The Effects of A Therapeutic Martial Arts Program on Youth in Residential
Psychiatric Treatment

Resubmitted to the A.D. Williams Committee May 1, 2003


Principal Investigator
Brian Hill, LCSW, Assistant Professor, VCUHS Department of Psychiatry
Team Leader and Therapist, Residential Treatment Program, Virginia Treatment Center for Children

Co-Investigator
Brian L. Meyer, Ph.D. Assistant Professor, VCUHS Department of Psychiatry
Executive Director, Virginia Treatment Center for Children

Background and Objectives
The efficacy of traditional martial arts programs applied toward mental health issues is well documented.  Over the
last twenty years, proliferating research has explored martial arts training (MAT) programs as alternatives to
traditional psychotherapy.  Therapeutic martial arts programs emphasize respect for the self and others, containment
of aggression, and conflict resolution.  Therapeutic MAT can be used as an adjunctive treatment to traditional
interventions, in the same way that psychotherapy and pharmacotherapy complement each other.  MAT can foster
and expose feelings through a physical mode, for example, which may then be addressed through other modes of
therapy.  

The research on therapeutic martial arts programs for adults has demonstrated a number of positive outcomes.  
Konzak and Boudreau (1984) pointed out that adults could use MAT as a means of self-help, and Weiser, Kutz, Kutz,
and Weiser (1995) suggested MAT be used as adjunctive psychotherapy due to its simultaneous intervention on
physical, interpersonal, and intrapsychic levels.   Other studies on adults (cf., Madden, 1995; Rothpearl, 1980;
Kurian, Caterino & Kulhavy, 1993) demonstrated that therapeutic MAT increases self-esteem and assertiveness, and
decreases anxiety, hostility, and aggressiveness.  

Although the literature supports the beneficial results of martial arts training (MAT) for adults, there is a much more
limited base regarding children and adolescents.  Most studies on youth have used school-based and existing for-
profit community MAT programs, with subjects identified as at-risk for or currently experiencing behavioral problems,
and youth identified with serious emotional disturbance.  Studies on youth typically focus on the effects of MAT on
aggression, self-esteem, and behavioral problems.  Early research with children and adolescents (Nosanchuk, 1981)
indicated that MAT increased self-control, assertiveness, self-esteem and self-confidence.  Importantly, Nosanchuk
and MacNeil later (1989) determined that “modern” or non-traditional MAT that discounts meditation, values of peace,
etc. actually increased aggressive behavior.  Further, increased dosage of treatment exposure (longer training time)
created a heartier result.  The 1989 study also identified key elements in conducting therapeutic MAT, including
instructor characteristics (restraint, parental figure, and faith in the student), values and ethics, and techniques
focusing on conflict resolution.

Other research has further demonstrated the efficacy of therapeutic MAT with children and adolescents.  Trulson
(1986) studied the effects of traditional MAT in his work with juvenile delinquent youth in a community setting, and
found significantly decreased aggressiveness and anxiety, and increased self-esteem and social skills.  Of note, the
youths’ MMPI scores for juvenile delinquency measured in the normal range post-treatment.  Those youths who
engaged in modern style MAT, which focuses on fighting and competition, showed no improvement.  In their work with
children of varying disabilities, including mental retardation and attention deficits, Gleser and Brown (1988) found that
the use of a martial arts program improved levels of physical and psychosocial functioning.  Reynes and Lorant
(2001) studied personality characteristics of children in MAT programs and found that children choosing martial arts
training were no more aggressive than their peers.  Zivin, et al. (2001) duplicated several school- and community-
related MAT programs, and demonstrated decreased violence and delinquency within as few as 30 sessions.

