pediatrics
January 2016, VOLUME137 /ISSUE 1

School-Based Mindfulness Instruction: An RCT

  1. Erica M.S. Sibinga, MD, MHSa,
  2. Lindsey Webb, MSa,
  3. Sharon R. Ghazarian, PhDa, and
  4. Jonathan M. Ellen, MDa,b
  1. aDivision of General Pediatrics and Adolescent Medicine, Department of Pediatrics, Johns Hopkins School of Medicine, and
  2. bAll Children’s Hospital Johns Hopkins Medicine, Saint Petersburg, Florida
  1. Dr Sibinga conceptualized, designed, and oversaw study procedures, including data collection, analysis, and interpretation, and drafted and revised the manuscript; Ms Webb assisted with and coordinated data collection, conducted data analysis, and contributed to the manuscript; Dr Ghazarian contributed to the study design and data collection procedures, conducted data analysis and interpretation, and revised the manuscript; Dr Ellen contributed substantively to the study design and procedures and to data interpretation and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted.

Abstract

BACKGROUND AND OBJECTIVE: Many urban youth experiencesignificant and unremitting negative stressors, including those associated with community violence, multigenerational poverty, failing educational systems, substance use, limited avenues for success, health risks, and trauma. Mindfulness instruction improves psychological functioning in a variety of adult populations; research on mindfulness for youth is promising, but has been conducted in limited populations. Informed by implementation science, we evaluated an adapted mindfulness-based stress reduction (MBSR) program to ameliorate the negative effects of stress and trauma among low-income, minority, middle school public school students.

METHODS: Participants were students at two Baltimore City Public Schools who were randomly assigned by grade to receive adapted MBSR or health education (Healthy Topics [HT]) programs. Self-report survey data were collected at baseline and postprogram. Deidentified data were analyzed in the aggregate, comparing MBSR and HT classes, by using regression modeling.

RESULTS: Three hundred fifth- to eighth-grade students (mean 12.0 years) were in MBSR and HT classes and provided survey data. Participants were 50.7% female, 99.7% African American, and 99% eligible for free lunch. The groups were comparable at baseline. Postprogram, MBSR students had significantly lower levels of somatization, depression, negative affect, negative coping, rumination, self-hostility, and posttraumatic symptom severity (all Ps < .05) than HT.

CONCLUSIONS: These findings support the hypothesis that mindfulness instruction improves psychological functioning and may ameliorate the negative effects of stress and reduce trauma-associated symptoms among vulnerable urban middle school students. Additional research is needed to explore psychological, social, and behavioral outcomes, and mechanisms of mindfulness instruction.

  • Abbreviations:
    CDI-S
    Children’s Depression Inventory–short form
    CFA
    confirmatory factor analysis
    CPSS
    Children’s Post-Traumatic Symptom Severity Checklist
    CRSQ
    Children’s Response Style Questionnaire
    CSE
    Coping Self-Efficacy Scale
    EFA
    exploratory factor analysis
    HT
    Healthy Topics
    MASC
    Multidimensional Anxiety Scale for Children
    MBSR
    mindfulness-based stress reduction
    PANAS
    Positive and Negative Affect Schedule
    RCT
    randomized controlled trial
    SCL-90-R
    Symptom Checklist-90-R
    STAXI
    State-Trait Anger Expression Inventory
  • What’s Known on This Subject:

    Many urban youth experience significant negative stressors. Mindfulness instruction reduces stress and improves psychological functioning in adults. Research in youth is promising but has been conducted in limited populations. Little is known about mindfulness instruction for low-income minority students.

    What This Study Adds:

    In this randomized active-controlled trial, school-based mindfulness instruction led to improved psychological functioning and lower levels of posttraumatic stress symptoms. High-quality mindfulness instruction merits consideration as primary prevention for mental and behavioral health problems in low-income, minority urban students.

