The Collaborative Lifestyle Intervention Program in Knee Osteoarthritis Patients (CLIP-OA) trial: Design and methods

Abstract

Being overweight or obese is a primary modifiable risk factor that exacerbates disease progression and mobility disability in older knee osteoarthritis (OA) patients. Lifestyle interventions combining exercise with dietary weight loss (EX+DWL) yield meaningful improvements in mobility and weight loss that are superior to EX or DWL alone. Unfortunately, community access to practical, sustainable weight management interventions remains limited and places knee OA patients at increased risk of mobility disability. The Collaborative Lifestyle Intervention Program in Knee Osteoarthritis patients (CLIP-OA), was a two-arm, 18 month randomized-controlled, comparative effectiveness trial designed to contrast the effects of an evidence-based, theory-driven EX+DWL intervention, personalized to patient needs and delivered by our community partners, with those of the Arthritis Foundation’s Walk With Ease (WWE) standard of care self-management program in the treatment of knee OA patients with overweight or obesity. The primary outcome of the CLIP-OA trial was mobility performance assessed using the 400-m walk test (400MWT). Secondary outcomes included weight loss, pain, select quality of life and social cognitive variables, and cost-effectiveness of intervention delivery. Findings from the CLIP-OA trial will determine the comparative and cost-effectiveness of the EX+DWL and WWE interventions on key clinical outcomes and has the potential to offer a sustainable medium for intervention delivery that can promote widely accessible weight management among knee OA patients with overweight or obesity. Trial Registration: NCT02835326.

1. Introduction

Knee osteoarthritis (OA) is a prevalent chronic degenerative disease that serves as a primary cause of mobility disability among older adults in the United States [1-4]. The joint damage and pain accompanying symptomatic knee OA is associated with muscle weakness, activity restriction, and limitations in basic mobility-related functional tasks [2,3]. Obesity is also now recognized as one of the primary modifiable risk factors for the pathogenesis and progression of symptomatic knee OA and is widely prevalent among older knee OA patients with estimates of approximately half of all patients being overweight or obese [5]. Weight gain compounds risk for mobility disability and loss of independence with advancing age among knee OA patients by exacerbating a deleterious cycle involving knee pain, inactivity, loss of muscle strength, and accelerated progression of functional decline. Consequently, knee OA patients with overweight or obesity are a vulnerable population with an urgent need of increased access to efficacious, cost-effective weight management interventions.

There is growing recognition that lifestyle interventions targeting modifiable risk factors for knee OA progression, such as obesity, represent an integral component of disease management efforts. In this regard, exercise (EX) is widely accepted as a safe, efficacious intervention and is now a standard component of contemporary treatment for knee OA [6-14]. EX is the cornerstone of the Arthritis Foundation’s (AF) social cognitive theory-driven, evidence-based Walk With Ease (WWE) program that is designed to promote the self-management of arthritis through walking. WWE represents the AF’s standard of care self-management program that is widely accessible and remains the primary focus of the AF’s national community-based self-management programming. The established success and longevity of the WWE program underscores the potential for dissemination of lifestyle self-management programs through the AF’s existing community-based infrastructure.

Although the benefits of WWE reinforce its utility in the self-management of knee OA, contemporary evidence in the behavioral weight management literature clearly illustrates that EX is necessary, but not sufficient, to produce meaningful weight loss and prevent mobility disability in older adults with overweight or obesity [15-17]. Modifying energy balance through increased physical activity and reductions in caloric intake via change in dietary behavior is key to successful behavioral weight management interventions [18-20]. Consequently, comprehensive lifestyle interventions targeting change in both EX and dietary weight loss (DWL) are integral for successful behavioral weight management in knee OA patients. Evidence from our own prior, randomized controlled lifestyle intervention trials demonstrates that a combined EX+DWL intervention is a feasible, efficacious approach that results in superior improvements in weight loss and mobility performance, relative to EX or DWL alone, among older adults with overweight or obesity [10,21-24]. Collectively, these findings suggest a synergistic benefit of combining EX+DWL that may be the optimal approach to promoting successful weight management and preventing mobility disability among knee OA patients with overweight or obesity.

Unfortunately, community access to practical, sustainable weight management interventions that address the public health threat posed by obesity and functional decline in older knee OA patients remains limited. From a translational perspective, effective community collaborations are needed to deliver widely accessible, cost-effective, scalable lifestyle interventions to knee OA patients. Consistent with this position, prior translational lifestyle weight management intervention trials conducted in collaboration with community partners to increase the impact, accessibility, and scalability of these approaches have yielded promising results [21,25-27]. Notably, findings from the Cooperative Lifestyle Intervention (CLIP) trial demonstrated that a community-based EX+DWL lifestyle intervention yielded superior changes in mobility performance and produced significantly greater weight loss relative to EX-alone or the successful aging interventions in the treatment of 288 older adults with obesity, cardiovascular disease or metabolic syndrome [21]. Building off the success of CLIP, the CLIP-II trial increased the translational significance by having the lifestyle weight management interventions delivered by community partners with study staff serving as trainers and advisors. CLIP-II was a 3-arm 18-month randomized controlled trial evaluating the independent and synergistic effects of DWL alone and in combination with aerobic training and resistance training in a sample of 249 older, adults with overweight or obesity, cardiovascular disease, and/or metabolic syndrome. Results revealed both combined treatment arms yielded significantly greater improvements in weight loss and mobility performance relative to DWL alone. Additionally, ancillary findings illustrated that combining resistance training and DWL resulted in superior preservation of lean body mass and that loss of fat mass was the most robust predictor of improved mobility performance [25,26].

