Abstract
The objective is to test the efficacy of cognitive-narrative therapy in the treatment of depression, post-traumatic stress disorder (PTSD), complex posttraumatic stress disorder (CPTSD) and borderline symptoms on a sample of women who suffered from intimate partner violence (IPV). Trial design is a longitudinal randomized controlled trial with a sample of 19 battered women allocated in two groups, a control group and treatment group, assessed twice at baseline before intervention, and at follow-up. The outcome measures were the Patient Health Questionnaire, International Trauma Questionnaire, PTSD and CPTSD Diagnostic Interview Schedule for International Classification of Diseases (ICD)-11, Conjugal Violence Exposure Scale (CVES), Life Events Checklist and Intervention Program Satisfaction Assessment Instrument. The treatment group received a four-session cognitive-narrative manualized intervention. There were no statistically significant differences between groups at baseline and follow-up, however, positive effect sizes ranging between 0.04 and 0.43 were found in depression, PTSD, and borderline, as well in some CPTSD dimensions when analyzing baseline-follow-up deltas between groups. There was also a negative effect size of –0.28 in the CPTSD total. This intervention is effective in the treatment of depression, PTSD and borderline and is an important tool in the treatment of these disorders.
Introduction
Intimate partner violence (IPV) is a violation of human rights and includes physical, emotional, or sexual aggression, stalking or economic aggression directed at a partner (Arroyo et al., 2016; Orang et al., 2018). This impacts the life of millions of women around the world, and between 10% and 69% of women have reported being physically assaulted by an intimate male partner at some point in their lives (World Health Organization [WHO], 1999).
Survivors can suffer from a wide range of physical, social, economic, and mental health issues. IPV is associated with PTSD, alcohol and drug abuse, anxiety, and depression (Coker et al., 2002; Dekel et al., 2020; Pico-Alfonso et al., 2006; Ruiz-Pérez & Plazaola-Castaño, 2005; Sullivan et al., 2016). In the United States, according to the National Intimate Partner and Sexual Violence Survey: 2015 Data Brief—Update Release, one in four women and one in 10 men have experienced contact sexual violence, physical violence, and/or stalking by an intimate partner and reported an IPV-related impact during their lifetime (Centers for Disease Control and Prevention, 2015). According to the Associação Portuguesa de Apoio à Vítima (2017, 2018), there were 87,730 cases of domestic violence in Portugal between 2013 and 2017, which translated into 36,528 cases of victim support; in 2018 this association also provided aid for 6928 victims of IPV, mostly women (86.3%) with a mean age of 43, 7.4% of them also had children who were directly affected.
Complex Posttraumatic Stress Disorder (CPTSD) is a pathology that was recently included in International Classification of Diseases (ICD)-11 (WHO, 2018) and that, according to the manual, can develop after exposure to an event or series of events of an extremely threatening or horrific nature, most commonly prolonged or repetitive events from which escape is difficult or impossible (e.g., torture, enslavement, genocide campaigns, prolonged domestic violence, repeated childhood sexual, or physical abuse). This disorder is characterized by the basic symptoms of PTSD, which means that for CPTSD to be diagnosed, the diagnostic criteria for PTSD must have been filled at some point during the course of the disorder. In addition to these there are factors as affective dysregulation, negative self-concept, and disturbed relationships (Hyland et al., 2017; Rocha, Rodrigues, et al., 2019), which, according to ICD-11 (WHO, 2018), are severe and persistent: (a) problems in affective regulation; (b) beliefs about oneself as diminished, defeated or worthless, accompanied by feelings of shame, guilt or failure related to the traumatic event; (c) difficulties in maintaining relationships and in feeling close to others. CPTSD symptoms also cause impairment in personal, family, social, educational, occupational, or other important areas of functioning. Many survivors lose economic security, and are also forced to leave their homes to escape the violence. They often go to domestic violence shelters so that they can be safe and recover their stability. Even being in a safe environment, these abrupt life changes and the cycles of resource loss exacerbate the risks and severity of PTSD and other mental health problems (Arroyo et al., 2016).
Mental health issues perpetuate the consequences of IPV on several levels and place a serious burden on an individual’s functioning. All direct mental health effects of IPV are relevant targets for interventions, however, treating the damage inflicted by IPV is a challenging task and traditional forms of psychotherapy may not be the most appropriate to treat these issues in such a specific population.
