1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information


To examine the efficacy of psychological interventions for rheumatoid arthritis (RA), and to determine whether self-regulation interventions demonstrate efficacy superior to that of other psychological treatments.


Only randomized controlled trials testing a face-to-face psychological intervention among patients with RA were included. Two independent investigators extracted pertinent study data, rated each study on a scale of methodologic quality, and assessed each treatment condition for its inclusion of 5 behavior-change techniques derived from self-regulation theory (goal setting, planning, self-monitoring, feedback, and relapse prevention).


Twenty-seven trials were included, and cumulative effect sizes were calculated for the 5 outcomes. Significant effect sizes (Hedges' g) were found at posttreatment for physical activity (0.45), pain (0.18), disability (0.32), depressive symptoms (0.23), and anxiety (0.17). At followup (range 2–14 months), significant effect sizes were obtained for physical activity (0.36), pain (0.13), disability (0.15), and depressive symptoms (0.32). Comparative analyses revealed that interventions utilizing more self-regulation techniques reduced depressive symptoms and anxiety significantly more than interventions utilizing fewer such techniques. Additionally, depressive symptoms were reduced significantly more among recently diagnosed RA patients than among those with longstanding RA.


Psychological interventions are beneficial for many patients with RA, particularly when it comes to increasing physical activity levels. Intervention techniques derived from self-regulation theory appear to play a role in reducing depressive symptoms and anxiety among patients with RA.


  1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information

Rheumatoid arthritis (RA) is a chronic autoimmune disorder that affects ∼1% of the general population (1). Common symptoms include pain, swelling, and tenderness in joints; morning stiffness; and functional limitations (disability). These symptoms are most frequently treated with a combination of medication and physical exercise (2, 3). In addition, however, several psychosocial and behavioral treatments have been developed to address these symptoms. Some of the most commonly implemented psychological interventions are stress management training to help patients cope with functional problems caused by their RA, cognitive–behavioral therapy (CBT) to teach patients methods to control their arthritis pain (4), and/or education to help patients make informed decisions about how to best self-manage their condition (5).

Among these psychological interventions for RA, the Arthritis Self-Management Program (ASMP) (6) is perhaps the most widely implemented. The ASMP combines elements of patient education and cognitive–behavioral interventions, such as educating patients about arthritis and the importance of physical activity, while engaging them in goal setting, action planning, and self-monitoring of physical exercise, as well as other self-management strategies. In several trials, the ASMP and other ASMP-based interventions have increased the practice of physical exercise, as well as reduced pain, depressive symptoms, and anxiety among patients with RA (6–9).

Several techniques utilized by the ASMP (goal setting, action planning, self-monitoring, and provision of feedback) closely reflect the basic tenants of self-regulation theory (SRT) (10, 11). SRT puts forth the idea that behavior is goal directed, and that by taking an active rather than passive role in the management of a chronic condition, patients can create their own pathways to goal achievement (10, 12). The techniques of goal setting, planning, self-monitoring, and feedback serve then to focus the attention of patients on the steps necessary to self-manage their condition.

Apart from the successes of the ASMP among arthritis patients, the ability of interventions based on self-regulation to improve behavioral, physical, and psychological outcomes has also been demonstrated among other populations with chronic diseases, e.g., coronary heart disease (13), asthma (14), and renal disease (15). For that reason, this meta-analysis will examine whether interventions that use more core self-regulation principles (goal setting, planning, self-monitoring, feedback, and relapse prevention) (16) will produce greater treatment gains for RA patients than interventions that use fewer of these techniques.

Several previous meta-analyses have shown that psychological interventions produce small significant effect sizes upon physical and psychological outcomes important in RA (4, 17, 18). However, to our knowledge, no meta-analysis has yet assessed the effects of psychological interventions upon levels of physical activity among patients with RA. The American College of Rheumatology guidelines for the management of RA (2) include strengthening and aerobic conditioning, and various studies have demonstrated the safety and efficacy of physical exercise at improving strength, disability, disease activity, pain, mobility, and aerobic capacity among RA patients (19–21). However, patients with RA remain less physically active than members of the general population, and 35–75% of RA patients do not comply with the physical activity recommendations of their rheumatologist or physical therapist (7, 22–24). Many RA patients, therefore, miss out on the assumed benefits of physical exercise, and place themselves at greater risk of developing other chronic illnesses associated with a sedentary lifestyle (25).

Low adherence to physical activity recommendations, and lowered physical activity in general, are influenced by a number of personal and environmental factors, including the high prevalence of sedentary work and leisure activities in Western societies, lack of access to safe or appropriate exercise settings, lack of time, knowledge, and motivation (including depression), and certain aspects of patient-provider interactions (26). As a result, psychological interventions that help patients to better plan for physical activity, increase motivation and problem solving skills, or improve aspects of patient care might lead to increased physical activity and prove beneficial in clinical practice. This review will examine the effects of psychological interventions upon physical activity levels among patients with RA.

The primary aims of this study were to determine the overall efficacy of psychological interventions of increasing physical activity, as well as of reducing pain, disability, depressive symptoms, and anxiety among patients with RA, and to determine whether interventions including more techniques derived from SRT produce greater treatment gains than those using fewer such techniques.


  1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information

Search strategy.

This meta-analysis (without protocol) included only randomized controlled trials (RCTs) published in peer-reviewed journals in either English or Dutch, and which tested face-to-face psychological interventions for adult humans with RA. To be included, studies must have reported data suitable for meta-analysis for at least 1 of these 5 outcomes: physical activity, pain, disability, depressive symptoms, and anxiety.

To find RCTs, searches were conducted for the years 1980–2008 within the electronic databases PsycINFO, Medline, and the central catalog of Dutch libraries. For the explicit search strategies for PsycINFO and Medline, see Supplemental Appendix A (available in the online version of this article at

Recovery of trials.

