|Home | About | Journals | Submit | Contact Us | Français|
The benefits of regular physical activity are particularly salient to persons with serious mental illness (SMI) who have increased prevalence of obesity, diabetes, and earlier mortality from cardiovascular disease.
The Activating Consumers to Exercise through Peer Support (ACE) trial will examine the effectiveness of peer support on adherence to a 4-month pilot exercise program for adults with SMI. Design, rationale and baseline data are reported. Baseline measures included: graded treadmill test; six-minute walk; height, weight and blood pressure; body composition; fasting blood; and self-reported psychiatric symptoms. Fitness levels were compared with national data and relationships among fitness parameters, psychological factors and cardiovascular disease risk factors were examined.
There were 93 participants and 18 peer leaders recruited from community psychiatry programs with an average age of 47 years (SD 10). There were no differences in demographics (76% female, 72% African American) or mental health symptoms between participants and peer leaders. Ninety-five percent of the sample had below average fitness levels for their age and sex with average MET levels of 5.9(SD 2.2) for participants and 6.2(SD 2.3) for peer leaders. Fitness evaluated during the treadmill test and the six-minute-walk were associated (rs = 0.36, p<.001). Lower MET levels were associated with a higher BMI (rs = −0.35, p<.001) and percent body fat (rs = −0.36, p <.001).
The uniformly low baseline cardiovascular fitness and the association of fitness with BMI and adiposity underscore the importance of suitably tailored programs to increase physical activity among adults with SMI.
Persons with serious mental illness (SMI) have elevated risk of cardiovascular disease (CVD), diabetes mellitus, and obesity even among users of medical care (Daumit, Pratt, Crum, Powe, & Ford, 2002; L. Dixon et al., 2000; Felker, Yazel, & Short, 1996; Sokal et al., 2004). Obesity is also a significant health concern among persons with SMI (Allison et al., 1999; Coodin, 2001; Daumit et al., 2003; L. B. Dixon et al., 2004; McElroy et al., 2002). Many psychotropic medications may compound these health concerns as they are associated with weight gain, and certain second generation antipsychotics in particular are associated with increased appetite, glucose intolerance, and lipid abnormalities (Allison, Mentore et al., 1999; Baptista, Kin, Beaulieu, & de Baptista, 2002; Gupta et al., 2003; Lean & Pajonk, 2003; Lindenmayer et al., 2003; Macritchie, Geddes, Scott, Haslam, & Goodwin, 2001). Furthermore, evidence among elderly with SMI suggests that the conventional antipsychotics are at least as likely to lead to premature mortality as the newer antipsychotics (Wang et al., 2005). Overall, persons with SMI carry a tremendous burden of medical comorbid illness and have increased risk of early mortality compared to age-matched peers without mental illness (Brown, 1997; Colton & Manderscheid, 2006).
There are few lifestyle changes that address as many of the aforementioned health concerns as regular physical activity. Physical activity has been linked with decreased risk of CVD, less insulin resistance, improved weight loss, and decreased mortality (Garber et al., 2011; U.S. Department of Health and Human Services, 2000). Moreover, the benefits of cardiovascular fitness have been realized across body mass index (BMI) classifications (Wei et al., 1999). Although there is a significant body of literature linking regular physical activity and increased cardiovascular fitness with a range of physical health benefits, these associations have received little attention with respect to adults with SMI. There have been lifestyle interventions that have targeted weight loss but few trials have targeted cardiovascular fitness changes among adults with SMI (Daumit et al., 2010; Faulkner & Biddle, 1999; Faulkner, Soundy, & Lloyd, 2003; Fogarty, Happell, & Pinikahana, 2004; McKibbin et al., 2006).