    Recent violence in Baltimore City has drawn attention to the acute and chronic problems experienced by many of the city’s poor and minority residents. Due to effects of multigenerational poverty, limited educational and economic opportunities, high levels of drug use and trade, and pervasive community violence, urban youth in Baltimore and many US cities are at increased risk for exposure to a variety of stresses, including early life stress, recurrent and chronic stress, and exposure to significant and/or recurrent traumas. As an example, it is estimated that 50% to 96% of urban youth directly witness violence within their community.1 The significant, recurrent, and chronic nature of these stressors may overwhelm the capacity to cope acutely and chronically,2 which is required for healthy development and positive trajectories. Any or all of these forms of stress may contribute to the state of toxic stress in which an individual’s ability to manage or cope with stress is overwhelmed on an ongoing basis. Pediatricians have been called to action to understand the complex and intertwined systems that are disrupted by stress,3 as well as to recognize that effective health approaches to mitigate the negative effects of toxic stress and trauma may be interventions that occur in the community, not only in medical settings.4

    In response to the pervasive exposure to stress and trauma, we have developed and previously tested a program of mindfulness instruction intended to reduce the negative impacts of stress and toxic stress among urban youth. Based on Jon Kabat-Zinn’s well-studied mindfulness-based stress reduction (MBSR) program for adults,5,6 we had previously adapted and evaluated the MBSR program for use with urban youth in clinic79 and school settings.10 In particular, we conducted a randomized controlled trial (RCT) of our school-based 12-week MBSR program compared with a 12-week active control (health education) program at a small middle school for city-dwelling boys. Compared with the active control program, MBSR participants had significantly less anxiety, improved coping, and borderline attenuation of the salivary cortisol increase associated with the academic term.10 We were interested in expanding the program to urban public schools with significant need. Through collaboration with Elev8 Baltimore, a nonprofit organization aimed at optimizing linkage and utilization of school-based activities for the benefit of students and their families, we found the complementary elements for implementation. Elev8 Baltimore had experience with and presence in the schools and understood firsthand the high level of students’ exposure to trauma and toxic stress.

    Implementation science has elucidated a number of essential considerations for program expansion. The consolidated framework for implementation research articulates the importance of the following elements: evidence-based intervention, the outer setting, the inner setting, the individuals involved, and the process.11,12 Attending to these elements of implementation and in collaboration with Elev8 Baltimore, we used the evidence-based MBSR program with two Elev8 Baltimore schools in which we could ensure high potential for student benefit, site-specific awareness of school culture, school administration buy-in, high-quality program instruction, high-quality program implementation staff, community partnership, infrastructure to support ongoing collaboration with school and Elev8 Baltimore, and structure and process for evaluation of implementation and program effect.

    Methods

    Participants and Study Design

    Students were eligible to participate in the study if they attended fifth through eighth grades, including special education, in either of two Elev8 Baltimore schools selected for this trial, during the 2012–2013 academic year. Conceptualized and presented as two complementary programs focused on overall student wellness, Healthy Topics (HT) and MBSR were incorporated into the school curriculum and delivered to all students during the “resource” class time for a portion of the school year, instead of classes such as advisory, art, or music. Students were randomly assigned by school and grade into either the intervention program or the active control program; thus, each school had both programs. Blinding to group assignment occurred at the data management, analysis, and interpretation levels. The study was approved by the Johns Hopkins School of Medicine Institutional Review Board.

    Intervention Program

    Our 12-week program is adapted from MBSR, a structured 8-week program of instruction in the cultivation of mindfulness, a practice of purposeful nonjudgmental attention to the happenings of the present moment.5 MBSR programs consist of 3 components: (1) didactic material related to mindfulness, meditation, yoga, and the mind-body connection; (2) experiential practice of various mindfulness meditations, mindful yoga, and body awareness during group meetings and encouragement of home practice; and (3) group discussion focused on the application of mindfulness to everyday situations and problem-solving related to barriers to effective practice.5,13,14 The MBSR program includes a number of formal and informal techniques, all of which share the goal of enhancing nonjudgmental present-focused awareness, aimed to reduce dysregulated focus on the past (ie, rumination) and worries about the future (ie, anxiety).

    Our previously-adapted 12-week, school-based MBSR curriculum10 was taught by 2 experienced MBSR instructors, both with long-standing personal meditation practices, >10 years’ experience teaching mindfulness, and MBSR instructor training through the University of Massachusetts Center for Mindfulness. The MBSR instructors met regularly to optimize program implementation. The adapted MBSR program remains consistent with typical MBSR programs for adults and with MBSR core content.13 In addition, the content, course structure, sequence of content presentation, and core mindfulness practices are maintained. Students attended an average of ∼80% (74%–85%) of the program sessions and anecdotally reported especially enjoying the mind jar and worry box activities.

