The term mindfulness refers to, at least, two different concepts: (a) a meta-cognitive exercise which involves bringing sustained and intentional attention to experiencing the present moment, while diminishing the emotional and cognitive reactivity generated by the experience or (b) a state of consciousness characterized by the detached observation of one’s thoughts and feelings [1
]. Mindfulness mainly comes from the vipassana meditation of Buddhist tradition and plays a central role within the framework of a conceptual and applied system whose ultimate aim is the cessation of suffering [2
]. However, it was introduced in the Western world with a secular approach, disassociated from any religious or cultural tradition, as a technique oriented toward promoting the quality of life of people with high stress levels associated with chronic illnesses of a physical type [3
]. Over the years, mindfulness has been extended for therapeutic purposes to other health conditions, mainly due to its beneficial effect on people who present psychopathological profiles and somatic diseases [4
]. This expansion has taken place through the implementation of several intervention protocols included under the umbrella term mindfulness-based interventions (MBIs).
Amongst the MBIs one can find two seminal protocols, Mindfulness-Based Stress Reduction (MBSR) [3
] and Mindfulness-Based Cognitive Therapy (MBCT) [5
]. Both are widely used and have remarkable empirical support [6
]. There is also a broad set of protocols that have been developed from MBSR and MBCT protocols for specific therapeutic purposes. Such protocols usually introduce variations regarding MBSR and MBCT in aspects such as program structure or pedagogical content. However, all MBIs have, as a common point, the systematic training in meditation practices for the development of mindfulness, which is at the heart of the intervention [8
]. Other distinctive aspects of MBIs have to do with the way they are administered: usually in group format and through one or more teachers who are not necessarily therapists or mental health professionals [9
]. These aspects differentiate MBIs from other therapies—the so-called acceptance-based approaches—in which mindfulness is used as a tool to promote psychological acceptance, but no systematic use of meditation practices is made [10
]. This is the case with the Acceptance and Commitment Therapy (ACT) [11
] or the Dialectic Behavioral Therapy (DBT) [12
], in which mindfulness is taught primarily using short exercises and informal practices which consist of daily activities with a focus on the present moment.
The popularity of MBIs has increased exponentially over the last 30 years, alongside the volume of research conducted on their effectiveness in the treatment of different psychological conditions [13
]. Nevertheless, evidence regarding their applicability in schizophrenia is still scarce. This is possibly partly due to a cautious reaction to a few early research papers, which found a relationship between meditation practice and the emergence of psychotic-type symptoms [14
]. These publications, mostly based on case studies, reported that extended periods of meditation practiced over several consecutive days, in combination with other factors—a history of previous psychotic episodes, sleep deprivation, or the discontinuation of psychiatric medication—could lead to the onset of psychotic disorders. In this regard, authors such as Kuijpers et al. [16
] have suggested that, under these conditions, meditation may act as a trigger of symptoms in individuals who have shown vulnerability. Despite this, some studies published in the early 2000s found that the application of therapies which included mindfulness practice resulted in improvements in patients with schizophrenia [17
]. Soon after, Chadwick [18
] developed the Person-based Cognitive Therapy for distressing psychosis (PBCT), the first intervention specifically designed for people with schizophrenia to incorporate mindfulness as a central element, combined with other therapeutic components derived from cognitive-behavioral therapy (CBT).
CBT is one of the most widespread and empirically supported treatments for schizophrenia [19
] used to specifically address psychotic symptoms, such as hallucinations and delusions, and which aims to modify patient held beliefs [20
] (pp. 115–118). Although there is a certain amount of controversy regarding its effectiveness in the treatment of schizophrenia [21
], on the whole, research yields positive results, especially in the decrease of positive symptoms [22
]. A crucial difference between CBT and MBIs is that the latter do not focus the intervention on the psychotic symptoms per se but rather on the relationship patients establish with psychotic-like experiences. People with schizophrenia learn through mindfulness to abandon their usual response to aversive psychotic experiences—e.g., control efforts and escape-avoidance coping—to adopt a response style fundamentally based on acceptance, which, although counterintuitive, may result in a reduction of the generated distress [25
To our knowledge, four meta-analyses have been published to date on the effectiveness of interventions, which include elements of mindfulness for people with psychosis: Khoury et al. [27
], with 13 studies; Cramer et al. [28
], with 8 studies; Louise et al. [29
], with 10 studies; and Jansen et al. [30
], with 16 studies. It is noteworthy that their results are, in general, heterogeneous. Thus, only one reported large effects on overall symptomology [30
], and two reported small effects [28
]. Regarding the characteristic symptoms of schizophrenia, only one found moderate effects on positive symptoms [28
] and two found modest effects on negative symptoms [27
]. However, it is relevant to note that the four meta-analyses cited included in their analyses both MBI’s and acceptance-based approaches (ACT or DBT), which, as mentioned, have significant differences between them. They all included studies in which participants had psychotic symptoms but not necessarily a diagnosis of schizophrenia spectrum disorder (e.g., major depression with psychotic symptoms). These methodological aspects also provide a plausible explanation for the observed variability in their respective results.
The inclusion in the previous meta-analyses of MBI’s and acceptance-based approaches is understandable given that both types of intervention foster the development of mindfulness—by different means—and because of the small number of available studies in which MBI’s were applied to people with schizophrenia. However, we consider that this situation has now evolved, thanks to the recent publication of high-quality studies in this respect. Consequently, we conducted a systematic review and meta-analysis of RCTs to examine the effectiveness of MBIs in improving clinical parameters related to schizophrenia.