A small literature describes the mental health benefits of regular physical activity among those with SMI. Gorczynksi and Faulkner summarized the evidence for mental health outcomes of physical activity in schizophrenia in a recent Cochrane review reporting that in two small trials, negative symptoms were improved (Gorczynski & Faulkner, 2010; Beebe 2005; Duraiswamy 2007). Two recent studies showed that depressive symptoms decreased after participation in an exercise intervention in persons with serious mental illness, (Daumit et.al, 2011; McKibbon et. Al., 2006); and other literature reviews have suggested that physical activity is associated with improved well-being in schizophrenia (Faulkner & Biddle, 1999; Holley, Crone, Tyson, & Lovell, 2010). In persons with primary depression or anxiety diagnoses, physical activity has been associated with decreased symptoms along with improvements in panic disorders (Paluska & Schwenk, 2000). Despite the many benefits of regular physical activity and evidence that physical activity promotion was well received by adults with SMI; most adults with mental illness remain less active than the general population and do not meet physical activity recommendations (Jerome et al., 2009; Ussher, Stanbury, Cheeseman, & Faulkner, 2007).
One way to increase physical activity among person with SMI would be the integration of exercise programs into health services that these individuals are already receiving (Richardson et al., 2005). A successful integration would likely target both environmental and individual determinants of physical activity. Environmental changes could include structured exercise programs or access to fitness facilities. Social support is a likely individual determinant of physical activity among those with SMI and increasing social support is a recommended strategy for increasing adherence in physical activity programs (Task Force on Community Preventive Services, 2001). Moreover, adults with severe mental illness have reported low levels of exercise related social support (Ussher, et al. 2007). Social learning theory suggested that peers can be influential role models and source of support as they are able to communicate in culturally appropriate ways and resemble the target population in appearance and values (Broadhead et al., 1998). Social support, specifically peer support, has been increasingly included in health promotion efforts among adults with SMI and has strong potential to benefit both the peer leaders and the peers (Davidson, Chinman, Sells, & Rowe, 2006; Fisher, 1988; Needle, Brown, Coyle, & Weissman, 1994; Solomon, 2004; Williams, 2007). Few studies have examined the association between social support and physical activity levels among adults with mental illness. One retrospective study suggested that men receive social support from exercise and sporting activities; however, there have been no reports of the role played by social support in adherence to structured physical activity programs (Carless & Douglas, 2008).
This pilot study was designed to examine the effect of peer support on adherence and subsequent changes in cardiovascular fitness associated with moderate intensity exercise program among adults with serious mental illness receiving outpatient mental health treatment. Our conceptual framework aimed to bring a social support component to a community psychiatry clinic setting in order to facilitate improved exercise participation. This paper provides the study rationale and design. In addition, this paper reports descriptive statistics of the study sample, compares baseline cardiovascular fitness levels to a national sample, and examines associations among cardiovascular fitness, demographics, psychological factors and cardiovascular disease risk factors.
The Activating Consumers to Exercise through Peer Support (ACE) Study is examining the effects of a group exercise program with and without peer support among adults with serious mental illness receiving outpatient mental health services. Both intervention groups are offered progressive group exercise classes designed to increase cardiovascular fitness among low active adults. These moderate intensity exercise classes are offered three times per week for four months and are led by a trained exercise instructor. We found group exercise classes were effective in promoting healthier lifestyles with a similar population (Daumit et al., 2010). Additionally group exercise classes are well suited for an outpatient setting as they present lower cost than individualized coaching and require minimal equipment. Those in the exercise plus peer support condition received support from peers who underwent training on how to act as a role model and encourage participation in the exercise class. The primary outcome is change in fitness after four months of exercise class.
The ACE study has three aims: 1) develop a culturally appropriate group exercise program and corresponding peer leader training for adults with SMI; 2) conduct a pilot study to examine the effect of peer support on adherence to the exercise program; 3) evaluate changes in mental and physical health of both the study participants and the peer leaders.
Participants were recruited from the Johns Hopkins Hospital Outpatient Community Psychiatry Program and the Outpatient Community Psychiatry Program at Johns Hopkins Bayview Medical Center which treat approximately 1500, and 4500 patients respectively. Recruitment occurred through targeted mass mailings, brochures placed in waiting rooms and brochures provided to clinicians who made them available to interested patients. The research staff encouraged referrals through presentations at staff meetings and individual meetings with clinicians. Clinicians also helped to identify potential peer leaders for the first cohort. Outpatient clinicians made recommendations by identifying consumers who had a history of keeping appointments; adhering to medications; were likely available during the hours classes were offered; and potentially interested in helping peers. The performance of participants and peer leaders during an active cohort also informed the selection of peer leaders in subsequent cohorts. The informal criteria for inclusion as a peer leader in subsequent cohorts included: regularly attending class; enthusiastic participation; ability to teach others (exercise routines, use of tracking sheets, etc). We did not specifically exclude diagnoses, but we recruited participants from two community psychiatry clinics that see almost exclusively persons with schizophrenia and schizophrenia spectrum conditions, bipolar disorders and major depression. Due to the nature and severity of illness in this sample along with the timing of the exercise classes (in the middle of the day when individuals without impairment are likely to be in school or working), we are describing the participants as having serious mental illness.
