Transdiagnostic treatment for emotional disorders
Emotion is one of the biggest sources of human distress. Symptoms of emotional difficulties, in particular depression and anxiety, are among the most common symptoms observed in psychiatric and clinical-psychological settings. Lifetime prevalence of depressive and anxiety disorders is estimated at 6.3 and 6.7 % in Japan [
1], and 16.6 and 28.8 % each in the U.S. [
2]. Depressive and anxiety disorders were respectively ranked as the second- and ninth-largest causes of global years lived with disability in 2013 [
3], and depressive disorders in particular are predicted to become the first cause of disease burden by 2030 [
4]. The economic burden of depressive disorder has been enormous ($11–18 billion in Japan and $173–211 billion in the U.S. in 2010) [
5‐
7]. Similarly, the economic burden of anxiety disorders was reported as $20.5 billion in Japan and $42.3 billion in the U.S. in 1990 [
8,
9]; and it should be noted that this figure for anxiety disorders in the U.S. could be greatly underestimated, because it did not consider long-term opportunity cost or the cost associated with comorbidity, which could bring total costs to the U.S. economy as a result of anxiety disorder to an estimated $100 billion [
10].
Depressive and anxiety disorders, which are commonly comorbid, are both conceptualized as disorders of emotion [
11]. Accumulated findings show shared psychopathology, etiology, neurobiological characteristics, and similar cognitive-affective, interpersonal, and behavioral-maintenance factors commonly observed among depression, anxiety, and trauma-related, obsessive-compulsive, and other emotion-related disorders [
11‐
15]. These findings leads to the idea of transdiagnostic conceptualization of these disorders [
16]; the term often used is “emotional disorders” [
15]. This broad category includes, at minimum, the
Diagnostic and Statistical Manual of Mental Disorders 5th edition (DSM-5) categories of depressive disorders, anxiety disorders, trauma-related disorders, and obsessive-compulsive disorders [
15,
17].
Cognitive behavioral therapy (CBT) has been shown to be an efficacious treatment for emotional disorders [
18,
19], and is recommended by treatment guidelines as a high-intensity treatments for depressive and anxiety disorders [
20‐
28]. Most CBT protocols have been developed to treat one specific disorder. Recently, however, the accumulated evidence of shared psychopathology has influenced not only the conceptualization of emotional disorders, but also their treatment. Transdiagnostic psychological treatment has been developed to treat depression, anxiety, and trauma-related, obsessive-compulsive, and other emotion-related disorders by targeting their shared psychopathology [
11,
29,
30]. The transdiagnostic approach has several potential strengths over the disorder-specific treatment: applicability to diverse disorders with comorbidity, simplification of treatment models for diverse emotional disorders, and ease of learning for novice therapists [
30,
31].
Though the transdiagnostic approach for emotional disorders has been proposed only relatively recently, uncontrolled/nonrandomized and randomized controlled trials (RCTs) have already been mounted to investigate its efficacy and feasibility. Transdiagnostic psychological treatments vary in format (individual, group, internet), theoretical orientation (transdiagnostic cognitive behavior therapy, transdiagnostic mindfulness- and acceptance-based treatments), strictness or flexibility of the application of the protocol to individual patients, and range of disorders covered (anxiety disorders only, both anxiety and depressive disorders). The latest meta-analysis has identified 47 clinical trials (31 RCTs) on the transdiagnostic approach, with a total of 3705 participants. That study reported the cumulative effect sizes of transdiagnostic CBT on self-report measures among patients with diagnosis of depressive and/or anxiety disorders, in comparison to control groups (waiting-list, supportive counseling, psychoeducation, treatment-as-usual, etc.): medium for anxiety (
n = 24,
g = .65, 95 % CI .51–.79), large for depression (
n = 22,
g = .80, 95 % CI .62–.98), and medium for quality of life (
n = 13,
g = .46, 95 % CI .34–.57) [
32,
33].
Despite the promise of transdiagnostic treatment, there are several limitations to it in its current state. First, the evidence of its efficacy is still inconclusive, because most existing trials have had a high risk of bias and because there is significant heterogeneity between the trials [
32,
33]. Newby et al. reported that only seven out of 31 RCTs were rated low risk for bias in all of five criteria (namely, random sequence, concealment, blinding of assessors, incomplete outcome data, selective reporting); and clinical trials for transdiagnostic treatment on anxiety and depression in comparison to control group(s) have been shown to have moderate heterogeneity (
I
2 = 51–83 %) [
32,
33]. More importantly, the latest meta-analysis did not examine the results for clinician-rated measures, because only 13 out of 31 RCTs reported an independent assessor rating for a structured clinical interview, and the interviews used varied from trial to trial. Second, no RCT for transdiagnostic CBT has been conducted in non-Western countries, which means that transdiagnostic treatment cannot yet be regarded as a transcultural treatment [
12]. Third, only eight out of 31 RCTs have included both depressive and anxiety disorders, which are the putative populations for transdiagnostic treatment, which means that, many trials included a sub-category of anxiety disorders only. It is necessary to include transdiagnostic clinical populations in clinical trials to examine the external validity (i.e., generalizability) of the findings. Fourth, few studies have examined the putative treatment mechanism that rationalizes the treatment as transdiagnostic (e.g., neuroticism, emotion regulation) [
12,
34]. Fifth, no study has examined the neuropsychological underpinnings of transdiagnostic treatment.
