Background Exercise reduces depressive and anxiety symptoms among the general population, and those with a chronic illness. Elevated depressive and anxiety symptoms are prevalent co-morbidities in Rheumatoid Arthritis (RA) therefore, addressing same through exercise may have an important impact on their health related quality of life (HQoL). Evidence does support the effect of exercise on these outcomes however, quantitative synthesis of evidence from randomised controlled trials (RCTs) of exercise effects, on these critically important symptoms in RA, has yet to been conducted.

Objectives To quantify the overall population effect of exercise on depressive and anxiety symptoms, fatigue, and pain, derived from available RCT’s, and to explore the extent to which participant and trial characteristics moderated the mean effect.

Methods Articles published before September 2017 were located by two independent reviewers using Google Scholar, PsycINFO, PubMed, and Web of Science. Trials included both randomization to exercise and non-exercise control using validated measures of depression and anxiety, assessed at baseline and post-intervention. Hedges’d effect sizes (95% CI) were computed and random effects models were used for all analyses. Sources of bias were also assessed independently by two reviewers using the Cochrane bias assessment tool for RCTs and Newcastle–Ottawa Quality Assessment Scale for non-RCTs.

Results Seventeen studies were included, with 1214 participants, of which 12 RCT’s contributed to the meta-regression analysis. Participants were aged 49±9 years and 83%±14% female. Exercise training consisted on average of 3±1 weekly sessions, 60±17 min per session, and 11±5 weeks in duration. Interventions were diverse with a mix of aerobic and/or resistance training including, 4 different types of Yoga, 2 dance based and 1 Tai-chi. Mean reported adherence was 87%±11%. For depression, 18 of 20 effects (90%) were >0. The mean effect size Δ was 0.20 (0.10–0.31; p<0.001). For anxiety, seven of seven effects (100%) were >0. The mean effect size Δ was 0.50 (0.27–0.74; p<0.001). Seven of 16 effects reduced pain by a mean effect delta of 0.04 (95% CI: −0.14 to 0.21; z=0.41; p>0.69). Six of 11 effects (54.5%) did not reduce fatigue with the mean effect delta being −0.01 (95% CI: −0.20 to 0.19; z=-0.09; p>0.93). Depressive or anxiety symptoms were not the primary outcome in any of the included trials.

Conclusions Pharmacologic interventions have improved the management of RA however, research indicates that exercise remains an important part of the overall treatment. This quantitative synthesis of evidence from RCTs of the effects of exercise shows significant small-to-moderate reductions in depressive and anxiety symptoms. It has been reported that the degree of depression and anxiety in people with RA is a preceding sign of physical disability that may appear later in life therefore, aiming to target both through exercise may help to improve HQoL. Future trials should focus on depression and/or anxiety as the primary outcome. Exercise prescription is a core skill for physiotherapists therefore, they should be confident in prescribing exercise to people with RA, who have depression and anxiety, as it significantly reduces their symptoms.

Disclosure of Interest None declared