Childhood maltreatment and clinical outcomes of bipolar disorder

Authors

  • L. Daruy-Filho,

    1. Developmental Cognitive Neuroscience Research Group, Postgraduation Program in Psychology – Human Cognition, Pontifical Catholic University, Porto Alegre
    Search for more papers by this author
  • E. Brietzke,

    1. Developmental Cognitive Neuroscience Research Group, Postgraduation Program in Psychology – Human Cognition, Pontifical Catholic University, Porto Alegre
    2. Bipolar Disorder Research Program, Psychiatry Institute, University of São Paulo, São Paulo, Brazil
    Search for more papers by this author
  • B. Lafer,

    1. Bipolar Disorder Research Program, Psychiatry Institute, University of São Paulo, São Paulo, Brazil
    Search for more papers by this author
  • R. Grassi-Oliveira

    1. Developmental Cognitive Neuroscience Research Group, Postgraduation Program in Psychology – Human Cognition, Pontifical Catholic University, Porto Alegre
    Search for more papers by this author

Rodrigo Grassi-Oliveira, Avenida Ipiranga, 6681, prédio 11, sala 936, Partenon, Porto Alegre, RS 90619-900, Brazil.
E-mail: rodrigo.grassi@pucrs.br

Abstract

Daruy-Filho L, Brietzke E, Lafer B, Grassi-Oliveira R. Childhood maltreatment and clinical outcomes of bipolar disorder.

Objective:  Adverse life events, especially early trauma, play a major role in the course and expression of bipolar disorder (BD). The aim of this article is to present a systematic review about the impact of childhood trauma on the clinical course of BD.

Method:  A computer-aided search was performed in Medline, ISI database, EMBASE, PsychInfo, Centre for Reviews and Dissemination, and Databases of Thomson Reuters at April 2011, supplemented by works identified from the reference lists of the first selected papers. Two investigators systematically and independently examined all articles, selecting those according inclusion and exclusion criteria.

Results:  Four hundred fifteen articles were identified, of which 19 remained in the review after exclusion criteria were applied. In general, childhood maltreatment predicted worsening clinical course of BD. After assessing the quality of the data and of the measurements, childhood maltreatment can be strongly associated to early onset of disorder, suicidality, and substance abuse disorder in patients with BD.

Conclusion:  Data suggest that childhood abuse and neglect are risk factors associated with worsening clinical course of BD. The conclusions should be interpreted with caution because all the studies included are cross-sectional and the majority are showing inconsistencies regarding childhood trauma as independent variable and how it is assessed.

Summations

  •  Childhood abuse and neglect are related to more severe bipolar disorder (BD). There is a body of evidence relating history of childhood maltreatment to rapid illness progression.
  •  Childhood physical abuse seems to be the strongest predictor of unfavorable BD outcomes, despite there is very few research looking for the impact of emotional abuse and neglect.
  •  It is very important to know which predictors play a role in accelerating staging process of BD to determine which interventions could help to prevent illness progression, including suicide prevention and child abuse prevention programs.

Considerations

  •  Despite the important role of early life stress in BD, there are few cross-sectional studies investigating its impact in clinical outcomes and no longitudinal study, which reduces the quality of evidence.
  •  The majority of studies considered different types of trauma and abuse as one phenomenon and/or did not distinguish the nature of maltreatment

Introduction

Since Kraepelin, the importance of environmental stressors in the individual variations of the clinical course of bipolar disorder (BD) has been considered (1). Among those stressors, childhood trauma has emerged as one of the most important factors associated with negative outcomes of psychiatric disorders (2), including BD (3).

Childhood abuse and neglect have been reported by 51% of patients with BD (4). Specifically, emotional abuse has been reported by 37% of bipolar patients, 24% reported physical abuse, 24% emotional neglect, 21% sexual abuse, and 12% physical neglect. In addition, one-third of those patients presented a combination of different types of trauma (4).

Early life stress has been suggested to mediate vulnerability to affective disorders (5), despite the exact mechanism of this association is not completely understood. In line with that, a variety of studies have been relating childhood maltreatment with disrupted neurodevelopment (6–9). Traumatic events during childhood are associated with long-term structural and functional brain alterations (10, 11) especially involving dysfunctions in prefrontal cortex, amygdala, and hippocampus (12). Theses changes have been taking place in the pathophysiology of BD and have been consistently related with its severity (13).

Although the association between BD and childhood maltreatment can be found in some studies, most of the times, it would be considered as secondary outcome and few of them investigated the direct influence of childhood maltreatment in the clinical outcomes (11, 14). One of the most important problems in the literature is that, despite the relevance of the topic, there is only one systematic review available (15). Today, those who decide to look for the impact of childhood abuse and neglect in BD will have to face a puzzling task.