Our initial search returned 288 relevant articles. After reviewing the abstracts, 54 articles that met the inclusion criteria remained. The reference lists of all review articles eliminated at this stage were then scanned, revealing 4 additional studies to be considered for inclusion; however, the full text of one of these articles was unavailable. The full texts of the remaining 57 articles were then reviewed and 30 of these were subsequently excluded for the following reasons: lack of a control group (n = 3), did not report on outcomes of interest (n = 3), did not provide separate data for different illness groups (n = 9), data provided were not suitable for meta-analysis and further data were unavailable (n = 13), and the study provided secondary analysis of data provided by another study (n = 2). Twenty-seven studies were finally included, some of which tested multiple treatment conditions. Figure 1 demonstrates how the recovered articles were scrutinized, and the included studies are summarized in Table 1.

thumbnail image

Figure 1. Flow of studies through the search and acquisition process. RA = rheumatoid arthritis; OA = osteoarthritis.

Download figure to PowerPoint

Table 1. Characteristics of included studies*
Author, year (ref.)CountryGroups used in analysis (no.)Sessions/lengthRA criteria (ref.)Intervention providerMeasureGSPLSMFBRPSR scoreStudy quality
  • *

    RA = rheumatoid arthritis; GS = goal setting; PL = planning; SM = self-monitoring; FB = provision of feedback; RP = relapse prevention; SR = self-regulation; ARA = American Rheumatism Association; VAS = visual analog scale; NR = not reported; HAQ = Health Assessment Questionnaire; HADS = Hospital Anxiety and Depression Scale; CBT = cognitive–behavioral therapy; STAI = State-Trait Anxiety Inventory; DACL = Depression Adjective Checklist; MD = medical doctor; AIMS = Arthritis Impact Measurement Scales; IRGL = Impact of Rheumatic Diseases on General Health and Lifestyle Questionnaire; BDI = Beck Depression Inventory; JP = joint protection; ACR = American College of Rheumatology; OT = occupational therapist; PT = physical therapist; ZDS = Zung Depression Scale; DS = Depression Scale; SW = social worker; ROM = range of motion; CES-D = Center for Epidemiologic Studies Depression Scale; SCL90-R = Symptom Checklist-90-Revised.

Appelbaum et al, 1988 (27)USRelaxation, cognitive pain management strategies (9)10 × 1 hr/ 6 wksARA (75)PhD studentVAS1.502104.524
  Symptom monitoring control group (9; active)           
Barlow et al, 2000 (8)UKGroup Arthritis Self-Management Program (114)6 × 2 hrs/ 6 wksNRLaypersonVAS, HAQ, HADS21020535
  Wait-list control group (77; passive)           
Bradley et al, 1984 (28)USGroup CBT with biofeedback, relaxation, education, GS (4)15/unknownARA (76)PsychologistVAS, STAI, DACL12001421
  No intervention control group (3; passive)           
Brus et al, 1998 (29)The NetherlandsGroup training in physical exercise, RA information, contracting (25)4 × 2 hrs/ 4 wksARA (77)MDAIMS, HAQ12010.54.533
  No intervention control group (30; passive)           
Evers et al, 2002 (30)The NetherlandsCBT with relaxation, GS, cognitive restructuring (30)10 × 1 hr/ 20 wksARA (77)PhD studentIRGL, BDI1.50.501.525.539
  Standard medical care (29; active)           
Hammond et al, 1999 (31)UKGroup education, contracting, JP, problem solving (17)4 × 2 hrs/ 12 wksACR (78)OTHAQ21.5220.5834
  Wait-list control group (18; passive)           
Hammond et al, 2001 (32)UKGroup education, contracting, JP, problem solving (63)4 × 2 hrs/ 4 wksNRPTVAS21.52207.541
  RA education control group (58; active)           
Helewa et al, 1991 (33)CanadaJP/aids use education, counseling (52)Unknown/ 6 wksARA (76)OTHAQ, BDI00000031
  Wait-list control group (50; passive)           
Helliwell et al, 1999 (34)UKGroup RA education, pacing, JP, exercise (43)4 × 2 hrs/ 4 wksARA (77)MDHAQ10.50102.532
  Wait-list control group (34; passive)           
Hill et al, 2001 (35)UKRA education, exercise, JP, coping (33)7 × 30 mins/6 mosARA (77)NursePain score00000035
  Standard medical care (30; active)           
Huiskes et al, 1991 (36)The NetherlandsGroup CBT/OT: GS, relaxation, homework, active coping, JP/exercise (21)10 × 2 hrs/ 10 wksARA (77)PhD studentIRGL10020324
  Wait-list control group (19; passive)           
Kaplan and Kozin, 1981 (37)USRA education, group counseling of RA problems (11)12 × 1.5 hrs/12 wksARA (76)CounselorZDS00000026
  RA education (17; active)           
Kraaimaat et al, 1995 (38)The NetherlandsGroup CBT: relaxation, GS, FB, active coping, homework, education (24)10 × 2 hrs/ 10 wksARA (77)PsychologistIRGL1001.502.536
  Wait-list control group (19; passive)           
Leibing et al, 1999 (39)AustriaGroup education: relaxation, pain management, SM (19)12 × 1.5 hrs/12 wksARA (77)MDVAS, HAQ, DS, STAI21202742
  Standard medical care (20; active)           
Lindroth et al, 1997 (40)SwedenGroup education: pacing, pain coping, exercise, relaxation (37)8 × 2.5 hrs/ 8 wksARA (77)Team: nurse, MD, PT, OT, dietician, SWVAS, HAQ21020532
  Wait-list control group (36; passive)           
Lundgren and Stenstrom, 1999 (41)SwedenGroup relaxation: Jacobson and guided imagery (33)20 × 30 mins/10 wksARA (77)PTVAS00000039
  No intervention control group (27; passive)           
Neuberger et al, 1993 (42)USSelf-instruction, practice of/contracting for ROM/JP (15)6 contacts/ 16 wksNRNurseVAS, CES-D20020420
  No intervention control group (11; passive)           
O'Leary et al, 1988 (43)USGroup CBT: GS, pain management, relaxation, self-reward (14)5 × 2 hrs/ 5 wksNRUnknownVAS, HAQ, ZDS22212926
  Self-help book (12; active)           
Parker et al, 1988 (44)US1-week clinic stay, group CBT: pain coping, relaxation, stress management (29)24 × 1 hr/ 6 mosARA (76)PhD studentVAS, AIMS11011.54.532
  Standard medical care (28; active)           
Parker et al, 1995 (45)USCBT: stress management, relaxation, pain coping strategies (44)10 × 1.5 hrs/10 wksARA (77)CounselorVAS, AIMS11011.54.537
  Standard medical care (44; active)           
Pradhan et al, 2007 (46)USGroup mindfulness-based stress reduction, home practice (28)8 × 2.5 hrs/ 8 wksARA (75)CounselorSCL90-R00000042
  Wait-list control group (32; passive)           
Radojevic et al, 1992 (47)USGroup CBT: relaxation, cognitive pain coping, family participation (15)4 × 2.5 hrs/ 6 wksARA (77)PhD studentAIMS, CES-D00201333
  No intervention control group (15; passive)           
Riemsma et al, 2003 (48)The NetherlandsGroup GS, contracting, FB, relaxation, exercise, partner participation (71)5 × 2 hrs/ 5 wksARA (77)NurseAIMS20.5020.5535
  Self-help book (76; active)           
Scholten et al, 1999 (49)AustriaGroup education: RA, pain, relaxation, exercise, JP, GS (38)9 × 1.5 hrs/ 2 wksARA (77)Team: psychologist, MD, PT, SWHAQ, BDI10201.54.528
  Wait-list control group (30; passive)           
Sharpe et al, 2001 (50)UKEducation/CBT: RA, relaxation, GS, pacing, problem solving (23)8 × 1 hr/ 8 wksARA (77)PhD studentVAS, HAQ, HADS11.52026.540
  Standard medical care (22; active)           
Shearn and Fireman, 1985 (51)USGroup stress management, relaxation, coping strategies (22)10 × 1.5 hrs/10 wksARA (76)PhD studentVAS, HAQ, CES-D00000027
  No intervention control group (26; passive)           
Taal et al, 1993 (7)The NetherlandsGroup education: RA, exercise, GS, problem solving (27)5 × 2 hrs/5 wksARA (77)Nurse or PTAIMS, HAQ20120531
  No intervention control group (30; passive)           