The principles inherent in traditional martial arts parallel certain mainstream psychological constructs, such as self
concept, social skills, and conflict resolution.   Fuller (1988) advocated the use of aikido, a Japanese martial art, as
perhaps most suited to work with people with mental health problems.  He cited the work of Madenlian (1979), who
compared an aikido MAT program with traditional group psychotherapy and found that students who participated in
aikido made greater gains on the Piers Harris Self Concept Scale than those who participated in group therapy.  
Fuller’s (1988) assertion that aikido matches the needs of psychotherapeutic programs makes sense.  The concepts
of blending, non-violence, peaceful conflict resolution, centering, and non-competitiveness inherent in aikido practice
make it an ideal modality to approach emotionally disturbed children and adolescents.  For those reasons, the
proposed research program will use aikido as the martial arts intervention.

The proposed study focuses on the effects of a therapeutic martial arts program as an adjunct treatment for severely
emotionally disturbed children and adolescents in residential treatment at the Virginia Treatment Center for Children
(VTCC).  The subjects reside at VTCC for an average of 5-6 months.  Behavioral progress and incidents of
aggression are monitored on an ongoing basis by nursing staff, and these data will be used in the research.  The
study will also measure the cognitive effects of treatment similar to those explored in earlier community and school-
based populations.   The goal of the study is to demonstrate how therapeutic martial arts training can reduce violence
and increase emotional well-being in troubled youth.  What makes this proposal new and innovative is fourfold:
·        The study focuses on a seriously disturbed population.  Existing research focuses on children and adolescents
in community and school settings.  The proposed study focuses on severely emotionally disturbed, treatment
resistant youth who reside in a long-term residential mental health treatment facility.
·        The research investigates a little-studied martial arts modality.  The study will use aikido as the martial art,
which focuses on pure self defense, blending with attacks, and peaceful conflict resolution.  Most existing research
has used more aggressive styles such as tae kwon do and karate, which teach punching and kicking.  Aikido does
not teach such aggressive techniques, which fits the needs of youth with pre-existing aggression and anger
management problems.
·        The research uses cognitive measures related to behavioral performance.   Prior research has used measures
of self-esteem and self concept.  This research examines not only self concept, but also self-control, since the latter
may relate more to the ability to contain aggression.  
·        The research uses precise behavioral measures.  Previous research has used anecdotal or subjective reports
of behaviors such as aggression.  The residential treatment setting includes precise observation and measurement of
behaviors by 24-hour nursing care staff.  The study will examine incidents of aggression, time out, seclusion and
therapeutic holds, as well as overall positive behaviors (as measured on daily point sheets).  Behaviors have not
been measured with this level of specificity in previous studies of therapeutic martial arts programs.

Objectives of the proposed study:
1)        To examine the effects of a therapeutic martial arts program on youth placed in residential psychiatric
treatment, specifically exploring aggression, self-control, and self concept.
2)        To conduct research on the effects of therapeutic martial arts in a controlled environment to increase validity
of the results.
3)        To complete a pilot study as a means of obtaining externally funded research (through NIMH or the CDC)
investigating alternative methods for treating violence and aggression in treatment resistant populations.  
4)        To identify salient variables and develop new hypotheses that will be explored in future research.

Hypotheses
1)        Participants in the therapeutic martial arts program will decrease documented behavioral incidents (time out,
aggression, seclusion, and therapeutic holds) more than controls.  
2)        Participants in the therapeutic martial arts program will increase their scores on measures of self-control and
self concept more than controls.
3)        Participants in the therapeutic martial arts program will increase their positive behaviors (as measured by
points earned) more than controls.
4)        Participants will maintain gains of training at one month follow up.

Methodology
Participants
Participant will be 40 males and females from a long-term child and adolescent residential psychiatric treatment
program.  They must be enrolled for a minimum of two weeks of orientation and observation on the unit.  Residents
who have mental retardation or Pervasive Developmental Delays will be screened out.  The ages of the subjects may
range from 8-17 years old.  VTCC patients typically have multiple Axis I DSM-IV diagnoses, as well as Axis II
diagnoses that include Learning Disorders.  Almost all have had serious problems with aggression, impulse control,
oppositional and defiant behavior, and at school.  VTCC has 30 residential treatment beds; consequently, we
estimate that 12 months of data collection will yield approximately 40 subjects.  A power analysis (Cohen, 1992)
indicates this is a sufficient number of subjects for the planned analyses when a = .05, b = .20, and the effect size is .
4 (medium to high).