Aims of the study

Considering the importance of childhood trauma in bipolar disorder (BD) and the very few studies about this topic, this article presents a rare systematic review investigating the association between childhood abuse and neglect and BD clinical course.

Material and methods

A search for English-written articles, published in the last 12 years, selection was performed in Medline, ISI database, EMBASE, PsychInfo, Centre for Reviews and Dissemination, and Databases of Thomson Reuters at April 2011. The key terms used were child maltreatment, childhood trauma, early life stress, psychological stress, emotional stress, and child abuse and neglect. They were cross-referenced separately with the terms bipolar disorder and manic-depressive disorder. The search criteria was the presence of key terms in any field of article.

The articles resulting from the search were systematically and independently examined by two investigators according to inclusion criteria and later, exclusion criteria as well. The following exclusion criteria were applied: i) reviews and theoretical articles, ii) articles with samples composed by BD and other mood disorder without discriminate analysis, iii) no clinical outcome as dependent variable, iv) articles without childhood trauma as independent variable, and v) republished data.

The authors also searched in the reference list of each included paper for additional studies not found in the first steps.

Quality of studies assessment

The methodological weight of studies (0–4) was measured using a quality score described by Fisher and Hosang (15), attributing one point to each one of the following characteristics: minimum of 100 patients with BD in sample; standardized diagnostic measure; standardized childhood trauma measure; and current mood control by standardized instruments.

Results

The search identified 415 papers. This list was screened by hand, and exclusion criteria were applied. The flow chart is shown in Fig. 1. Six studies were selected at references search. Final search resulted in 20 studies, analyzing the effect of childhood trauma in the clinical course of BD. Republished data about the results of the Stanley Foundation Bipolar Treatment Outcome Network Study (16) were analyzed and only the article by Leverich et al. (23) was included. All 19 final studies were retrospective studies. A summary of these studies is shown in Table 1.

Figure 1.

 Flow chart of systematic review.

Table 1.   Summary of selected studies
StudySampleDiagnostic measureTrauma assessmentResultsQuality score (15)
  1. BD, bipolar disorder; K-SADS, St. Louis Kiddie Schedule for Affective Disorders and Schizophrenia; PTSD, posttraumatic stress disorder; ADHD, attention deficit hyperactivity disorder; OCD, obsessive–compulsive disorder; SCID, Structured Clinical Interview; CTQ, Childhood Trauma Questionnaire; SADS-L, Shcedule for Affective Disorders and Schizophrenia—Lifetime; BDI, Beck Depression Inventory; HAM-D, Hamilton Scale for Depression; YMRS, Young Mania Rating Scale.