Although these findings suggest that this approach holds considerable promise, the utility of implementing the CLIP model of offering lifestyle weight management interventions by leveraging community partnerships and their existing infrastructure in program delivery has yet to be determined in the treatment of knee OA patients with overweight or obesity. Furthermore, the extent to which the EX+DWL intervention may be a superior, cost-effective alternative to the widely accessible WWE remains unclear. Thus, there is a critical need to systematically investigate community-based delivery of weight management interventions for knee OA patients with overweight or obesity and to evaluate the comparable efficacy and cost-effectiveness of this approach with standard of care self-management programming that is currently widely available for this patient population.

To address this pressing issue, the Collaborative Lifestyle Intervention Program in knee OA patients (CLIP-OA) was designed to determine the comparative effectiveness of a community-based EX+DWL program relative to the AF’s WWE program in producing improvements in mobility and weight loss among older knee OA patients with overweight or obesity. To enhance the translational significance of the trial, the interventions were offered leveraging existing community infrastructure and delivered by community partners with our staff serving as trainers and advisors. The primary aim of the CLIP-OA trial was to compare the effects of the EX+DWL and WWE interventions on mobility performance assessed using the 400-m walk test (400MWT). Secondary aims were to compare the effects of the two intervention arms on weight loss, pain, and quality of life outcomes and evaluate the cost-effectiveness of intervention delivery. It was hypothesized that the EX+DWL would result in significantly greater improvement in mobility performance. Additionally, whereas delivery of both interventions will be similar in cost, it was hypothesized that the EX+DWL intervention would yield superior improvements in weight loss, pain, and quality of life relative to the WWE intervention.

2. Methods

2.1. Trial design and participants

CLIP-OA was a single-blind, two-arm, 18-month comparative effectiveness trial designed to compare community-based Exercise and Dietary Weight Loss (EX+DWL) versus Walk With Ease (WWE) interventions in the treatment of older, knee OA patients with overweight or obesity. The institutional review board of Ohio State University approved all trial procedures. All participants provided written informed consent. Recruitment efforts focused on direct solicitation methods via zip-code based direct mailings, clinical collaboration with the Department of Rheumatology, and community recruitment involving advertisements in newsletters and presentations. Participants were randomly assigned to either the EX+DWL or WWE arms. Outcome assessments were completed at baseline, 6, 12, and 18-month follow-up visits and conducted by research staff who were blinded to treatment arm assignment. Collaborative partnerships were established with community centers/fitness facilities that either currently offered, or planned to begin offering, the AF’s WWE program. In collaboration with center/facility leadership, instructors were identified that were interested and well-qualified to serve as intervention leaders. Each of these individuals was employed as center wellness directors, coordinators, and/or fitness trainers at their respective facilities. The community instructors, who were trained using standardized training modules, materials and procedures used in the investigative team’s prior trials [10,21-25], delivered both the WWE and EX+DWL interventions at four community center sites. The CLIP-OA trial was designed to recruit 227 participants across 11 waves and four community partner sites.

2.2. Eligibility

The inclusion criteria were designed to enroll a representative sample of older, knee OA patients with overweight or obesity and risk for mobility disability that could benefit from lifestyle intervention programs (Table 1). In order to be included in the CLIP-OA trial, participants had to meet the following criteria: (a) community-dwelling women and men aged ≥50 years of age; (b) physician diagnosed knee OA documented by radiographic tibiofemoral OA of one or both knees; (c) overweight or obesity with BMI ≥ 25; (d) self-reported difficulty walking ¼ mile, climbing stairs, lifting or carrying groceries, or performing other household activities due to knee pain; (e) inactive lifestyle defined as currently participating in <20 min of weekly structured moderate intensity exercise occurring in bouts >10 min; (f) all participants must be free of severe heart or systemic disease or any other serious medical condition that would make moderate intensity exercise participation unsafe; (g) planned stability of residence for the duration of the trial; (h) receive physician medical consent from their treating rheumatologist and/or primary care physician; and (i) willing to accept randomization and complete all assessment and intervention procedures.

Table 1.

Inclusion and exclusion criteria.

