Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arthritis Rheum. Author manuscript; available in PMC 2010 July 15.
Published in final edited form as:
PMCID: PMC2710408

Comparison of Two Methods of Conducting Fit and Strong!



Fit and Strong! is an award winning, evidence-based multiple component physical activity/behavior change intervention. It is a group and facility-based program that meets for 90 minutes three times per week for eight weeks (24 sessions total). We originally tested Fit and Strong! using Physical Therapists (PT) as instructors but have transitioned to using nationally Certified Exercise Instructors (CEI) as part of an effort to translate Fit and Strong! into community based settings. This paper presents findings from our test of the impact of this shift in instruction type on participant outcomes.


We used a two group design. The first 161 participants to sequentially enroll received instruction from PTs. The next 190 sequential enrollees received instruction from CEIs. All participants were assessed at baseline, at the conclusion of the eight-week Fit and Strong! program, and at 6 months.


We saw no significant differences by group on outcomes at eight weeks or 6 months. Participants in both groups improved significantly with respect to lower-extremity strength, aerobic capacity, pain, stiffness, and physical function. Significant differences favoring the PT-led classes were seen on two of five mediators, self-efficacy for exercise and barriers adherence efficacy. Participant evaluations rated both types of instruction equally highly, attendance was identical and no untoward health events were observed or reported under either instruction mode.


We conclude that outcomes under the two types of instruction are remarkably stable. These findings justify the use of CEIs in the future to extend the reach of Fit and Strong!.

Given dramatic growth in the number of older adults in the U.S. and the high prevalence of sedentary lifestyles among them, the need to efficiently and effectively translate and diffuse evidence-based physical activity programs specifically designed for this population is paramount. The need to diffuse evidence-based programming for older adults with osteoarthritis (OA) is particularly acute. Osteoarthritis is not only the most common chronic condition among older adults, but is a major barrier to participation in physical activity, and is a known risk factor for disability and institutionalization.1,2 Fit and Strong! is an evidence-based physical activity/behavior change program that effectively targets this high risk group. It addresses documented strength and aerobic deficits in this population and is inexpensive and simple to replicate.3,4

Fit and Strong! is a multiple component physical activity/behavior change intervention. It is a group and facility-based program that meets for 90 minutes three times per week for eight weeks (24 sessions total). The first 60 minutes consist of a multiple-component exercise program that incorporates flexibility/balance, aerobic walking and/or low impact aerobics and lower-extremity strength training using elastic exercise bands and adjustable ankle cuff weights. The remaining 30 minutes of each session are devoted to group problem-solving and education using a curriculum designed to facilitate arthritis symptom management, self-efficacy (SE) for exercise, and commitment to lifestyle change.5 In Week 6, participants meet with instructors to develop an individualized physical activity plan of their choice that can include home-based exercise or an ongoing group/facility-based program (or combination of the two) with the goal of maintaining 20 minutes of flexibility, 20 minutes of aerobic and 20 minutes of resistance training a minimum of 3 times per week. This plan becomes a physical activity maintenance contract that each participant signs at a graduation ceremony on the last day of class.

We tested the efficacy of Fit and Strong! in a randomized trial with 215 treatment and control participants.5,6 Relative to controls, treatment participants experienced statistically significant improvements in SE for exercise (p=0.001), exercise participation (p=0.000), and lower-extremity stiffness (p=0.018) at the conclusion of Fit and Strong!, 8 weeks from baseline. These benefits were maintained at six months when several other outcomes also were significant: increased time adherence efficacy (p=0.001), reduced lower-extremity pain (p=0.040), and a marginally significant increase in SE for arthritis pain management (p=.052). Despite a substantially smaller sample size at 12 months, significant treatment effects were maintained on SE for exercise (p=0.006) and exercise participation (p=0.001), that were accompanied by marginally significant reductions in lower-extremity stiffness (p=0.056) and pain (p=0.066) No adverse health effects were reported. Effect sizes for SE for exercise and for exercise participation were 0.798 and 0.713, and 0.905 and 0.669, respectively, in the treatment group at six and 12 months.

