Osteoarthritis (OA) is a prevalent and disabling source of chronic pain for which African Americans (AAs) bear a disproportionate burden. The purpose of this study was to test a patient-centered, non-invasive intervention to improve pain outcomes and reduce disparities in AA and White (WH) Veterans with knee OA. The intervention consisted of activities designed to help Veterans develop positive psychological skills (e.g., kindness, gratitude), the health benefits of which are well-documented.
The primary aim of this study was to evaluate the impact of a positive psychological intervention on pain and physical functioning in AA and WH Veterans with knee arthritis through a 2-site randomized, controlled, clinical trial. We hypothesized that patients randomized to a positive skill-building program would experience improved pain and functioning compared to patients who were randomized to an affectively neutral control program, and that these improvements would be larger for AA than for WH Veterans. The secondary aim was to identify variables that mediate the effects of the PA program on pain and functioning. We hypothesized that the effects of the positive program would be mediated by psychosocial variables known to be associated with arthritis outcomes or racial differences in arthritis outcomes.
A randomized 2-arm design was used to compare the effects of a 6-week positive skill-building program with that of a neutral control program on pain and functioning at 1-, 3-, and 6-months post-intervention among AA and WH Veterans with knee arthritis. The original target sample of 180 AA and 180 WH primary care patients with knee pain consistent with arthritis from VA medical centers (VA Pittsburgh Healthcare System and Corporal Michael J. Crescenz VA Medical Center) were consented and enrolled (SPA1 cohort). Due to accelerated recruitment, an additional 157 patients (SPA2 cohort) were enrolled to increase power to detect sex differences in secondary analyses. Of the 517 patients that were consented and enrolled, all were 50 years or older, were non-Hispanic white or African American, received primary care at a participating site, had frequent pain characteristics of symptomatic knee OA with a pain level of 4 or higher on a 0-10 scale, and could speak, read, and write in English. Exclusion criteria included: self-reported serious problems with hearing, eyesight, or memory and an inability to complete study-related telephone calls and program activities that involved reading and writing, diagnosed with any type of arthritis other than OA or degenerative arthritis, treated for cancer in the past 3 years, had a steroid injection for knee pain or had a knee replacement in the past 3 months, plans to have a knee replacement in the next 6 months, no reliable telephone number, or answering 2 or more items incorrectly on a 6-item screener for cognitive impairment. Eligible participants completed an in-person baseline assessment of study outcomes, mediators, and control variables and were randomized to a 6-week positive or control program.
The positive program consisted of completing at-home activities (1 per week) that have been shown to increase positive psychological skills such as gratitude, kindness, and mindfulness. The control program consisted of similarly structured, but affectively neutral, activities. Both groups received weekly telephone calls from trained interventionists to assess program completion and clarify instructions. Outcomes and proposed mediating variables were assessed via telephone surveys at 1-, 3-, and 6-months post-intervention. The primary outcomes were self-reported pain and functional difficulty measured using the WOMAC Osteoarthritis Index (possible range 0-100; higher scores=worse). Hypothesized mediators included positive/negative affect, satisfaction with life, depressive symptoms, arthritis self-efficacy, pain coping, pain catastrophizing, perceived discrimination, global stress, and social support. We tested the study hypothesized 3-way interaction between treatment group, race, and time using linear mixed models that allow repeated measures on the continuous outcomes for each participant. Similar models were used to test lower order interactions between treatment group and time and the main effect of time. We also conducted exploratory subgroup analyses to elucidate our primary findings.