    Active Control Program

    HT is an age-appropriate general health program structured to match MBSR, including number and length of classes, location, didactic and experiential instruction, and group size. The HT program covers age-appropriate topics such as nutrition, exercise, body systems, adolescence, and puberty. HT was adapted from the Glencoe Health Curriculum (McGraw Hill, 2005) and is designed to control for the effects of a positive adult instructor, peer group experience, attention, and time. It was taught by 3 trained, experienced health instructors. Students attended an average of ∼80% (76%–88%) of the program sessions.

    Implementation

    Our community partner, Elev8 Baltimore, had 3 years of experience with school administration and staff at both schools. In discussion with Elev8, school administrators agreed to incorporate MBSR and HT programs into the school curriculum and work with the program staff to facilitate scheduling and implementation. Program staff worked closely with Elev8 staff and program instructors to enhance program implementation, school support, and the community partnership. Both schools offered other programs to middle school students focused on mentorship, homework assistance, and after-school activities, but no other programs directly targeted stress.

    Measures

    Data were collected by program staff (not instructors) during class time over 2 sequential days at baseline and after completion of the 12-week programs. Makeup data collection for absent students was conducted on a case-by-case basis over the following few days. Valid, reliable, age-appropriate self-report surveys were used to measure outcomes of interest, including mindfulness, psychological functioning, and trauma symptoms.

    Mindfulness was measured with the 10-item Children’s Acceptance and Mindfulness Measure,1517 which had good reliability in our sample (α = 0.74 and 0.73). The widely used Perceived Stress Scale18 showed low reliability within our sample (α = 0.41 and 0.29); by using exploratory factor analysis (EFA) methods, we extracted 2 factors: positive/coping (4 items) and stress (6 items), which had improved reliability (α = 0.64–0.75).

    Psychological symptoms measured were depressive symptoms (Children’s Depression Inventory—Short Form [CDI-S]19); paranoid ideation, hostility, somatization, by using the Symptom Checklist-90-R (SCL-90-R)20; and anxiety (Multidimensional Anxiety Scale for Children [MASC]).21 The CDI-S had good reliability within our sample (α = 0.84). The SCL-90-R showed acceptable to excellent reliability in our sample (hostility α = 0.86 and 0.85; somatization α = 0.87 and 0.90; paranoid ideation α = 0.67 and 0.71). The MASC is a measure of anxiety, with excellent reliability in our sample (α = 0.89).

    We assessed mood and emotion regulation with the following measures: Positive and Negative Affect Schedule (PANAS),22 the Differential Emotions Scale (DES),23 the Aggression scale,24 and the State-Trait Anger Expression Inventory (STAXI-2).25 The PANAS yields 2 factors: positive affect and negative affect, with good reliability in our sample (positive affect α = 0.81 and 0.89; negative affect α = 0.84 and 0.87). The DES yields several factors found to have acceptable reliability and validity in youth26 and adequate reliability in our sample (α = 0.61–0.81). The Aggression Scale has sufficient reliability and validity in similar populations27 and excellent reliability in our sample (α = 0.92). The STAXI is a measure of anger expressivity, from which we used 2 subscales: temperamental expressivity and reactive expressivity. The measure has been found to have adequate reliability and validity in African American youth28 and was reliable with our sample (α = 0.78–0.80).

    Coping was measured by using the Children’s Response Style Questionnaire (CRSQ),29 the Brief COPE,30 and the Coping Self-Efficacy Scale (CSE).31 The CRSQ measures 3 types of reactions: rumination, problem solving, and distraction with adequate reliability (rumination α = 0.86 and 0.87; problem solving α = 0.68 and 0.73; distraction α = 0.67 and 0.70). The Brief COPE measures 14 coping approaches with 2 items each. Given the many subdomains, confirmatory factor analysis (CFA) method was used to identify typically positive coping approaches (16 items: use of instrumental support, active coping, distraction, venting, positive humor, use of emotional support, positive reframing, planning, and religion) and typically negative coping approaches (9 items: behavioral disengagement, denial, substance abuse, negative humor, and self-blame). The factors demonstrated adequate reliability (α = 0.73–0.88). Finally, the CSE creates an overall variable of coping self-efficacy with excellent reliability in this sample (α = 0.96).

