Effects of a Weight and Pain Management Programin Patients With Rheumatoid Arthritis With Obesity

A Randomized Controlled Pilot Investigation

Tamara J. Somers, PhD; James A. Blumenthal, PhD; Caroline S. Dorfman, PhD; Kim M. Huffman, MD, PhD; Sara N. Edmond, PhD; Shannon N. Miller, BPH; Anava A. Wren, PhD; David Caldwell, MD; Francis J. Keefe, PhD


J Clin Rheumatol. 2022;28(1):7-13. 

In This Article

Patients and Methods


Participants (N = 50) were recruited from August 2009 to April 2011 through the Division of Rheumatology at Duke University Medical Center and from advertisements (i.e., flyers, brochures) in the surrounding community. All patients provided written informed consent prior to participation. Procedures and protocols were approved by the Duke University Medical Center Institutional Review Board, and the trial has been registered (NCT04246827). Following informed consent, participants provided a history and underwent a physical examination with the study rheumatologist (D.C.) to evaluate study inclusion/exclusion criteria. Participants also completed a treadmill test with electrocardiogram monitoring to ensure safe participation. Participants with cardiovascular abnormalities (e.g., evidence of myocardial ischemia, abnormal blood pressure response to exercise) were excluded. Eligible participants then completed the initial assessment and were randomized to either the treatment or standard care conditions by a study member who did not interact with participants using a random allocation computer software program.

To be included in the study, patients had to (1) have met the American College of Rheumatology criteria for RA, (2) have obesity (BMI >28 kg/m2), (3) have had RA for at least 2 years, and (3) have self-reported RA pain in the last 2 weeks. Patients were excluded if they (1) had a significant rheumatic disorder other than RA or another organic disease that would affect functioning (e.g., chronic obstructive pulmonary disease, cancer); (2) had a significant medical condition that would increase their risk of adverse experiences during the study (e.g., history of myocardial infarction); (3) were already involved in a regular exercise program and/or PCST program; (4) had abnormal cardiac response to exercise (e.g., exercise-induced deep vein thrombosis (DVT), abnormal blood pressure response); (5) were cognitively impaired (e.g., dementia, mental deficiency); (6) were pregnant at enrollment or initiation of intervention; or (7) were younger than 18 years or older than 85 years. Patients with RA tend to have increased fat mass and decreased muscle mass, making a BMI of 28 kg/m2 or greater a more appropriate marker of obesity22 than the traditionally defined BMI for individuals with obesity of 30 kg/m2 or greater.[23]

Enhanced Lifestyle Behavioral Weight Management Group (Treatment)

Those randomized to the treatment group participated in an enhanced lifestyle behavioral weight management intervention that included instruction in pain coping skills and traditional behavioral weight management strategies. This program was delivered weekly over 12 weeks in 90-minute group sessions. The intervention followed a session-by-session, manualized protocol and was delivered by clinical psychologists with prior experience in PCST. The intervention was designed to (1) enhance participants' abilities to control and decrease pain by increasing use of adaptive coping strategies (e.g., distraction, relaxation, and cognitive restructuring) and (2) teach participants strategies for gradually decreasing calorie and fat intake through permanent lifestyle changes. Participants completed a second assessment following intervention completion and received $40 for completing each study assessment. Supplemental Table 1 (http://links.lww.com/RHU/A363) provides a session-by-session outline of the intervention. To teach PCST and behavioral weight management techniques, participants were presented with a simplified version of Melzack and Wall's[24] gate control theory during the first session to show that pain is a complex experience affected by thoughts, feelings, and behaviors. Each of the 12 sessions included both pain coping skills (i.e., progressive muscle relaxation, mini practices, activity-rest cycling, pleasant activity scheduling, changing negative cognitions, and imagery) and lifestyle behavioral weight management strategies (i.e., goal-setting strategies for weight loss, strategies for increasing activity through lifestyle changes and formal exercise, and information about good nutrition).

Treatment group patients also participated in exercise sessions 3 times per week for 12 weeks at the Duke Center for Living fitness facility. Exercise sessions were scheduled concurrently with the group sessions described above and were supervised by exercise physiologists specifically trained in the study exercise protocol. Exercise sessions incorporated strength training to build muscle mass, as well as aerobic conditioning to decrease fat mass. Each exercise session began with a 10-minute warm-up period, with general flexibility and isometric strengthening of postural muscles. Aerobic exercise progressed from an initial intensity and duration of 55% of heart rate reserve for 15 minutes to 70% of heart rate reserve for 30 minutes. Heart rates and Borg scale ratings of perceived exertion were monitored 3 times during each exercise session. Sessions concluded with 10 minutes of cool-down consisting of light stretching and deep breathing exercises.

Standard Care Group

The control condition was designed to serve as a standard care or usual treatment comparison group. Patients assigned to this condition continued to receive their routine RA care, including regular appointments with their rheumatologist and appointments as needed for symptom flares or other RA-related problems. Individuals in the standard care group did not receive a study intervention. Patients in the standard care group completed assessments at time intervals corresponding to the beginning and end of the treatment period and received $40 for completing each study assessment.