Participation in the study will be voluntary.  All participants and their parents/legal guardians will sign informed
consent forms.  All staff involved in the project will sign confidentiality forms.  Subjects will be given an identification
number, and collected materials will be identified by number.  All materials will be kept in a locked file cabinet, and will
be destroyed upon completion of the study.  Results will be reported in aggregate form only.

Procedures
From the population of the residential program, twenty subjects will be randomly selected and divided into a treatment
group (Group A) and a wait-list control group (Group B).  In addition to typical residential treatment services, the
treatment group will participate in a therapeutic martial arts program for ten weeks, twice weekly for 45-60 minutes per
session, while the wait-list group will receive the regular treatment services. When Group A completes the program,
Group B will become the treatment group and a new wait list group (Group C) will be randomly selected from newly
admitted residents.  This process will continue until Group D has completed treatment. Preliminary data for subjects
who either drop out of the intervention or are discharged prior to completion will be compared with preliminary data
for subjects who complete the intervention to determine whether there are any differences between the two groups.  
The instructor will be the principal investigator, who is a clinician as well as a qualified martial arts teacher.

Every group will be administered the self-control and self concept measures pre-treatment, post-treatment, and at
one month follow up.  Objective behavioral indexes of aggression, time out, seclusion, and restraint will be tallied for
the two weeks prior to treatment and every two weeks thereafter through four weeks following the completion of
treatment.  

The subjects will take part in a structured therapeutic martial arts program using the Japanese art of aikido.  Aikido
literally means “the path of harmonizing with the energy of the universe” (Ai= harmony, ki=energy, do=path).  The
martial arts groups will consist of stretching warm-up, followed by daily review of behavioral expectations, including
the philosophy of traditional martial arts training.  The instructor will be assisted by a uke, a trained assistant who
demonstrates how to take falls, to increase the safety of the participants.  Participants will progress from simple solo
practices in which they learn to fall safely if pushed, to partner practices which allow them to learn escapes from
grabs, to defenses against punches.  Same sex dyads will practice the physical techniques of self-defense to prevent
the risk of inappropriate touching. Concurrently, participants will learn anger management techniques, how to diffuse
and handle criticism, ways of resolving conflict without physical intervention, and basic safety awareness.  Each class
will end with a meditative relaxation exercise. The program will culminate in a ceremony recognizing students’
achievement with a patch for their uniforms, both of which they will be allowed to keep.

Measures

Discrete behaviors
Discrete behaviors are directly observed and documented on behavioral program point sheets every half-hour by 24-
hour nursing staff.  They are trained to recognize and provide consequences for negative behavior.  The following
behaviors are tallied to determine progress on a daily basis:
·        Incidents of aggression (hitting, kicking, biting, pushing, etc.) towards peers or staff
·        Time Out (removal of subject to a quiet room for up to five minutes)
·        Out of control or dangerous behaviors requiring seclusion in a locked room
·        Out of control or dangerous behaviors requiring therapeutic holding (physical restraint) by staff

Self-Control
Children’s Perceived Self-Control Scale (CPSC)
This 11-item instrument was developed by Humphrey (1982) to measure children’s perceptions of their self-control
from a cognitive-behavioral perspective.  The instrument measures three aspects of self-control: interpersonal,
personal, and self-evaluation.  The instrument includes an overall measure of self-control, as well as individual
subscale indexes.  The CPSC has been reported (Humphrey, 1982) to have satisfactory reliability (.71 for total
scores, Interpersonal Self-Control (ISC) =.63, Personal Self Control (PSC) =.63, Self Evaluation (SE) =.56), but no
data on internal consistency is available.  Evidence for concurrent validity has been minimal, but naturalistic
observations correlate highly with ISC.