    We measured posttraumatic symptoms by using the Children’s Post-Traumatic Symptom Severity Checklist (CPSS).32 Unfortunately, an administrative error led to the inadvertent omission of the CPSS from the baseline survey; we have CPSS data from postprogram only. Because all other measures were comparable between groups at baseline, we infer that CPSS was as well. The measure has adequate reliability and validity in socioeconomically disadvantaged youth33 and excellent reliability in our sample (α = 0.94). Furthermore, 2 factors were generated from the CPSS: depressive symptoms (10 items) and reexperiencing symptoms (6 items). Both factors were found to have good reliability (depressive α = 0.92, reexperiencing α = 0.85).

    Sample Size

    A priori power calculations demonstrated power >80% with a sample of at least N = 90 based on previous work with small to moderate effects (β ranging from 0.38 to 0.51; ΔR2 ranging from 0.04 to 0.54) for associations between MBSR participation and outcomes of coping and psychological symptoms. Thus, ample power was present for this study.

    Data Analysis

    Preliminary data analysis included EFA and CFA to determine the most appropriate factor structure for each measure. EFA employed principal components analysis methods with varimax rotation. CFA methods used maximum likelihood extraction methods with varimax rotation and were used for all measures to examine the factor structure. EFA methods were used when CFA results suggested potential deviation from expected factor structures among established scales. Where applicable, EFA and CFA results were presented in the Measures section above. Preliminary analyses examined descriptive statistics (eg, means, median, proportions) for all study variables of interest. Given that data collection at both time points was held over 2 class periods, missing data due to attrition/absence occurred. Although we provided classes to 400 students at 2 schools, baseline data were collected from 300 students, with subsequent data collection sessions ranging from 292 to 300 students (72.8%–74.8%). Missing data analyses demonstrated that missingness was due largely to attrition/absence and did not present in an obvious pattern. Given that data were collected at the aggregate level (not linked by individual student over time), it was not possible to examine participant characteristics in relation to missing data over the course of the study. Initial analyses examined potential differences in participant characteristics and demographics across intervention groups at baseline by using independent samples t tests for continuous variables and χ2 analyses for categorical variables. To test the overall potential intervention effect, multivariate linear regression models were examined. A binary grouping variable (intervention versus control) was the main predictor with outcome variables of interest at follow-up. All models included gender, age, and school as covariates. Each outcome variable of interest was examined in a separate model due to collinearity. Given the data structure of students within grades and within schools, mixed effects models were examined with random effects for grade. The random effect for grade was not significant in any models and was thus removed from further analyses. We hypothesized that MBSR would improve participants’ psychological functioning.

    Results

    Three hundred students in the fifth through eighth grades participated and provided survey data from 2 urban elementary/middle schools. Study participants were 50.7% female and 99.7% African American (Table 1). Approximately 99% of participants were eligible for free or reduced meals. When comparing HT and MBSR groups at baseline, χ2 tests revealed no significant differences in gender (P = .99), ethnicity (P = .32), age (P = .45), nor any study variables of interest (P = .10–.99). The lack of significant differences at baseline suggests that randomization resulted in balanced study arms.

    TABLE 1

    Baseline Characteristics, n (%)

    On the basis of our randomization scheme, we conducted 14 separate classes of 21 to 37 students each at 2 schools. Multivariate models demonstrated significant differences between MBSR and HT program participants after implementation of the 12-week programs; compared with HT, students who had participated in the MBSR program showed better psychological functioning and coping. As shown in Table 2, MBSR participants reported lower levels of depressive symptoms (β = –0.16, P = .02), self-hostility (β = –0.14, P = .02), somatization (β = –0.13, P = .03), negative affect (β = –0.19, P = .003), negative coping (β= –0.13, P = .04), and rumination (β = –0.13, P = .03). Importantly, MBSR students also showed significantly lower levels of posttraumatic stress symptoms (β = –0.15, P = .02), including in both subdomains of depressive (β = –0.13, P = .03) and reexperiencing (β = −0.17, P = .008) symptoms. No significant adverse events were reported.