Both participants and peer leaders had to be at least 18 years of age and walk without an assistive device. Exclusion criteria were the same for peers and peer leaders including: 1) known contraindication to weight loss (e.g., malignancy, liver failure, history of anorexia nervosa); 2) pregnancy; 3) weight > 182 kg; 4) inability to walk to participate in exercise; 5) symptoms of angina or cardiovascular event within the past 6 months; 6) active substance abuse; 7) contraindication to treadmill testing; 8) an abnormal treadmill test; 9) non-English speaking or 10) were planning to leave the Baltimore area within the next 6 months. An additional exclusion criterion was inability to provide consent as determined by a brief standardized questionnaire addressing the study goals, participant responsibilities and associated risks. A two-stage consent process was used with the first consent obtained to conduct screening procedures and a second consent obtained after baseline data collection and prior to randomization. Both participants and peer leaders completed six screening visits. The trial received approval from the Johns Hopkins Medicine Institutional Review Board, and informed written consent was obtained from all participants and peer leaders.
Data collection included interviewer administered questionnaires, graded treadmill testing, body composition assessment, and fasting blood testing. Participants received $10 for completing the graded treadmill test. Each primary care physician was sent a fax letter with a brief description of the study highlighting the exercise treadmill test and the moderate intensity exercise program. Physicians were asked to respond if they were aware of any contraindications to participation for their individual patient(s) in either the treadmill test or a moderate intensity exercise class.
Data collection occurred at baseline and at the end of the four-month intervention for both intervention participants and peer leaders. Age, sex, race, marital status and education were self-reported; all questionnaires were interviewer administered. Primary diagnosis and medications (e.g., psychotropic and beta blockers) were determined by chart review.
Aerobic fitness was measured during a peak treadmill exercise test using a modified Balke protocol (American College of Sports Medicine, 2000). Participants warmed up for three minutes at 1.7 mph and zero grade. The test included three minute stages starting at a speed of 2.5 mph and zero grade and the speed remained constant while the percent grade increased by 2.5% per stage for the first nine stages. The speed increased 0.5 mph per stage during stages 10-15 while the grade remained at 20%. Perceived rate of exertion was monitored using the OMNI walk/run scale which included graphic representation of a person walking/running at different exertion levels (Utter et al., 2004). This scale has been shown to correspond with heart rate and oxygen uptake during treadmill testing (Utter et al., 2004). Tests were terminated upon volitional fatigue, achieving predicted maximal heart rate or presence of clinical indicator for termination (Gibbons et al., 2002). Maximal time on treadmill has been strongly correlated with maximal oxygen consumption (Pollock et al., 1976; Pollock et al., 1982). Peak METs were reported based on workload associated with maximal stage achieved. Longer time on treadmill and lower heart rate responses indicate higher cardiovascular fitness.
Cardiovascular fitness was also assessed at a submaximal effort using the six-minute walk test (Guyatt et al., 1985). Each participant walked an indoor course for six-minutes and the total distance walked was measured. Greater distances walked during the six-minute walk test correspond with higher levels of fitness (Cahalin, Mathier, Semigran, Dec, & DiSalvo, 1996; Guyatt et al., 1985).
Trained study staff completed standardized measures of weight on a calibrated scale with the participant in light indoor clothes without shoes. Height was measured using a calibrated stadiometer, and body mass index (kg/m2) categories were based on National Institutes of Health cutpoints (National Institutes of Health, 1998).
Waist circumference was measured using a Gulick tape measure at a horizontal plane one cm above the navel. The iliac crest can be difficult to identify on overweight and obese individuals so the navel was used as a reference for the waist circumference measurement. This technique has been used in a range of clinical trials and epidemiological studies (Appel et al., 2003; Appel et al., 2011; Bild et al., 2002; Svetkey et al., 2008).