In an effort to minimize harm to participants, we originally developed and tested Fit and Strong! using physical therapists (PTs) who had a substantial amount of experience working with older adults. In 2003, we obtained an NIA grant (R01 AG23424) to replicate Fit and Strong! at seven senior centers in Chicago and to compare different methods of fostering maintenance of physical activity after Fit and Strong! ended. Most evidence-based programs need to be adapted when they are diffused into the community.7,8 We were concerned that reliance on a program led by PTs could limit the reach of Fit and Strong! in poor and underserved communities. Therefore, we used the current study as an opportunity to refine Fit and Strong! training materials and to shift from using PT instructors to using certified exercise instructors (CEIs). We made this shift in the middle of the study, which enabled us to compare outcomes under the two different instruction types. This paper compares the same outcomes collected for 161 PT-led participants to190 CEI-led participants.

Patients and Methods


We used a two group pre/posttest design to evaluate the impact of this adaptation in the mode of instruction of Fit and Strong!. We hypothesized that participants in both modes of instruction would have comparable outcomes at 8 weeks and 6 months. Participants were assessed at baseline (prior to inclusion in the program), at the eight-week conclusion of Fit and Strong!, and at 6 months from baseline.


The study was conducted at local senior centers. Participants were community-dwelling older adults who were recruited by newsletters, local media announcements, and presentations to senior groups. Importantly, participants were recruited from the same sources and received the program in the same settings under both modes of instruction.

Inclusion/Exclusion Criteria

Persons were considered ineligible if they were under 60, currently participated in an exercise program, had undergone uncomplicated hip or knee surgery within the previous six months or complicated surgery within the past year, had received steroid injections within the previous three months, had moderate to severe cognitive impairment, had rheumatoid arthritis, or had diabetes or blood pressure that was not under good control.

Potential enrollees were also examined by the study rheumatologist to determine clinical presence of OA of the hip or knee and to rate degree of functional significance using a modified version of the American College of Rheumatology (ACR) Functional Class.9 Clinical criteria for the presence of knee OA were knee pain plus at least 3 of the following 6 clinical findings: morning stiffness < 30 minutes’ duration, crepitus on active motion, tenderness of the bony margins of the joint, bony enlargement on examination, and a lack of palpable warmth of the synovium.10 Persons were classified as having hip OA if pain was present in combination with either 1) hip internal rotation greater than or equal to 15 degrees, pain present on internal rotation of the hip, morning stiffness of the hip for 60 minutes or less OR 2) hip internal rotation < 15 degrees, and hip flexion less than or equal to 115 degrees (sensitivity 86% and specificity 75%).11



We used the 10-item Short Portable Mental Status Questionnaire to screen for presence of moderate to severe cognitive impairment.12

The following mediators and outcomes were assessed at baseline, at the end of the 8 week Fit and Strong! program, and at 6 months for all participants.



SE to perform self-management tasks was assessed using three subscales developed by Lorig et al.: SE for exercise (3 items), SE for pain management (5 items), and SE for other symptom management (6 items).14,15 All three subscales have 10-point response formats and are scored by calculating an overall subscale mean.

We used two scales developed by McAuley and colleagues to measure self-efficacy for exercise adherence.16 The “barriers” adherence scale (13 items) measures self-efficacy to adhere to exercise in the presence of barriers. The “time” adherence measure (6 items) requires respondents to rate their self-efficacy to continue participating in regular exercise over a period of six months. Both scales have 0-100 point response formats and are scored by calculating an overall mean score, with higher scores indicating higher levels of self-efficacy.



Attendance was monitored and documented at each session.

