A randomized controlled evaluation of a secondary school mindfulness program for early adolescents: Do we have the recipe right yet?

Highlights

• We investigated the .b mindfulness program for a second time in early adolescents.

• We tightened adherence to the manualised curriculum.

• Parental involvement was added in one arm of the RCT design.

• We found no differences between the mindfulness groups with/without parental involvement and the control group.

• Further research is required to identify optimal age and content of school-based mindfulness programs.

Abstract

Objective

Mindfulness is being promoted in schools as a prevention program despite a current small evidence base. The aim of this research was to conduct a rigorous evaluation of the .b (“Dot be”) mindfulness curriculum, with or without parental involvement, compared to a control condition.

Method

In a randomized controlled design, students (M_age 13.44, SD 0.33; 45.4% female) across a broad range of socioeconomic indicators received the nine lesson curriculum delivered by an external facilitator with (N = 191) or without (N = 186) parental involvement, or were allocated to a usual curriculum control group (N = 178). Self-report outcome measures were anxiety, depression, weight/shape concerns, wellbeing and mindfulness.

Results

There were no differences in outcomes between any of the three groups at post-intervention, six or twelve month follow-up. Between-group effect sizes (Cohen's d) across the variables ranged from 0.002 to 0.37. A wide range of moderators were examined but none impacted outcome.

Conclusions

Further research is required to identify the optimal age, content and length of mindfulness programs for adolescents in universal prevention settings.

Trial registration

ACTRN12615001052527.

Introduction

Mindfulness presents as a promising transdiagnostic approach for mental health disorders, given its potential to counteract a number of shared risk factors for anxiety, depression and eating disorders (Johnson, Burke, Brinkman, & Wade, 2016a). Robust evidence exists in adults for the benefits of mindfulness-based interventions (MBIs) across this group of pathologies (Khoury et al., 2013). More recently, MBIs have been enthusiastically embraced in schools and are widely disseminated (Semple, Droutman, & Reid, 2017), but there are insufficient methodologically robust studies to make definitive conclusions about efficacy.

In mainstream secondary schools, only three large randomized controlled trials (RCTs) of MBIs have been conducted. Raes, Griffith, Van der Gucht, and Williams (2014) tested an 8-week MBCT-informed curriculum (N = 408, M_age 15.4 years; mixed sex; external facilitator) finding improvements in depression at post intervention and 6-month follow-up (Cohen's d ≥ 0.25). Atkinson and Wade (2015) investigated a 3-session mindfulness intervention with a body image focus (N = 347, M_age 15.7 years; female; external facilitator), with improvements across a range of eating disorder risk factors at 6 months (d ≥ 0.47), but no improvements in negative affect. A third study evaluated the manualized .b (“Dot be”) Mindfulness in Schools curriculum, which had previously shown promising results in a controlled study (Kuyken et al., 2013; N = 522, M_age 14.8 years, mixed sex, class teacher delivery), demonstrating reductions at 3 months for depression, stress and wellbeing (d ≥ 0.25). The replication RCT (Johnson et al., 2016a; N = 308, M_age 13.6 years, mixed sex, external facilitator) showed no improvements across a wide range of outcomes at post-intervention or 3-month follow-up (d < 0.28).

Several hypotheses for the lack of replication of the .b curriculum exist. First, that the ideal dosage or active ingredients necessary to successfully translate adult MBIs for youth remain unknown. Second, although an early adolescent group was deliberately targeted, prior to the escalating stressors of mid-late adolescence (Kuyken et al., 2013), it may be that older adolescents respond better. Third, inadequate program adherence in the replication trial may have impacted results i.e., the curriculum was shortened by one lesson, students were not supplied with a user friendly version of the home practice manual, and an external facilitator was used (Johnson et al., 2016a). Therefore, the main aim of the current study was to conduct a tighter replication of the .b curriculum. A secondary aim was to test whether increased “dose” might be achieved by inviting parents to take part in the intervention, to stimulate discussion of mindfulness at home together and remind students to do home practice. Three small controlled trials of MBIs (Bögels et al., 2008, Semple et al., 2010, van der Oord et al., 2012) have included parents in MBIs for children, evidencing medium to large effect size improvements in attention, behavior problems and anxiety in these clinical samples. However, there have been no experimental comparisons that isolate the effect of parental involvement, nor has this been tested in community samples. We predicted that our outcome measures would show improvement at 12 month follow-up (the longest to date in a youth MBI study) in the mindfulness group with parental involvement compared to the mindfulness group without, due to higher levels of home practice compliance, and that both of these groups would show improvement compared to the control group.