Inclusion Diagnosed with knee OA: physician documented radiographic tibiofemoral OA of one or both knees; Overweight or Obese: BMI ≥ 25 kg/m2; Knee Pain and Risk for Mobility Disability: self-reported difficulty walking ¼ mile, climbing stairs, lifting or carrying groceries, or performing other household activities due to knee pain; Sedentary: sedentary lifestyle defined as currently participating <20 min of weekly structured moderate intensity exercise occurring in bouts >10 min; Health Status: all participants must be free of severe heart or systemic disease that would make moderate intensity exercise participation unsafe; Age: ≥ 50 years of age; Stability of Residence: does not plan to move out of the Columbus, OH metropolitan area for the duration of the study; Consents: willing to give an informed consent and sign a HIPAA authorization form; Physician Consent: treating rheumatologist and/or primary care physician provide medical consent for participation; and Agreeableness: willing to accept randomization and complete all assessment and intervention procedures.

Exclusion Health Status: any serious medical condition that precludes safe participation in an exercise program; Study Staff Judgment: inability to complete the study protocol, in the opinion of the study staff, due to frailty, illness, or other reasons; Severe symptomatic heart disease: evidence of unstable angina, symptomatic congestive heart failure, or exercise induced complex ventricular arrhythmias; or MI or cardiovascular procedure within the last 3 months.

2.3. Recruitment and screening

A total of 227 older, symptomatic knee OA patients with overweight or obesity were recruited and randomly assigned to receive either the EX+DWL intervention (n = 110) or WWE (n = 117). Participants were recruited through three approaches: 1) direct solicitation using zip code coordinated direct mailings, 2) clinical recruitment using patient referrals and direct mailings to all knee OA patients treated by rheumatology physicians, and 3) community recruitment using advertisements in community newsletters and mailings, presentations and advertisements to local support groups, and online research recruitment through ResearchMatch and StudySearch. After completing a brief phone screening, individuals interested in participating and meeting all eligibility criteria were scheduled for the baseline screening visit. Recruitment and baseline assessments for all participants in the trial began in November of 2016 and was completed in January of 2020.

2.4. Trial assessments

All assessments were conducted by trained study staff using standardized operating procedures that were blinded to group assignments. The study assessment procedures and follow-up schedule are provided in Table 2.

Table 2.

List and timing of outcome measures.

Assessment procedures	Phone screening	Baseline visit (SV1)	3 week re- assess	F-6	F-12	F-18
Telephone Screen/Demographics	X
Consent		X
HIPPA Waiver		X
MD name, address, fax (for clearance letter)		X
Usual & Emergency Contact Information		X		Update	Update	Update
Medication Questionnaire		X		X	X	X
Medical History		X		X	X	X
Modified Community Healthy Activities Model Program for Seniors Physical Activity Survey (CHAMPS)		X		X	X	X
Body Composition		X		X	X	X
Weight		X		X	X	X
Height		X		X	X	X
BMI		X		X	X	X
BodPod/DXA		X		X	X	X
Mobility-Related Self Efficacy of 400Meter Walk Test		X		X	X	X
400Meter Walk Test		X		X	X	X
Mobility-Related Self Efficacy of Stair Climb Test		X		X	X	X
Stair Climb Test		X		X	X	X
Mobility-Related Self Efficacy of Lift & Carry Test		X		X	X	X
Lift & Carry Test		X		X	X	X
Pepper Assessment Tool for Disability (PAT-D)		X		X	X	X
Satisfaction with function and appearance		X		X	X	X
Multidimensional self-regulatory self-efficacy (MSES)		X	X	X	X	X
Weight efficacy lifestyle questionnaire		X	X	X	X	X
Goal commitment and goal difficulty - Exercise		X		X	X	X
Goal commitment and goal difficulty – Weight Loss		X		X	X	X
Physical Activity Self-Regulation Scale (PASR-12)		X		X	X	X
Diet Self-Regulation Scale (DSR-12)		X		X	X	X
Pain acceptance		X		X	X	X
Western Ontario & McMaster Universities Osteoarthritis Index (WOMAC)		X		X	X	X
Rand Medical Short Form-12 Health Survey		X		X	X	X
Experiencing Close Relationships		X		X	X	X
Quality of Well-Being scale - Self-Administered (QWB-SA)		X		X	X	X
Healthcare Utilization Questionnaire (HUQ)		X		X	X	X
Group Climate Questionnaire-Short (GCQ-S)				X	X	X
Social Provisions				X	X	X
3 Day Food Record		X		X	X	X
Food Frequency Questionnaire (FFQ VioScreen)		X		X	X	X
ActivPal 7 Day Log		X		X	X	X
Direct Costs Tracking		X		X	X	X

2.5. Primary outcome: mobility performance

The trial’s primary outcome was mobility performance assessed using the 400-m walk test (400MWT). The 400MWT is a valid, reliable measure of mobility performance, which has been successfully implemented in multiple prior lifestyle intervention trials conducted by the investigative team [10,21-25]. The 400MWT was completed in a corridor with two cones spaced 20 m apart. Individuals are instructed to walk as quickly as they can and the time to complete 10 laps around the 40-m course is recorded as the performance measure. A script is used to standardize instructions to all participants. Participants could stop and rest if necessary and are given a maximum of 15 min to complete the test. The 400MWT was selected as the trial’s primary outcome based on compelling evidence that it is a robust, prognostic indicant of mobility disability and risk of loss of independence and mortality in older adults with obesity [21,24,28-30].