Exercise Maintenance: CHAMPS

We used the Community Healthy Activities Model Program for older adults (CHAMPS) measure to assess maintenance of physical activity. The CHAMPS has 41 items that assess participation in leisure-time, moderate and vigorous physical activity, and non-exercise activities (e.g., reading, church attendance, volunteering). The CHAMPS provides frequencies of exercise participation and estimates of weekly caloric expenditure, and is valid, reliable, and sensitive to change.20

Performance Measures

Functional Lower-Extremity Muscle Strength

We used the Timed-Stands test to functionally assess lower-extremity muscle strength and endurance.17 Participants are asked to sit in a straight back chair and rise with their arms folded. 10 If they successfully rose, they were asked to stand and sit down five times as quickly as possible. Raw scores were transformed into a rate per minute to assess change in those who were unable to perform the test at any point.

Functional Exercise Capacity

We used the six-minute distance walk test to measure functional exercise capacity.18 Participants were asked to walk for 6-minutes, accompanied by research staff who measured distance walked in feet.


We used the Western Ontario and McMasters University Osteoarthritis Index (WOMAC) self-report instrument to examine lower-extremity pain (5 items), stiffness (2 items), and physical function (17 items).19

Participant Program Evaluations

All Fit and Strong! participants completed a 75-item program evaluation that assessed participants’ satisfaction with the physical environment, exercise program, group discussion sessions, manual, and overall satisfaction with the instructor and program.

Independent Variables

Participant demographic variables included participants’ age, race, sex, income, and education.

Instructor Type

The primary independent variable --instructor type --was coded 1 for PT-led groups and 0 for CEI-led groups. PTs were required to have a Bachelor’s or Master’s degree in Physical Therapy, current PT licensure (APTA), liability insurance, and experience working with older adults with OA. They were recruited through postings at hospitals, clinics, and universities, and referrals. Six PTs conducted the first 12 sessions of Fit and Strong!

CEIs were recruited through online postings, instructor certification websites, and referrals. We sought CEIs who had experience in group fitness instruction and in working with older adults. The instructors were required to have some type of nationally recognized certification (e.g., Aerobic and Fitness Association of America, American College of Sports Medicine) and possess liability insurance. Certification programs vary by organization, but are typically offered in a classroom format, manual-based for home study, or online. The courses include information on anatomy and physiology, appropriate technique, health behavior change strategies, injury prevention, principles of group exercise, instruction guidelines, class formats, and working with special populations. Although not a requirement, many CEIs had college degrees in a health-related field (e.g., Exercise Physiology, Nutrition). Six CEIs instructed the next nine Fit and Strong! sessions.

PTs and CEIs participated in a half-day training before beginning instruction. The training was led by the PI, the study PT, and the project manager. Training included an overview of how to work with older adults and people with OA, the background of the study, a detailed description of the components of Fit and Strong!, and implementation issues. The PT training devoted more time on how to lead a class and engage participants in group exercises by varying routines. The CEI training devoted more time to describing the symptoms of OA and the types and intensity of exercises that are appropriate for older adults with OA. Project staff, including the Principal Investigator and the study PT, conducted site visits to assist the instructors and monitor the fidelity of the program. Adherence to the exercise routine, appropriateness of modifications to exercises for participants of varying levels of ability, and facilitation of the group problem-solving and education component were assessed. Feedback was provided to instructors following each site visit. Site visits were conducted at a minimum on the first day of class, at the midpoint, and the last day of class.


We examined overall changes in study outcomes using a random effects model to account for repeated measures. Data analysis involved one between subjects factor (PT vs. CEIled class) and one within subject factor (time). We treated time non-linearly by including indicator variables for the two measurement points. Because we had repeated measures data and different numbers of respondents by instructor type over time, we analyzed the data using a random intercept model which assumes that data are missing at random (MAR) conditional on covariates. A simple random effects model for the data can be written as:


where the interaction terms, Instruct*Time2, Instruct* Time6 test whether the two instructor types changed differently over time. We included one covariate, Arthritis Functional Class, in order to control for baseline disease severity


Table 1 shows baseline characteristics of participants by instructor type. Across both groups, participants had a mean age of 71 (not shown), the majority were female, white, had annual incomes less than $30,000, had at least a high school education, and had Class 2 ARA functional class scores. Approximately 64% of the total sample also reported presence of hypertension, 24% reported diabetes, and 15% reported other cardiovascular disease. Importantly, no significant differences were noted by instructor type on any of the demographic or disease measures.