Total body fat was measured via Dual Energy X-Ray Absorptiometry (DEXA), specifically the Lunar Prodigy DEXA manufactured by General Electric (CT, USA). The DEXA assesses total and regional body composition by analyzing images in the frontal plane to determine bone, fat and fat-free mass.
Blood pressure was measured with an OMRON HIM-907 sphygmomanometer three times during screening visits 1 and 3 by trained staff using a standard protocol (Appel et al., 2003). Participants rested quietly for five minutes in the seated position and the appropriately sized cuff was determined based on arm circumference. Three measurements were obtained with a minute in between each measurement and the readings were averaged. Following standardized procedures, fasting (10–12 hour fast) serum samples were obtained to determine lipid levels, glucose and fasting insulin levels (Appel et al., 2003).
Exercise self-efficacy was assessed when the participants indicated their confidence in exercising 3 days per week at a moderate intensity for six different lengths of time (i.e. 5, 10, 20, 30, 40, 50 minutes). Participants indicated their degree of confidence on completing each of the six bouts using an eleven point scale that start at 0% and increased by 10 up to 100%. Scores on the five items were averaged. This self-efficacy measure was constructed in accordance with guidelines recommended by Bandura and has shown to be predictive of both short and long term exercise behavior (Bandura, 1997; Bandura, 2001; McAuley & Courneya, 1993; McAuley, Lox, & Duncan, 1993; McAuley, Jerome, Elavsky, Marquez, & Ramsey, 2003). General self-efficacy was assessed using the General Self-efficacy Scale (Schwarzer, 1993). Participants indicated the extent they believe they can handle different challenges (e.g. “If I am in trouble, I can usually think of a solution) using a four point scale (1, not at all true - 4, exactly true) and scores are summed across the ten items. This scale has been used with samples of adults with mental illness (Schwarzer, 1993; Yanos, Primavera, & Knight, 2001). General social support was measured by the Medical Outcomes Study (MOS) Social Support Survey which is a brief instrument developed for patients with chronic conditions (Sherbourne & Stewart, 1991).
The Center for Epidemologic Studies Depression Scale (CES-D) was used to measure depressive symptoms (Radloff, 1977). The CES-D has been validated in SMI populations, and can identify depression in people with schizophrenia (Weissman, Sholomskas, Pottenger, Prusoff, & Locke, 1977). The Symptom Checklist-90 (SCL-90) was used to measure severity of psychiatric symptoms (Derogatis, Lipman, & Covi, 1973). The physical functioning and mental functioning composite scores of the MOS SF-12 were used to assess health status. These measures have been validated among adults with SMI (Salyers, Bosworth, Swanson, Lamb-Pagone, & Osher, 2000).
After obtaining informed consent, baseline measures, and enrollment consent, participants were randomly assigned to a treatment group. Participants were assigned to a treatment group based on a random number generation list that was stratified by sex. After the group assignment was made, the participants met with the intervention director to learn about opportunities and expectations associated with their group assignments.
The control group was offered moderate intensity low-impact group exercise classes three days per week for 16 weeks. Group exercise classes were held on-site at the outpatient centers. Trained exercise instructor lead the classes and were supervised by the study interventionist who took attendance. The classes were progressive with respect to length of class and complexity of the exercise routine and included a warm-up, moderate intensity low impact aerobics and a cool down. They started at a level appropriate for a sedentary individual with no experience in a group exercise class and gradually progressed in order to build participant confidence in their ability to complete 45 minutes of exercise three days per week. Examples of adapting the group exercise protocol for this population included: opportunities to practice more challenging movements during warm up at a slow pace; cueing which includes verbal direction, modeling by the instructor and hand directions; options to modify intensity and movements based on participant limitations; and emphasis on returning to a calm state both physiologically and psychologically during cool down. Due to the heterogeneity of the study population and the open door policy for participantsto rejoin the program after extended absences, particular emphasis was placed on providing options for different intensity levels. This included traditional techniques for changing intensity (e.g. inclusion/exclusion of arm movements) and prompting for participants to take a break if needed. The structure and format of these classes were informed by our previous efforts promoting physical activity among adults with mental illness (Daumit et al., 2010).