Coding and data extraction were conducted by 2 independent coders using an a priori developed data extraction form. For the complete coding form, see Supplemental Appendix B (available in the online version of this article at Articles were coded for the following features: type of intervention tested, country/year of the study conducted, type of RA diagnostic criteria, provider of the intervention, number of sessions, total time of patient-provider contact in hours, and the type of control group used for comparison. We also assessed each study for its stated aims, outcome measures used, and the average age and disease duration of its participants.

Study quality and risk of bias.

A 29-item version of the Cochrane Collaboration Depression Anxiety, and Neurosis Review Group (CCDAN) scale (52), as adapted by Lackner et al (53) to suit the evaluation of psychological trials, was used to assess the quality of the included studies. Each of the 29 items was scored as a 0, 1, or 2 according to the criteria inherent in the scale (Table 2 and Supplemental Appendix B, available in the online version of this article at Any discrepancies were discussed by the 2 coders until a consensus was reached. Four study quality items (blinding of assessors, concealed treatment allocation, inclusion of dropouts, and manualized treatment), as well as whether a study's control contained an active component, were considered the largest risks of bias among the included studies.

Table 2. Consensus ratings of methodologic quality for included studies (ref.)*
Methodologic criterion(27)(8)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45)(46)(47)(48)(49)(50)(51)(7)
  • *

    1 = done and/or reported to some extent; 2 = adequately done and/or adequately reported; 0 = not done and/or not reported.

Clear objectives122222212222222202222222222
Sample size020110222200002100121021011
Trial duration222211202121222211221122222
Power calculation010002220000010200002100200
Allocation method222222222222222221222222222
Allocation concealment020000012100000000102000200
Treatment clearly described121122211120221212012222102
Manualized treatment010021100000220000000210101
Representative sample110111111212122111111110211
Inclusion criteria211222021212222211222122222
Exclusion criteria020122220101120100122120202
Described demographics210122221012222200222121112
Assessor blinded000101111100110101001000200
Treatment compliance200202202201100211222200012
Treatment side effects201120211200120201002110100
Dropout information012122211222212202222111111
Outcome measures222222222222222222222222222
Between-group comparisons020222212122222221222221221
Dropout inclusion010011002000000000001000100
Well-presented results222120220120220011002222220
Appropriate analyses120122211221221221212222222
Justified conclusions222222222222222221222222222
Interests declared022222220202222222222222222
Allegiance to therapy000000000000000000000000000
Followup duration222211202010122201221022122
Cointervention avoided000020020200010102000000000
Drug use assessed000120202202122202022200100
Treatment credibility000000000000100000220110000
Consecutive subjects000200000000022020020002200
Total quality243521333934413132352426364232392026323742333528402731
Self-regulation principles.

Each treatment condition from the included studies was assessed for the presence of the 5 core self-regulation principles (goal setting, planning, self-monitoring, feedback, and relapse prevention) using the following a priori validated procedure. Each self-regulation principle was assigned a score of 0, 1, or 2 based on the extent to which that principle was a part of the intervention (Supplemental Appendix B, available in the online version of this article at Both the published intervention descriptions and any secondary references cited therein were used to determine the final score. The authors of studies that inadequately described their treatment conditions, according to our coding of study quality (i.e., scored <2 on item 7 of the modified CCDAN scale), were contacted when possible for more information regarding the content of their interventions.

To increase the transparency of the coding process for later comparison, the coders were instructed to note which source (original article or specific cited reference), page, and paragraph contained the text on which their coding decision was based. In the case of minor discrepancies (0 versus 1 or 1 versus 2), the mean of the 2 ratings was used, and in the case of major discrepancies (0 versus 2), the 2 coders discussed the items and referred back to their notes to reach a consensus.

At the end of the coding process, a total self-regulation score was calculated by summing the 5 principle scores for each treatment condition. Total self-regulation scores could therefore range from 0–10. The included studies were then dichotomized at the median to allow for subsequent comparisons of the effects of studies high in self-regulation versus those low in self-regulation. For studies that tested multiple treatment groups against the same control group, only the treatment group with the highest total self-regulation score was included in order to maintain between-studies independence.


