Mindfulness-Based Interventions for Parents: A Systematic Review of Target Groups, Effects, and Intervention Features

A preliminary search in PROSPERO showed no protocol or systematic review (published or in process) that aimed to synthesize the features that are most frequently used in effective MBIs for parents as well as the groups in which they are applied and the specific effects.

A systematic search of papers published in peer-reviewed journals before December 2023 was conducted in three electronic databases: PubMed, Web of Science, and Scopus. The search was conducted independently by the first and second authors (BC and AC) without language or study design restrictions. However, although the search was initially unrestricted, only papers written in English and randomized controlled trials (RCTs) were included. In all databases, the search consisted of combinations of the following terms:

[(mindful* OR meditation OR mindful parenting interventions OR mindful parenting programs OR mindful parenting OR mindfulness-based OR MBSR OR “mindfulness-based stress reduction” OR MBCT OR “mindfulness-based cognitive therapy” OR MBCBT OR “mindfulness-based cognitive behavioral therapy”) AND (parent* OR mother* OR father* OR maternal* OR paternal*) AND (programs OR intervention OR interventions OR treatment OR therapeutic OR outcomes OR findings OR efficacy OR effectiveness OR “randomized controlled trial” OR RCT OR “controlled clinical trial”)].

A secondary search was conducted by analyzing the references of the papers included in this study as well as systematic reviews/meta-analyses relevant to the research question.

To be included in this review, studies had to meet the following criteria based on PICOS (participants, interventions, comparators, outcomes, and study design). With regard to the participants, studies were included if the intervention was delivered to parents of children (0–18 years old). The participants could be mothers, fathers, or both parents (independent of their gender identity and sexual orientation). Adoptive/foster families and kinship caregivers were also included. Additionally, studies of parents and/or children with or without a clinical diagnosis or mental health professional referral (i.e., clinical and non-clinical samples; Burgdorf et al., 2019) were considered. Studies whose intervention included sessions where children were present were also included. Studies that delivered a parallel arm of the intervention for other populations (e.g., children and teachers) were excluded. With regard to the intervention format, individuals and/or groups as well as face-to-face or/and online interventions were included. Interventions that had components other than mindfulness-based were included, but parental interventions in which mindfulness-based components were not the core component of the intervention and mindfulness-informed interventions (e.g., DBT and ACT) were excluded. Regarding the control groups, all types were considered (i.e., participants receiving another or no treatment, treatment as usual [TAU], and waitlist control [WLC]). Studies that assessed parental psychology (e.g., parenting stress) and/or parenting outcomes (e.g., mindful parenting) were included. Studies that assessed children’s psychological outcomes (e.g., children’s behavioral problems) were also included. Studies in which parents' outcomes were measured only as mechanisms of change were included. Furthermore, only studies that described at least one of each of the following three types of features were included: dissemination, format, and content (themes, formal mediation practices in session, homework practice). Finally, with regard to the study design, only studies with a randomized controlled trial (RCT) design published in peer-reviewed journals were considered.

In the first phase of the screening, after the first author (BC) defined the search strategy, the first and second authors (BC and AC) conducted the paper search. Following the completion of the search in the three electronic databases, the list of references was sent to the Rayyan bibliographic reference management program, which was used to eliminate all duplicate papers. Based on the eligibility criteria, the first and second authors (BC and AC) reviewed the titles and abstracts of the studies. The authors employed a data codification form designed for this review, which included the name of the first author, year of publication, title, language, decision (include/exclude/unclear), and notes (doubts and study goals). The full text of any potentially relevant or ambiguous papers were obtained, and irrelevant papers were excluded. In the second phase, two authors (BC and AC) conducted a full and independent paper review. Disagreements were resolved through discussion between the two authors, and when consensus could not be reached, the third author (SA) and/or the last author (RP) were consulted to achieve a 100% agreement rate. Finally, the reference lists of the included papers were reviewed, and those that met the inclusion criteria were included. The final tables were reviewed by all authors.

