. χ2 tests and ANOVAs were used for bivariate analyses, followed by multivariate multinomial logistic regression analyses to assess main effects and interactions in predicting remission status at FU1.

Bivariate tests demonstrated significant differences in remission status as a function of chronicity level, depressive and obsessive-compulsive symptom severity and social support. Regression analyses revealed that increased baseline OCD severity reduced the odds for both partial and full remission at FU1 (OR = .87, 95%CI = .82–93, P < 0.001). Increased depressive severity at baseline reduced the odds for partial remission at FU1 (OR = .95, 95%CI = .91–.98, P < 0.01). Interactions analyses demonstrated that the adverse effects of OCD severity on partial remission disappeared at the presence of secure attachment (IOR = 1.11, 95%CI = 1.05–1.24, P < 0.05) and high social support (IOR = .88, 95%CI = .78–98, P < 0.05).

The contribution of clinical severity is critical for understanding the prognosis of OCD. The interpersonal context of OCD patients may mitigate the unfavorable effect of severity on outcome, thus should be addressed in treatment.

The authors have not supplied their declaration of competing interest.

There is an urgent need for an effective psychological treatment for patients with obsessive compulsive disorder (OCD) with poor insight, since this disorder is associated with severe suffering and a low quality of life. The inference based approach (IBA), a new psychotherapy for OCD specifically targets insight in OCD. In a randomized controlled multicentre trial, the effectiveness of IBA was compared to the effectiveness of CBT for treating patients with OCD with poor insight. In this study, 24 sessions of IBA were tested versus 24 sessions of CBT. Ninety patients with a main diagnosis of OCD with poor insight according to the DSM-IV criteria participated in the study. The primary outcome was reduction of the obsessive-compulsive symptoms.

In both conditions, a significant OCD symptom reduction was reached, but no condition effects were established. Post hoc, in a small subgroup of patients with the worst insight (n = 23), it was found that the patients treated with the IBA reached a significantly higher OCD symptom reduction than patients treated with CBT [estimated marginal mean = –7.77, t(219.45) = –2.4, P = 0.017]. Of patients treated with IBA, 41.9% were responder and 20.9% completely recovered. Of the patients treated with CBT, 42.6% were responder and 12.8% recovered.

Patients with OCD with poor insight improve significantly after psychological treatment. The results of this study suggest that both CBT and the IBA are effective treatments for OCD with poor insight. The IBA might be more promising than CBT for patients with more extreme poor insight.

The authors have not supplied their declaration of competing interest.

There is substantial evidence that even minor perturbation of thyroid function plays a significant role in clinical course and treatment outcome in depressive disorder; however the same is not yet clear in bipolar disorders.

To study the relationship between pretreatment thyroid profile and response to treatment with lithium along with other predictors of response to treatment with lithium in cases of bipolar mood disorder.

This study was conducted in the indoor facilities of a regional Institute of Mental Health, Tezpur, India in the year of 2012. Forty-five consecutive indoor patients diagnosed with bipolar mood disorder using DSM-IV-TR criteria were selected. On day 1, blood was collected for thyroid profile and BPRS 24 item scale version 4.0 was applied. They were started on lithium monotherapy and only lorazepam was used on S.O.S basis. On day 30, the BPRS was applied again to check the response to treatment, statistical analysis was done using SPSS version 16.

The mean percentage fall of the BPRS score was 40%, with the maximum fall in the subscale of grandiosity and minimum for depression. Age, illness duration, substance use, family history second or later episodes were negatively correlated with treatment response. Pretreatment T4 level was positively correlated, while pretreatment TSH level was negatively correlated with the treatment response.

Lithium monotherapy proved to be a good agent for first episode of bipolar Mood disorder patients with manic symptoms and pretreatment T4 and TSH level were predictors of treatment response.

The authors have not supplied their declaration of competing interest.

The study gathered information in order to draw useful conclusions to describe bipolar patients and their clinical management. The data collection was conducted as part of RENDiBi epidemiological study.

The statistical analysis of the collected data will be essential to understand the possible changes in drug treatment, through the help offered by a parameter, Polarity Index (PI), the numerical expression of the efficacy profile of a drug, very useful especially in the long-term management.