Meta-analyses were conducted using Comprehensive Meta Analysis software (Biostat) (54). The analyses were weighted by trial size, and the DerSimonian and Laird random-effects model was used (55). For each of the assessed outcomes, a separate meta-analysis was conducted to determine the cumulative effect sizes (Hedges' g) (56) at both posttreatment and followup. The statistical consistency (heterogeneity) of included studies was examined using the I2 statistic (57).

Additional analyses.

Comparative subgroups analyses were used to examine the effects of categorical study variables (treatment type, dichotomized self-regulation score, within-study risk of bias, and dichotomized disease duration of study participants) upon treatment effects for each outcome. Meta-regressions were subsequently used to examine the effects of continuous study variables (patient-provider contact time, total study quality, and sample size) upon treatment effects for each outcome. All analyses involving subgroup comparisons and meta-regression were conducted using SPSS, version 16.0, for Windows (SPSS) (58). Publication bias was then assessed by visually examining funnel plots for asymmetry.


  1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information

Analyses of outcomes at posttreatment.

Cumulative effect sizes and heterogeneity statistics obtained from posttreatment data are presented in Figures 2, 3, and 4 for each of the assessed outcomes (physical activity, pain, disability, depressive symptoms, and anxiety).

thumbnail image

Figure 2. Forest plot of posttreatment data for physical outcomes. SR = self-regulation; 95% CI = 95% confidence interval.

Download figure to PowerPoint

thumbnail image

Figure 3. Forest plot of posttreatment data for psychological outcomes. SR = self-regulation; 95% CI = 95% confidence interval.

Download figure to PowerPoint

thumbnail image

Figure 4. Forest plot of posttreatment physical activity data. SR = self-regulation; 95% CI = 95% confidence interval.

Download figure to PowerPoint

Analyses of outcomes at followup.

Physical activity.

Four studies provided followup data on physical activity for periods varying from 10–14 months after baseline. These studies produced a small but significant effect size of g = 0.361 (P = 0.020; 95% confidence interval [95% CI] 0.058, 0.665). The effect size for this subset of studies at posttreatment was g = 0.471 (P = 0.009; 95% CI 0.116, 0.826).


Thirteen studies provided followup data for pain. The followup periods ranged from 2–14 months posttreatment. The effect size for this data from baseline to final followup was not significant at g = 0.127 (P = 0.069; 95% CI −0.010, 0.265). The effect size for this subset of studies at posttreatment was g = 0.189 (P = 0.006; 95% CI 0.054, 0.325).


Twelve studies provided followup data for disability, with followup periods ranging from 2–14 months posttreatment. The effect size for this data from baseline to final followup was g = 0.145 (P = 0.047; 95% CI 0.002, 0.288). The effect size for this subset of studies at posttreatment was g = 0.417 (P = 0.001; 95% CI 0.179, 0.655).


Twelve studies provided followup data for depression with followup periods ranging from 2–14 months posttreatment. The effect size for this data from baseline to final followup was g = 0.318 (P < 0.001; 95% CI 0.160, 0.475). The effect size for this subset of treatments from baseline to posttreatment was g = 0.279 (P = 0.015; 95% CI 0.054, 0.504).


The cumulative effect size for anxiety at final followup was not significant at g = 0.122 (P = 0.200; 95% CI = −0.065, 0.308, k = 8). At posttreatment, this subset of 8 treatments had a cumulative effect size of g = 0.121 (P = 0.201; 95% CI −0.064, 0.306).

Further analyses.

Interrater reliability.

Calculation of interrater reliability yielded a Cohen's kappa of 0.68 for study quality coding and of 0.78 for self-regulation score coding, both of which are satisfactory (59, 60).

Comparative subgroup analyses.

For each outcome variable, comparative subgroup analyses separately examined the effects of the categorical study variables treatment type (CBT, patient education, or stress management), median-dichotomized self-regulation score (high score of ≥4.5 versus low score of <4.5), control type (active versus passive), adequacy of blinding (adequate versus inadequate), allocation concealment (concealed versus not concealed), inclusion of withdrawals/dropouts (included versus not included), and dichotomized average disease duration of participants (disease duration of ≥10 years versus disease duration of <10 years) upon effect sizes.

To assess normality of the effect size distributions, Shapiro-Wilk tests were conducted on the effect sizes of each outcome. For the outcomes pain, disability, depressive symptoms, and anxiety, the tests were nonsignificant (meaning that the effect size distributions for these outcomes could be considered normal). Subsequently, Welch t-tests were used to compare subgroups for these outcomes. The results of these comparative analyses are presented in Table 3. Since the Shapiro-Wilk statistic was significant for the physical activity outcome, normality could not be assumed and no comparative analyses were conducted.

Table 3. Comparative analyses assessing the effects of study and treatment characteristics upon effect size, separated by outcome*
Subgroup analysesPhysical activityPainDisabilityDepressive symptomsAnxiety
  • *

    k = number of studies included per subgroup per outcome; g = Hedges' g effect size; CBT = cognitive–behavioral therapy; md = missing data; nc = no comparison possible; ns = not significant (P > 0.1); DD = mean disease duration of patient sample.

 CBTmdmd 100.15 80.37 90.25 70.14 
 Patient education50.45nc90.15ns60.29ns50.29ns30.21ns
 Stress managementmdmd 20.40 20.2920.16mdmd
 High40.47nc150.20ns130.36ns110.40≤ 0.0190.22≤ 0.05
Treatment allocation               
 Not concealed50.45180.20120.17140.2380.14
 Includedmdmdnc40.12ns50.06≤ 0.0540.34ns30.24ns
 Not included50.45180.21120.41150.2080.12
 Active20.39nc110.20ns80.33ns70.40≤ 0.1060.23ns
 ≥10 years20.34nc110.17ns80.24ns90.14≤ 0.0550.08ns
 <10 years10.4880.2080.4360.5540.33
Meta–regression analyses.

A regression line was fit to the study quality data to determine if study quality has increased over time within this body of research. A line with βs = 0.74 (P < 0.001) was calculated, indicating that methodologic quality of studies has generally increased over time.

