PSYCHOLOGICAL BENEFITS OF YOGA ON FEMALE PRISONER 1

Psychological Benefits of Trauma-Focused Yoga Intervention on Female Prisoner Population

Yoika Danielly

University of San Francisco

Author Note

Special thanks to James Fox, director of the Prison Yoga Project (PYP) for all his support and assistance, and to Angie Still, PYP’s volunteer yoga instructor who taught the classes for this study. My gratitude to Dr. Colin Silverthorne, Department of Psychology, University of San Francisco, for his guidance and edits.

Abstract

Until recently, practitioners of Yoga have suggested it has many benefits but little research has been conducted to test these possible benefits. For this research, as part of the Prison Yoga Project, female inmates at two correctional facilities in South Carolina served as subjects. Inmates were selected from those who applied to be allowed to participate in a ten-week trauma focused yoga program. To create control and experimental groups, inmates who requested to participate in the yoga class were randomly assigned to either be in the class (Experimental group, n = 33) or placed on a waitlist (Control Group, n =17). Inmates on the waitlist joined the next class so all who wanted could participate. A variety of psychological measures were tested and data was collected from both groups before the class began and again at the end, ten weeks later. To assess the changes from preintervention to postintervention, mixed design ANOVA’s tests were conducted. Inmates in the yoga group reported a significant decrease in depression (p <.05) and stress (p <.01) and improved self awareness (p <.02). No significant effect was reported on anxiety, rumination and self-control in the treatment group. Although not substantial, anxiety scores did decrease and self-control scores improved for the yoga group. Inmates in the control group reported a worsening or no change on these measures. Rumination stayed about the same for both groups. The results suggest that Yoga is an inexpensive intervention that could benefit both inmates and prison staff by reducing negative behaviors and possibly mental health problems. Future studies should include male participants and measure the effect of a longer yoga program.

The Effects of Trauma-Focused Yoga Intervention on Female Prisoner Population

The United States comprises 5% of the world population, yet it has 25% of the total prisoners in the world (APA, 2014). During the past four decades, the number of incarcerations has increased dramatically (Guerino, Harrison & Sabol, 2011). This change has been largely attributed to the closing of mental asylums in the 1960s, which left the mentally ill residents out on the street with no community resources to treat them. Instead of getting treatment within the mental health system, many people are now caught up in the criminal justice system; leading to prisons having many inmates who are mentally ill. Another factor that contributed to the increase in incarcerations was the War on Drugs which began in the early 70s and shifted the US policies for treating illicit drug use and distribution to a classification of major criminal acts (Golembeski & Fullilove, 2005). In addition, harsher sentencing laws, such as mandatory minimum sentences and the three strikes law which ended discretion in sentencing of the judicial system for certain offenses, also contributed to the overcrowding of prisons.

Female inmates have typically been the minority population in correctional facilities. However, women are currently outpacing men as the fastest growing prison population (Messina, Burdon, Hagopian & Prendergast, 2006; Drapaski, Yuman, Stuewig & Tangney, 2009; Sabol, Minton & Harrison, 2007). This shift has encouraged researchers to look at the factors related to helping women in the criminal justice system, but the results have not necessarily translated to interventions tailored directly to the particular needs of females (Covington, 1998). The vast majority of female incarcerations are due to non-violent offenses usually related to alcohol, other drugs or property crimes (Chesney-Lind & Bloom, 1997; Watterson, 1996). Another difference with female inmates is that unlike their male counterparts who often deal with their anxiety by working out, women tend to fear physical exposure through exercise, and instead cope with their anxiety by oversleeping, eating or using prescription pills (LeBlanc, 1996). A large number of female inmates also report having experienced emotional, physical and/or sexual abuse in their lives prior to incarceration (Byrd & Davis, 2009; Wolff, Blitz, & Shi, 2007; Wolff & Knight, 2007). Studies have shown that early exposure to trauma can cause an individual to engage in high-risk behavior and contribute to delinquency, substance abuse and criminality among women (Harris, Putnam, & Fairbank, 2004; Ullman et al., 2005). While being incarcerated is a trauma in and of itself, investigations conducted by the Human Rights Watch and Amnesty International found that women were often subjected to sexual abuse, rape and/or unwarranted voyeurism to supervision by male correctional officers, verbal abuse, and use of threat or force by an abuser which they could not avoid (Thomas, 1996; Amnesty International, 1999). The actual extent of abuse is probably a lot higher than is reported given that in institutional settings abuse is even less likely to be reported by women than in communities due to fear of retaliation (Smith, 2001).