At baseline and posttreatment, participants completed a physical examination with the study rheumatologist (D.C.) to assess disease activity and severity. The physician participated in standardized training methods and was blinded to other study assessment measures. In addition, the physician was not made aware of participants' study condition. Participants provided demographic (e.g., gender, race, relationship and employment status, etc.) and medical information (e.g., disease duration, recent flares, etc.) and completed self-report measures (e.g., physical functioning, eating behavior, pain, weight self-efficacy, arthritis self-efficacy). Participants also completed the 6MWT, and waist circumference was obtained.


Feasibility was measured by examining study accrual, intervention adherence, and study attrition.

Primary Outcomes

Weight. Participants were weighed by the study coordinator at the time of the physical examination with the study rheumatologist. Weight was measured in pounds without shoes or jackets and in the standing position. Weight was converted from pounds to kilograms by dividing weight in pounds by 2.2046.

Pain. Four items from the Brief Pain Inventory were used to assess usual, worst, least, and average pain in the last week. For each item, patients rated their pain using a scale ranging from 1 to 10, with 1 representing "no pain" and 10 representing "pain as bad as you can imagine."[25] Items were averaged to create a total score.

Secondary Outcomes

Physical Function. Physical function was assessed by physician rating, patient self-report, and the 6MWT. First, the study rheumatologist completed a visual analog scale for pain assessment (VAS) to assess disease severity after conducting a physical examination and medical chart review. The VAS was a horizontal line 100 mm in length, with anchors ranging from 0 "no disease activity" to 100 "high disease activity." The rheumatologist was asked to mark a notch on the line to indicate the patient's disease severity, and the VAS score was determined by measuring the distance in millimeters from 0 to the notch. Previous studies, including our own and others, have assessed disease severity in a similar fashion.[26,27]

Second, participants completed the physical functioning scale of the revised Arthritis Impact Scales (AIMS-2). The AIMS-2 assesses the health status of arthritis patients.[28,29] Scores range from 0 to 10, with 10 representing reduced physical functioning.

Finally, physical functioning was also assessed using the 6MWT, an objective evaluation of functional exercise capacity.[30] The 6MWT was performed indoors. Participants were asked to walk along a long, flat, straight, enclosed corridor as quickly as possible for 6 minutes. Scores are reflective of the distance walked during the 6 minutes.

Eating Behavior. The Binge Eating Scale is a 16-item self-report instrument used to measure dietary habits[31] and assesses both behavioral (e.g., eating large quantities of food) and cognitive (e.g., guilt and fear about being unable to stop eating) manifestations of binge-eating tendencies. Each item contains 3 or 4 multiple-choice statements. Weights assigned to response options range from 0 to 3, and a scaled score is obtained by summing the weights across the 16 items. Higher scores indicate greater severity of over eating.

Self-efficacy for Weight Control and Eating Behaviors. The Weight Efficacy Lifestyle Questionnaire was used to determine self-efficacy for weight control.[32,33] The 20-item measure asks participants to provide information about how certain they are that they can resist overeating across 5 situations (i.e., positive emotions, negative emotions, food availability, social context, and physical discomfort). Response choices range from 0 = "not at all confident" to 9 = "very confident." Items were averaged to obtain a global measure, with higher scores reflecting greater self-efficacy for resisting eating.

Arthritis Self-efficacy. The Arthritis Self-efficacy Scale (ASES) assesses patients' confidence in their abilities to control arthritis pain and minimize disability.[34,35] The ASES comprised 3 subscales: (1) self-efficacy for pain control, (2) self-efficacy to perform functional tasks, and (3) self-efficacy to control other symptoms. Responses for each item range from 10 to 100, where 10 is "very uncertain" and 100 is "very certain." A total score was calculated as the average across scores on the 3 subscales.

Data Analysis

Descriptive statistics were calculated for sociodemographic and medical data for all participants completing the baseline assessment (N = 50), who were randomized to 1 of the 2 study arms. A χ 2 test or analysis of variance, as appropriate, was used to test for differences in baseline characteristics of those randomized to the treatment (n = 29) or standard care (n = 21) groups.

Descriptive statistics were used to characterize feasibility data (i.e., accrual, adherence, and attrition). Intent-to-treat analyses used all available data from patients randomized to either the treatment or standard care arms (n = 50) regardless of protocol adherence, initiation of treatment, and/or subsequent withdrawal from treatment or assessment. Given the pilot nature of this study, we did not conduct formal statistical analyses to test between-group differences.[36] Instead, we computed descriptive statistics and paired t tests to examine within-group changes from baseline to follow-up on study outcome variables. The magnitudes of the effect sizes were calculated and defined according to standard convention for Cohen d (small = 0.2, medium = 0.5, large = 0.8).[37] Outcome analyses were conducted using both a last value carried forward approach and complete case analysis. The results of both analytic strategies were comparable, and we present the results of the complete case analysis. Given that this was a pilot feasibility study, outcome analyses were exploratory, and thus, results should be interpreted with caution.[38]