IPV victims, especially those living in a shelter home and going through a life-changing and dependence-breaking process have a complex set of needs and problems, such as safety, conflict management, mental illness, parenting skills, economic challenges, difficulties finding jobs, and juridical needs of many sorts. Brief therapies are thus often indicated for IPV survivors in shelter situations, as they might not stay for a long time, and even for those who find other solutions or are in uncertain living conditions (Arroyo et al., 2016; Orang et al., 2018).
Meta analyses have found that several psychotherapies have positive effects in areas such as PTSD, depression, anxiety, and substance/alcohol abuse. These therapies include cognitive-behavioral therapy, interpersonal psychotherapy, relapse prevention and relationship safety (RPRS) groups, dialectical behavior therapy, brief counseling and others; the number of sessions required for therapy ranged from a minimum of four to a maximum of 12 weeks of treatment (Arroyo et al., 2016).
Narrative exposure therapy (NET), which is an evidence-based short-term treatment that aims to treat and diminish the repercussions of exposure to accumulative, repetitive, and multiple types of trauma, mostly among victims of war and torture (Schauer et al., 2011), has also been found effective in the treatment of women who have suffered IPV. Orang and colleagues (2018) found effect differences between the treatment as usual (TAU) group and the treatment group (TG) in a sample of Iranian women exposed to IPV, with between-group effect sizes ranging from 0.34 and 0.43 in 3 and 6 months follow-ups (FUs) for PTSD, depression, and perceived stress.
Cognitive-narrative therapy, developed by O. Gonçalves and Machado (1999), uses a constructivist approach focused on deconstructing personal episodic narratives and constructing new ones with multiple meanings and coherences. This therapy perspective emphasizes the power of the word and sees the individual as the builder of their experiences through storytelling (O. Gonçalves, 2002b). This therapy has been tested in randomized trials measuring its efficacy in populations suffering from grief, and those who suffered a termination of pregnancy after a positive prenatal diagnosis, and it demonstrated positive effects in preventing and treating complicated grief, perinatal grief, depression, and anxiety symptoms (Andrade et al., 2017; Azevedo et al., 2017; Barbosa et al., 2013; Rocha et al., 2018).
Following these positive results, the objective of this randomized trial is to test the efficacy of cognitive-narrative therapy in the treatment of depression, PTSD, CPTSD, and borderline symptoms on a sample of women who suffered from IPV.
Method
Trial Design
A cognitive-narrative intervention program for IPV, with a total of four sessions of 60 min each, was applied.
The selection criteria were: (a) to be more than 18 years old; (b) to have been a victim of IPV.
This trial was divided into four parts: (a) randomization of the participants into a control group (CG) and a TG; (b) application of a PTSD, CPTSD, BPD, and depression symptom evaluation protocol; (c) implementation of the cognitive-narrative intervention; (d) FU evaluation using the same protocol as in Part One.
This research has Ethics Committee approval: reference 31/CE-IUCS/2019.
Intervention: Manualised cognitive-narrative program for IPV victims
This manual is based on the programs developed by O. Gonçalves et al. (1997), O. Gonçalves (2002a), and Barbosa et al. (2013), and was adapted to better address the type of trauma suffered by IPV victims.
The intervention is a brief therapy divided into four weekly sessions of 60 min, which seeks to construct new meanings and coherences for the traumatic experiences the victims went through.
The objective of the first session, “Recalling,” is to evoke an episode related to the IPV and to clarify its meaning; to achieve this, the patient is first invited to describe the history of the relationship with the aggressor and to make a timeline of the violence, and then to describe an IPV-related episode with as many details as possible.
In the second session, “Emotional and Cognitive Subjectivation,” the objective is to describe the episode and to structure the patient’s experience with a sense of authorship, coherence and the diversity of emotional and cognitive contents which were not valued in the previous session. To achieve this, the patient is asked to narrate the emotions and cognitions/thoughts associated with the episode described in “Recalling.” Initially emotions are activated, and then in a second phase the cognitive component is explored. Associations are made between emotions and thoughts.
The objective of the third session, “Metaphorisation,” is to explore different meanings of the chosen episode and to find a unifying metaphor/title (e.g.,: “Ripping the sheet”—breaking the relationship with the aggressor, which symbolizes the beginning of a new life). The patient is asked to build a metaphor for the episode and the emotions/cognitions explored in the previous sessions, and to explain its meaning. It is important for the therapist to use techniques such as paraphrasing, silences, and meaning reflections.