Some individuals who have experienced trauma develop maladaptive coping strategies such as distraction and avoidance, which may serve to minimize the traumatic stress in the moment, but also may prevent them from processing the traumatic experience and leaving post traumatic symptoms (Whittlesey, 1999). Another maladaptive coping skill is that of disassociation from the body, especially for those individuals who have experienced physical and sexual abuse (Dale et al., 2011)). When the individual no longer pays attention to their body, they may disregard somatic symptoms that can lead to worse physical health (Van der Kolk, 2006; Van der Kolk & Ducey, 1989). In a continued attempt to ameliorate the traumatic symptoms, some individuals develop addictive behaviors, substance abuse, become defiant, develop destructive, aggressive behavior towards themselves or others, and tend to have poor self-control and low self-esteem (Dale et al., 2011). Ultimately, all these behaviors in combination with their social environment may lead to criminal acts and result in incarceration (Filipas & Ullman, 2006; Huan-Zhang, 2008; Johnson et al., 2003; Ullman 2005).

There are various treatments for individuals who have experienced trauma such as cognitive behavioral therapy (CBT). However, for individuals who have experienced prolonged and various types of trauma, or complex trauma, they are least responsive to treatment (Ford & Kidd, 1998). Research shows that trauma is remembered in the body (Van der Kolk, 1996). For individuals who have disconnected from their bodily experiences, physically-oriented interventions to address physiological symptoms related to trauma and the way trauma is remembered in the body is essential (Van der Kolk, 1996).

Studies have shown that yoga as an intervention helps increase self-awareness, and connectedness to the body (Baptiste, 2002; Iyengar et al., 2005). Several studies with vulnerable and clinical samples have shown yoga to be effective in reducing perceived stress, anxiety, depression, and PTSD which are all symptoms of trauma (Sharma & Haider, 2013; Kozasa et al., 2008; Michalsen et al., 2005; Vadiraja et al., 2009; Teller & Balkrishna, 2012). Yoga has also been associated with improved emotion regulation (Gootjes, Franken & Van Strien, 2011) and self-efficacy (Franzblau, Smith, Echevarria, & Van Cantfort, 2006). West (2011) using trauma-sensitive yoga with individuals with PTSD showed a reduction in symptoms and frequency of dissociative symptoms. Emerson et al (2009) reported the findings of a pilot study that compared Hatha yoga, the physical postures associated with yoga, to Dialectical Behavioral Therapy (DBT) treatment for trauma survivors. Yoga participants reported greater improvement in symptoms when compared to DBT treatment. Given that similar mental health problems are found in the prisoner population, yoga has been incorporated into inmate recreational programs in some correctional facilities. However, based on a review of the literature, few randomized studies have been conducted, and no trauma-focused yoga intervention study has been conducted in prisons.

The Prison Yoga Project (PYP) is a nonprofit organization that uses a trauma-informed, mindfulness based teaching methodology to address minimizing the impact of embodied trauma in prison populations. Their trauma-informed approach to yoga was developed to address the needs of participants for whom inappropriate behaviors were related to certain commands or poses. Instructors are trained to not use certain phrases or give commands to get in a certain yoga pose. Rather, they use a protocol that ensures the participant feels safe, invited, and gently encourages them to start paying attention to their body, what feels good and what does not, as well as feeling empowered in making choices when it comes to their body. Many correctional facilities in the United States currently offer trauma-focused yoga provided by PYP trained, volunteer instructors.

This study investigated the effects of a 10-week trauma-focused PYP Hatha Yoga intervention with female inmates in two correctional facilities. The researchers tested whether inmates, who are exposed to PYP’s trauma-focused yoga, had an improvement in their symptoms of traumatic stress and other factors. It was hypothesized that female inmates who participated in a PYP yoga class for 10 weeks would report a decrease in perceived depression, anxiety, stress, and an improvement in self-control, rumination and non-judgement of inner experience at post-intervention, compared to the control group who did not take the yoga class at the same time.

