Behavioral Incentives to Increase Exercise in Cocaine Dependent Women

While physical activity has been associated with decreased craving and increased abstinence
rates in smokers, few published studies have examined the effects of exercise on recovery
from other drugs such as cocaine. One factor that has impeded such research has been low
levels of patient compliance with exercise protocols. One robust strategy for promoting and
maintaining behavior change is contingency management or Behavioral Incentives (BI). BI
delivers incentives (prizes, vouchers) contingent upon target behaviors such as cocaine
abstinence (Higgins et al., 1994); treatment attendance (Svikis et al., 1997) and other
pro-social behaviors (Kirby et al., 1998). While the literature is replete with studies
demonstrating the benefit of BI compared to control conditions (Stitzer & Petry, 2006), the
translation of BI methods from research to clinical practice has met with some resistance.
Contributing factors include philosophical differences (e.g., counselors feel extrinsic
reinforcement undermines recovery) and practical barriers (e.g., monetary costs of
incentives may be prohibitive). The latter concern was addressed by Petry (2005) who
developed the "fish bowl" method, which uses escalating variable ratio procedures to reduce
per patient costs of BI with similar effect sizes.

As a Stage 1 behavioral therapies development grant (Rounsaville et al., 2001), the primary
aim of this research is to pilot test a BI intervention designed to promote regular physical
activity in a sample of women receiving inpatient treatment for SUDs. The target behavior,
physical activity, will be objectively defined as 30 minutes of observed treadmill walking
at any intensity 3 days/week at Level 1, and 30 minutes of higher intensity physical
activity that meets ACSM criteria for moderate exercise Level 2. Specifically, a pilot
randomized clinical trial will compare rates of physical activity over a 6 week study period
in a sample of N=50 women with Cocaine Dependence. All participants will complete baseline
assessment, attend a 45-min Health and Fitness (HF) education group, followed by random
assignment to either the experimental (BI) or control (C) groups, with equal daily access to
on-site treadmills. Those randomized to BI, however, will also be eligible 3 days/week, to
receive incentives for completing 30 minutes of treadmill walking.

Incentives will be dispensed using Petry Fish Bowl methods. Women assigned to the BI group
will receive behavioral incentives (in the form of gift cards or prizes) for completing
their scheduled exercise sessions (Level 1), and have the opportunity to earn "bonus" draws
for meeting moderate intensity exercise criteria, as specified by ACSM (2007, revised)
guidelines (Level 2). In Level 1, the number of tokens participants can draw from the gym
bag (present study equivalent to "fishbowl") at each visit will be linked to exercise
session attendance, with consistent attendance resulting in greater number of draws. In
Level 2, the number of tokens a participant can draw from the gym bag will be linked to the
intensity of the participant's exercise. Specifically, beginning in Week 2 of the study,
participants will be re-evaluated for exercise capability. Those who pass the safety screen
will be encouraged to exercise at a higher intensity, and those who do not will be
re-evaluated until they obtain safety clearance. Once cleared, if the participant meets
criteria for moderate exercise during her scheduled session, then she will be allotted a
"bonus" draw (Level 2) for meeting Level 2 criteria, in addition to the escalating number of
draws she would received for scheduled exercise session completion (Level 1).

Scheduled treadmill walking will be monitored and recorded for both BI and C group women.
Follow-ups will occur at study midpoint and completion (3 and 6 weeks post-randomization,
respectively), and at 4 weeks post-discharge. Assessments will focus on drug craving, mood,
stress, motivation/self-efficacy, and physical health and well being. The investigators
hypothesize that women in the BI group will complete more treadmill sessions and spend more
time treadmill walking than those in the C group. As a Stage 1b therapy development RCT,
study data will be used for effect size estimation in preparation for Stage 2 RCT. This
dissertation proposal will provide benchmark data on the utility of BI for promoting
physical activity. Further, it will promote exercise compliance, allowing scientists to
better evaluate potential benefits of physical activity on treatment outcomes in women with
SUDs.