Table 1
Participant Baseline Demographic Characteristics by Instruction Mode


Similar proportions (16% of PT-led participants (N=19) and 18% (N=27) of CEIled participants) failed to complete Fit and Strong!. The three most common reasons for class drop out were health concerns (n=22), time/schedule conflicts (n=14), and caregiving responsibilities (n=3). No significant differences between PT and CEI-led class participants were found for Fit and Strong! class attendance (20.0 classes, 83%, vs. 19.1 classes, 80%, respectively).

Posttest Attrition

Eight week posttests were obtained on 73% of PT group participants (N=117) and 79% of CEI group participants (N=151); at 6 months, posttests were obtained on 58% of PT-led participants (N=93) and 68% of CEI-led participants (N=129). A logistic regression model compared those who remained in the study (responders) to those who left (non-responders).22 We regressed a variable indicating continued participation in the study on each of the baseline values of the outcome variables, a dummy variable for instructor type, and the interaction of the two. These analyses demonstrated no significant differences between responders and non-responders on any of the outcome measures (Table 2). Importantly, the comparison in attrition rate between responders and non-responders by instructor type was not significant (p=0.142). The analyses also found no statistically significant differences between responders and non-responders on demographic characteristics, arthritis severity, or on the baseline value of the outcome measures. However, non-responders had slightly worse scores on the outcome measures at baseline.

Table 2
Attrition: logistic regression comparing participant responders and non-responders

Program Evaluations

No significant differences between groups were found on responses to the program evaluations. Across all domains assessed, participants in both conditions were overwhelmingly positive about Fit and Strong!. One hundred percent of PT-led and 99% of CEIled participants rated their overall reaction to the Fit and Strong! instructor as “excellent” or “good.” Ninety-eight percent of participants in both instruction groups felt that they benefited from Fit and Strong!, and 99% of participants in both types of instruction incorporated information learned during Fit and Strong! into their lifestyle. There were considerable site differences in responses to questions about environment (e.g., temperature, parking), but no differences related to instructor type.

Baseline, 8-week, and 6 month findings

Raw scores for mediators and outcomes are shown broken down by instruction type in Table 3. Table 4 presents pre-posttest findings on mediators and outcomes for the entire sample of participants (main effects are shown first in Table 4 and then are broken down by type of instruction). Examination of scores for the total sample (Table 4) revealed that participants as a whole changed significantly at 8 weeks relative to baseline on two mediators: SE for other symptom management and barriers adherence efficacy. SE for other symptom management increased significantly, while barriers adherence SE declined significantly between baseline and 8 weeks and this decline was maintained at 6 months. No significant pre-post differences were seen on SE for exercise, pain management or time adherence. Participants as a whole also improved significantly at 8 weeks on frequency of participation in all and moderate exercise activities, caloric expenditure on all and moderate exercise activities, timed sit-stand, 6-minute distance walk, stiffness, and physical function. These benefits were maintained at 6 months.

Table 3
Baseline, 8-week, and 6 months outcomes over time by instructor typea
Table 4
Random Effects Model: Baseline, 8-week, and 6 month Outcomes by time and instructor type a,b


Lorig Self-Efficacy Scales

A significant difference (p=0.001) was seen favoring PT-led class participants at 8 weeks on the Lorig Exercise Efficacy Scale. This difference was not maintained at 6 months. PT-led participant scores increased by about one point at 8 weeks; whereas CEI-led group scores declined by four-tenths of a point. At six months, PT-led participant scores decreased almost to baseline levels while CEI-led participant scores increased to baseline levels. The interactions between instructor type and time on the Lorig Pain and Symptom Management Scales at 8 weeks and 6 months were not significant.