Information sessions followed group exercise classes during the first few weeks. These sessions covered topics such as appropriate clothing and footwear, importance of hydration, importance of compliance with medicine, and knowing when to take a break. When conducting the intervention, we gave particular consideration to potential cognitive deficits of the participants and all intervention materials were at a 5th to 8th grade reading level. Although we had planned on providing handouts associated with these brief educational sessions we found that participants did not keep the handouts. After the first cohort we stopped providing handouts and increased the extent the group exercise leader provided verbal reinforcement of these points during the 16 week exercise class.
The group exercise plus peer support group was offered group exercise classes following the same frequency, format and procedures as the control group. In addition, those in the group exercise plus peer support condition were assigned to a peer leader who attended class and met weekly with the participant. The main goal of the peer support was to encourage and motivate participants to attend and participate in the exercise classes. Each peer leader mentored 1–4 participants. The mentoring was theoretically based and focused on providing social support, role modeling, building self-efficacy, self-monitoring, goal setting and problem solving. The intervention director was present during interactions between the peer leaders and the participants, and provided follow-up or debriefing as needed.
The peer leaders attended the group exercise classes with the participants and worked with them in completing a tracking sheet after each exercise class. This was done in the group exercise room with the intervention director present. Tracking sheets were maintained in a file box stored in a locked closet on site. The tracking sheet was used to record the day of the week, the scheduled minutes of exercise, minutes actually exercised, plan for getting to exercise class on time (ex: Leave house at 11am to take 11:20 bus to get to class by 12pm) and a space for a star if the weekly goals were achieved. Gold stars were used in a reinforcement program in which participants could save points and redeem them for a range of small items (pedometers, t-shirts, gym bags, gym socks). Once a week the peer leaders and participants used the tracking sheet to summarize each participant’s progress for the week and helped to set goals and identify corresponding strategies for the upcoming week. This was done in a group setting with the intervention director present. Progress was charted both individually and for the class as a whole. Total time spent exercising among both peer leaders and participants was translated into miles walked and tracked on a map of the United States. The group chose destination goals and worked as a group to reach these destinations. The peer leaders also lead efforts to encourage participation among those who missed exercise class. This included writing notes to participants who had prolonged absences to encourage participation and get well cards for those who were ill. The intervention director would mail these notes to participants and the intervention did not facilitate an exchange of home addresses or telephone numbers among the participants.
The first cohort of potential peer leaders was identified with assistance from the therapists and interviewed by data collectors to determine their eligibility for training. The initial peer leader cohort completed 10 weeks of training, which included group exercise class three times per week. These classes served to increase the peer leader’s confidence with the exercise format and progression. Peer leaders also received training on methods for providing peer support, being a role model, self-monitoring, planning, positive interactions, and problem solving. The training included role modeling from the intervention director who met individually with the peer leader trainees to model an appropriate mentoring session. During the second half of the training the peer leaders mentored each other. During the intervention, the intervention director had weekly training sessions (approximately 60 minutes) with the peer leaders as a group. This meeting was used to provide refresher training, addressed challenges the peers experienced during the previous week, and helped to prepare the peer leaders for the following week. The peer leader training was based on the study team’s experience with informal peer leaders during group exercise activities among adults with mental illness and from effective peer outreach models (Daumit et al., 2010; Latkin, 1998).
After the first cohort, peer leaders were selected from previous peer leaders and participants from earlier cohorts. In cohorts 2-4 the length of peer leader training was decreased based on the experience of the peer leaders. Peer leaders were compensated for their time and could earn up to $400 for their services. They were paid monthly based on their attendance and participation.
The main results paper will examine the effect of peer support on adherence to the exercise program and corresponding changes in fitness. The primary outcome will be change in cardiovascular fitness (i.e. time on treadmill, heart rate response and maximum METs) from a peak treadmill exercise test. We will examine changes from baseline to follow-up and between group comparison of these changes. Cardiovascular fitness was selected as the primary outcome as it is associated with lower rates of cardiovascular disease and all-cause mortality independent of BMI classification. (Wei et al., 1999).
