{PSOCIAL}

Borderline personality disorder (BPD) is a mental health problem involving difficulties in four main dimensions: (a) interpersonal instability, such as intense relationships and a strong fear of abandonment; (b) cognitive disturbances such as dissociation, identity disturbance and obsessive and paranoid ideas; (c) emotional and affective dysregulation; and (d) behavioural problems, including impulsivity and self-injurious and suicidal behaviours (American Psychiatric Association [APA], 2013). The prevalence of this pathology ranges from 0.50% to 5.90% of people (Grant et al., 2008; Tomko et al., 2014), although the data may vary depending on the assessment methods used and where the studies are conducted. In Spain, the prevalence is 0.017% of those surveyed, with a higher prevalence in young people than in adults, and a significant percentage of these individuals also suffer from another mental health problem (Aragonès et al., 2013). Although people with BPD manifest great suffering (Slotema et al., 2019), especially due to affective instability (Zimmerman et al., 2017) and emotional regulation problems (Fitzpatrick et al., 2021), they often seek treatment for other mental health problems or repeated suicide attempts (Gunderson et al., 2018). People with BPD often present anxiety disorders, mood disorders, stress disorders, other personality disorders, neurodevelopmental disorders such as attention deficit hyperactivity disorder and substance use disorders (SUD) (Grant et al., 2008; Tomko et al., 2014).

The literature indicates that a large proportion (25%) of people with BPD also suffer from a SUD (Trull et al., 2018), with 70.60% of them presenting alcohol use disorder, 62.70% cocaine use disorder, 45.80% cannabis use disorder, 26.80% opiate use disorder and 22.20% sedative use disorder (González et al., 2019). This comorbidity is associated with a worse course and outcome of the disorders (Kienast et al., 2014).

To date, specific treatments have been proposed for BPD and SUD individually (Meuldijk et al., 2017; Lo Coco et al., 2019), but few studies have addressed the joint treatment of both conditions and shown high rates of improvement (Trull et al., 2000). This group is particularly vulnerable due to their difficulties with emotional regulation, and the high incidence of self-injurious and suicidal behaviour.

This qualitative systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standard (Shamseer et al., 2015). The protocol for the systematic review was registered in PROSPERO (CRD42021228860) (Booth et al., 2012).

The aim of the present study was to determine the effectiveness of psychological treatments applied to individuals with co-morbid BPD and SUD by means of a systematic review according to PRISMA standards (Shamseer et al., 2015). After an exhaustive search, few studies that met the specified inclusion and exclusion criteria were obtained, and 12 studies were finally analysed. These studies focused on assessing the effects of different psychological and psychiatric treatments in people with co-morbid BPD and SUD. Our results show that there is a wide variety of approaches to the treatment of people with co-morbid BPD and SUD, and that they generally share a common problem: low adherence to treatment. However, the treatment most frequently referred to in the literature is Mentalization-Based Treatment. This treatment is based on the fact that the symptoms of BPD occur when the person is unable to mentalise, makes decisions and performs behaviours without reflecting, and is detached from reality. This skill is developed in childhood when living in a stimulating and emotionally safe environment. Treatment combining individual and group sessions is therefore generally directed towards increasing reflective thinking, evaluation of one's own and others' emotional situations and flexibility (Bateman & Fonagy, 2004; Choi-Kain et al., 2017). The studies reviewed indicated that this treatment was effective for BPD symptoms, but not for SUD symptoms (Kaltenegger et al., 2020; Lana et al., 2016; Philips et al., 2018a; Philips et al., 2018b).

Dialectical Behaviour Therapy, an individual therapy that aims to improve social skills, emotional regulation, frustration tolerance and attention to the present (Linehan, 1993; May et al., 2016), proved useful over time in reducing BPD and SUD symptoms, suicidal behaviours and consumption, improving management of emotions, and increasing quality of life (Flynn et al., 2019; Van den Bosch et al., 2005).

In the studies in which it is used (Ball et al., 2011; Ball, 2007; Ball et al., 2005) Dual-Focus Schema Therapy reduces the severity of BPD and SUD symptoms, the results are not sustained over time and there is a high level of experimental mortality. In contrast, General Psychiatric Management achieves adequate therapeutic alliance and symptom reduction (Penzstadler et al., 2018), which helps to improve adherence. Integrative Borderline Adolescent Family Therapy was shown to be effective in reducing adolescent and family distress and adolescent substance use, especially among adolescents who also presented major depression (Santisteban et al., 2015). Finally, Dynamic Deconstructive Psychotherapy was shown to be effective in reducing suicidal behaviour, substance use, hospitalisation, healthcare attendance, and dissociative, depressive and BPD-associated symptoms (Gregory et al., 2008).