    TABLE 2

    Multivariate Postprogram Comparisons

    Discussion

    In our school-based RCT of an adapted MBSR program compared with an active control (HT) for vulnerable urban youth, MBSR participants showed significant improvements in psychological symptoms, coping, and a reduction in posttraumatic symptoms. In particular, they had lower levels of depressive symptoms, self-hostility, somatization, negative mood, negative coping approaches, and posttraumatic symptoms. This rigorous school-based RCT, informed by implementation science, supports the hypothesis that the MBSR program is effective primary prevention for the negative effects of toxic stress and trauma, and ultimately beneficial for urban youth.

    Research on mindfulness for children and youth is beginning to emerge,34 but few randomized active-controlled trials of school-based mindfulness instruction exist in the literature and populations are limited. Britton conducted a RCT in a private independent school comparing a mindfulness meditation class with an active control (n = 101), finding that mindfulness was associated with reductions in thoughts of self-harm.35 A recent trial of a mindfulness program compared with a social responsibility control in a mostly middle-class population showed that mindfulness led to positive outcomes in psychological symptoms, cognitive control, interpersonal outcomes, and stress physiology.36 Although these findings are promising, little information is available on mindfulness instruction for low-income, urban, minority populations.

    Important aspects of implementation were addressed in this trial: an evidence-based program; Elev8 Baltimore’s school-based staff and experience with each school’s students, staff, and administration; dedicated staff to coordinate program implementation and data collection and link with teachers and school staff; dedicated program instructors; and reasonably accepting administration. Furthermore, our RCT used specific strategies to ensure the delivery of high-quality program content and process in both study arms: instructors were trained, experienced, and dedicated mindfulness and health education instructors who brought expertise regarding program content and delivery; the mindfulness curriculum was previously tested and had been adapted from the evidence-based MBSR program8,10; communication between program and Elev8 staff was supported by weekly meetings; and the community partnership functioned effectively to navigate challenges that arose related to logistics, school administration, and implementation.

    There are a number of limitations of this study. These include variability of student session engagement and attendance, no information regarding outside student mindfulness exposure and/or practice, missing data, variability in school administration support for programs, and variability in classroom teacher support for programs.

    However, the trial also has a number of notable strengths. The RCT study design with an active comparison group (controlling for positive adult instructor and group activity) and comparable groups at baseline provides a high level of confidence that the improvements seen in the MBSR arm are due specifically to the mindfulness aspects of the intervention, as opposed to baseline differences and/or other nonspecific intervention effects. Furthermore, because this is a group-based prevention program, program costs may well be offset by decreased need for behavioral and mental health interventions.

    Given the unmanageable toxic stress and trauma experienced by many urban youth, these findings are important and timely. Efforts to improve the circumstances in which urban youth live are essential and impactful. This study provides additional support for efforts to include high-quality mindfulness instruction to enhance students’ capacity to manage the inevitable stress and trauma they will face, as well as in trauma-informed approaches. Depressive and posttraumatic stress symptoms have been linked with impaired academic performance and attendance.37,38 By providing high-quality mindfulness instruction during childhood, improvements in psychological symptoms, coping, and posttraumatic symptoms have the potential to shift life trajectories in meaningful ways, including academic performance, mental and physical health, and quality of life.

    Conclusions

    As we continue to learn that many adult diseases have their roots in childhood exposure to stress and trauma, it is essential to intervene with primary prevention strategies to reduce their negative effects among children and youth. This trial provides convincing evidence that high-quality school-based MBSR instruction for youth in urban public schools is feasible, acceptable, and leads to improvements in psychological symptoms, coping, and posttraumatic stress symptoms. Improvements in these domains may ultimately reduce the negative impact of stress and trauma experienced in childhood and adolescence and lead to significant positive shifts, when imagined over the life course.

    Acknowledgments

    We thank Nicole A. Johnson and Elev8 Baltimore for seeing the potential of this work to reduce student suffering, as well as the school administration and teachers, participating students, program instructors, and program staff.

    Footnotes

      • Accepted September 30, 2015.
    • Address correspondence to Erica M.S. Sibinga, MD, MHS, Johns Hopkins School of Medicine, Center for Child and Community Health Research, Suite 4200, Mason F. Lord Building, Center Tower, 5200 Eastern Ave, Baltimore, MD 21224. E-mail: esibinga{at}jhmi.edu
    • This trial has been registered at www.clinicaltrials.gov (identifier NCT02493218).

    • FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

    • FUNDING: Supported by Elev8 Baltimore.

    • POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

    References