The present systematic review and meta-analysis included ten studies and a total of 1094 participants. The results showed that MBIs combined with TAU are effective for the treatment of schizophrenia when compared with both TAU control groups and active treatment control groups—mostly psychoeducation groups—under the same time and frequency conditions as the MBIs. The MBIs generated moderate to large effects in reducing overall symptomatology and small to moderate effects in reducing both positive and negative symptoms in pretest–posttest comparisons. The results suggested that these effects were not mediated by the variables age, gender, adherence to treatment, methodological quality, specific treatment protocol, duration of the treatment, or type of control group.
Another relevant result is that the MBIs generated large magnitude effects on the psychosocial functioning level and moderate magnitude effects on the level of disease awareness. However, the data for these results came from four studies in which the MBIs included psychoeducational elements aimed explicitly at their improvement. Despite this, we consider that this result is noteworthy from a clinical point of view since improvements in the levels of psychosocial functioning and awareness of illness are highly relevant objectives in the treatment and recovery of people with schizophrenia [62
Moreover, although only four studies analyzed changes in the mindfulness variable in the pretest–posttest comparisons, all of them reported large magnitude changes. This result is relevant since it points to a possible relationship between mindfulness and the therapeutic changes found in these trials after the application of MBIs. Future studies should explore this matter further.
Overall, the findings of this review differed from those reported by the four meta-analyses published to date on the effectiveness of interventions that include elements of mindfulness for people with psychosis [27
]. In the reviews by Cramer et al. [28
], Louise et al., [29
] and Jansen et al. [30
], treatments produced small to moderate effects on overall symptomatology, which is in line with our findings; Khoury et al. [27
] did not analyze this outcome. Cramer et al. [28
] and Jansen et al. [30
], again in line with our findings, reported small to moderate effects of MBIs on positive symptoms; however, in Khoury et al. [27
] and Louise et al. [29
], the changes in this variable did not reach significance. Regarding negative symptoms, only Jansen et al. [30
] reported—in agreement with our findings—small but significant effects. In any case, due to the methodological differences already discussed, caution must be exercised when comparing the results presented in this review with those of the mentioned meta-analyses.
An interesting finding of this review was that, although MBIs provided, on the whole, better clinical results than psychoeducation groups, this advantage was not usually reflected in the posttest measurements but rather in the follow-up measurements, six months after the end of the intervention. In the scrutiny of the data, it was be observed that this result was not so much due to MBIs’ higher retention capacity of therapeutic effects but rather to a steady increase of their effects over time. In this regard, it was observed that MBIs generated greater effects on overall symptomatology in the comparisons between the pretest and follow-up measurements (moderate–large effects) than in the comparisons between the pretest and posttest measurements (large effects). We believe this finding could provide a future line of research.
None of the studies reviewed reported any harmful effects related to the implementation of MBIs, which suggests that this type of intervention is safe for people with schizophrenia when structured protocols of intervention are followed. Thus, the present review did not find empirical evidence to support the risks suggested in previous investigations [16
] regarding the possible exacerbation of psychotic symptoms as a result of mindfulness training. The reviewed trials shared certain elements of mindfulness practice which may have a positive impact on their therapeutic safety: (a) application within the context of a structured protocol, (b) training facilitated by instructors, (c) group practice, and (d) short-duration session practice (4 of the 9 studies specify that the duration of the practice sessions was between 10 and 30 minutes). A recent study has, in fact, found that these three factors—structured protocol, group, and short-duration practice—were associated with a lower occurrence of the unwanted side effects linked to mindfulness practice in the general population [65
The effects obtained from MBIs in this review were comparable to those obtained in meta-analytic studies regarding the effectiveness of CBT in the reduction of positive and negative symptoms [23
]. Nevertheless, the effectiveness of MBIs in the reduction of psychotic symptoms is noteworthy, given that the objective of these interventions is not to modify this type of symptoms—an aspect on which CBT does tend to focus [68
]—but rather to manage the psychological discomfort they may cause by modifying the relationship that the patient establishes with them. On the other hand, this review found an average dropout rate in MBIs of 15.6%, similar to the dropout rate reported in efficacy trials of CBT, such as the one conducted by Burns et al. [22
], where it averaged 14%. However, a recent meta-analysis on the efficacy of low-intensity CBT (<16 contact hours) reported a considerably lower dropout rate of 5.53% [23
]. Now, considering these results, in which both therapeutic approaches—MBIs and CBT—are similar in terms of efficacy and safety, the choice between one or the other could be based—as suggested in a previous review [70
]—on factors concerning their effectiveness in terms of cost and time. In both senses, MBIs tend to be highly efficient because they are applied to groups, unlike CBT which tends to be applied to individual patients [23
] as recommended by relevant clinical practice guidelines for the psychosocial treatment of schizophrenia [71
Despite the relevance of the described findings, they must be considered with caution, given the limitations of the present review. In five of the ten studies, we found certain aspects that affect their methodological quality (see Supplementary Materials, Table S2
) and which, therefore, threaten the validity of their findings. Additionally, consideration should be given to the high degree of heterogeneity that we found in the analyses and which constitutes a source of bias. This heterogeneity could be related to the differences between the applied MBI protocols. Although the applied MBIs shared mindfulness practice as the core element and guiding theme of the interventions and by extension the change of relationship with psychotic symptoms, they were heterogeneous regarding the psychoeducational contents which are, in fact, active ingredients that interfere with the results’ interpretation, so that they cannot be attributed to mindfulness with absolute certainty. On the other hand, the studies were conducted in a variety of geographic locations with sample participants from different ethnic backgrounds, which is also a considerable source of heterogeneity. In this regard, four of the reviewed studies [55
] used samples exclusively made up of recently diagnosed volunteers—in the last five years—which can affect the possibilities of generalizing the results—e.g., to include individuals who feel little motivation toward the intervention or who have suffered the illness for a longer period of time.