2.6. Secondary outcomes

2.6.1. Anthropometric battery

Body weight was measured to the nearest 0.1 kg using a calibrated and certified digital scale. Height without shoes was measured to the nearest 0.1 cm using a stadiometer. BMI was calculated using the body mass and height values. Body composition was assessed using dualenergy X-ray absorptiometry scans (DXA; Lunar iDXA GE Medical Systems, Madison, WI). DXA is well-established as the gold standard measure of body composition [31]. Whole-body measures and regional measures of body composition were estimated. Body composition was also assessed using the Bod Pod (Life Measurement Inc., Concord, CA). The body weight and composition outcomes were important variables to compare EX+DWL and WWE interventions on weight loss outcomes.

2.6.2. Functional battery

In addition to the primary outcome (400MWT), the timed Stair-Climb Test (ST) and the Lift and Carry Test (LCT) were used as objective measures of physical function [32-36]. The ST is a valid, reliable functional performance measure that involves ascending and descending one flight of stairs with 8 steps as quickly as possible [32-34]. The LCT assesses how quickly and easily a person can lift and carry a 10 pound weight over 10 m [32-34]. Prior to each performance test, assessments of Mobility-Related Self-Efficacy (MRSE) assessed participants’ confidence in their ability to complete incrementally more challenging amounts of each of the 400MWT, ST, and LCT performance tests [35,36]. Self-reported functional limitations were measured with the Pepper Assessment Tool for Disability (PAT—D) [37].

2.6.3. Patient-reported outcomes

Self-reported physical function and pain symptoms were assessed using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) [38]. Participants indicated on a scale of 0 (none) to 4 (extreme) the amount of difficulty they experienced performing basic physical function tasks in the past 48 h due to knee OA. The physical function subscale consists of 17 items that are summed to produce a physical function score ranging from 0 to 68 with higher scores indicating poorer function. Participants also indicated the pain severity they experienced during the past 48 h due to knee OA on a scale ranging from 0 (none) to 4 (extreme). The WOMAC pain subscale consists of 5 items and total scores range from 0 to 20 with higher scores indicating greater pain. Health-Related Quality of Life (HRQOL) was assessed using the Rand Medical Outcomes Study 12-item Short Form Health Survey (SF-12) [39,40]. The SF-12 is a generic measure of HRQOL which consists of two norm-based composite scales (mental health and physical health) and eight subscales (physical functioning, mental health, role-physical, role-emotional, bodily pain, general health, vitality, and social functioning).

2.6.4. Physical activity and dietary behavior

Assessments of physical activity participation were obtained using device-assisted and self-report measures. The ActivPAL accelerometer (PALtechnologies, Scotland, UK) was used to obtain an objective assessment of physical activity participation [41]. Participants wore the ActivPAL on their right thigh with a waterproof sleeve and adhesive patch for seven consecutive days following the completion of each testing visit. Participants recorded the times they took off and re-applied the ActivPAL using a self-monitoring log. The ActivPAL provides assessment of minutes of light, moderate, and vigorous physical activity participation as well as calculating total daily steps taken [30]. Participants were provided verbal and written instructions on how to wear the accelerometer. Monitors were returned to trial staff via the U.S. Postal Service. The ActivPAL was selected to capture physical activity and sedentary behaviors in this trial due to the device’s advantages of a) reducing wear time concerns via the feature of affixing directly it to the patient’s thigh; and b) established accurate assessment of posture given prior data from our team’s IMPACT trial demonstrating high volumes of sedentary time and light physical activity participation [24]. Self-reported physical activity was assessed using the Modified Community Healthy Activities Model Program for Seniors (CHAMPS) Questionnaire. The CHAMPS is a 41-item measure developed specifically for the assessment of physical activity in older adults. The CHAMPS measure has well established validity and reliability and yields estimates of total minutes of physical activity and energy expended per week in all physical activities of moderate or higher intensity [10,24,42]. Both of these measures have been used successfully in prior lifestyle intervention trials among knee OA patients and older adults at risk for mobility disability [21,24,28-30]. Dietary intake was assessed using a 3 day-food log including one weekend day and two weekdays [43,44]. Nutrient analysis of all food records was conducted using the Minnesota Nutrition Data System for research [45]. VioScreen, an online dietary analysis of 90-day nutrition tracking, was also used. The VioScreen electronic food frequency questionnaire requires participants to estimate consumption of a wide variety of foods (Viocare, Princeton, NJ). Scoring yields estimates of macro- and micro-nutrient intake, as well as intake by specific food groups for both group and individual analyses. Additionally, participants completed weekly food logs using MyFitnessPal (or paper logs for a limited number of patients who were uncomfortable using the electronic tracking app or website), which provided daily and weekly summaries of estimated calorie and macronutrient intake.