McAuley Barriers and Time Exercise Adherence Efficacy

A significant difference favoring the PT-led classes was seen on the barriers adherence self-efficacy scale (p=0.001) at 8 weeks and 6 months. Barriers adherence declined significantly by 13 points across both groups between baseline, 8 weeks, and 6 months. However, less decline was experienced in the PT-led group (7% vs 22%, respectively). The interaction between instructor type and time on the time adherence self-efficacy scale was not significant at 8 weeks or 6 months.


Exercise Maintenance (CHAMPS)

Participants in both groups reported significant increases in weekly frequency of participation in all and moderate physical activities, and in caloric expenditure at 8 weeks and 6 months. The interactions between instructor type and time were not significant for any of the four CHAMPS measures at 8 weeks or 6 months.

Functional Lower-Extremity Muscle Strength (Timed-Stand)

Participants in both PT and CEI-led classes demonstrated significant improvement on the timed-stand measure between baseline and 8 weeks (p=0.000) and baseline and 6 months (p=0.000). The interaction between instructor type and time was not significant for this measure.

Functional Exercise Capacity (Six-minute Distance Walk)

Participants in both instructor types demonstrated significant improvements at 8 weeks (p=0.021) and 6 months (p=0.000). A slightly larger improvement occurred among CEI-led class participants (90.73 greater vs. 70.16 greater). The interaction between instructor type and time for the 6-minute distance walk scores was not significant at 8 weeks (p=0.749) or 6 months (p=0.47


Pain improved slightly in both groups between baseline and 8 weeks. Joint stiffness decreased among participants in both PT and CEI –led classes at 8 weeks and 6 months as did functional impairment at the same time points. As evidenced by the interaction between instructor type and time, no significant differences were seen between PT and CEI-led participants on the WOMAC pain, stiffness, or physical function subscales.

Adverse Health Outcomes

No adverse health outcomes were reported by participants in either type of instruction.


Although the need to adapt evidence-based programs in order to maximize their diffusion into community settings is widely recognized, little evidence has been reported in the literature regarding the impact of adaptations on program participants.23,24 Lorig and colleagues evaluated implementing the Chronic Disease Self-Management Program (CDSMP) using different modes of delivery and different levels of instructors. The first Lorig study randomized participants to either the traditional, small-group version of CDSMP or to an Internet-based version of the program.25 At one year, participants in both groups experienced significant improvement in health status. The second study compared lay-taught and professional- taught arthritis self management courses in order to assess the impact of different levels of instructors.26 Participants under both modes of instruction improved significantly on knowledge compared to a no-treatment control group. However, professional-led participants had a significantly larger improvement in knowledge at four months; whereas, those in the lay-led groups improved significantly relative to professional-led participants on the practice of relaxation. The results of both studies suggest that similar outcomes can be achieved with varying levels of instructor education and modes of delivery.

Aside from the Lorig studies, we could not find any studies that compared outcomes achieved by varying levels of professional leaders or by different levels of instructors of exercise interventions. This paper addresses this gap in the literature by providing findings from our recent adaptation of the type of instruction used in Fit and Strong! Findings from our analyses show that when participants were aggregated across both types of instruction they improved significantly with respect to exercise participation, caloric expenditure, lower-extremity stiffness, physical function, and on performance measures that assessed lower-extremity strength and aerobic capacity.

Importantly, when findings were examined separately by instructor group, no adverse events were reported by participants, attendance levels were virtually identical, and both types of instructors were highly rated by participant evaluations. Superior results were experienced by the PT-led group participants on two of five SE measures used as mediators in this study: SE for exercise and barriers adherence efficacy. Exercise SE scores increased among PT-led participants and decreased slightly for CEI-led participants at 8 weeks, but at six months no differences were seen by instruction type. Barriers adherence efficacy scores decreased under both modes of instruction, but PT-led participants experienced less decline than CEI-led participants (4.5 point vs. 14.8 point decrease).