Additionally, meta–regression analyses examined the effects of the continuous study variables of study quality, hours of patient-provider contact time, and total sample size upon the effect sizes of each outcome. For the physical activity outcome, larger sample size was significantly associated with smaller effect sizes. The meta–regression analyses revealed no other significant associations (Table 4).

Table 4. Meta–regression analyses assessing the effects of study and treatment characteristics upon effect size, separated by outcome*
Meta-regressionsPhysical activityPainDisabilityDepressive symptomsAnxiety
  • *

    βs = standardized beta coefficient; k = number of studies included per subgroup per outcome; ns = not significant (P > 0.1).

Study quality−0.025ns0.0122ns−0.0317≤ 0.100.0119ns0.0211ns
Contact time0.034ns0.0020ns0.0216ns−0.0216ns−0.0210ns
Sample size−0.015≤ 0.05−0.0022ns−0.0017ns0.0019ns0.0011ns

Publication bias.

To examine whether publication bias may have affected our cumulative effect sizes, funnel plots of effect size versus SEM were examined for each outcome. Visual inspections revealed some asymmetry for physical activity, but not for the other outcomes. Additionally, fail-safe numbers (FSNs) were calculated for each outcome. Based on the limit (5k + 10) set forth by Rosenthal (61), publication bias is unlikely for the disability outcome (FSN = 147), but not for the others (physical activity FSN = 24, pain FSN = 80, depression FSN = 61, and anxiety FSN = 3).


  1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information

The benefits of physical exercise for most patients with RA are well documented. Despite this, however, a high percentage of patients with RA lead a sedentary lifestyle (7, 19–24). Based on the set of studies included here, psychological interventions appear to have a positive impact upon physical activity levels among patients with RA. A cumulative effect size (Hedges' g) of 0.45 was found for 5 patient education interventions that reported on physical activity, which was most commonly assessed as self-reported exercise sessions per week. This effect size represents a gain of ∼1 exercise session per week for the treatment groups over the control groups. These results, however, do not provide any insight into the types of exercise these patients increased, or whether this general increase in physical exercise leads to the same improvements in pain and disability demonstrated by more structured exercise programs (62, 63). Since RA is a chronic and progressive condition, longitudinal data are crucial in determining whether increased physical activity leads to a sustained (rather than deteriorating) radiologic condition, the maintenance of functional ability, and, ultimately, reduced health care utilization.

The set of psychological interventions included in this analysis had small effects upon the physical and psychological outcomes of RA. The effect sizes (Hedges' g) calculated for pain (0.18), disability (0.32), depressive symptoms (0.23), and anxiety (0.17) are significant, small, and similar to those calculated in previous meta-analyses examining psychological treatments for RA (17, 18) and mixed arthritis populations (4). These small effect sizes occur in addition to what is achieved by standard care alone, thereby illustrating the possible utility of such treatments in practice. However, since all outcomes were measured using self-report questionnaires that vary in their respective sensitivities to change, the relationship between statistical and clinical significance should be examined on a case-by-case basis. Furthermore, when examining these results, it should be noted that several studies yielded negative effect sizes for various outcomes, indicating that psychological treatments are not beneficial for all patients or for all outcomes (8, 31, 36, 38, 42, 51).

To compare the efficacy of several categories of psychological treatment, we conducted a series of comparative subgroup analyses. Confirming the results of previous researchers (17, 18), neither CBT, patient education, nor stress management interventions produced effect sizes that were significantly greater than the others. As previously noted (64), CBT, patient education, and stress management interventions utilize overlapping techniques derived from multiple theoretical backgrounds (Table 1). The similarity of their treatment effects is therefore not surprising.

In an attempt to isolate a specific set of techniques (as opposed to over-reaching treatment modalities) and to assess its combined impact upon RA outcomes, we assessed each included intervention for the use of 5 techniques derived from SRT: goal setting, planning, self-monitoring, feedback provision, and relapse prevention. Comparative analyses subsequently revealed that studies that utilized more of these self-regulation techniques reduced depressive symptoms and anxiety significantly more than those utilizing fewer such techniques.

Since several of the studies from this analysis excluded clinically depressed or anxious patients, and the studies' baseline means generally indicated that patients reported only mild to moderate anxiety and depressive symptoms, we must question whether these findings also hold for RA patients with clinical levels of depression or anxiety. Since highly distressed patients have been shown to have difficulty setting realistic goals and engaging in goal-directed behaviors (65), self-regulation interventions are most likely to be effective among RA patients with subclinical levels of anxiety and depression. Before beginning any behavioral or exercise program, RA patients with clinical levels of psychological distress should receive evidence-based treatment for those symptoms in order to optimize the chances of success in that program.

Among patients with subclinical levels of depression and anxiety, self-regulation interventions perhaps reduce psychological symptoms through increases in arthritis self-efficacy, or through one's perceived ability to influence or control various aspects of arthritis (8). Techniques that typify self-regulation interventions (goal setting, self-monitoring, and receiving feedback) have each been linked in previous research to increased self-efficacy, which has in turn been linked to improvements in psychological variables among patients with RA (66). Achievement of behavioral goals, which are believed to directly affect the course of arthritis, empowers patients and may subsequently reduce worry and negative thoughts about living with RA. This relationship is supported by research in other chronic disease populations as well, in which perceived control and self-efficacy have been inversely linked to depression and anxiety (67–69).

Since depressive disorder is 2–3 times more prevalent among patients with RA than among members of the general population (70), and since it has been linked to reduced physical activity adherence (23, 71, 72), it is an important target of intervention within this population. Our results, and those of previous researchers (18), suggest that psychological interventions reduce depressive symptoms most effectively among recently diagnosed RA patients. This is perhaps the case because depressive symptoms among recently diagnosed patients stem from factors that are more alterable by psychological interventions. One might expect recently diagnosed patients to experience depressive symptoms in response to uncertainty about the future, or as a reaction to the unknown course of the illness with which they have just been diagnosed; whereas patients with more longstanding diagnoses may experience depressive symptoms stemming from the pain and functional limitations more common in the later stages of RA (73). Future research should investigate other patient characteristics (e.g., personality or illness perceptions) that might also predict the success of psychological treatments for RA.