Unified protocol for the transdiagnostic treatment of emotional disorders (UP)
The UP is one of the most empirically well-supported transdiagnostic psychological treatments [
33,
35]. It was developed based on the recognition of diagnostic overlap across emotional disorders (i.e., comorbidity), non-specificity of treatment response to comorbid symptoms by diagnosis-specific treatments, latent structure of emotional disorders (e.g., negative affectivity), etiology, and commonality of findings of affective neuroscience on emotional disorders (e.g., reductions in connectivity between the amygdala and ventromedial prefrontal cortex) [
11,
13]. Evidence relating to individual UP has been reported in various forms: theoretical paper [
11,
13], case report [
36], pilot clinical trial [
37], and RCT with waiting-list [
38]. Moreover, the feasibility of UP has been extended to individual and group formats for depressive and anxiety disorders [
39], individual format for bipolar disorder [
40], borderline personality disorder [
41], and emotional disorders in adolescents [
42].
The existing evidence, however, should be regarded as preliminary at best [
12]. Findings by RCT to date are limited to one trial each for individual- and group-format UP, with small sample sizes [
38,
39]. Further, there is clear lack of evidence regarding UP efficacy for depressive disorders and for Asian populations. Previous pilot studies for individual UP have examined only three patients with the principal diagnosis of major depressive disorder [
36‐
38], and there is no RCT of UP efficacy for depressive disorders. Comparative clinical trials have been limited to one in the U.S. and one in Brazil [
39].
Thus, we aimed to examine the efficacy of UP for Japanese patients with either or both depressive and anxiety disorders. Though such a broad category of participants constitutes a heterogeneous clinical population in terms of the standard diagnostic criteria (i.e., DSM-5 or
International Classification of Diseases 10th revision; ICD-10) and leads to some methodological challenges unique to transdiagnostic treatment [
43], we nevertheless selected this population because the unified protocol was originally developed to treat heterogeneous clinical populations transdiagnostically.
Treatment mechanism and process for enhancing the effectiveness of implementation of unified protocol
Though this clinical trial is mounted primarily to test the efficacy of the unified protocol on the primary outcome (i.e., emotional symptoms), it is also important to examine the treatment mechanism, process, and neurological correlates [
12,
32]. Furthermore, because the UP was developed based on the findings on the shared underlying process of psychopathology across a wide range of emotional disorders (depressive, anxiety, trauma-related, and obsessive-compulsive disorders), it is important to examine how this shared process actually relates to the outcome of the treatment. These examinations are expected to not only allow theoretical refinement of UP, but also sharpen its clinical implementation. We focus on four different variables to examine the treatment mechanism; neuroticism, emotion regulation, emotion exposure, and anxiety sensitivity. These variables are considered to be transdiagnostic mediators/moderators of the treatment. Neuroticism here refers to the shared aspects of temperament commonly observed between depressive and anxiety disorders [
44,
45]. Deficits of emotion regulation are also a putative transdiagnostic treatment mechanism of CBT [
19,
34,
46]: through UP, patients will learn to engage adaptive emotion regulation and disengage maladaptive emotion regulation. The core module of the unified protocol is emotion exposure [
35]; theoretically, the other modules will be expected to function to maximize the therapeutic effect of emotion exposure [
11], and hence the patient’s ability to expose themselves to their own emotion should be one of the key focuses in terms of treatment mechanisms [
47,
48]. Finally, anxiety sensitivity is regarded as the transdiagnostic process affecting the (non) maintenance of emotional disorders [
49].
The effects of psychotherapy depend on its process: who delivers the protocol, how it is delivered, the degree to which the patient understands the concept or practices the skill, etc. [
47,
50,
51]. Past studies suggest that strict adherence to CBT protocols does not linearly predict the treatment outcome [
52,
53]. Hence, it is not enough to monitor adherence to the treatment protocol; one must also assess various aspects of the implementation or processes of the psychotherapy. In this clinical trial, we will collect some of the most important process measures regarding the implementation of the unified protocol. These include treatment expectancy [
54,
55], therapeutic relationship [
56‐
58], homework compliance [
47,
59‐
61], and patient comprehension of treatment rationale. Examination of the relationships between these process measures and treatment outcome is expected to inform and help optimize the implementation of UP for patients with emotional disorders.
Neurological correlates regarding the unified protocol
To the best of our knowledge, no study has yet examined neurological change after the intervention of transdiagnostic treatment; nevertheless, the existence of transdiagnostic neurological commonalities across emotional disorders has been suggested. Some researchers have developed the transdiagnostic neural model [
62,
63] to reflect this. Recent studies have been focused more on the similarity of neural systems between depression and anxiety disorders, in addition to the differences [
64]; this focus is consistent with the view taken by the Research Domain Criteria project of the U.S. National Institute of Mental Health [
65]. In fact, neurobiological studies have shown some similarity in changes from before to after CBT across emotional disorders [
66]. Considering that UP is a transdiagnostic treatment partly based on findings from the field of affective neuroscience [
13], transdiagnostic change at the neurological level is indeed expected across emotional disorders. Moreover, recent investigation of the domain of neuropsychological perspective has predicted treatment efficacy using neurological measures; for example, Ball et al. [
67] showed that the random forest model using pre-treatment neuroimaging data predicted treatment response of generalized anxiety disorder and panic disorder patients with good test characteristics. Hence, we will correct these neuropsychological data as an ancillary study.