2.6.5. Social cognitive process measures

As noted in the introduction, both intervention arms have a foundation in Social Cognitive Theory. Accordingly, multiple theory-based measures were assessed including the Physical Activity Self-Regulation scale [46], Diet Self-Regulation Scale, Satisfaction with Function and Appearance Scale [47], Multiple Exercise Self-regulatory Self-Efficacy scale [48], the Weight Efficacy Lifestyle questionnaire [49], Goal Commitment and Goal Difficulty [50], Group Climate Questionnaire, Social Provisions, Experiencing Close Relationships, and Pain Acceptance [21].

2.7. Cost-effectiveness analysis

The incremental cost-effectiveness of the EX+DWL intervention in comparison to the WWE intervention was calculated as the ratio of the additional cost per participant per increment of effectiveness. Effectiveness was quantified using both the primary outcome of the study (400MWT) and Quality Adjusted Life Years (QALYs). Costs were calculated from healthcare organization and societal perspectives. The time horizon of analysis was 12 months. The sensitivity of estimates used to calculate intervention costs and cost per health care contact were varied bi-directionally to test the robustness of findings. (i.e., wages, benefits, facilities cost).

2.7.1. Effectiveness outcomes

The incremental improvement in 400MWT for the EX+DWL arm compared to the base case (WWE arm) was calculated per second and per minimally significant change of 25 s [51]. Earlier studies indicate that after 12 months, EX+DWL participants improved 12 s more than EX alone, and 7% more EX+DWL participants were 25 s faster [21]. The Quality of Well-Being scale - Self-Administered (QWB-SA) was used to calculate QALYs. The QWB-SA is a generic, preference-based measure of HRQOL that has been used in a number of large NIH clinical trials including DPP, NETT, and The LIFE Study [52-64].

2.7.2. Direct costs and health care utilization

Direct costs were tracked in accordance with guidelines published by Neumann et al. [57]. Local per unit costs were replaced with national averages when possible (i.e., wages, benefits rate, facilities cost/sq. ft.) to provide a more generalizable national estimate of intervention costs. Costs were standardized to the year in which interventions begin (201×) using the U.S. Consumer Price Index [58]. Personnel costs were estimated by multiplying intervention staff member time spent on intervention activities by the US Bureau of Labor Statistics national mean hourly wage by position [59]. Intervention training and preparation time was included. Fringe benefits were added using the average fringe benefit rate for the US [59]. Facility costs were calculated by multiplying facility time and square footage used by the national average retail rate for facilities. Miscellaneous costs include intervention materials, telephone charges, office supplies, refreshments, and participant incentives. Direct costs to participants, such as costs for athletic equipment, transportation, and the opportunity cost of participation were included in the societal perspective using methods described by Hatziandreu et al. [60] and national average hourly wage rates [59].

2.7.3. Health care costs

Health care costs were estimated using the Healthcare Utilization Questionnaire (HUQ) [61]. Self-reported healthcare utilization has been shown to be reliable and valid [62] when the length of recall is limited to 3 months [63]. Participants list the number of times each type of utilization occurs (hospital inpatient or nursing home days, emergency room or urgent care visits, ambulance use, physician, osteopath or nurse practitioner visits, home care, inpatient and outpatient surgeries, medical supplies, medications) during the previous 3 months. The cost per healthcare contact came from national means from the Medical Expenditure Panel Survey (MEPS) [64]. Participants were instructed to exclude any research-related health care contacts.

2.7.4. Adherence

Adherence to the exercise and dietary interventions was monitored based on attendance to sessions and completion of weekly summary logs. Exercise interventionists and community instructors routinely monitored the logs for completeness and reminded patients to complete these on a weekly basis.

2.8. Interventions

The EX+DWL and WWE interventions were delivered by the same community partner instructors, with our staff serving as trainers and advisors, using our established standardized protocols that were successfully implemented in the previous CLIP trials [21,25]. Whereas both arms receive 36 total intervention contacts, there is differential sequencing in delivery of these contacts. The 36 EX+DWL group contacts were delivered across the 18-month intensive, transition, and maintenance phases of treatment as we have found this is a theory-driven aspect of the intervention design that is integral to promoting maintenance of behavior change and weight loss in prior trials. Consistent with the established timeframe of traditional delivery for the standard of care WWE programs, 36 WWE intervention contacts were provided across 3 months. Therefore, the differential delivery sequencing for the 36 contacts represents a key difference between the approaches and is a critical aspect to be considered in evaluating the comparative effectiveness of the EX+DWL and WWE arms. All participants received an 18-month membership to their participating community center to provide equal access to facilities necessary for maintenance of physical activity participation central to each intervention.

