Introduction
Suicide is one of the leading causes of death worldwide, and the World Health Organization attributes over 800,000 deaths per year to suicide [1]. It has tragic and drastic effects, and, globally, for every person who dies by suicide, 20 more people attempt to take their life but do not die [1]. Many are bereaved in the aftermath of a suicide. Family members, friends, and those close to the deceased are at an increased risk of suicide themselves. Therefore, understanding and implementing methods for reducing suicide are critical, as is the prioritization and recognition of suicide as a solvable major public health problem.
For the purpose of this review, suicide is defined as the act of deliberately killing oneself, and deliberate self-harm (DSH) is defined as any nonfatal suicidal behavior, such as intentional self-injury, poisoning, or self-harm with or without a fatal intent. Suicide and DSH are preventable, and therapeutic approaches that specifically target suicidal ideation (SI) provide successful results [2-4]. Cognitive behavioral therapy (CBT) can reduce SI and treat depression and insomnia [5]. The use of CBT as an intervention for suicidality has been tested in a number of randomized controlled trials (RCTs), with some evidence for its efficacy [6-8]. CBT for suicide prevention and CBT for suicidal patients have also had positive effects on SI [2,3].
An evidence base for the “third wave” of CBTs, such as dialectical behavioral therapy (DBT), mindfulness-based CBT (MBCT), and acceptance and commitment therapy (ACT), has been established over the last 15 years [9]. DBT is highly effective in treating presentations of self-harm among those with borderline personality disorder and is frequently delivered as group therapy [10]. MBCT has become an acceptable alternative to CBT. In MBCT, the practice of mindfulness activities is considered a useful addition to standard CBT activities [11]. Given the strong evidence for the efficacy of third-wave therapies for other indications [9-12], it is worth examining whether ACT shows promise in the area of suicide prevention.
ACT attempts to increase psychological flexibility mainly by targeting experiential avoidance—the tendency to avoid unwanted thoughts or emotions [13]. The 6 core processes of ACT are as follows: (1) acceptance of uncomfortable private experiences (thoughts, feelings, or physical sensations); (2) cognitive defusion/distancing from one’s own uncomfortable thoughts; (3) being present (directing attention to present events and experiences rather than focusing on the past or future); (4) self-awareness in the present moment through the “observing self;” (5) identification of personal values; and (6) commitment to action in line with the identified values. Since the publication of Acceptance and Commitment Therapy in 1999 by the treatment’s cocreators—Steven Hayes, Kirk Strosahl, and Kelly Wilson [13]—the number of ACT-based RCTs has increased [12]. Unsurprisingly, ACT has been used to treat common mental health conditions, such as depression, anxiety, addiction, and stress, as well as physical conditions, such as chronic pain [12,14,15]. However, trials that examine the efficacy of ACT in targeting SI/DSH continue to lack.
There is good reason to hypothesize that ACT may be effective in reducing SI and DSH by improving psychological flexibility [16]. Some of the predominant psychological frameworks that attempt to explain suicide, particularly the entrapment/cry of pain model, include escape from pain as a key factor [17-19]. Escape, referred to as experiential avoidance, is one of the key target areas of ACT treatment. The application of mindfulness skills, acceptance of distress, and defusion from distressing thoughts may improve an individual’s ability to live with the discomfort of severe emotional pain. Finally, the identification of personal values and taking up of positive action aligned with these values may lead to an integrated individual, thereby improving wellbeing.
To date, reviews of ACT for SI or DSH have not been published. A 2014 meta-analysis of the efficacy of ACT examined 60 RCTs that focused on psychiatric disorders, somatic disorders, and stress at work [12]. Of these studies, none examined SI or DSH. A 2016 meta-analyses and metaregression of studies that examined the effectiveness of psychotherapy in reducing suicidal attempts and nonsuicidal self-injury rates included 32 RCTs, none of which included ACT as a treatment despite the brief mention of its promise [20]. Hacker et al [21] highlighted the efficacy of ACT in systematic reviews but were critical of the broad range of presentations reviewed and the lack of specificity on common mental health problems. We have attempted to overcome this limitation with our focus. Given that some authors have proposed that ACT treatment has the potential to be an effective intervention for suicidal individuals [22,23], the available evidence must be reviewed. For this review, we aimed to examine (1) whether ACT is an effective treatment for SI or DSH and (2) the suitability of reported SI, DSH, and other measures in determining the efficacy of ACT.