The current analyses report descriptive statistics on baseline characteristics. We used an established cut-point for depressive symptoms (CES-D < 16) where higher CES-D scores indicate higher levels of depressive symptoms (Radloff, 1977). Similarly, higher SCL-90 scores indicate higher severity of psychological symptoms and previous reports identified 0.91 as an appropriate cut-point between community populations and clinical samples (Holi, Marttunen, & Aalberg, 2003). The following criteria were used to identify hypertension: blood pressure above 140/90 mmHg, antihypertension medications or positive response on medical history. Hyperlipidemia was defined as LDL > 130 mg/dL or use of cholesterol lowering medications. Diabetes was identified by glucose > 126 mg/dL, diabetes medications or positive response on medical history.
These analyses also examined the 1) association among cardiovascular fitness measures; 2) comparison of fitness levels to a national sample; 3) association among cardiovascular fitness measures, race, sex, and BMI; 4) association among cardiovascular fitness measures and other physical measures; and 5) association of cardiovascular fitness measures with psychosocial variables, self reported health, depression and psychiatric symptoms. If participants transitioned into a peer leader role, their peer leader data was not used in comparisons to ensure independence of the data. Time on treadmill was compared to a national sample matched on sex and age (Thompson, 2009). Spearmen rank-order correlations were reported for all bivariate associations. Chi-square tests and Student’s t-tests were used to examine between group differences. All analyses were conducted using SAS, version 9.2 (SAS Institute Inc, Cary, North Carolina), and all p values were two-sided.
As seen in Figure 1 there were 169 individuals screened and 93 were randomized to either the exercise (n = 44) or the exercise plus peer support (n = 49) conditions. There were four cohorts recruited in series across the two year study. Additionally there were 18 peer leaders including some individuals who transitioned from being a participant to being a peer leader (n = 4). Cohort 1 included 12 peer leaders and 19 participants in the exercise plus peer support condition. In cohorts 2–4 there were 3–4 peer leaders per 10 participants. Both the participants and the peer leaders were predominantly female (76% and 83% respectively) and African American (72% and 89% respectively). The average age was 47 years and 40% of the participants and 60% of the peer leaders had completed some college (Table 1). A larger percentage of the peer leaders (77.8%) were taking antidepressants compared to the participants (50.8%). There were no other differences in mental health status and symptoms between participants and peer leaders.
The most frequently identified primary diagnosis was depression (44–57%) followed by bipolar disorder (25–33%). Participants were taking an average of 2.2(SD 1.1) psychotropic medications. Among the sample 51% had elevated depression symptoms based on the CES-D and 37% had mental health symptoms above what is expected from community samples. Among participants, 68% had hypertension, 52% had hyperlipidemia and 36% had diabetes. Seventy-six percent of the participants were obese and 15% were overweight.
At baseline there were no differences in average MET level between treatment conditions (Exercise 6.3(SD 2.1); Exercise plus Peer Support 5.6(SD 2.4)) and peer leaders 6.2(SD 2.3). When time on treadmill was matched by sex and age to a national sample (data not shown), 88% of both the men and women had fitness levels corresponding with the lowest decile and only 8% of the men and 4% of the women had above average fitness levels (Thompson, 2009). There were significant correlations (ps < 0.05) between the four different fitness measures derived from the treadmill test including total time on treadmill, maximal MET level achieved, and heart rate response at stages 1 and 2 (Table 2). There were also significant correlations (ps < 0.05) between fitness as measured by the treadmill test and performance on the six-minute walk test as higher levels of fitness as measured during the treadmill test corresponded with longer distances covered during the six-minute walk.
As shown in Table 3 men had higher levels of fitness than women (ps = 0.002–0.015). Compared to African Americans, Non-African Americans had higher fitness levels based on heart rate response at stage 2 and distance walked on the six-minute walk (ps = 0.002–0.031). Compared to obese individuals, overweight participants had higher levels of fitness based on heart rate response on the treadmill test (ps = 0.001–0.007).
All fitness measures were significantly correlated with measures of adiposity including BMI, waist circumference and percent body fat (rs = 0.16–0.53). As seen in Table 4, a lower heart rate response was associated with both lower levels of glucoses and insulin. Better performance on the six-minute walk was associated with lower diastolic blood pressure and better insulin functioning.