2.9. EX + DWL intervention

Consistent with procedures in our prior lifestyle intervention trials [21,25,28,30] our evidence-based EX+DWL intervention was delivered by our community partners with our study staff as trainers/advisors and was comprised of personalized EX and dietary intake prescriptions and group-mediated cognitive behavioral counseling designed to promote independent adherence to EX and dietary behavior change. The theoretical foundation of the lifestyle intervention is based on Bandura’s Social Cognitive Theory [35] and the group dynamics literature [65,66] focusing upon the use of established cognitive behavioral strategies to enhance self-efficacy and foster self-regulation of physical activity and dietary intake, while harnessing group dynamics as an agent of behavior change to increase motivation to develop, practice, and master behavioral self-regulation. The intervention delivery structure across the 18-month trial is divided into three phases: Intensive (Months 1–6), Transition (Months 7–12), and Maintenance (Months 13–18). During the Intensive phase, participants met at the community center for weekly sessions. At the beginning of each in-person session, patients were asked to weigh-in on a digital scale. In addition, there were three group sessions and one small-group/individual session per month. During the Transition phase, frequency tapered to bi-weekly meetings consisting of one group and one small-group/individual session per month. No formal group sessions were held during the Maintenance phase, with only individual sessions scheduled as necessary, resulting in participants receiving a total of 36 intervention contacts with study staff during the three phases of the lifestyle intervention. The structure of these phased contacts was designed to progressively wean patients from the dependency on community staff toward independent self-regulation of exercise, physical activity, and dietary intake. All supervised sessions involved 60 min of EX followed by 60 min of group-based weight management education and counseling.

The primary goal of the EX component was to shape physical activity participation so that within 3-months of the onset of the study, participants are independently engaging in at least 30 min of moderate intensity activity on most, if not all days, of the week for a total weekly accumulation of at least 150 min. The structure of EX involved approximately 1 h consisting of aerobic and resistance training, where both were individualized so that duration, volume, and intensity were gradually increased based on each patient’s exercise tolerance and capacity. The primary mode of aerobic EX was walking. However, patients also had access to alternative, non-weight bearing modes, such as a stationary cycle or elliptical trainer, and were encouraged to integrate their preferred mode as needed on days where they experience activity-related exacerbation of pain symptoms. Across the aerobic training, patients gradually progressed from 10 to 30 min/session performed at a rating of perceived exertion of 12–14 (Moderately Hard). The resistance training involved 1–3 sets of 8RM–12RM repetitions at a rating of perceived exertion ranging from 12 to 15 of multiple exercises (leg extension, leg curl, leg press, chest press, lat pull-down, overhead press, triceps extension, bicep curl, and calf raises). Consistent with the procedure for aerobic activity, resistance exercise volume and intensity were also gradually increased based on individual exercise tolerance and capacity. Weekly written self-monitoring logs were used to document EX adherence and participants were also provided a FitBit to objectively self-monitor and promote physical activity [67].

The DWL component was based on recommendations provided by the USDA/USDHH Dietary Guidelines for Americans and the NIH National Cholesterol Education Program Therapeutic Lifestyle Changes Diet for High Cholesterol [68,69]. The DWL component’s manualized intervention content was modeled after previous successful interventions including our prior CLIP trials and LookAHEAD that focus on self-regulation [21,25,70]. The primary objectives of the DWL intervention were to shape eating habits that decrease caloric intake and optimize dietary patterns to achieve a healthier weight, elicit a gradual 1–2 lb./week weight loss in the intensive phase, and progress to a total weight loss of 7–10% and/or achieve the upper limit of standardized age-appropriate healthy weight. The aims of the transition phase depended on individual patient progress toward target weight loss while the focus moved toward sustaining a healthier weight in the maintenance phase. During the maintenance phase in months 7–12, participants were encouraged to continue DWL as long as their BMI did not fall below 20 kg/m². However, the intervention content focus progressively shifted from weight loss to promoting weight maintenance, the primary target for the maintenance phase. Although general calorie range goals were provided based on patients’ baseline weights (<250lbs: 1200–1500 kcal/day; >250lbs:1500–1800 kcal/day), caloric intake was personalized to a specific range that safely and effectively facilitated progress of the desired gradual weight loss. Dietary composition focused upon personalized guidance toward a diet which included foods rich in whole grains, a variety of vegetables and fruits per day, and limited consumption of processed foods which are high in fat, are high in added sugars, and are low in nutrient density. Intake guidance consisted of approximately 45–55% of kcal of complex carbohydrates, <30% kcal of fat, and particular emphasis placed on at least 1 g/kg body weight of protein intake to aid in offsetting muscle loss. Dietary intake was monitored by the study staff and Registered Dietitian and was personalized to each patient’s intake needs. Participants completed weekly food logs using MyFitnessPal (or paper logs for a limited number of patients who were uncomfortable using the electronic tracking app or website), to provide daily and weekly self-monitoring summaries of dietary intake.

DWL sessions provided weekly behavioral lessons and education tailored to patient needs, and counseling on cognitive and behavioral lifestyle strategies. The topics of the manualized behavioral lessons were integrated from prior CLIP and LookAHEAD trials [21,25,70]. Consistent with the self-efficacy and agency aspects of social cognitive theory, DWL sessions integrated systematic group-mediated counseling on core behavioral strategies (self-monitoring; adaptive goal setting; social problem solving; lapse and relapse prevention) to foster self-regulation of EX+DWL behaviors, while harnessing the dynamics of the supportive group learning environment to facilitate the learning, development, and practice of self-regulatory skills. This approach is structured to create a progressive group counseling effect that enhances self-efficacy and supports self-regulation with sequenced contacts using a phased increase in the ratio of personal responsibility to independently manage EX+DWL behavior. Additionally, systematic social support was cultivated to reinforce a continuous social and individual problem-solving approach to overcoming barriers through peer-initiated solutions designed to empower participants to exert greater control over their behavior, cognitions, and environment.