Methods
This review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines [24].
Search Strategy and Selection Criteria
The following electronic databases were systematically searched: MEDLINE, PubMed, EMBASE, PsycINFO, SCOPUS, Cochrane Central Register of Controlled Trials, and the Cochrane Database of Systematic Reviews. A comprehensive set of search terms was identified with reference to terminology used in previous review papers for ACT [9,12] and suicide prevention [4,25]; these terms were combined with MeSH terms relevant to each of the databases. Search terms included “acceptance and commitment therapy” or “acceptance-based therapy” and either “suicide,” “ideation,” “assisted suicide*,” “attempted suicid*,” “self-injurious behavior,” “self-mutilation,” “self-harm,” “self-poison*,” “self-inflicted wounds,” “drug overdose,” “overdose,” or “parasuicid*.” In addition, the reference lists of all included studies and relevant reviews were examined to identify additional relevant publications.
Despite the emergence of ACT in the late 1990s, no date restrictions were placed on searches that were completed on December 11, 2017. A study was eligible for inclusion if it satisfied the following criteria: (1) The study used ACT as an intervention. Multicomponent therapy types were excluded because of our interest in examining the efficacy of ACT when used as a standalone therapeutic intervention. Interventions could be delivered to individuals or groups or through technology. (2) The study assessed suicidal behavior by using a validated measure or structured reporting. Suicidal behavior was defined in its broadest terms and ranged from SI to the various forms of self-harm indicated in the search terms. (3) The study is an original peer-reviewed article published in English. Given the recency of third-wave interventions, study design was not restricted to RCTs. Instead, all research designs were included (eg RCTs, quasiexperimental, pre–post, single group, and case studies). Finally, the age of participants was unrestricted. Appendix 1 provides search term details.
Selection Process
After the removal of duplicates, 2 researchers (JT and JN) independently reviewed the relevance of all titles and abstracts that were returned by the search. Studies considered irrelevant by both the reviewers were excluded. The full texts of the remaining articles were then independently examined by the same 2 authors to confirm eligibility. Included articles and reasons for exclusion were compared to achieve consensus and, where necessary, disputes were settled by a third researcher (FS).
Data Extraction
One author (JT) extracted the study characteristics and outcome variables, which were independently checked by JN. The following variables were extracted: author name, publication year, sample type, control group details, program format, participant age, program length, and follow-up interval. Outcome data on SI/DSH, depression, and psychological flexibility (acceptance and mindfulness) were also extracted.
Risk of Bias
The study quality and risk of bias of the included RCTs were assessed using the Cochrane Collaboration “Risk of Bias” tool [26].
Discussion
Principal Findings
This review aimed to understand if ACT can successfully reduce SI or self-harm and to examine the suitability of the measures used in the included studies. The review found few empirical investigations on ACT that specifically target the reduction of SI/DSH, with only 1 RCT among the 5 studies. This is the first review we are aware of that has examined the impact of ACT on the reduction of SI/DSH.
All 5 studies examined SI by using various measures, and only the 2 case studies examined DSH. All studies reported a reduction in SI, and both case studies reported reductions in DSH over the course of the interventions. The degree of the reduction in SI varied: Tighe et al [27] reported a nonsignficant 30% reduction in SI scores, and Walser et al reported a significant 20.5% reduction in the prevalence of SI among participants at follow-up [32]. Ducasse et al [36] reported significant reductions in all SI measures at 1-week and 3-month follow ups [36]. However, the effective evaluation of ACT treatment was diminished given that the ACT component of the participants’ therapy was provided as an add-on to treatment-as-usual. The authors of this trial concluded that ACT might help reduce the intensity and frequency of SI by increasing acceptance and valued action and by reducing risk factors, such as hopelessness and psychological pain [36]. The 2 case studies measured and reported a reduction in SI and DSH [44,45].