Alvarez et al., 2011 (47)102 adult outpatients, 40 with BD diagnosis (20 BD patients with history of any kind of abuse)Not informedTraumatic Life Events QuestionnaireHistory of psychological abuse and domestic violence witness were associated with number of admittances in bipolar patients1
Romero et al., 2009 (25)446 youth (7–17 years old) outpatients and inpatients with BD (92 with history of abuse)Not informedK-SADS interviewHistory of physical abuse was associated with longer duration of illness, PTSD, and psychosis. History of sexual abuse was associated with PTSD. History of any type of abuse was associated with longer duration of the illness, suicide attempt, greater rates of PTSD, conduct disorder, and substance use disorder. Childhood abuse was not associated with suicidal ideation, self injury behavior, hospitalization, BD onset age, and rates of comorbidity with ADHD, OCD, and anxiety disorders2
Savitz et al., 2008 (48)49 bipolar adults and 61 controlsSCID (DSM-IV)CTQBipolar participants with history of psychosis had higher levels of self-reported childhood sexual abuse than the participants without history of psychosis and unaffected individuals2
Carballo et al., 2008 (31)168 adults outpatients with BDSCID (DSM-III-R)Clinical interviewHistory of childhood abuse, in addition with family history of suicidal behavior, was associated with suicidality, impulsivity, history of aggression, early onset of bipolar illness and first hospitalization, and comorbidity with substance use disorder and borderline personality disorder. There was no association with hostility and history of stressful life events2
Garno et al., 2008 (30)100 adults with BDSCID (DSM-IV)CTQChildhood emotional abuse, emotional neglect, and physical abuse were associated with trait aggression [Brown–Goodwin Aggression scale (BGA)]. Sexual abuse and physical neglect was not associated with BGA scores3
Goldstein et al., 2008 (27)249 adolescents with BDK-SADS (DSM-IV)Clinical interviewSexual and physical abuses were associated with presence of substance use disorder2
Maguire et al., 2008 (17)60 outpatients with BDClinical diagnosisTrauma History Questionnaire and CTQTrauma in childhood was related to trauma in adulthood, but not to PTSD diagnosis. Quality of life, inter-episode depressive symptoms, number of admissions, and duration of admission were not related to childhood trauma alone, just when childhood and adulthood trauma was associated1
McIntyre et al., 2008 (14)381 adult outpatients with BD (68 with history of abuse)Chart review for DSM-IV-TR diagnostic criteriaChart review for childhood abuse (sexual or physical)History of childhood abuse was associated with lifetime suicidality (ideation or attempts), comorbid psychiatric disorders, and concomitant medication. Childhood abuse was not associated with age of illness onset, cycling, dysphoric mania/hypomania, recurrent depressive or manic episodes, hospitalizations, and substance abuse1
Grandin et al., 2007 (22)155 bipolar spectrum adultsSADS-LChildren Life Events ScaleHistory of childhood maltreatment (prior to or after the age of onset) was associated to higher BDI scores. Negative emotional events predicted the age of first episode onset for the patients with BD4
Dienes et al. 2006 (18)64 adult outpatients with BDSCID (DSM-IV)Clinical interviewEarly sexual adversity and neglect were associated with earlier age of BD onset1
Rucklidge, 2006 (19)24 individuals with BD, aged between 13 and 17 years old (13 with history of childhood trauma)K-SADSK-SADS interviewHistory of trauma was not associated with psychosocial functioning: anger regulation (Novaco Anger Scale), hopelessness (Hopelessness Scale for Children), self-esteem (Rosenberg Self-Esteem Scale), and locus of control (Norwicki–Strickland Internal–External Control Scale)2
Brown et al., 2005 (24)330 adult inpatients with BDSCID (DSM-IV)Clinical interviewHistory of any childhood abuse was associated with larger number of involuntary hospitalizations, disability pension, PTSD, panic disorder, and alcohol use disorder. Participants with only sexual abuse report had fewer side effects from psychiatric medications and alcohol use disorder. History of only physical abuse was associated with rapid cycling pattern, PTSD, and alcohol use disorder. Participants with sexual and physical abuse had lower quality of life, higher comorbidity with PTSD, and alcohol use disorder2
Garno et al., 2005 (4)99 adult patients with BDSCID (DSM-IV)CTQHistory of severe childhood abuse was associated with more depressive episodes, HAM-D, and YMRS rates, and an earlier age of illness onset. Lifetime suicide attempts were associated just with sexual abuse. Abuse was not associated with hospitalization, but number of manic episodes3
Goldberg & Garno, 2005 (29)100 adult outpatients with BD (51 with history of severe childhood abuseSCID (DSM-IV)CTQHistory of childhood abuse was associated with prevalence of PTSD3
Marchand et al., 2005 (32)66 young outpatients with BD (35 with history of abuse or neglect)Chart review for DSM-IV-TR diagnostic criteriaChart review for childhood eventsHistory of sexual abuse and neglect were associated with worst final response to BD treatment. History of physical abuse was associated with more hospitalization, delay on diagnosis, and worst final response to BD treatment. History of foster care/adoption was associated with more residential treatment. Maltreatment was not associated with duration of BD treatment0
Neria et al., 2005 (20)109 adult inpatients with BD (30 with history of childhood assaultive trauma)SCID (DSM-III-R)Clinical interviewPatients with history of childhood trauma had higher scores of General Health Questionnaire (GHQ) and less happiness (Happiness Scale)2
Hammersley et al., 2003 (21)96 adult outpatients with BDSCIDChild Maltreatment History Self-Report and direct referencesAn association was found between reports of any trauma and the presence of auditory hallucinations2
Leverich et al., 2003 (26)648 adult patients with BDSCID-P (DSM-IV)Clinical interviewEarly physical and sexual abuses were associated with suicide attempts with an addictive effect of the occurrence of both physical and sexual abuse2
Leverich et al., 2002 (23)631 adult outpatients with BDSCID-P (DSM-IV)Clinical interviewHistory of either physical or sexual abuse was associated with an early age of illness onset, faster cycling frequencies, increased incidence of suicide attempts, and severity of mania. The abused participants had a longer duration of time ill until the entry to treatment and an increased mean number of lifetime Axis I disorders. These patients had a higher incidence of negative psychosocial stressors occurring before the onset of the first and recent affective episodes2

Variables analysis

Different variables were considered as BD clinical outcomes and they are listed in Table 2. Most articles used the concept of childhood abuse (physical, sexual, or emotional) and neglect (emotional or physical) as independent variable, but others used trauma (17–21) or maltreatment (22) to define the same events.