Administration of a first detection card (demographic data, medical history) and five scales (CGI-BP, Mood Insight Scale, YMRS, HDRS) and a structured interview (MINI). The parameters analyzed were: polarity prevalence, ratios efficiency (IE) (values indicating the effectiveness of treatment compared to manic components and/or depressive), treatment and PI.

The degree of correlation between PI and IETot is positive and statistically significant. The correlation between PI and IEm is statistically significant; the correlation is however not significant between PI and IEd; treatment with antipsychotics alone has increased PI, while the one with mood stabilizers has lesser; treatment with antipsychotics has increased PI in patients with predominantly polarity than those with manic depressive prevailing polarity.

There is a correlation between PI and effectiveness on manic symptoms and it is statistically significant (as already evident in the literature). The PI is numerically higher in the treatment of the subject with manic polarity, in agreement with previous studies that associate to the more effective drugs used for the management of manic recurrences a higher PI.

The authors have not supplied their declaration of competing interest.

Rapid cycling (RC) worsens the course of bipolar disorder (BD) being associated with poor response to pharmacotherapy. Previous results about clinical variables potentially associated with RCBD were discordant or unreplicated.

An early diagnosis should be the goal to properly treat RCBD patients.

To compare clinical variables between RC and non-RC bipolar patients and to identify related risk factors.

A sample of 238 bipolar patients was enrolled from three different community mental health centers. Descriptive analyses were performed on total sample and patients were compared in terms of sociodemographic and clinical variables according to the presence of RC by multivariate analyses of variance (MANOVAs, continuous variables) or χ2 tests (qualitative variables). Binary logistic regression was performed to calculate odds ratios.

Overall, 28 patients (11.8%) had RC. The two groups were not different in terms of age, age at onset, gender distribution, type of family history, type of substance use disorder, history of antidepressant therapy, main antidepressant, psychotic symptoms, comorbid anxiety disorders, suicide attempts, thyroid diseases, diabetes, type of BD, duration of untreated illness, illness duration, duration of antidepressant treatment and GAF scores. In contrast, RC patients had more often a history of obstetric complications (P < 0.05), obesity (P < 0.05) and a trend to hypercholesterolemia (P = 0.08). In addition, RC bipolar patients presented more frequently lifetime MDMA misuse (P < 0.05) than patients without RC.

Obesity and obstetric complications are risk factors for the development of RC in BD. Lifetime MDMA misuse may be more frequent in RC bipolar patients.

The authors have not supplied their declaration of competing interest.

Patients with subthreshold bipolar disorder (sub-BP) experience severe clinical courses and functional impairments, which are comparable to those with bipolar I and II disorders (BP-I and -II). Nevertheless, lifetime prevalence, socioeconomic correlates and diagnostic overlaps of bipolar spectrum disorder (BPS) have not yet been estimated in the general population of South Korean adults.

This study aimed to estimate the lifetime prevalence, correlates and diagnostic comorbidities of BPS using a validated screening instrument in the nationwide general population of South Korea.

A total of 3013 adults among the 2011 Korean Epidemiologic Catchment Area survey (KECA-2011) completed face-to-face interviews using the Korean versions of the composite international diagnostic interview 2.1 and mood disorder questionnaire (K-CIDI and K-MDQ).

The lifetime prevalence of BPS in the South Korean adults was measured to be 4.3% (95% CI 2.6–6.9). Nearly 80% of the subjects with BPS were codiagnosed with other DSM-IV nonpsychotic mental disorders: 35.4% (95% CI 24.2–48.5) for major depression and dysthymic disorder, 35.1% (95% CI 27.7–43.3) for anxiety disorders and 51.9% (95% CI 40.5–63.1) for alcohol and nicotine use disorders. Younger age (18–34 years) was the only sociodemographic predictor of BPS positivity (P = 0.014) and the diagnostic overlap patterns were different between men and women.

Positivity for BPS was estimated to be much greater than the prevalence of DSM-IV BP in South Korea. Most of the respondents with BPS were diagnosed with other major mental disorders and this might be related with mis and/or underdiagnosis of clinically relevant Sub-BP.

The authors have not supplied their declaration of competing interest.

Many hypotheses have been proposed about development of bipolar disorder including inflammatory processes due to the external and endogenous factors. There are strong evidences that immunological dysfunction is present in bipolar disorder. In the pathophysiology of bipolar disorder, there are many data that support the inflammatory hypothesis.