Finally, the objective of the fourth session, “Projection,” is to promote the construction of, and experimentation with, other possible ways of organizing the episode. To achieve this, the patient is asked to find an alternative and more positive metaphor for the episode. After the alternative metaphor is found, its meaning, its cognitive and emotional differences, and its representation of a more adaptive functioning are discussed. Future projects are also reflected, as the victims of IPV are trying to start a new life chapter.
Therapists
The therapists were trained on the manualised intervention (through theoretical sessions and roleplay), monitored and supervised by the provider of the training and other professionals. The therapists also evaluated themselves using a self-evaluation checklist with all the aspects to be addressed throughout the four sessions of the intervention.
Both therapists have Masters Degrees in Psychology and experience working with IPV victims.
Sample
The sample was composed of 23 adult women who had suffered from IPV (Table 1). All participants were women, volunteers, and were at a special protection shelter house (SH) for IPV victims in complex crisis situations, with high traumatic exposure, ongoing criminal court processes as victims, and who were considered at risk of violence in their own family context. There were other women in the SH, but some left before the protocols could be applied and one victim declined to participate. The time and circumstances under which each participant might leave the SH could not be predicted.
Sociodemographic and Clinical Characteristics of TG (n = 8) and CG (n = 11).
Note. TG = treatment group; CG = control group; IPV = intimate partner violence; LEC = Life Events Checklist; CVES = Conjugal Violence Exposure Scale.
Twelve people in this sample were randomly allocated into the CG (TAU, but without any specific intervention targeting those symptoms) and 11 into the TG. Due to their unpredictable circumstances, one person in the CG and three in the TG were not assessed, and we were not able to apply the FU protocol with two of the CG participants.
Randomization
Due to the specific sample conditions, randomization was performed using the Research Randomiser (randomizer.org), a generator of aleatory numbers with an initial set of 30, which provided a randomization ratio of 1:1. These 30 numbers represented a list of the possible number of participants who could be sheltered in the SH during the time span of this investigation. Taking this into account, 23 participants, the residents of the SH at the time of the randomization, were randomly and blindly allocated into two groups: 12 to the CG and 11 to the TG (Figure 1).

Flow of participants through each stage of the trial.
Outcome Measures
This trial’s main outcome measure is depression symptoms and secondary outcomes are PTSD, CPTSD, and BPD symptoms. The protocol consisted of five instruments and an informed consent.
The Patient Health Questionnaire-9 (PHQ-9; Kroenke et al., 2001 Portuguese version by Ferreira et al., 2018), which has internal consistency values ranging between .747 and .869, and was used to evaluate depression, which consists of nine questions assessing the presence of the depression symptoms described in Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association, 2013) in the previous 14 days. These symptoms are: depressed mood, anhedonia, sleep problems, tiredness or lack of energy, change in appetite or weight, feelings of guilt and/or worthlessness, concentration problems, slow thinking or thoughts, and suicidal thoughts. Depression is diagnosed if five or more of the nine symptoms have been present in the last 2 weeks (Kroenke et al., 2001).
The ICD-11/International Trauma Questionnaire (ITQ; Rocha et al., 2019) was used to evaluate PTSD, CPTSD, and borderline symptoms. This instrument has internal consistency values ranging from .839 to .882. Initially the participant is asked to select an episode of a traumatic event or stressful life experience and to indicate how long ago it happened. This questionnaire is divided into three sections, the first of which evaluates PTSD, through a set of seven questions answered on a Likert-type-style scale (0–4), which must be answered according to what the person felt in the last month. The second section evaluates CPTSD through a set of 17 questions answered on a Likert-type-style scale (0–4). Unlike the questions in the previous section, these should be answered according to how the participant normally feels and not according to what they felt in the last month. The third section evaluates borderline personality disorder through a set of 14 questions answered on a binomial scale (0–1, Yes or No).
Other instruments
The CVES Research Version (Rocha, Moreira, et al., 2019) was used to evaluate the level of exposure to conjugal violence. This scale has an internal consistency value of .872 and consists of seven questions that address the dimensions of verbal abuse: violation of freedom or dignity, threat or intimidation, physical abuse, sexual abuse, economic violence, and privacy control/disrespect. Another item allows the participant to add another type of violence suffered that is not addressed in the seven previous questions.
