**3.6 Safety of PRT for RA patients**

For many years, intensive weight-bearing exercise was considered inappropriate for RA patients due to concern that this unaccustomed stress on the joints would exacerbate inflammation, pain, and joint damage (e.g. Sutej & Hadler, 1991). Even today, many rheumatologists and their multidisciplinary teams retain these anachronistic beliefs and advise patients to avoid strenuous physical pursuits in order to protect their joints and conserve their energy (i.e. the strategy of "pacing") (for further discussion on this see Metsios et al., 2007; Munneke et al., 2004). This is despite the unanimity of research findings that exercise training, including resistance training (Table 1), irrespective of the intensity employed, is safe in RA patients. In fact, although most studies report no changes in disease activity following resistance training, findings of improvements are not uncommon; e.g. reductions in: erythrocyte sedimentation rate (ESR; Hakkinen et al., 1994, 1997, 1999), morning stiffness (Ekdahl et al., 1990), number of tender and swollen joints (Ritchie articular index; Ekdahl et al., 1990; Hakkinen et al., 1994, 1997; van den Ende et al., 1996), self-reported joint count (Komatireddy et al., 1997), pain (Komatireddy et al., 1997; McMeeken et al., 1999; Rall et al., 1996b), and Disease Activity Score (DAS28, DAS4; Hakkinen et al., 1999, 2001, 2004a). High-intensity exercise even appears to be safe in patients with active disease; van den Ende et al. (2000) randomly allocated RA patients admitted to hospital for RA flares to perform either HI exercise (isokinetic and isometric strength training) or LI exercise (ROM and isometric exercises). After 24 weeks of training (3x's/week), improvements in DAS were observed for both groups with a trend toward greater improvement in the HI patients.

Adherence to PRT over prolonged periods also provides no cause for concern. Hakkinen et al. (2001) in an RCT comparing 2 years of strength training to conventional physiotherapy (ROM exercises), found that although DAS28 improved significantly for both groups, the strength training group enjoyed greater benefit. Similarly, de Jong et al. (2003) in their 2 year RCT (the RAPIT trial) also identified reductions in disease activity (DAS4) in their HI exercise (including strength training) group; albeit, this time with no difference between the exercise and control ("usual care") groups.

In a broader investigation of immune responses to PRT in RA patients, Rall et al. (1996c) detected no effects of 12 weeks HI training on peripheral blood mononuclear (PBMC) subpopulations, or stimulated proliferation of TNF-α, interleukin (IL)-1β, IL-2, IL-6, or prostaglandin E2, or delayed type hypersensitivity skin response.

Although reassuring effects on joint counts, systemic inflammation, pain, and more generalised disease activity are provided by studies of strength training interventions in RA patients, relatively few studies have assessed the effects of training on radiographic joint damage. An exception to this was the RAPIT trial. Initially, reports from this investigation (de Jong et al., 2003; Munneke et al., 2005) raised concerns by suggesting that high intensity exercise exacerbated joint damage progression in large joints with extensive pre-existing damage. Results from an 18 month follow-up study (de Jong et al., 2009), however, have seen the investigators retract this conclusion. Instead, they are now confident that longterm, intense weight-bearing exercise does not cause further damage to large joints, even those already extensively damaged. This revised interpretation thus accords with the verdict they had previously made with regard to the small joints of the hands and feet (de Jong et al., 2003). This general conclusion of training not increasing radiological progression of joint damage agrees with the findings of others (Hakkinen et al., 1994, 2001, 2004b; Nordemar et al., 1981). In the earliest of these studies, Nordemar et al. (1981) found that RA patients who had performed 4-8 years of resistance exercises for the legs had reduced joint damage in these limbs relative to non-exercising disease-matched controls. Whilst in the other studies, all by Hakkinen's group (1994, 2001, 2004b), no acceleration in joint damage was detected by x-ray in RA patients performing long-term (up to 5 years; Hakkinen et al., 2004b), regular, HI PRT relative to patients receiving standard care.