The present study has a number of limitations. First, the large cumulative effect size found for physical activity increases must be taken with caution due to the small number of studies that reported on this outcome (k = 5). Although the cumulative effect size was itself significant, and 2 of the 5 included studies demonstrated large increases in physical activity (7, 40), more studies examining this relationship are required before any firm conclusions can be drawn.

Second, since physical activity data were collected by all included studies using self-report measures, the tendency for people to overestimate their own levels of physical activity, particularly after repeated measurements (74), may have inflated the results. In future research, using a physical activity diary might allow for more accurate measurement of physical activity. It should also be mentioned that self-report measures used by the included studies to measure disability, anxiety, and depressive symptoms may not fully correspond with clinical or objective measurements of these same variables.

Third, publication bias might also have affected our results, since the FSNs for all outcomes besides disability failed to exceed the limits set forth by Rosenthal (61). Although a visual inspection of funnel plots did not reveal any obvious asymmetry, it is possible that the existence of unpublished studies with negative results could render our cumulative effect size estimates nonsignificant.

Finally, for some studies included in this analysis, our self-regulation coding process had to rely solely on the intervention descriptions provided in the published articles. In these circumstances, it is possible that our coding process underreported the use of self-regulation principles within those studies. Conversely, since there are no guarantees that all techniques mentioned in an intervention description were actually applied during treatment, it is possible that for some interventions we have overreported their use of self-regulation principles. In the future, researchers of psychological interventions should strive to fully and accurately report on the techniques they have utilized (64) and provide at least some assessment of therapist adherence to an intervention protocol.

This meta-analysis demonstrates the efficacy of psychological interventions as adjuncts to standard care for patients with RA, not only because of their effects upon physical and psychological variables, but also because of their apparent ability to increase physical activity among these patients. It appears, however, that most effects of psychological treatment seem to dissipate after the treatment (contact) period ends. Future research should therefore address how these small symptom improvements and behavioral changes might be increased and better maintained over time. Finally, our findings implicate the use of self-regulation interventions to target mild anxiety and depressive symptoms within this population, particularly among patients with more recent diagnoses. Future longitudinal research should examine whether self-regulation interventions that improve psychological status produce carry-over effects upon pain, disability, or physical activity.


  1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Mr. Knittle had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Knittle, Maes, De Gucht.

Acquisition of data. Knittle.

Analysis and interpretation of data. Knittle, Maes, De Gucht.


  1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information

We would like to thank Rebecca Schouten for her assistance with coding the included studies.