In this study, to clarify the etiology of bipolar disorder, based on the inflammatory process hypothesis, it is aimed to measure and evaluate serum 15d-PGJ2 and PPARγ, anti-inflammatory cytokine IL-4 levels in patients with bipolar disorder.

This study was performed at Ankara Numune Training and Research Hospital. Ninety-five patients are included in the study that were in their mania or remission periods and meet the DSM-V criteria for bipolar disorder. Forty-four healthy volunteers are included in the study as well. Serum IL-4, 15d-PGJ2, PPARγ levels are measured in both groups. Young Mania Scale, Hamilton Depression Scale, demographic data form were given to patient group.

In our study, 15d-PGJ2, PPARγ levels were found statistically significantly lower in patients with bipolar disorder compared to healthy controls.

There are differences in anti-inflammatory prostaglandin levels in patients with bipolar disorder who are in their mania period when compared to healthy controls and patients in their remission period. This does not show any significance according to smoking and gender. This implies that inflammation markers could be a good candidate to determine trait markers, which will provide an insight for preventing patient from mania period or prognosis after the diagnosis of bipolar disorder.

The authors have not supplied their declaration of competing interest.

Impulsivity is a key feature of both bipolar disorder (BD) type I (BDI) and type II (BDII).

Structural neuroimaging studies help clarifying brain mechanisms underpinning the regulation of impulsivity in BDI and BDII.

To address the question whether grey matter (GM) alterations relate differently with impulsivity in BDI and BDII.

We assessed 54 euthymic outpatients, diagnosed with BDI (n = 28) or BDII (n = 26) according to DSM-IV-TR criteria. They underwent a 3 T magnetic resonance imaging (MRI) investigation. GM brain volumes were analyzed on a voxel-by-voxel basis using Statistical Parametric Mapping 8. The Barratt Impulsiveness Scale (BIS), version 11A, was used to assess trait impulsivity.

BDI and BDII patients present an inverse relationship between impulsivity and GM volume in two cerebral areas: the right cerebellum (right crus I) and the interface between the left angular gyrus and the left inferior parietal cortex (Brodmann Area 39, 7, 40). More specifically, a negative relationship for BPI and a positive relationship for BPII were found in both areas.

Results suggest that the different diagnosis between BDI and BDII could have a significant effect on GM changes according to impulsivity severity and point up the importance of considering the BP subtype distinction in neuroimaging studies on this topic.

The authors have not supplied their declaration of competing interest.

There is growing evidence about neuroinflammation in the aetiopathogenesis of bipolar disorder. Early diagnosis and intervention strategies are thought to be excessively important lately.

To check neuroinflammation levels in early stage bipolar disorder and explore the associations with clinical variables.

We aimed to evaluate inflammation and neurodegeneration findings in early stage bipolar disorder.

Serum interleukin 1-receptor antagonist (IL-1Ra), interleukin 6 (IL-6), tumor necrosis factor-alpha (TNF-α), high sensitive C reactive protein (hs-CRP), S100B and neuron specific enolase (NSE) levels were assessed by enzyme-linked immunosorbent assays in a total of 30 patients with bipolar disorder in the early stage and compared with 30 matched healthy controls. The clinical symptoms were rated using Montgomery Asberg Depression Scale, Young Mania Rating Scale, Positive and Negative Syndrome Scale and Clinical Global Impression Scale.

Among the patients with bipolar disorder, 14 (% 46.6) were in a manic/hypomanic state and 12 (% 40) were in a euthymic state. Serum IL-6 levels were significantly higher (P = 0.018), TNF-α and S100B levels were significantly lower in the early stage group (P < 0.001 and P = 0.03, respectively). After repeated analysis with only drug-naive patients, the results showed no difference. There was a positive and significant correlation between TNF-α levels and CGI, MADRS scores (all P < 0.05); NSE levels and MADRS scores (P < 0.05).

This study supported the association of early stage bipolar disorder with inflammation and neurodegeneration. IL-6 may be a potential biomarker. Thus, early diagnosis and intervention may be crucial to prevent progressive neuroinflammation and neurodegeneration in early stages of disorder.

The authors have not supplied their declaration of competing interest.