The Life Events Checklist (LEC; Weathers et al., 2013) was adapted to the Portuguese population through the research version by Rocha et al. (2013). LEC is a self-report scale built to evaluate potential traumatic events in a participant’s life. It assesses exposure to 16 life events that may potentially result in PTSD or distress, and includes an additional item that evaluates the presence of any non-rated traumatic events in the 16 previous questions.
After the first protocol application, therapy and FU were concluded, the Intervention Program Satisfaction Assessment Instrument (IASPI; Rocha, 2019), which has an internal consistency value of .903, was also applied. This instrument has 19 items, two dichotomous assessments of whether the participants found the intervention program useful or not, specific questions about the positive and negative components of the program, with 11 Likert-type items on a scale from zero to 10, and a final question asking which component was more important, whether the intervention prompted any changes in them and which parts of their lives improved following the sessions.
Statistical Method
We used independent sample t-tests and effect sizes to assess the differences between groups at the level of the symptoms of depression, PTSD, CPTSD, and borderline. Considering the sample size, there are procedures to control bias, comparing outcome results at baseline. If there is bias due to sample size randomization, the interception between groups and time can be obtained by calculating delta values (FU minus baseline) to clarify the changes in the outcome measure results between groups and time. We used Cohen’s d value to calculate the effect size.
Results
Between Groups Analysis for the Main Outcome Measure: Depression
Comparing the evolution of the participants in both the TG and the CG (Table 3) show that there are no statistically significant differences, but there is a positive Cohen’s d effect size of 0.43 (Table 2).
Cohen’s d Effect Sizes for Depression, PTSD, CPTSD, and BPD Variations Between FU and Baseline.
Note. PTSD = posttraumatic stress disorder; CPTSD = complex posttraumatic stress disorder; BPD = borderline personality disorder; FU = follow-up; PHQ = Patient Health Questionnaire.
Baseline and FU Values for TG (n = 8) and CG (n = 11).
Note. TG = treatment group; CG = control group; PHQ = Patient Health Questionnaire; PTSD = posttraumatic stress disorder; CPTSD = complex posttraumatic stress disorder; BPD = borderline personality disorder; FU = follow-up.
Between Groups Analysis for the Secondary Outcome Measures: PTSD, CPSD, and BPD Symptoms
Comparison of the PTSD baseline and FU values (Table 3) between groups shows that there are no statistically significant differences, but there is a small positive effect size of 0.19.
Analysis of the CPTSD total evolution between groups at baseline and FU (Table 3) demonstrates no statistically significant differences. The CPTSD total delta shows a small negative effect size of –0.28 (Table 2).
To make the best possible assessment of the CPTSD symptom evolution between groups, and to better understand the negative effect on complex PTSD, we analyzed the evolution between groups across all the dimensions of this outcome measure. Analysis of CPTSD is thus divided into CPTSD negative self-concept, CPTSD emotional numbing, CPTSD impulsivity control, CPTSD affective dysregulation and CPTSD impairment, and CPTSD total.
CPTSD negative self-concept delta also shows a medium negative effect size of –0.38 (Table 2). Similarly, the CPTSD affective dysregulation delta also shows a medium negative Cohen’s d effect size of –0.52 (Table 2).
Contrary to the previous two dimensions, CPTSD emotional numbing and CPTSD impulsivity control and impairment deltas show positive effect sizes. Comparison of the CPTSD emotional numbing baseline and FU values (Table 3) between groups, show no statistically significant differences, however, there is a small positive effect size of 0.04. Participants reveal high levels of satisfaction with the program (Table 4).
IASPI Frequencies (N = 7).
Note. IASPI = Intervention Program Satisfaction Assessment Instrument.
The CPTSD impulsivity baseline and FU values comparison between groups shows no statistically significant differences, but there is a medium positive effect size of 0.37.
Finally, there are no statistically significant differences when comparing the BPD symptoms baseline and FU values (Table 3) between groups, however, there is a small positive effect of 0.18.
Discussion
IPV has been associated with many mental health issues, such as an increase in depression, anxiety, PTSD, substance abuse (Coker et al., 2002; Dekel et al., 2020; Pico-Alfonso et al., 2006; Ruiz-Pérez & Plazaola-Castaño, 2005; Sullivan et al., 2016) which together with the traumatic experiences, crisis situations, and life changes these victims go through, make IPV victims a priority population for intervention.