The instructors were trained to deliver the EX+DWL intervention using the standardized procedures successfully implemented in prior trials [21,25]. The group-mediated discussions in each behavioral DWL session were facilitated by the community-center instructor in a systematic manner using the following session structure: 1. Sharing of progress (10–15 min): the group leader welcomed the participants and begins by asking them to share thoughts about their progress over the past week or month. Thoughts included their progress with goals, barriers, and how they overcame them, or praise for accomplishments; 2. Presentation of topic (5 min): group leader summarized thoughts from the introductory sharing of progress segment and transitions into leading presentation on the new topic of the day from the manualized DWL intervention content; 3. Facilitated group discussion (25–30 min): the group leader guided the participants in an interactive discussion of the topic or behavioral strategy for that session. Participants were asked to participate in several ways: helping the group define and relate to the topic, giving examples related to the topic, problem solving, and actively listening and responding to ideas/thoughts of other group members. Participants were encouraged to discuss their lived experience with practicing and mastery of the self-regulatory skills. Emphasis in the discussion was placed on promoting social problem-solving through sharing of peer-initiated solutions to barriers and/or challenges in applying self-regulation of exercise and dietary behaviors; 4. Summary/Takeaways (5 min): the group leader summarized key points of the session, encouraged engagement of the participants, and answered any questions; and 5. Mindfulness Reflections/Sharing of Weekly Goals (5 min): participants reflected upon the key points of the session on the last page of each session in their DWL manual. Sharing of goals was encouraged among participants to assist in shaping behavioral process and facilitate social support and accountability for the upcoming week’s weight loss goals. Intervention fidelity was monitored by the principal investigator by using a standardized instructor/participant behavioral coding system successfully implemented in the prior CLIP trials [21,25].

2.10. Walk With Ease (WWE) arm

The AF’s WWE program is also a Social Cognitive Theory-based intervention designed to promote the self-management of arthritis through exercise. WWE is a six-week program involving three sessions per week (18 total sessions) each lasting approximately 60 min in duration. The WWE intervention was delivered by the same community instructors leading the EX+DWL intervention. All participants completed two consecutive WWE classes for a total of 12 weeks of program exposure (36 contacts). This is a common practice in the routine delivery of the WWE program and also equated the total number of contacts between the two treatment arms in the CLIP-OA trial. During the final week of the 12 weeks of group-based WWE program, participants were trained and transitioned to the self-directed version of the WWE to facilitate independent maintenance of EX participation following the cessation of the group program. Consistent with the EX+DWL arm, all participants in the WWE arm received a membership to the community center where the intervention was being delivered for the duration of the trial to provide necessary access for aiding the promotion of independent maintenance of long-term physical activity participation across the 18-month follow-up period.

WWE is a multi-component program comprised of: (1) walking; (2) stretching and strengthening exercises; and (3) health education lectures and motivational tips/tools. All three components were addressed in each of the 36 program sessions. WWE sessions begin with a 10 min health education module focusing upon basic exercise and arthritis self-management education. The exercise component is comprised of a 5 min warm-up with stretching, 10–35 min of walking, and a 5 min cool-down. All exercise was tailored to the needs and personal exercise tolerance/capacity of each individual. Participants were encouraged to progressively increase their walking duration across the program. Strengthening exercises were taught once during the program and encouraged to be completed at home twice/week during the intervention. All participants were provided with a WWE workbook including self-tests, behavioral contract, walking log, and weekly motivational tips. The safety, feasibility, and efficacy of the WWE has been established and the program has been widely implemented in AF community-based settings during the last 15 years [71]. In summary, both the EX+DWL and WWE arms are delivered by the same instructors, received an identical total of 36 intervention contacts, and are provided free access to the community center throughout the duration of the trial in order to support the independent maintenance of physical activity across the 18 month follow-up period.

2.11. Statistical analysis

The primary aim of the CLIP-OA trial was comparing the EX+DWL and WWE on mobility performance (400MWT) at 18 months. Secondary aims were to compare the effects of the EX+DWL and WWE interventions on weight loss, pain, and quality of life and evaluate the cost-effectiveness of each intervention arm. Comparisons were made by mixed-model repeated measures analysis of covariance with a subject random effect to account for the fact that multiple measurements within a participant over time are not independent. Covariates used in the analysis of end points included (1) gender (as a fixed effect), (2) site and wave within each site (as fixed effects), and the baseline (pre-randomization) value of the end point being analyzed. Site is included to adjust for potential differences between sites, and wave within site is included to adjust for the grouped intervention. Analysis of group differences adjust for the baseline levels to reduce the residual variance, to adjust for chance imbalances in important prognostic factors not used in the stratification, and to enhance the acceptability of the results. All tests of hypotheses were two-sided and reported p-values were considered statistically significant at the 0.05 alpha level. Analyses of secondary outcomes will use similar models.

