Study 1
Fishbein and colleagues (2015) conducted a randomized controlled trial to examine whether high-risk adolescents who participated in the Be BOLD Yoga Intervention demonstrated a decrease in substance use and subsequently experienced positive effects with regard to emotional and physical health. The study authors hypothesized that adolescents who participated in the yoga intervention would report a decrease in use of various substances (alcohol, marijuana, illicit drugs), and that these students would experience positive changes in cognitive and affective processes. The authors also theorized that participants in the treatment group would demonstrate improved psychophysiological responses to stress, reflected in heart rate and respiratory sinus arrhythmia.
This study took place in a mid-sized city at a nontraditional public school, which was for students who were at high risk of dropping out. These students had mostly been failing in traditional schools; were 1 to 2 years behind in credits; and struggled with absenteeism, academic issues, and personal/family problems. To be eligible for the study, adolescents had to be students at this school and older than 14. Their fitness level also had to be assessed. Students were not permitted to participate in the study if they had any health conditions that could increase the risk of side effects from the yoga intervention such as dizziness, heart trouble, or breathlessness. Students were also excluded from participating if their blood pressure exceeded 140/90 at rest.
The initial sample comprised 104 students, but 19 students dropped out of the study following the pretest. The final sample at posttest comprised 85 students. These 85 participants were randomly assigned to either the treatment group (n = 45), which received the yoga intervention, or the control group (n = 40), which carried on with care as usual and went to their regularly scheduled classes. There were three different cohorts of students recruited: one was from the middle school program from within the school, and two were from the high school program. This study took place across the following three periods: 1) fall of 2010, 2) spring of 2011, and 3) fall of 2011. Participant ages ranged from 14 to 20 years old (the average age was 16.7 years), and there were slightly more females in the sample (54 percent). In terms of race/ethnicity, 59 percent of the participants were African American, 17 percent were Hispanic, 9 percent were white, and 14 percent were multiracial.
The posttest was conducted after 20 sessions of the program had been completed. A variety of measures were used pre- and posttest to determine the effectiveness of the intervention. The Drug Use Screening Inventory–Revised (DUSI–R) was used to assess the extent of substance use, including marijuana, alcohol, and illicit drug use. The Abbreviated Dysregulation Inventory (ADI), a 31-item scale, was used to measure emotional, behavioral, and cognitive ability. The Brunel Mood Scale, a 24-item scale, was also used to assess negative moods across six different levels: 1) anger, 2) confusion, 3) depression, 4) fatigue, 5) tension, and 6) vigor. To measure coping skills, the researchers used the Response to Stress Questionnaire (RSQ), a 57-item measure that looks at both voluntary and involuntary coping skills. For this study, the researchers used the involuntary coping scale, which concentrates on responses such as disengagement, rumination, and emotional numbing. They also used the Five Facet Mindfulness Questionnaire (FFMQ), a 39-item scale that measures mindfulness across five dimensions: 1) observing, 2) describing, 3) acting with awareness, 4) non-judging of inner experience, and 5) non-reactivity. Finally, they used the Behavior Assessment Scale for Children (BASC-2) to measure externalizing behaviors and social competency. Teachers in the school rated each student pretest and posttest on this 46-item scale.
To determine the effectiveness of the yoga intervention, a multilevel model, equivalent to a two-time-point longitudinal growth model, was used. The principal estimate of the effect was the group (treatment versus control) by time (pretest to posttest). The models controlled for age, and age and gender were included as covariates. The study authors did not conduct subgroup analyses.