Table 2.   Summary of bipolar disorder course measures
  1. This table shows all the clinical variables tested by the 19 studies selected to analysis.

Aggression and self-injury behavior
Axis I and II comorbidity
Current mood status
Cycling
Delay of diagnosis
Disease onset
Duration of disease
Hospitalization and other intensive treatment history
Inter-episode symptoms
Life quality
Number and severity of mood episodes
Psychosis
Psychosocial functioning
Suicidality
Treatment response

Childhood physical abuse.  Physical abuse seems to be the strongest predictor of unfavorable BD outcomes. It is associated with early onset of the disorder (23), and delay in diagnosis (17) and first treatment (23). Patients reporting physical abuse tend to present a rapid-cycling pattern (23) with an odds ratio (OR) of 1.96 (24).

In addition, mood episodes are more likely to present psychotic symptoms (OR = 2.3) (25) and to be associated with suicide attempts (23, 26). Some authors found that those patients present more severe manic symptoms and course with more hospitalizations (23).

Regarding psychiatric comorbidities, childhood physical abuse was associated with high risk of posttraumatic stress disorder (PTSD) in patients with BD (OR = 2.6–10.4) (27) and substance abuse was also described (24, 27).

Childhood sexual abuse.  Presence of childhood sexual abuse was related to a very similar impact on BD clinical outcome of physical abuse, characterized by an early onset (18, 23), delay on treatment, and a rapid-cycling pattern (23). Suicidality also appeared as a frequent outcome associated with such variable (OR = 2.27–3.32) (4, 26). The episodes were marked by higher frequency of psychosis (28) and a increased severity of manic symptoms (23).

High rates of psychiatric comorbidities were associated with childhood sexual abuse like PTSD (OR = 4.9–7.8) (25, 29) and substance abuse (23, 24, 27).

Childhood emotional abuse and neglect.  Few studies analyzed separately emotional abuse and/or neglect. Garno et al. (30) found them as predictors of trait aggression. Childhood neglect was also associated with an early onset of BD (18).

Nonspecific childhood abuse.  The majority of studies considered different types of trauma and abuse as one phenomenon and/or did not distinguish the nature of maltreatment. Despite Childhood Trauma Questionnaire (CTQ) could discriminate childhood sexual, physical, or emotional abuse and physical or emotional neglect, some authors have been using them in accordance with presence or absence of ‘severe childhood abuse’ regardless of the nature of maltreatment (4). These authors found that severe childhood abuse was significantly associated with early onset of BD, with more depressive episodes during life and with increased manic/depressive symptoms severity.

Carballo et al. (31) analyzed two risk factors as predictors to BD clinical outcomes: childhood abuse (physical or sexual) and family history of suicidal behavior. Authors did not discriminate the nature of childhood abuse in their analyses, but both variables were associated with early onset of BD, early hospitalization, increased suicidality, impulsivity, aggression, and comorbidities.

Another study (24) found that history of childhood sexual abuse or physical abuse was associated with higher frequency of comorbidities such as PTSD (OR = 3.39), panic disorder (OR = 1.01), and alcohol abuse (OR = 2.28) and hospitalizations (OR = 2.37). In addition, combined forms of abuse (physical and sexual abuse) were strong predictors of lower quality of life (OR = 0.97) in patients with BD.

Childhood physical or sexual abuse was found to be positively associated with suicidality (14, 25), polypharmacy (14), worst response to treatment (32), hallucinations (21), depression severity (22), conduct disorder (25), substance abuse (25), and trauma-related disorders (17, 29).

Quality of studies

Only one of the 19 studies (22) included current mood status as a covariate in their analysis, to minimize the potential bias of symptoms in childhood maltreatment memories.

Fifty-eight per cent of studies utilized the Structured Clinical Interview for Diagnosis for BD diagnosis. Three studies utilized non-standardized measures and two not declared the method of BD diagnosis.

For trauma assessment, 47% of studies utilized non-standardized interviews. CTQ was the instrument more frequently used by five studies.

Pondered syntheses of the results are shown in Table 3.

Table 3.   Pondered results of selected studies
BD outcomesPhysical abuseSexual abuseEmotional abuse/neglectNonspecific early abuse/trauma
  1. PTSD, posttraumatic stress disorder.