Self-Concept
Piers Harris Children’s Self-Concept Scale (PHCSCS)
The PHCSCS is well known, written on a 3rd grade reading level, designed for 7-18 year olds.  The PHCSCS was on
1,183 students grade 4-12, and showed test-retest reliability coefficients from.42 to .96, with a median test-retest
reliability of .73.  This test reported internal consistency coefficients ranging from .88 to .92 for the total score and
from .73 to .81 for the cluster scales (Forgan, 2000).

Data Analysis
Data will be analyzed to provide summary statistics for the sample (e.g., ranges, means, and standard deviations for
key demographic and outcome variables).  Time series analyses, holding age, gender, and time constant, will be
used to compare individual scores on outcome variables measured across time.  Repeated Measures Analysis of
Variance will be used to compare group means on outcome variables at prior to treatment, after treatment, and at
one month follow-up.  The level of significance, a, will be set at .05, while b will be set at .20.  These analyses will
determine whether treatment subjects have changed significantly over time in self-control, self concept, positive
behavior, and behavioral incidents (aggression, time out, seclusion, and restraint) compared to controls.  If sufficient
subjects are available, additional analyses will be performed to indicate whether specific age or gender groups differ
in the magnitude of their change over time. Measures obtained at one month follow up will be compared to determine
if treatment effects are maintained over time.  All statistical analyses will be conducted using SPSS, version 11 (or
updated versions, as needed).

Incidents of aggression, time out, seclusion, and restraint, as well as self-control and self concept scores, will be
compared within individuals over time, with measurements taking place prior to treatment, after treatment, and at one
month follow-up.  These same measures will be compared between groups prior to treatment, after treatment, and at
one month follow-up.

Limitations
The study is limited by a small pool of participants (N = 40), due to time constraints in the length of the study and the
capacity of the residential treatment program.  Additional limitations include the large range of participant ages and
the mixed gender construction of the treatment and control groups.  The control group members will not be matched
with the treatment group due to the capacity of the residential treatment program.  The use of the principal
investigator as the therapeutic martial arts group leader is not anticipated to cause measurement bias, since the
research assistant will collect the self-control and self concept measures, and the behavioral measures are rated by
multiple nursing staff on three different shifts.


Timeline

Month 1                Submit IRB Proposal
Months 2-4                Group A Treatment; Waitlist B Control; and Data Collection
Months 4-6                Group B Treatment; Waitlist C Control; and Data Collection
Months 7-9                Group C Treatment; Waitlist D Control; and Data Collection
Months 9-11                Group D Treatment; Group E Control; and Data Collection
Month 12                Data Analysis and Interpretation
Months 13-15        Report and manuscript preparation; prepare proposal for external funding

Future efforts:

The room for the Therapeutic Martial Arts Program will be renovated in May 2003 through a grant awarded by the
MCVH Auxiliary.  The costs of renovation, including mirrors and mats, are included in the Auxiliary funding, enabling
the program to begin this summer, perhaps as early as June 2003.  Funding the current proposal will allow the
investigators to gather pilot data for a proposal to either the National Institute of Mental Health or the Centers for
Disease Control regarding the reduction of youth violence in treatment resistant populations.  

Several avenues will be pursued in future research.  First, the questions of optimal dosage frequency and length of
treatment need to be addressed to design the most effective therapeutic martial arts intervention.  Second, expanded
funding would allow a larger sample size so that results for children and adolescents of differing ages and genders
can be compared.  Long term follow-up to determine if gains are maintained after discharge would be valuable.  It is
important to determine if therapeutic martial arts training has comparable effects on other treatment resistant
populations such as juvenile delinquents and students in alternative schools.  A large sample size would also allow
the development of a structural model to demonstrate how therapeutic martial arts training mediates the propensity
towards aggression, how it acts upon the moderator variables of self-control and self concept, and to what degree.  
This pilot study therefore has the potential of initiating a long line of clinical research.  
Literature Review

References

Curriculum
Therapeutic Martial Arts