To assess the risk of bias of each study included in this review, the Joanna Briggs Institute Critical Appraisal Tool for RCTs (Barker et al., 2023) was used. The domains of this tool included bias related to selection and allocation; administration of the intervention/exposure; assessment, detection, and measurement of the outcome; and participant retention. The risk of bias in the included papers was independently assessed by the first and second authors (BC and AC). Any disagreements were resolved through discussion. Interrater agreement was calculated with Cohen’s kappa coefficient, considering k < 0.00 as poor, k ≤ 0.20 as slight, k ≤ 0.40 as fair, k ≤ 0.60 as moderate, k ≤ 0.80 as substantial and k > 0.81 as almost perfect agreement (Landis & Koch, 1977).

A narrative analysis of the reported study findings was conducted following PRISMA guidelines (Page et al., 2021). Each included study was synthesized using the previously described data-collection form developed by the authors of this review. In some cases, the authors contacted the researchers of the studies for further information. Even in the absence of a response, the studies were included in the review if they met the inclusion criteria. Due to large variation in the participant samples and in the outcome measures of the included studies as well as the lack of fidelity of existing research, it was deemed inappropriate to perform a meta-analysis.

Figure 1 shows the process of study selection and exclusion. The database searches identified 10,014 studies. After duplicates were removed, a total of 8,194 records were screened. Based on the title and abstract screening, 105 studies were retained, and their full text was retrieved and assessed for eligibility. Of these, 85 studies were excluded for the reasons shown in Fig. 1. After the eligibility assessment, we identified 20 papers published in peer-reviewed journals that we considered eligible for this systematic review. These 20 studies included 15 different interventions since Chaplin et al. (2021a) and Chaplin et al. (2021b) described the same intervention as the studies of Neece (2014), Neece et al. (2023), and McGregor et al. (2020). However, these studies described the same intervention in different samples and with different therapeutic goals, so they were considered separately in the current review. The interrater agreement for the selection of papers was almost perfect (k = 0.840, p < 0.001).

Table S1 (Supplementary Material) presents the risk of bias assessments for each included study. With regard to internal validity, 13 studies presented a moderate to high risk of bias related to selection and allocation. Most of the studies (n = 12) did not have an allocation concealed to groups. All studies presented a high risk for bias related to the administration of the intervention/exposure, and 12 studies showed a moderate to high risk of bias related to the assessment, detection, and measurement of the outcome. The risk of bias related to participant retention was low across all studies. With regard to statistical validity, only four studies presented a moderate risk, while the others (n = 16) presented a low risk. Since the risk of bias in general is influenced solely by internal validity (Barker et al., 2023), the studies were generally rated as having a moderate to high risk of bias. Finally, the interrater agreement for the risk of bias assessment of the studies was almost perfect (k = 0.889, p < 0.001).

As shown in Table 1, the studies were conducted between 2013 and 2023 in a range of countries, with studies from the U.S. being most common (n = 6) (e.g., Chaplin et al., 2021a; Dykens et al., 2014; Neece et al., 2023). With regard to the study population, most studies (n = 11) included only mothers in their sample. In three of these studies (Chaplin et al., 2021a, b; Dykens et al., 2014), parental dyads were included in the intervention; however, only data from mothers were analyzed. In the studies in which fathers were included (n = 8), most of the participants were mothers. The average sample size of the included studies was 102 parents, ranging from 15 (Ferraioli & Harris, 2013) to 292 (Fernandes et al., 2022b). With regard to the parents’ mental health status, only one study included a clinical sample (i.e., recurrent depression) (Mann et al., 2016). Within the non-clinical samples, the inclusion criteria of seven studies required parents to present moderate to high levels of parenting stress. Regarding the children of the target parents, in most studies, the sample was also non-clinical (n = 12). In studies with clinical samples, children were diagnosed with attention deficit hyperactivity disorder (n = 2; Behbahani et al., 2018; Liu et al., 2021), developmental disabilities (i.e., autism spectrum disorders; n = 6; e.g., Dykens et al., 2014; Ferraioli & Harris, 2013), or substance abuse (n = 1; Chaplin et al., 2021a).