Methods

Procedure

Female, adult inmates currently serving time at one of two correctional facilities in South Carolina (referred to as facility A and facility B throughout this study) were recruited by the Prison Yoga Project teaching staff. An announcement was made about a new yoga class being offered, and about voluntary participation in the study. All interested participants were invited to attend a meeting prior to the first yoga class where they were given a consent form to read and sign, and a set of questionnaires to complete. The same measures were completed at post-intervention. Due to the low educational level of some inmates, the instructions and questions were read out loud by the same researcher to all participants. Participants were randomly assigned to the yoga class or control group, with the understanding that those placed in the control group would be allowed to be in the next yoga class after the first 10-week intervention was complete, because of class size limitations. The same PYP yoga instructor conducted the classes at both facilities.

Participants

Sixty-two participants consented to participate in this study. The final sample was 50 due to 3 participants being released during the course of the study, 2 participants being placed in lock up at time of post-intervention assessment, 5 quit attending yoga class and were unavailable to complete the assessment, and 2 declined to participate post-intervention. Thus, a total of 50 participants were included in this study with sample sizes at facility A (treatment group n=18; control group n=10) and facility B (treatment group n=15; control group n=7). Participants ranged in age from 23 to 70-years-old (M=37.92, SD=10.188). Their sentences ranged from minimum of two years to life sentences. The large majority of participants (54%) described themselves as Caucasian, followed by African-American (38%), Hispanic (6%), and Other (2%). Most of the participants reported having a high school diploma (28%) or GED (26%) as their highest education (see Table 1).

______

Insert Table 1 about here

______

Measurements

The Depression Anxiety and Stress Scale (DASS-21; Henry & Crawford, 2005) was used as a measurement of depression, anxiety and stress. This is a 21-item scale using a 4-point likert type scale range from “Never” to “Almost Always”. The 10-item Perceived Stress Scale (PSS; Cohen, & Williamson, 1988) was used as a measurement of stress using a 5-point likert type scale range from “Never” to “Very Often”. The Rumination Reflection Questionnaire (RRQ; Silvia, Eichstaedt, & Phillips,2005) was used as a measurement of rumination. This is a 6-item survey using a 5-point likert type scale range from “Strongly disagree” to “Strongly Agree”. The Brief Self Control Survey (BSCS; Tangney, Baumeister, & Boone, 2004) was used as a measurement of self-control. This is a 10-item scale using a 5-point likert type scale range from “Not at all like me” to “Very much like me”. The Barratt Impulsiveness Scale (BIS-11; Patton et al, 1995) was used as a measurement of impulsivity. For purposes of this study, the subscale of self-control was used. This is a 6-item subscale using a 4-point likert type scale ranging from “Never” to “Almost Always”. The Five Facet Mindfulness Questionnaire (FFMQ; Baer, R. A et al, 2006) was used to measure self-awareness. This is a 39-item questionnaire with a 5-point likert type scale range from “Never or rarely true” to “Very often or always true”. For this study, only the non-reactivity to inner experience subscale (7 items) was included to maintain the length of the questionnaire at a reasonable amount. Participants also provided their age, ethnicity, highest education completed, and sentence length.

Results

To assess the impact of yoga intervention on participants’ self-reported scores on each of the measures for depression, anxiety, stress, rumination, self-control and non-reactivity to inner experience, mixed-design analysis of variancetests were conducted with time (pre-intervention and post-intervention) as a within-subject factor, treatment group (treatment or control group) as the between–subjects factors. Descriptive analysis was used to report the demographics of the sample participants.

The research questions concerned whether there was a change in the PYP yoga treatment group scores over time (pre-intervention and post-intervention) when compared to the control group. It was hypothesized that there would be a decrease in symptoms of self-reported depression, anxiety, stress, and improvement in self-control, rumination and self-awareness from pre-intervention to post-intervention for the yoga treatment group participants. The results are presented for each measure below.