Brain-derived neurotrophic factor (BDNF) may be involved in the pathogenesis of bipolar disorder (BD). The functional BDNF Val66Met polymorphism (rs6265) is associated with secretion of BDNF. The current study aimed to explore the correlation between changes of plasma BDNF and cognitive function after 12 week of treatment, considering the influence of the BDNF val66Met polymorphism. The correlation of changes of plasma BDNF with quality of life (QOL) was explored.

First diagnosed patients with BD were recruited. Symptom severity, plasma BDNF levels were examined during weeks 0, 1, 2, 4, 8, and 12. QOL, Wisconsin Card Sorting Test (WCST) and the Conners’ Continuous Performance Test (CPT) were assessed at baseline and endpoint. The genotype of the BDNF Val66Met polymorphism was determined. The change of cognitive function and QOL measures over 12 weeks were reduced by factor analysis. Pearson's correlation was used to investigate the association between change of plasma BDNF levels with cognitive function and QOL.

Five hundred and forty-one BP patients were recruited. Three hundred and fifty-five (65.6%) patients completed the 12-week follow-up. A significant negative correlation was found between changes of plasma BDNF level with factor 1 (WCST) (r = −0.25, P < 0.001). After further stratification according to subtypes of BD and the BDNF genotypes, above significant correlation was found only in those with BP-I and the BDNF Val66Met Val/Met genotype (r = −0.54, P < 0.008).

We conclude that changes in plasma BDNF significantly correlated with changes in WCST in patients with BD; such correlation is moderated by the BDNF Val66Met polymorphism and subtype of BD.

The author has not supplied his declaration of competing interest.

Cortical inhibition (CI) is a neurophysiological outcome of the interaction between GABA inhibitory interneurons and other excitatory neurons. Transcranial magnetic stimulation (TMS) measures of CI deficits have been documented in both symptomatic and remitted bipolar disorder (BD) suggesting it could be a trait marker. The effects of medications and duration of illness may contribute to these findings.

To study CI in BD.

To compare CI across early-course medication-naive BD-mania, remitted first episode mania (FEM) and healthy subjects (HS).

Symptomatic BD subjects having < 3 episodes, currently in mania and medication-naive (n = 27), remitted FEM (n = 27; YMRS < 12 and HDRS < 8) and 45 HS, matched for age and gender, were investigated. Resting motor threshold (RMT) and 1-millivolt motor threshold (MT1) were estimated from the right first dorsal interosseous muscle. Paired-pulse TMS measures of short (SICI; 3ms) and long interval intracortical inhibition (LICI; 100ms) were acquired. Group differences in measures of CI were examined using ANOVA.

Table 1.

Symptomatic mania patients had the highest motor thresholds and the maximum LICI indicating a state of an excessive GABA-B neurotransmitter tone. Remitted mania patients had deficits in SICI indicating reduced GABA-A neurotransmitter tone. Putative changes in GABA-A neurotransmitter system activity with treatment may be investigated in future studies. CI has received less attention in BD as compared to schizophrenia and is a potential avenue for future research in this area.

The authors have not supplied their declaration of competing interest.

A major concern in patients with BIPOLAR Disorder (BD) is the high frequency of suicidality. It is important to determine the subgroup of patients particularly exposed to this risk.

To explore sociodemographic and clinical characteristics of BD patients with a history of suicide attempts and compare them to those who never attempted suicide.

This is a retrospective, cross-sectional, descriptive and comparative study on 100 patients followed in our department and diagnosed with BD type I according to DSM 5. Demographic and clinical data was compared across the groups: suicide+ (S + ) and suicide–(S-).

Overall, 77 patients never attempted suicide (S−), whereas 23 had made at least one suicide attempt (S+). Females represented 61,9% of the S+ group. S+ patients had more relatives with psychiatric illness (82,6% vs. 57.1%) and affective illness (43.5% vs. 33,8%). The mean diagnostic delay was 6.61 years in the S+ group vs. 4.58 in the S− group, with 78,3% of S+ patients first receiving another diagnosis than BD. S+ patients had significantly more depressive episodes and manic episodes with mixed features. Quality of intervals was worse in S+ patients. Anxiety comorbidity was significantly higher in the S+ group (52.2% vs. 13%; P = 0,000). Hyperthymic temperament was significantly associated with the S− group.