Our findings of an impact on outcomes minus or with a diminution of strength in mediators are somewhat unusual and counterintuitive to theory. It is possible that PT-led participants differentially improved on SE for exercise and experienced less decline in barriers adherence efficacy because the clinical training of PTs is geared to therapeutic exercise. The clinical training for PTs may put more emphasis on individual mastery of the exercises in order to achieve independent function. However, no differences on adherence, pain, stiffness, physical function, or performance outcomes were seen by instructor type at 6 months despite the differences seen in these two SE variables. The differential SE findings may also indicate that the training for the CEIs should be expanded to accommodate more emphasis on techniques to foster SE, while still addressing the special exercise needs of older adults with OA. Partially in response to these findings we have increased the CEI training time from a half to a full day. Additionally, the significant decline in barriers adherence efficacy among participants in both conditions may reflect an unrealistic degree of optimism about ability to overcome barriers to exercise that is present at baseline among sedentary volunteers. Actual participation may cause participants’ confidence to decrease to a lower, but potentially more realistic, level. We plan to explore this finding further in future quantitative and qualitative work.

It is important to note that the findings that we report here were obtained under ideal dissemination circumstances. We were able to monitor instructors and classes reasonably closely because instructors were employed directly by the study team and sites were clustered in a relatively small geographic area. This proximity enabled us to provide regular feedback to instructors during the program, and at the beginning and end of each iteration. These training sessions provided structured opportunities to discuss and correct problems that arose, share lessons learned, and to serve as “boosters” to reiterate core components, goals, and objectives of Fit and Strong!. Monitoring instructors and treatment fidelity will be more challenging when disseminating Fit and Strong! across the country.

Despite these limitations, we conclude that these results of no differences by group on the outcomes tested are strongly positive findings. PT input into the program design and initial implementation was critically important in fostering the efficacy and safety of the program for older adults with lower-extremity osteoarthritis, but the transition to using CEIs is beneficial for community translation. Fit and Strong! is less expensive to provide using CEIs (~$20/hr for CEIs vs. ~$45/hr for PTs) and these instructors are more likely to be available, especially in rural and underserved urban areas. We look forward to other opportunities to examine further adaptations to Fit and Strong! with older adults in other parts of the country in the future.


This study was funded by the National Institute on Aging (Grant #: R01 AG23424). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Aging or the National Institutes of Health. We gratefully acknowledge the many participants who took part in the study. We also acknowledge the involvement and significant contribution of the following organizations and individuals: Chicago Department of Senior Services, Greater Chicago Chapter of the Arthritis Foundation, National Arthritis Foundation, and the UIC Survey Research Lab.

Funded by: National Institute on Aging, National Institutes of Health (R01 AG23424)