  1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information
  • 1
    Lawrence RC, Helmick CG, Arnett FC, Deyo RA, Felson DT, Giannini EH, et al. Estimates of the prevalence of arthritis and selected musculoskeletal disorders in the United States. Arthritis Rheum 1998; 41: 77899.
  • 2
    American College of Rheumatology Subcommittee on Rheumatoid Arthritis Guidelines. Guidelines for the management of rheumatoid arthritis: 2002 update. Arthritis Rheum 2002; 46: 32846.
  • 3
    Kennedy T, McCabe C, Struthers G, Sinclair H, Chakravaty K, Bax D, et al, for the British Society for Rheumatology Standards, Guidelines and Audit Working Group (SGAWG). BSR guidelines on standards of care for persons with rheumatoid arthritis. Rheumatology (Oxford) 2005; 44: 5536.
  • 4
    Dixon KE, Keefe FJ, Scipio CD, Perri LC, Abernethy AP. Psychological interventions for arthritis pain management in adults: a meta-analysis. Health Psychol 2007; 26: 24150.
  • 5
    Lorig K, Konkol L, Gonzalez V. Arthritis patient education: a review of the literature. Patient Educ Couns 1987; 10: 20752.
  • 6
    Lorig KR, Mazonson PD, Holma HR. Evidence suggesting that health education for self-management in patients with chronic arthritis has sustained health benefits while reducing health care costs. Arthritis Rheum 1993; 36: 43946.
  • 7
    Taal E, Riemsma RP, Brus HL, Seydel ER, Rasker J, Wiegman O. Group education for patients with rheumatoid arthritis. Patient Educ Couns 1993; 20: 17787.
  • 8
    Barlow JH, Turner AP, Wright CC. A randomized controlled study of the arthritis self-management programme in the UK. Health Educ Res 200; 15: 66580.
  • 9
    Lorig K, Gonzalez VM, Ritter P. Community-based Spanish language arthritis education program: a randomized trial. Med Care 1999; 37: 95763.
  • 10
    Maes S, Karoly P. Self-regulation assessment and intervention in physical health and illness: a review. Int Rev Appl Psychol 2005; 54: 26799.
  • 11
    Carver CS, Scheier MF. On the structure of behavioral self-regulation. In: BoekaertsM, PintrichPR, ZeidnerM, editors. Handbook of self regulation. 1st ed. London: Academic Press; 2000. p. 4280.
  • 12
    Bandura A. Social foundations of thought and action: a social cognitive theory. 1st ed. Englewood Cliffs (NJ): Prentice-Hall; 1986.
  • 13
    Sniehotta FF, Scholz U, Schwarzer R. Bridging the intention-behaviour gap: planning, self-efficacy, and action control in the adoption and maintenance of physical exercise. Psychol Health 2005; 20: 14360.
  • 14
    Clark NM, Gong ZM, Wang SJ, Lin X, Bria WF, Johnson TR. A randomized trial of a self-regulation intervention for women with asthma. Chest 2007; 132: 8897.
  • 15
    Christensen AJ, Wiebe JS, Moran PJ, Ehlers SL, Lawton WJ. Effect of a behavioral self-regulation intervention on patient adherence in hemodialysis. Health Psychol 2002; 21: 3937.
  • 16
    Michie S, Hardeman W, Fanshawe T, Prevost AT, Taylor L, Kinmonth AL. Investigating theoretical explanations for behaviour change: the case study of ProActive. Psychol Health 2008; 23: 2539.
  • 17
    Riemsma RP, Kirwan JR, Taal E, Rasker JJ. Patient education for adults with rheumatoid arthritis. Cochrane Database Syst Rev 2002; 3:CD003688.
  • 18
    Astin JA, Beckner W, Soeken K, Hochberg MC, Berman B. Psychological interventions for rheumatoid arthritis: a meta-analysis of randomized controlled trials. Arthritis Rheum 2002; 47: 291302.
  • 19
    Van den Ende CH, Vliet Vlieland TP, Munneke M, Hazes JM. Dynamic exercise therapy in rheumatoid arthritis: a systematic review. Rheumatology (Oxford) 1998; 37: 67787.
  • 20
    Hakkinen A, Hannonen P, Nyman K, Lyyski T, Hakkinen K. Effects of concurrent strength and endurance training in women with early or longstanding rheumatoid arthritis: comparison with healthy subjects. Arthritis Rheum 2003; 49: 78997.
  • 21
    Stenstrom CH. Home exercise in rheumatoid arthritis functional class II: goal setting versus pain attention. J Rheumatol 1994; 21: 62734.
  • 22
    Feinberg J. The effect of patient-practitioner interaction on compliance: a review of the literature in rheumatoid arthritis. Patient Educ Couns 1988; 11: 17187.
  • 23
    Brus H, van de Laar M, Taal E, Rasker J, Wiegman O. Compliance in rheumatoid arthritis and the role of formal patient education. Semin Arthritis Rheum 1997; 26: 70210.
  • 24
    Bradley LA. Adherence with treatment regimens among adult rheumatoid arthritis patients: current status and future directions [review]. Arthritis Care Res 1989; 2: S339.
  • 25
    Chakravarthy MV, Joyner MJ, Booth FW. An obligation for primary care physicians to prescribe physical activity to sedentary patients to reduce the risk of chronic health conditions. Mayo Clin Proc 2002; 77: 16573.
  • 26
    O'Grady M, Fletcher J, Ortiz S. Therapeutic and physical fitness exercise prescription for older adults with joint disease: an evidence based approach. Rheum Dis Clin North Am 2000; 26: 61746.
  • 27
    Appelbaum KA, Blanchard EB, Hickling EJ, Alfonso M. Cognitive behavioral treatment of a veteran population with moderate to severe rheumatoid arthritis. Behav Ther 1988; 19: 489502.
  • 28
    Bradley LA, Young LD, Anderson KO, McDaniel LK, Turner RA, Agudelo CA. Psychological approaches to the management of arthritis pain. Soc Sci Med 1984; 19: 135360.
  • 29
    Brus HL, van de Laar MA, Taal E, Rasker J, Wiegman O. Effects of patient education on compliance with basic treatment regimens and health in recent onset active rheumatoid arthritis. Ann Rheum Dis 1998; 57: 14651.
  • 30
    Evers AW, Kraaimaat FW, van Riel PL, de Jong AJ. Tailored cognitive-behavioral therapy in early rheumatoid arthritis for patients at risk: a randomized controlled trial. Pain 2002; 100: 14153.
  • 31
    Hammond A, Lincoln N, Sutcliffe L. A crossover trial evaluating an educational-behavioural joint protection programme for people with rheumatoid arthritis. Patient Educ Couns 1999; 37: 1932.
  • 32
    Hammond A, Freeman K. One-year outcomes of a randomized controlled trial of an educational-behavioural joint protection programme for people with rheumatoid arthritis. Rheumatology (Oxford) 2001; 40: 104451.
  • 33
    Helewa A, Goldsmith CH, Lee P, Bombardier C, Hanes B, Smythe HA, et al. Effects of occupational therapy home service on patients with rheumatoid arthritis. Lancet 1991; 337: 14536.
  • 34
    Helliwell PS, O'Hara M, Holdsworth J, Hesselden A, King T, Evans P. A 12-month randomized controlled trial of patient education on radiographic changes and quality of life in early rheumatoid arthritis. Rheumatology (Oxford) 1999; 38: 3038.
  • 35
    Hill J, Bird H, Johnson S. Effect of patient education on adherence to drug treatment for rheumatoid arthritis: a randomized controlled trial. Ann Rheum Dis 2001; 60: 86975.
  • 36
    Huiskes CJ, Kraaimaat FW, Brons MR, Bijlsma JW. The effect of behavior therapy and occupational therapy in patients with rheumatoid arthritis. Gedragstherapie 1991; 24: 25368.
  • 37
    Kaplan S, Kozin F. A controlled study of group counseling in rheumatoid arthritis. J Rheumatol 1981; 8: 919.
  • 38