Depressive polarity, anxiety comorbidity, as well as diagnostic delay seem to elevate the risk for suicidality in bipolar patients. Suicidality should be closely monitored in patients with these characteristics.

The author has not supplied his declaration of competing interest.

Euthymic patients with bipolar disorder (BD) experience residual symptoms. Interestingly, residual symptoms appear to impact the natural course of BD and represent potential predictors of recurrence and functional impairment.

The study aimed to analyse the relationship between residual depressive symptoms, sleep disturbances and cognitive impairment as determinants of psychosocial functioning in a large sample of euthymic BD patients.

We performed a cross-sectional study of 468 BD outpatients in clinical remission for at least 6 months. Bipolar Depression Rating Scale (BDRS), Pittsburgh Sleep Quality Index (PSQI) scale, Visual Analogic Scales (VAS) evaluated cognitive impairment and functioning assessment short test were used to assess residual symptomatology and functioning of patients. We evaluated functioning with. Structural equation modelling (SEM) was used to describe the relationships among the residual depressive symptoms, sleep disturbances, perceived cognitive performance and functioning.

SEM showed good fit. This model revealed that residual depressive symptoms (path coefficient = 0.37) and perceived cognitive performance (path coefficient = 0.27) were the most important features significantly related to psychosocial functioning. Sleep disturbances were indirectly associated with functioning via residual depressive symptoms and perceived cognitive performance (path coefficient = 0.23).

This study contributes to a better understanding of the determinants of psychosocial functioning during the interepisodic periods of BD patients. These findings should have implications for the improvement of functioning of BD patients in a personalized approach to treatment.

COI: Dr. Samalin reports personal fees and nonfinancial support from Astra-Zeneca, Bristol Myers Squibb, Janssen, Lundbeck, and Otsuka.

The authors L. Boyer, A. Murru, I. Pacchiarotti, M. Reinares, C.M. Bonnin, C. Torrent, V. Norma, P. Corinna, I. de Chazeron, M. Boucekine, P.A. Geoffroy, F. Bellivier, P.M. Llorca, E. Vieta have have not supplied their declaration of competing interest.

Suicidality is a major health concern with as yet unclear neurobiology.

To identify emotional correlates of suicidality in bipolar disorder I (BD-I).

To detect subcortical structural morphology changes associated with suicide attempts.

We enrolled 30 patients with BD-I of which 15 had history of suicide attempts, and 15 healthy controls (HCs) with no such history. Groups were defined according to suicide attempt history and psychopathology. Subcortical gray matter volumes were obtained from 3 T structural MRI scans using FreeSurfer. Intergroup differences were investigated through ANOVAs followed by post hoc Fischer's least significant difference.

HCs had larger left hippocampal and left accumbens volumes than both BD-I attempters and nonattempters. BD-I attempters had smaller left hippocampi and larger left amygdala than both nonattempters and HCs.

Differences were observed in emotional processing mediating neural circuitries, with BD-I attempters showing opposite patterns to nonattempters between amygdala and hippocampus.

The authors have not supplied their declaration of competing interest.

Relatives’ coping strategies – an essential element for the recovery of patients with severe mental disorders – are grouped in problem-oriented and emotion-focused. The former include practical strategies to deal with the stressful situation and are associated with a better long-term outcome of patients and relatives; the latter are psychologically driven and are associated with a worse outcome. It has been reported that psychoeducational family intervention (PFI) can improve problem-oriented coping strategies, while few data are available on relatives of patients with bipolar disorder.

To assess the impact of the PFI on promotion of problem-oriented coping strategies adopted by relatives of patients with bipolar I disorder.

This study was conducted in 11 Italian mental health centers. Patients and their relatives were allocated to the experimental group receiving PFI or to the control group (waiting list). Before starting the intervention and at the end of the PFI, coping strategies were assessed using the family coping questionnaire.

Of the 139 recruited families, 72 families were allocated to the experimental group and 67 to the control group. Relatives from the experimental group reported a significant improvement in problem-oriented coping strategies, such as positive communication (P < .01) and searching for information (P < .05). On the other hand, a reduction in collusion (P < .0001), avoidance (P < .01) and resignation (P < .001) were found at the end of the intervention.

PFI is effective in promoting the coping strategies in relatives of patients with bipolar I disorder and it should be given routinely in mental health centers.

The authors have not supplied their declaration of competing interest.