1. Penninx BWJH, Rejeski WJ, Pandya J, Miller ME, Di Bari M, Applegate WB, Pahor M. Exercise and depressive symptoms: A comparison of aerobic and resistance exercise effects on emotional and physical function in older persons with high and low depressive symptomology. J Gerontol B Psychol Sci Soc Sci. 2002;57B:P124–P132. [PubMed]
2. Dunlop D, Hughes SL, Edelman P, Singer R, Chang R. Impact of joint impairment on disability and disability-specific domains at four years. J Clin Epidemiol. 1998;51:1253–1261. [PubMed]
3. Minor MA, Hewett JE, Weber RR, Anderson SK, Kay DR. Efficacy of physical conditioning exercise in patients with rheumatoid arthritis and osteoarthritis. Arthritis Rheum. 1989;32:1396–1405. [PubMed]
4. Semble EL, Loeser RF, Wise CM. Therapeutic exercise for rheumatoid arthritis and osteoarthritis. Semin Arthritis Rheum. 1990;20:32–40. [PubMed]
5. Hughes SL, Seymour RB, Campbell R, Pollak N, Huber G, Sharma L. Impact of the fit and strong intervention on older adults with osteoarthritis. Gerontologist. 2004;44:217–228. [PubMed]
6. Hughes SL, Seymour RB, Campbell RT, et al. Long-term impact of Fit and Strong! on Older Adults with osteoarthritis. Gerontologist. 2006;46:801–814. [PubMed]
7. Glasgow RE, Vogt TM, Boles SM. Evaluating the Public Health Impact of Health Promotion Interventions: The RE-AIM Framework. Am J Public Health. 1999;89:1322–1327. [PubMed]
8. Greene LW, Glasgow RE. Evaluating the Relevance, Generalization, and Applicability of the Research: Issues in External Research and Translation Methodology. Eval Health Prof. 2006;29:126–153. [PubMed]
9. American Rheumatism Association. Dictionary of the Rheumatic Diseases. Vol. 1. New York: Contact Association International; 1982.
10. Altman R, Asch E, et al. Development of criteria for the classification and reporting of osteoarthritis. Arthritis Rheum. 1986;29:1039–1049. [PubMed]
11. Altman R, Alarcon G, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum. 1991;34:505–514. [PubMed]
12. Kahn RL, Goldfarb AI, Pollack M, Peck A. Brief objective measures for the determination of mental status in the aged. Am J Psychiatry. 1960;117:120–124. [PubMed]
13. Hughes SL, Edelman P, Chang RW, Singer RH, Schuette P. The GERI-AIMS, Reliability and validity of the Arthritis Impact Measurement Scales adapted for elderly respondents. Arthritis Rheum. 1991;34:856–865. [PubMed]
14. Lorig KR, Chastain RL, Ung E, Shoor S, Holman HR. Development and evaluation of a scale to measure perceived self-efficacy in people with arthritis. Arthritis Rheum. 1989;32:37–44. [PubMed]
15. Lorig K, Stewart A, et al. Outcome Measures for Health Education and other Health Care Interventions. Thousand Oaks, CA: Sage Publications; 2006.
16. McAuley E, Lox C, Duncan TE. Long-term maintenance of exercise, self-efficacy, and physiological change in older adults. J Gerontol B Psychol Sci Soc Sci. 1993;48:218–224. [PubMed]
17. Guralnik JM, Ferrucci L, Simonsick EM, Salive ME, Wallace RB. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995;332:556–561. [PubMed]
18. Guyatt GH, Sullivan MJ, et al. The 6-minute walk, A new measure of exercise capacity in patients with chronic heart failure. Can Med Assoc J. 1985;132:919–923. [PMC free article] [PubMed]
19. Bellamy N. Pain assessment in osteoarthritis, Experience with the WOMAC osteoarthritis index. Semin Arthritis Rheum. 1989;18:14–17. [PubMed]
20. Stewart AL, Mills KM, King AC, Haskell WL, Gillis D, Ritter PL. CHAMPS physical activity questionnaire for older adults: Outcomes for interventions. Med Sci Sports Exerc. 2001;33:1126–1141. [PubMed]
21. Allison PD. Fixed Effects Regression Methods for Longitudinal Data Using SAS. Cary, NC: SAS Institute Inc; 2005.
22. Hedeker D, Gibbons RD. Longitudinal Data Analysis. New York, NY: Wiley-VCH; 2006.
23. Greene LW, Glasgow RE. Evaluating the Relevance, Generalization, and Applicability of the Research: Issues in External Research and Translation Methodology. Eval Health Prof. 2006;29:126–153. [PubMed]
24. Glasgow RE, Klesges LM, Dzewaltowski DA, Bull SS, Estabrooks P. The future of health behavior change research: What is needed to improve translation of research into health promotion practice? Ann Behav Med. 2004;27:3–12. [PubMed]
25. Lorig KR, Ritter PL, Laurent DD, Plant K. Internet-based chronic disease self-management: a randomized trial. Med Care. 2006;44:964–71. [PubMed]
26. Lorig K, Feigenbaum P, Regan C, Ung E, Chastain RL, Holman HR. A comparison of lay-taught and professional-taught arthritis self-management courses. J Rheumatology. 1986;13:763–7. [PubMed]