Analyses of the main outcomes of self-reported pain and functional difficulty have been completed for (N=360). Of these participants, 287 (79.7%) completed at least 5 weekly calls during the 6-week intervention period, and 239 (66.4%) reported entirely or partially completing at least 5 correct activities. A total of 85.0%, 85.8% and 86.1% completed the 1-, 3-, and 6-month post-intervention follow-up surveys, respectively. Adherence and follow-up rates did not differ significantly by treatment group or by race (p-values>0.05). Participants at baseline reported mean pain and functional difficulty scores of 48.8 (standard deviation [SD]=17.6) and 46.8 (SD=18.1), respectively. The hypothesized 3-way interaction between treatment group (positive vs. neutral control program), race (non-Hispanic White vs. non-Hispanic African American), and time (baseline vs. 1-month, 3-month, and 6-month post-intervention) was not significant for either outcome. Models testing only the treatment group by time interaction and main effect of time revealed significant main effects of time, but no 2-way interactions. Both pain and functional difficulty decreased significantly over time (pain: M=48.8 vs. 44.5 vs. 43.6 vs. 42.4 at baseline and 1-, 3-, and 6-months, respectively, overall test for time: Chi-Squared (3)=49.50, p<0.001; functional difficulty: M=46.8 vs. 43.9 vs. 43.4 vs. 42.9 at baseline and 1-, 3-, and 6-months, respectively, overall test for time: Chi-Squared (3)=22.11, p<0.001). Subgroup analyses indicated that restricting analyses to patients who had greater disease or symptom severity at baseline, who showed greater adherence to the intervention protocol (i.e., reported completing at least 5 of 6 assigned intervention activities), or who completed all follow-up assessments yielded the same findings as in the intent-to-treat analyses. The subgroup analyses ruled out insufficient disease severity, poor adherence, and study drop out as explanations for the lack of differences across treatment groups. Given that the intervention did not have the hypothesized effects on pain and functional difficulty, we did not conduct the planned mediation analyses. Instead, we examined each of the proposed mediators as outcomes and determined that they were unchanged by the intervention. Analyses focusing on gender differences are planned for the SPA2 cohort data.
This large 2-site randomized controlled trial tested a theoretically and empirically-driven psychosocial intervention to reduce well-documented racial disparities in pain and functioning in Veterans with arthritis. The study fills important gaps in the literature by testing the use of positive psychological interventions (PPIs) in older Veterans with chronic pain, testing for racial differences in response to PPIs, and comparing the long-term effects of PPIs to those of a strong control group. Unfortunately, the study does not support PPIs as a stand-alone treatment for pain among white or African American Veterans with knee OA. To ensure that Veterans with chronic pain benefit from PPIs, additional effort would be needed to identify specific PPI components that resonate with this population. The potential additive effect of incorporating PPIs into more comprehensive pain treatment regimens should also be considered.
- Tighe CA, Youk A, Ibrahim SA, Weiner DK, Vina ER, Kwoh CK, Gallagher RM, Bramoweth AD, Hausmann LRM. Pain Catastrophizing and Arthritis Self-Efficacy as Mediators of Sleep Disturbance and Osteoarthritis Symptom Severity. Pain medicine (Malden, Mass.). 2020 Mar 1; 21(3):501-510.
- Vina ER, Hausmann LRM, Obrosky DS, Youk A, Ibrahim SA, Weiner DK, Gallagher RM, Kwoh CK. Social & psychological factors associated with oral analgesic use in knee osteoarthritis management. Osteoarthritis and cartilage. 2019 Jul 1; 27(7):1018-1025.
- Hausmann LRM, Youk A, Kwoh CK, Ibrahim SA, Hannon MJ, Weiner DK, Gallagher RM, Parks A. Testing a Positive Psychological Intervention for Osteoarthritis. Pain medicine (Malden, Mass.). 2017 Oct 1; 18(10):1908-1920.
- Hausmann LRM, Ibrahim SA, Kwoh CK, Youk A, Obrosky DS, Weiner DK, Vina E, Gallagher RM, Mauro GT, Parks A. Rationale and design of the Staying Positive with Arthritis (SPA) Study: A randomized controlled trial testing the impact of a positive psychology intervention on racial disparities in pain. Contemporary clinical trials. 2018 Jan 1; 64:243-253.
- Hausmann LR. Using Positive Psychology to Improve Pain and Functioning in Veterans with Arthritis. Spotlight on Pain Management [Cyberseminar]. VA HSR&D. 2018 Mar 6.
Musculoskeletal Disorders, Health Systems
Treatment - Comparative Effectiveness
Behavioral therapy, Comparative Effectiveness, Disparities, Ethnicity/Race, Outcomes - Patient, Pain, Symptom Management