BD course
 Early onset of disorder•• (23)•• (23) ••••••••• (4, 22, 31)
 Rapid cycling•••• (23, 24)•• (23)  
 Early hospitalization   •• (31)
 More hospitalizations   •• (24)
Clinical features
 Psychosis•• (25)•• (48) •• (21)
 Suicidality•••• (23, 26)••••••• (4, 23, 26) •••• (25, 31)
 Impulsivity   •• (31)
 Aggressivity••• (30) ••• (30)•• (31)
 More depressive episodes   ••• (4)
 Manic symptoms severity•• (23)•• (23) ••• (4)
 Depressive symptoms severity   ••• (4)
Comorbidity
 PTSD•••• (24, 25)•• (25) •••• (24, 25, 29)
 Conduct disorder   •• (25)
 Substance abuse disorder•••• (24, 27)•••• (24, 27) •••••• (24, 25, 31)
 Borderline personality disorder   •• (31)
 Panic disorder   •• (24)
 Nonspecific comorbid disorder•• (23)•• (23)  

Discussion

The results of selected studies showed that childhood abuse and neglect are related to early onset of the disorder, suicide behaviors, and substance abuse disorder among patients with BD. The course of BD in patients reporting childhood trauma includes higher number of mood episodes, more severe symptoms, higher number of psychiatric comorbidities, along with aggression and impulsivity, and worsening response to treatment.

Some aspects are especially important in the evaluation of these results. First, the childhood maltreatment assessment: only half of studies (n = 9) used direct clinical interview for assessment of child abuse and neglect. The CTQ (33) was used by five studies (27%), and K-SADS Interview (34) by two. Few other instruments were Trauma History Questionnaire (35), Child Maltreatment History Self-Report (36), and Children’s Life Events Scale (37). For the majority of studies, the concept of childhood maltreatment was based on patterns measured by those instruments, picturing heterogeneity of definition regarding early trauma independent variables. In addition, studies that used clinical interview or chart review for assessment of history of childhood trauma and/or abuse are limited by difficulty to assess sexual abuse (38) and the fragility of constructs describing what would be considered abusiveness. Second, child maltreatment is a robust risk factor for substance abuse (39, 40) and it is also associated to impaired emotional self-regulation (41); despite that, none of the studies in this systematic review explicitly investigated the impact of drug abuse as a mediator of the impact of childhood maltreatment in clinical outcomes of BD. This is a very important issue considering that distal and proximal stressors, recurrence of episodes, and drug abuse can increase responsivity (sensitize) to themselves and cross-sensitize to the other, thus driving illness progression (42). Future studies should necessarily investigate the role of substance abuse, especially cocaine, in interaction with clinical variables and childhood trauma in BD. Third, the current review focused only the impact of childhood maltreatment in clinical outcomes; however, there are plenty of evidence connecting it to neuroimmunoendocrine dysregulation and cognitive impairment in BD (11). Therefore, early life stress could impact clinical outcomes through neurobiological and cognitive alterations.

Considering such findings, it is plausible to suggest that childhood adversity may be related to increased vulnerability to BD symptoms during development. Keeping this in mind is crucial, as high-risk longitudinal studies have highlighted the heterogeneity of mood disorders during development and provide evidence of the evolution of psychopathology across the time (43, 44). Because the studies included in this review evaluated childhood trauma retrospectively, conclusions about cause and effect are not possible and issues about memory bias should be considered. Because of that, an alternative explanation for the association between childhood trauma and BD must also be taken into consideration. This second hypothesis is that children with hereditary or environmental vulnerability to BD could also be more abused and neglected.

Another implication related to childhood trauma in BD is the highly heterogeneous symptomatology, which could explain delay in diagnosis. As presented earlier, BD clinical outcomes associated with childhood maltreatment include more severe impairments in emotional self-regulation. Suicide attempts, aggressiveness, impulsivity, and mood dysregulation are independently associated with history of childhood trauma (45, 46); therefore, the emergence of BD could be precipitated by these behavioral and emotional patterns.

Usually, clinical course, environmental factors, family history, and treatment response are not considered in the current diagnostic classification systems (44). Assuming that childhood abuse and neglect could accelerate the evolution of BD or may impact its pathophysiology, the incorporation of early life trauma in diagnostic and treatment decisions may assist in BD treatment planning and prognosis, emphasizing the imperative need of early intervention.

In conclusion, results of this review corroborate the importance of systematically investigating the history of childhood abuse and neglect in BD. Particularly, it is very important to know which predictors play a role in accelerating BD course to determine which interventions could help to prevent illness progression, including suicide prevention and child abuse prevention programs. Nonetheless, further follow-up studies are needed, as well as well designed cross-sectional ones.

Acknowledgements

The authors thank Bruno Kluwe Schiavon and Cristiane da Silva Fabres for their support during abstract selection.

Declaration of interest

None.

Ancillary