Cognitive-narrative therapies are demonstrated to be successful in the treatment of complicated grief, depression and anxiety symptoms (Azevedo et al., 2017; Barbosa et al., 2013; S. Gonçalves et al., 2016), but there have not been any investigations studying its efficacy specifically on IPV victims for either PTSD or complex PTSD, and thus the relevance of an investigation that also studies all these variables.
It is important to note that there is bias in this sample, as the TG has higher baseline scores than the CG in all the studied variables except for CPTSD deregulation.
This therapy proved to be effective in reducing depressive symptoms in this group, which is in line with the results obtained with bereaved samples. As the applied intervention is a very short therapy of four sessions, an effect size of 0.43 is a very relevant reduction of the depressive symptoms in such a short time span, and makes it an important tool for work with these victims, and in reducing the possible consequences.
We also find a smaller but positive effect size on PTSD. This kind of decrease in the symptoms of such an incapacitating and impactful psychopathology is important, and shows that these symptoms can be reduced in a short period of time, thus improving the life of people suffering from them. Considering the number of traumatic experiences IPV victims go through, an effect size of 0.19 is important, shows that this intervention is useful in crisis situations, and can be a prime instrument for those who need to make an intervention quickly and in a short period of time.
Complex PTSD has a negative effect size. This was surprising and needs to be addressed in future research. This is the first known randomized controlled trial (RCT) controlling CPTSD symptoms and it raises new questions. Analysis of the procedure, the intervention and the sample characteristics, suggests that a possible explanation for this negative effect is the memory activation that occurs during therapy, which leads to a short-term increase in awareness of such difficult experiences. All the sessions invite memory focused activity meaning making, which may create an overly intensive confrontation in participants. This might generate some CPTSD symptoms to increase over a short time period and the positive effects might only be fully obtained after a few months of FU. Also, there is a negative baseline bias, having the TG more traumatic exposure, which can also explain this negative evolution. This randomized trial was the first to include this variable, and therapies might need to be adjusted to better address CPTSD symptoms and pursue longer FU. The negative self-concept and affective dysregulation dimensions of CPTSD may have specifically increased with the procedures 1 month after.
The emotional effect of recalling and reliving the traumatic experiences these victims went through is extreme, and the intervention process exacerbates this, which might explain the negative scores in CPTSD affective dysregulation. The effects of this therapy regarding the CPTSD dimension might only be assessable in a longer FU process, which allows all the work initiated by the participant during the intervention process to have its full effect.
CPTSD negative self-concept had a negative effect size, which might be explained by the increased awareness of their emotions and cognitions that these victims have after the intervention. Having better awareness is an improvement in the mental health of the participants, but before it can contribute to a better self-concept it can decrease it by making the participants more aware of all of what they have been through, their current situation and the consequences that it had on them as women, mothers and persons. In future interventions this potential confrontation should be addressed more carefully, in consideration of the deep emotional and traumatic experiences involved.
There are also positive effects for CPTSD regarding emotions of impulsivity, impairment, and emotional numbing, creating a faster decrease in these specific problems. The benefits of this therapy are also shown in the positive effect it has on BPD symptoms. Nevertheless, there are some limitations to take into consideration. The major limitation of this research is the short FU time and the impossibility of assessing the sample after 3 or 6 months, as well as sample size.
Sheltered victims often need to move to other areas of residence, change their habits or move forward with their lives, which might prevent them of continuing treatments. Although prolonged therapies are often the most indicated in this cases, cognitive-narrative therapy provides a good alternative when there is a need to act fast and efficiently or in crisis situations. The effects found for this therapy improve the tool box of psychological treatments available for IPV victims, adequate to such complex setting. Having positive treatments paves the way to improve awareness for victim’s mental health impairment that is overly neglected. Safety of victims is a priority, however, the inclusion of mental health interventions such this in IPV victims support protocols could benefit the reconstruction of life projects and reduce significantly future risks. In the future, it would be interesting to continue this line of research with a wider sample and long-term FU periods and using more ambitious outcome measures like positive readjustment to new life projects, well-being, and relations. There is also a need to explore the nuances of the effects in all the symptoms.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: DEBORA-PI-3RL-IINFACTS-2019, from CESPU, Portugal.