The types of control groups included no treatment (n = 3) (e.g., mindfulness workshop for control group parents who were interested in mindfulness, after their participation in the study; Lo et al., 2017), treatment as usual (n = 5) (e.g., receiving their usual mental health services, including regular visits to a child psychiatrist; Liu et al., 2021), waitlist control (n = 7), and receiving another intervention (n = 5) (e.g., Parent Education Program; Chaplin et al., 2021a, b). In three of the included studies, there were no dropouts in either the intervention or control groups (e.g., Kakhki et al., 2022). Among the studies with dropouts, most presented the rates separately for the intervention group and the control group (n = 12). In the intervention group, the dropout rate ranged from 0% (Behbahani et al., 2018; Perez-Blasco et al., 2013) to 40% (Ferraioli & Harris, 2013), while in the control group, it ranged from 0% (Burgdorf et al., 2022) to 38.5% (Perez-Blasco et al., 2013).

The narrative analysis allowed us to systematize several intervention features of the included studies.

This study presents some limitations that should be considered in the interpretation of the results. First, the results regarding the intervention’ effects should be interpreted with caution due to the methodological quality of the RCTs; that is, all studies were judged to have at least a moderate to high risk of bias. This was due to the lack of randomization as well as the subjective reporting of most outcomes in each study. High-quality RCTs should be conducted to enhance evidence of the beneficial effects of MBIs for parents. Second, studies relied mostly on self-report and hetero-report measures from parents about children. It would be helpful to integrate multi-informant (including both parents) and multi-method assessments to reduce measurement errors. Third, it may be possible that publication bias was established by excluding grey literature and publications in languages other than English. Moreover, it was decided that performing a meta-analysis was unsuitable due to the significant diversity in the included studies' participant samples and outcome measures as well as the lack of fidelity of existing research.

Despite these limitations, this study has important strengths. First, MBIs for parents seem to have effects in both clinical and non-clinical groups, which reinforces the preventive potential of these interventions. Since mindfulness was a common factor in all interventions, this leads us to emphasize the importance of mindfulness as a psychological mechanism and the need to practice it. Furthermore, almost all studies assessed the effects of the intervention on parental stress levels. However, we also observed that the only intervention in which there was no reduction in parental stress levels (Cheung et al., 2022) was the one that included only formal meditation and psychoeducation about mindfulness as content. This leads us to hypothesize that although mindfulness is the core component of MBIs, for MBIs to have an effect on parental stress levels, they should address other types of activities (e.g., informal meditation, perspective-taking activities). Additionally, there was no facilitator in the study that showed no reduction in parental stress, and the duration and frequency of the sessions were very different from those of all other interventions. In the future, it could be useful to conduct a meta-analysis to understand whether there are differences in the effectiveness of interventions based on the number of facilitators or their level of training as well as whether there are other intervention features that may interfere with effectiveness, such as duration, frequency, and content. Conducting a meta-analysis with these characteristics could help to clarify the results of the current study. Based on our mapping of the target groups, the effects, and the features of the interventions, we concluded that regardless of the features of the intervention and the target groups, the effects were similar. We also verified that there were a variety of curricula for the same target group. Although all of these factors appeared to have effects on the outcomes measured, all of the studies presented a moderate to high risk of bias. It is important to perform a meta-analysis to determine whether it is feasible to develop a single, extensively validated umbrella program that can be feasibly utilized for parents of children with various needs. These findings could guide us to develop more effective interventions adapted to other target groups for which these interventions have not yet been applied (e.g., adoptive or socially disadvantaged families). Future studies should include both parents (i.e., the parental dyad, regardless of gender) in the intervention so that the effects of the intervention on the outcomes of both parents can be analyzed. To do so, it is necessary to assess both parents’ preferences for the dissemination, format, and content features of interventions. By taking this information into account, it would be possible to develop interventions tailored to the preferences of both fathers and mothers and to retain both parents in the intervention, which would contribute to the construction of more effective interventions. Finally, by helping to develop more effective MBIs for parents, this systematic review can inform the development of mental health policies for both parents and children in multiple parenting contexts.