    Kraaimaat FW, Brons MR, Geenen R, Bijlsma JW. The effect of cognitive behavior therapy in patients with rheumatoid arthritis. Behav Res Ther 1995; 33: 48795.
  • 39
    Leibing E, Pfingsten M, Bartmann U, Rueger U, Schuessler G. Cognitive-behavioral treatment in unselected rheumatoid arthritis outpatients. Clin J Pain 1999; 15: 5866.
  • 40
    Lindroth Y, Brattstrom M, Bellman I, Ekestaf G, Olofsson Y, Strombeck B, et al. A problem-based education program for patients with rheumatoid arthritis: evaluation after three and twelve months. Arthritis Care Res 1997; 10: 32532.
  • 41
    Lundgren S, Stenstrom CH. Muscle relaxation training and quality of life in rheumatoid arthritis. Scand J Rheumatol 1999; 28: 4753.
  • 42
    Neuberger GB, Smith KV, Black SO, Hassanein R. Promoting self-care in clients with arthritis. Arthritis Care Res 1993; 6: 1418.
  • 43
    O'Leary A, Shoor S, Lorig K, Holman HR. A cognitive-behavioral treatment for rheumatoid arthritis. Health Psychol 1988; 7: 52744.
  • 44
    Parker JC, Frank RG, Beck NC, Smarr KL, Bueschler KL, Phillips LR, et al. Pain management in rheumatoid arthritis patients: a cognitive behavioral approach. Arthritis Rheum 1988; 31: 593601.
  • 45
    Parker JC, Smarr KL, Buckelew SP, Stucky-Ropp RC, Hewett JE, Johnson JC, et al. Effects of stress management on clinical outcomes in rheumatoid arthritis. Arthritis Rheum 1995; 38: 180718.
  • 46
    Pradhan EK, Baumgarten M, Langenberg P, Handwerger B, Gilpin AK, Magyari T, et al. Effect of mindfulness-based stress reduction in rheumatoid arthritis patients. Arthritis Rheum 2007; 57: 113442.
  • 47
    Radojevic V, Nicassio PM, Weisman MH. Behavioral intervention with and without family support for rheumatoid arthritis. Behav Ther 1992; 23: 1330.
  • 48
    Riemsma RP, Taal E, Rasker JJ. Group education for patients with rheumatoid arthritis and their partners. Arthritis Rheum 2003; 49: 55666.
  • 49
    Scholten C, Brodowicz T, Graninger W, Gardavsky I, Pils K, Pesau B, et al. Persistent functional and social benefits 5 years after a multidisciplinary arthritis training program. Arch Phys Med Rehabil 1999; 80: 12827.
  • 50
    Sharpe L, Sensky T, Timberlake N, Ryan B, Brewin CR, Allard S. A blind randomized controlled trial of cognitive-behavioural intervention for patients with recent onset rheumatoid arthritis: preventing psychological and physical morbidity. Pain 2001; 89: 27583.
  • 51
    Shearn MA, Fireman BH. Stress management and mutual support groups in rheumatoid arthritis. Am J Med 1985; 78: 7715.
  • 52
    Moher D, Cook DJ, Jadad AR, Tugwell P, Moher M, Jones A, et al. Assessing the quality of reports of randomised trials: implications for the conduct of meta-analyses. Health Technol Assess 1999; 3: 198.
  • 53
    Lackner JM, Morley S, Dowzer C, Mesmer C, Hamilton S. Psychological treatments for irritable bowel syndrome: a systematic review and meta-analysis. J Consult Clin Psychol 2004; 72: 110013.
  • 54
    BioStat. Comprehensive meta-analysis: version 2. Engelwood (NJ): BioStat; 2005.
  • 55
    DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 1788.
  • 56
    Hedges LV. Distribution theory for Glass's estimator of effect size and other related estimators. J Educ Stat 1981; 6: 10728.
  • 57
    Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003; 327: 55760.
  • 58
    Statistical package for the social sciences: version 16.0.2. Chicago: SPSS; 2008.
  • 59
    Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas 1960; 20: 3746.
  • 60
    Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 15974.
  • 61
    Rosenthal R. Meta-analytic procedures for social research. London: Sage; 1991.
  • 62
    De Jong Z, Munneke M, Zwinderman AH, Kroon HM, Jansen A, Ronday KH, et al. Is a long-term high-intensity exercise program effective and safe in patients with rheumatoid arthritis? Results of a randomized controlled trial. Arthritis Rheum 2003; 48: 241524.
  • 63
    Westby MD. A health professional's guide to exercise prescription for people with arthritis: a review of aerobic fitness activities. Arthritis Rheum 2001; 45: 50111.
  • 64
    Abraham C, Michie S. A taxonomy of behavior change techniques used in interventions. Health Psychol 2008; 27: 37987.
  • 65
    Endler NS, Kocovski NL. Self-regulation and distress in clinical psychology. In: BoekaertsM, PintrichPR, ZeidnerM, editors. Handbook of self regulation. 1st ed. London: Academic Press; 2000. p. 56999.
  • 66
    Smarr KL, Parker JC, Wright GE, Stucky-Ropp RC, Buckelew SP, Hoffman RW, et al. The importance of enhancing self-efficacy in rheumatoid arthritis. Arthritis Care Res 1997; 10: 1826.
  • 67
    Fortune G, Barrowclough C, Lobban F. Illness representations in depression. Br J Clin Psychol 2004; 43: 34764.
  • 68
    Allan JL, Johnston DW, Johnston M, Mant D. Depression and perceived behavioral control are independent predictors of future activity and fitness after coronary syndrome events. J Psychosom Res 2007; 63: 5018.
  • 69
    Paschalides C, Wearden AJ, Dunkerley R, Bundy C, Davies R, Dickens CM. The associations of anxiety, depression and personal illness representations with glycaemic control and health-related quality of life in patients with type 2 diabetes mellitus. J Psychosom Res 2004; 57: 55764.
  • 70
    Dickens C, McGowan L, Clark-Carter D, Creed F. Depression in rheumatoid arthritis: a systematic review of the literature with meta-analysis. Psychosom Med 2002; 64: 5260.
  • 71
    Ziegelstein RC, Fauerbach JA, Stevens SS, Romanelli J, Richter DP, Bush DE. Patients with depression are less likely to follow recommendations to reduce cardiac risk during recovery from a myocardial infarction. Arch Intern Med 2000; 160: 181823.
  • 72
    Minor MA, Brown JD. Exercise maintenance of persons with arthritis after participation in a class experience. Health Educ Q 1993; 20: 8395.
  • 73
    Katz PP, Yelin EH. Activity loss and the onset of depressive symptoms: do some activities matter more than others? Arthritis Rheum 2001; 44: 1194202.
  • 74
    Slootmaker SM, Schuit AJ, Chinapaw MJ, Seidell JC, van Mechelen W. Disagreement in physical activity assessed by accelerometer and self-report in subgroups of age, gender, education and weight status. Int J Behav Nutr Phy 2009; 6: 17.
  • 75
    Pinals RS, Masi AT, Larsen RA, and the Subcommittee for Criteria of Remission in Rheumatoid Arthritis of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Preliminary criteria for clinical remission in rheumatoid arthritis. Arthritis Rheum 1981; 24: 130815.
  • 76
    Ropes MW, Bennett GA, Cobb S. Jacox R, Jessar RA. 1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 1958; 9: 1756.
  • 77
    Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 31524.
  • 78
    Hochberg MC, Chang RW, Dwosh I, Lindsey S, Pincus T, Wolfe F. The American College of Rheumatology 1991 revised criteria for the classification of global functional status in rheumatoid arthritis. Arthritis Rheum 1992; 35: 498502.

Supporting Information

  1. Top of page
  2. Abstract
  8. Acknowledgements
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

ACR_20251_sm_appendix.doc69KSupplementary data

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.