______

Insert Table 2 about here

______

Depression

Self-reported depression scores had a significant main effect for time F(1,20) = 4.93, p <.05. The main effect by the group was not significant F(1,20) = .83, p >.05. The interaction between treatment and time was significant, F(1,20) = 4.93, p <.05. The depression symptoms reported for participants in the treatment condition decreased from pre-intervention (M= 32.13, SD = 11.55) to post-intervention (M= 25.43 , SD=8.30) with Cohen’s d = 67 suggesting medium practical significance. The depression symptoms reported for participants in the control condition showed no change from the pre-intervention (M= 25.43, SD=6.19) to the post-intervention (M=25.43, SD=6.63) with Cohen’s d = 0 suggesting small practical significance. As hypothesized, yoga did decrease depression symptoms in the treatment group.

Anxiety

Self-reported anxiety scores did not have a significant main effect of time F(1,20) = 3.42, p >.05. The main effect by the group also was not significant F(1,20) = .13, p >.05. The interaction between treatment and time was not significant, F(1,20) = 1.63, p >.05. The anxiety symptoms reported for participants in the treatment condition decreased from the pre-intervention (M=29.33, SD=10.41) to the post-intervention (M=24.67, SD=7.84) with Cohen’s d = .51 suggesting medium practical significance. The anxiety symptoms reported for participants in the control condition showed no change from pre-intervention (M=28.86, SD=9.65) to post-intervention (M=28.00, SD=8.64) with Cohen’s d = .10 suggesting small practical significance. As hypothesized, yoga did decrease anxiety symptoms in the treatment group.

Stress

Different measures of stress were used at the two facilities based on its applicability of the questions and delayed approval by one of the facilities. Self-reported stress scores in the Depression Anxiety Stress scale utilized in Facility A had a significant main effect for time F(1,20) = 8.04, p <.01. The main effect by treatment group was significant F(1,20) = .69, p <.05. The interaction between treatment and time was significant, F(1,20) = 5.78, p <.05. The stress symptoms reported for participants in the treatment condition decreased from the pre-intervention (M=36.80, SD=12.23) to the post-intervention (M=29.87, SD=9.46) with Cohen’s d = .63 suggesting medium practical significance. The stress symptoms reported for participants in the control condition increased from pre-intervention (M=37.43, SD=9.91) to post-intervention (M=36.86, SD=8.47) with Cohen’s d = .06 suggesting small practical significance. As hypothesized yoga did improve the stress symptoms in the treatment group, while there was an increase in stress symptoms for the control group over time.

Self-reported stress scores utilizing the Perceived Stress Scale at Facility B did not have a significant main effect by time F(1,26) = .30, p >.05. The main effect by treatment group was significant F(1,26) = 5.38, p <.05. The interaction between treatment and time was significant, F(1,26) = 8.13, p <.01. There was a decrease in symptoms of stress for participants in the treatment condition from the pre-intervention (M=29.28, SD=4.39) to post-intervention (M=25.44, SD=4.50) with Cohen’s d = .86 suggesting large practical significance. The stress symptoms reported for participants in the control condition increased from the pre-intervention (M= 30.80, SD=7.41) to the post-intervention (M=33.40, SD=8.54) with Cohen’s d = -.33 suggesting small practical significance. As hypothesized, there was an improvement of stress symptoms for the treatment group, and a slight increase in stress for the control group over time.

Self-Control

Different measures of self-control were used at the two facilities based on its applicability of the questions and delayed approval by one of the facilities.Self-reported self-control scores from the Brief Self Control Scale in Facility A did not have a significant main effect by time F(1,20) = 2.45, p >.05. The main effect by treatment group was not significant F(1,20) = 1.20, p >.05. The interaction between treatment and time was not significant, F(1,20) = 2.28, p >.05. The self-control scores for participants in the treatment condition showed no change from the pre-intervention (M=3.56, SD=.68) to the post-intervention (M=3.64, SE=.69) with Cohen’s d = -.12 suggesting small practical significance. The self-control scores reported for participants in the control condition showed an increase in scores from the pre-intervention (M=3.04, SD=.69) to the post-intervention (M=7.40, SD=11.45) with Cohen’s d = -.54 suggesting medium practical significance, and revealing worse self-control after 10 weeks for the control group. Contrary to the hypothesis, there was no significant change in self-control for the treatment group at Facility A. However, the self-control levels remained about the same for the treatment group in comparison with a decrease in self-control for the control group after 10 weeks.