Although this review focuses on SI and DSH, the secondary measures of depression, acceptance, and mindfulness are crucial for evaluating the effectiveness of ACT. Among the 5 studies included in this review, 3 reported outcomes for depression. This review supports the existing body of evidence that highlights the effectiveness of ACT in this common presentation [12]. Tighe et al [27] found that depression scores in the intervention group significantly reduced by 42% relative to that in the waitlist group. Walser et al [32] reported a 32% and 40% reduction in the depression scores of the cohort of veterans with and without recorded SI at baseline, respectively. The pre–post study conducted by Ducasse et al [36] lacked a control group but showed significant reductions for all measures, including depression. The positive results of ACT on depression support recent research that highlights the potential of ACT as an intervention for common mental health conditions, such as depression and anxiety [12,46]. Furthermore, the reductions in depression and SI reported by the included studies support previous research demonstrating that reductions in depression can lead to reductions in SI [47,48]. The use of the BDI-II by Walser et al [32] to measure depression and the presence of SI provided limited data on SI/DSH. The use of focused SI/DSH measures in future studies would likely enrich analyses. Notably, most measures were not common between studies. However, 2 of the 5 studies included the AAQ and the FFMQ, which are key measures of acceptance and mindfulness (key goals of ACT treatment). Including these measures in future studies on ACT would be useful because their analyses could show potential associations with SI/DSH. The use of established standardized measures of SI/DSH in RCTs and case studies would also improve the evidence base for ACT.
The 5 studies included 3 distinct cohorts of participants. The mHealth RCT by Tighe et al [27] targeted Aboriginal and Torres Strait Islander youth. Walser et al [32] targeted veterans of the United States Army, and Ducasse et al [36] targeted French outpatients with a recent suicide attempt. In the 2 included case studies, 3 of the 5 participants were suffering from severe distress in hospital settings. Across these diverse populations, ACT was positively associated with a reduction in SI. The methodological quality of the studies was low, with just 1 (Tighe et al) [27] including a control group. The 2 studies that reported positive results for SI through case studies (N=2 and N=3) had limited participants and lacked robust research frameworks from which to draw conclusions [44,45]. Both studies report impressive reductions in SI for all the 5 individuals studied; however, the studies had numerous limitations, including lack of specificity in the measurement of SI and self-harm reports (apart from hospital ward reports). Details regarding treatment-as-usual, such as medication administration to participants, were not provided. Notably, the only RCT included in this review delivered therapy through a self-help mobile app [27]. Whether this is indicative of an increasing adoption of mHealth is worth considering [49-51].
All 5 studies showed that ACT is associated with symptom changes. This result provides a rationale for the largescale systematic evaluation of the efficacy of ACT as an intervention for suicidal behavior. Tighe et al [27] reported significant reductions in depression and distress on standardized measures and a 30% reduction in SI scores. This reduction, however, was nonsignificant between groups. All 5 participants whose case studies were presented improved significantly with marked reductions in SI/DSH [44,45]. In addition to the significant results for SI, the 2 pre–post studies showed improvements on the AAQ, a measure of psychological flexibility [32,36]. Walser et al [32] showed that their improvements on the AAQ are associated with a reduction in SI scores across time. Similarly, Ducasse et al [36] found a significant correlation between SI and AAQ scores, such that increased acceptance is associated with reduced SI. Accepting reality and reducing avoidance are fundamental aims of ACT, and this therapeutic work is aligned with the theories of suicide that focus on an individual’s sense of entrapment and/or desire to escape their current reality [17-19]. Further trials are needed to test how increases in experiential acceptance and mindfulness affect SI/DSH scores and whether targeting these factors might increase reductions in SI/DSH.
Conclusion
The number of studies included in this review is too small to support the claim that ACT can effectively assist in the reduction of SI/DSH. Given the limited research that has been conducted on this topic to date, the efficacy of ACT in reducing SI or DSH requires further testing, particularly through controlled trials. The early evidence presented in this review suggests that the potential mechanisms of action, such as changes in experiential avoidance and mindfulness, should receive focus.