A summary of the results shows a wide variety of intervention proposals that generally lead to a reduction in the severity of the symptoms experienced and an improvement in the different variables evaluated, with the most common improvements being related to a reduction in suicide attempts and in hospitalisation and medical care. The results regarding substance use are not so conclusive, and it is difficult to ascertain the best intervention for reducing substance use in people with comorbid BPD.

Despite the contributions of our work, it should be noted that the studies that met the inclusion criteria have small samples, ranging from 17 to 41 participants with co-morbid BPD and SUD, with total samples ranging from 46 to 105 participants (if all the participants are taken into account, regardless of their condition). However, these samples are very heterogeneous, due to the fact that they include patients with different substance use issues, with different types of severity, and with other comorbid pathologies. In addition, the participants come from diverse institutional settings, ranging from institutionalised patients to homeless people to people subject to justice measures. This makes it difficult to compare treatments and conclusions. On the other hand, all the studies selected participants by convenience and there was a high dropout rate in some of them, although attendance was financially rewarded (Ball, 2007; Ball et al., 2005; Ball et al., 2011). We therefore believe that the sample characteristics and the sampling in these studies complicate the generalisability of the results to the general population. Furthermore, few studies report the drop-out rate of participants; and when they do, it is usually quite high. More studies are needed to continue investigating the situation of people with co-morbid BPD and SUD, using probability sampling that adequately represents the characteristics and needs of this population. Similarly, the sample size should be increased and randomised controlled trials should be conducted in order to draw higher quality conclusions. However, access to people with co-morbid BPD and SUD is complex, given the complexity of the situations they generally face, and the continuity of treatment is therefore also difficult.

In terms of the variables and instruments taken into account in the research studied, as well as the data analysis, all the studies use psychometrically appropriate instruments. However, few of them use robust statistics (Gregory et al., 2008; Lana et al. 2016, Santiesteban, et al. 2015) which enable a consistent analysis of the results obtained. Moreover, some of them do not take into account the analysis of masked variables, such as the severity of the disorders, traumatic experiences, family relationships or social support, or even the assessment of self-injurious behaviours or previous attempts at medical or psychological treatment, among other variables. Furthermore, some of them do not specify the number of dropouts (Flynn et al., 2019; Kaltenegger et al., 2020), or who the therapists were or their sociodemographic variables or profession (Ball et al., 2005; Ball et al., 2011; Flynn et al., 2019; Kaltenegger et al., 2020; Lana et al., 2016; Van den Bosch et al., 2005), despite the fact that all the studies considered in this review are published in impact journals. However, all of them use longitudinal designs, and most of them use randomised studies (Ball et al., 2005; Ball et al., 2011; Ball, 2007; Kaltenegger et al., 2020; Penzenstadler et al., 2018; Philips et al., 2018b; Van den Bosch et al., 2005). On the other hand, only one of the selected studies was conducted in the Spanish population (Lana et al., 2016), and as such we believe that research is still needed to clarify the effectiveness of these interventions in this group in our context.

Future research could also take into account important adjustment variables such as emotional bonds, resilience or emotional intelligence, as studies have generally focused on reducing risk factors, but have not addressed the increase in protective factors. In addition, future research should consider how to reach people in particularly vulnerable situations, such as homeless people, pregnant women and minors with co-morbid BPD and SUD. These at-risk groups often make it more difficult to provide them with treatment or fail to maintain the therapeutic alliance and adhere to therapy. Finally, other studies could apply Eye Movement Desensitisation Reprocessing in this group of patients, as it has been shown to be useful in people with BPD and in people with SUD (Carletto et al., 2018; Slotema et al., 2019).

Systematic reviews in co-morbid BPD and SUD are very scarce, and this applies particularly to those focused on effective interventions for their treatment. As a result of a systematic search of the literature based on precise and systematic inclusion criteria, this review extends knowledge beyond the conclusions of narrative reviews. In addition, our review included two blinded reviewers throughout the process, as well as the rate of agreement between them. Although some literature is available on the topic under investigation, it is not sufficient to draw any conclusions, and we propose that more studies be conducted, using larger samples. It is also essential to develop strategies to reduce experimental mortality, as this was the main problem or limitation in the research reviewed.

The main results of our study indicate that there is a diversity of treatment alternatives for people with co-morbid BPD and SUD, in individual, group or combined settings, carried out by different professionals including psychologists, psychiatrists, nurses and social workers, which have been shown to be particularly effective with the symptoms associated with BPD. We therefore believe that the present review provides professionals responsible for the health of individuals with a guide to the most extensively studied psychological interventions in the literature for co-morbid BPD and SUD. It facilitates decision-making regarding which therapy to apply for patients with these characteristics, as well as providing a summary of the main benefits of these interventions.