In the primary analyses, all randomized subjects were included in their original study group regardless of the final mode of intervention or the extent of compliance with the study protocol; that is, the primary analysis was an “intent to treat” analysis. Because differences in attrition rates among the study arms may complicate the analysis of the primary outcome measures, attrition rates across the study arms were compared. If a particular arm had a markedly higher dropout rate, we attempted to identify baseline covariates that predict attrition in secondary analyses. If such covariates were identified, supplementary analyses of outcome measures stratified on these covariates were conducted.

2.12. Power calculations

We used data from the CLIP trials [21,25] to estimate residual standard deviations after adjustment for the same baseline variables we used in CLIP-OA. In the CLIP trials, we also delivered the intervention in a group setting and we used data from these prior trials to estimate the intraclass correlation coefficient (ICC) within a group for 400MWT as 0.0200. With approximately 224 people randomized in two treatment groups, in 4 waves, and at 4 sites, this results in group sizes per wave of 7 (=224/2/4/4) and a design effect of 1 + (7–1)*0.002 = 1.012. Note that the use here of 224 is conservative. Allowing for up to 15% loss to follow-up and the lack of independence led us to propose a sample size of 224 that would result in an effective sample size of 188 (=224* (1–0.15)/1.012). For 400MWT, we estimated an adjusted residual (after adjustment for baseline and other covariates) standard deviation of 25.2 s. With an effect of at least 12 s, the sample provided 90.1% power using a two-sided test at the 5% level and 188 evaluable participants. With a difference of 11 (10) seconds, the sample provided 84.5% power. These calculations assumed a total of 17 degrees of freedom, for gender and baseline, and 15 degrees of freedom for wave/site covariates.

3. Discussion

It is well-established that overweight and obesity heightens the risk of mobility disability and loss of independence among older knee OA patients. Unfortunately, access to practical, effective, community-based lifestyle weight management interventions remains limited for older, knee OA patients. The present trial, CLIP-OA, was designed to determine the utility of community-based delivery of weight management interventions in the treatment of knee OA and evaluate the comparable efficacy and cost-effectiveness of this approach with standard of care OA self-management programming that is currently widely available in communities.

CLIP-OA was the first large scale randomized controlled comparative effectiveness trial to evaluate the effects of our evidence-based EX+DWL intervention in comparison with those of the AF’s standard of care WWE program in the treatment of older knee OA patients with overweight or obesity. To enhance the translational significance of the trial, the interventions are offered leveraging existing community infrastructure and delivered by community partners with our staff serving as trainers and advisors. If the EX+DWL intervention yields superior improvements in key clinically-relevant outcomes relative to the AF’s WWE program, these findings could shift the current approach implemented within community-based self-management approaches resulting in significantly greater access to effective weight management for older, knee OA patients with overweight or obesity. There is also an urgent need to provide physicians and health care professionals with concrete resources for lifestyle weight management recommendations for this population. In this regard, findings from the present trial could facilitate greater integration of successful weight management approaches in the routine care and self-management of knee OA patients. Collectively, evidence from the CLIP-OA trial also addresses a pressing need to evaluate the utility of creating partnerships that develop and implement sustainable, economically viable lifestyle interventions that promote successful weight management and prevention of mobility disability among aging OA patients with overweight or obesity.

Older knee OA patients with overweight or obesity are a rapidly growing, vulnerable population at heightened risk of mobility disability and loss of independence. Consequently, there is an urgent need to develop efficacious, sustainable, and widely accessible weight management interventions targeting the prevention of obesity and mobility disability in this priority population. Consistent with the methods of prior CLIP trials [21,25,26], EX+DWL personalizes exercise prescription and dietary intake to the individual patient needs and integrates manualized behavioral weight management content with group-mediated behavioral counseling to promote the development of self-regulatory skills and social support for the adoptions and maintenance of behavior change. The personalized exercise prescription capitalizes on the unique advantages of resistance exercise to sustain muscle mass during caloric restriction, enhances resting and total energy expenditure, and contributes to improving mobility for tasks that depend on weight transfer (i.e., stair-climbing) [5,22] while also leveraging the benefits of energy costs accompanying increased aerobic exercise and purposeful physical activity. The CLIP-OA trial addresses this need and increases the translational significance of this approach by leveraging existing community infrastructure and having the EX+DWL and WWE delivered by community partner instructors with our staff serving as trainers and advisors. If the CLIP-OA trial findings demonstrate that EX+DWL is a cost-effective intervention that results in superior improvements in mobility and key clinical outcomes relative to WWE, the AF offers a cost-effective, sustainable medium for the delivery of this improved model to promote widely accessible, practical community-based lifestyle weight management interventions to older knee OA patients with overweight or obesity nationwide. In summary, the findings of the CLIP-OA trial could inform a meaningful shift in the approach to community-based knee OA self-management and ultimately yield increased access to efficacious weight management for knee OA patients with overweight or obesity nationwide.