There were no significant relationships between fitness and mental health status, general self-efficacy, or symptoms (data not shown). Exercise Self-efficacy was associated with time on treadmill (rs = 0.28, p < 0.05), METS (rs = 0.29, p < 0.05), heart rate response at stages 1 (rs = −0.18, n.s.) and 2 (rs = −0.19, n.s.), and the six-minute fitness walk (rs = 0.32, p < 0.05). There was a modest yet statistically significant association between distance walked on the six-minute walk test and general social support (rs = −0.24, p = 0.015).
Regular physical activity has the potential to provide both physical and mental health benefits for adults with SMI. Unfortunately, few adults with SMI engage in sufficient physical activity to realize these benefits (Jerome et al., 2009; Ussher et al., 2007). This pilot study will investigate the effect of peer support on adherence to a group exercise program in this population. Social support is an important determinant of physical activity and there is increasing evidence that peer support is an effective adjunct in behavior change among adults with SMI. (Davidson et al., 2006; Fisher, 1988; Needle et al., 1994; Solomon, 2004; Williams, 2007). The ACE trial will investigate this promising theory based approached to increasing adherence to exercise classes.
Although the health benefits of increased cardiovascular fitness are well documented there have been few studies examining this association among adults with SMI. The low fitness levels of this sample substantiate the need for physical activity programming among this group. At baseline females and African Americans had lower fitness than males and non-African Americans suggesting that within this sample sex and ethnicity may be associated with health. The design of the trial will allow us to examine the association between changes in cardiovascular fitness and other changes in physical health by sex and race.
The six-minute walk test was appropriate for use in the field with this population. Strong associations with laboratory measures of fitness supported the validity of the test. The walk test holds potential for program evaluation, yet the ability of the test to detect changes in response to an intervention needs verification. The ACE trial will provide an opportunity to examine the sensitivity of this measure to identify changes in cardiovascular fitness among adults with SMI.
The extant literature reviewing intervention studies indicated that high levels of physical activity are associated with better mental health among adults with SMI (Faulkner & Biddle, 1999; Gorczynski & Faulkner, 2010; Lawlor & Hopker, 2001). The current cross-sectional results did not support these findings. Caution is warranted when interpreting the current results given the uniformly low fitness levels of our participants. Results from this pilot study will allow us to examine the association among changes in fitness and changes in these psychosocial and mental health measures. These baseline data supported the relationships between self-efficacy for exercise and fitness levels but not for the association among general efficacy and fitness. These findings validate the theoretical association between self-efficacy and behavior in this population and highlight the importance of using behavior specific efficacy measures.
There was a range of diagnoses among the participants. This limits our ability to draw conclusions about the effectiveness of the treatment among any specific group, yet increases the generalizability of the findings to the heterogeneity of diagnoses found in outpatient clinics. Moreover there were few statistically significant differences between participants and peer leaders even though the latter were selected based on their perceived ability to act as role models. The small sample size and heterogeneity of the sample may limit our ability to detect differences between these groups. Additionally, this was a descriptive paper and we did not adjust for multiple comparisons. Interpretation of the reported differences should take this into consideration.
In summary, the ACE trial was designed to examine the impact of peer support on adherence to group exercise program among adults with mental illness. The study has a number of strengths including standardized physical measurements and exercise classes designed to increasing aerobic fitness in this population (Daumit et al., 2010). The results should help improve our understanding of the association among cardiovascular fitness and the physical and mental health of those with serious mental illness. We expect study results will be salient to planning future physical activity interventions in mental health outpatient clinics.
We provide rationale and design for a pilot exercise study among adults with serious mental illness.
Adults with serious mental illness had uniformly low cardiovascular fitness.
Among adults with serious mental illness, cardiovascular fitness correlated with BMI.
This study was supported by National Institute of Mental Health Grant 5R34MHO078613 and by Grant Number UL1 RR 025005 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research. The authors were responsible for study design and conduct as well as the analyses. We would like to thank both Johns Hopkins Hospital Outpatient Community Psychiatry Program and the Outpatient Community Psychiatry Program at Johns Hopkins Bayview Medical Center for their assistance with this study. We give our thanks to the participants.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest
The authors have no conflicts of interest to declare.