GP or PN were calculated for those with multisite joint pain compared with those with single site joint pain. Cox-regression was used to obtain RR adjusting for age, gender, BMI and social deprivation. Results: Of 15 083 (53%) respondents, average age was 63.9 years (11.2sd) and 54% were female. There were 11 928 participants with joint pain, of which 4677 (39%) reported consulting a GP and 888 (7.4%) a PN. The crude RR (95% CI) for consulting with multisite vs single site joint pain was 1.89 (1.77, 2.00) and 2.61 (2.19, 3.17) for GP and PN consultations respectively. When adjusted for potential confounders the RR reduced to 1.76 (1.63, 1.89) and 2.45 (2.01, 2.97). The AR of consulting with multisite compared with single site joint pain was 21.6% (19.7%, 23.2%) and 5.5% (4.6%, 6.3%) for GP and PN consultations respectively. Conclusion: This study highlights the increased prevalence of consultations with a GP or PN attributable to multisite peripheral joint pain. The findings indicate that if all joint pain in the study population had been restricted to a single site there would have been 21.6% fewer people consulting their GP and 5.5% fewer people consulting a PN with peripheral joint pain in primary care in the last 12 months. The management of peripheral joint pain should consider strategies for the prevention or delay of the onset of multisite joint pain, such as developing the skills of GPs and PNs, which could in turn potentially reduce the burden of consultations for older adults with peripheral joint pain. Disclosure statement: The authors have declared no conflicts of interest. 41. LEVELS OF PHYSICAL ACTIVITY IN OLDER ADULTS WITH OR WITHOUT SELF-REPORTED JOINT PAIN: A CROSS-SECTIONAL SURVEY Robert Smith 1 , Emma Healey 1 , Gretl McHugh 2 , Ebenezer Afolabi 1 and Krysia Dziedzic 1 1 Research Institute for Primary Care & Health Sciences, Keele University, Staffordshire, 2 School of Nursing, Midwifery & Social Work, The University of Manchester, Manchester, UK Background: The National Institute for Health and Care Excellence (NICE) recommend that physical activity (PA) should be a core treatment for all adults with OA. Joint pain is one of the main symptoms experienced by individuals with OA and yet in the UK the levels of PA in adults with joint pain are unknown. The objectives were to: (i) describe the levels of PA in older adults with and without joint pain, and (ii) describe the uptake of NICE recommended PA for joint pain and OA. Methods: Cross-sectional analyses of a population survey mailed to 28,443 community dwelling older adults aged 45 years and over were conducted as part of the MOSAICS study. Participants were asked to report if they had joint pain in four specific sites (foot, knee, hip or hand) over the last 12 months. Levels of PA were collected using the Short Telephone Activity Rating (STAR) questionnaire. Participants were categorized as physically inactive, somewhat active or regularly active. Participants were divided into two mutually exclusive groups; those with and without self-reported joint pain. In adults with joint pain, questions on using PA approaches as a treatment for their joint pain were assessed. Results: Of 14 212 responders to the STAR questionnaire, the mean age was 63.6 (11.1 S.D.) years, mean BMI was 26.9 (4.7 S.D.) kg/m 2 , 54.2% were female and 11 310 (79.6%) participants reported joint pain. Table 1 displays the levels of PA. Participants with self-reported joint pain were less likely to be physically active compared with participants with no reported joint pain (OR ¼ 0.75, 95% CI 0.68, 0.77). In participants with joint pain, only 3667 (32.4%) reported trying PA approaches as a treatment for their joint pain in the last 12 months. However, only 868 (7.7%) participants reported having received a prescription of PA, in line with NICE guidelines, from the NHS. Conclusion: Levels of PA appear lower in older adults with joint pain compared with those without. Uptake of PA as a treatment for joint pain was also found to be low, particularly within NHS, despite NICE recommending PA approaches for joint pain and OA. Disclosure statement: The authors have declared no conflicts of interest. 42. PREDICTION OF FUTURE PAIN BY DAS28-P IN PATIENTS WITH EARLY RHEUMATOID ARTHRITIS: THE ERAN COHORT Luke Harries 1 , Daniel F. McWilliams 1 , Adam Young 2 , Patrick D. W. Kiely 3 and David A. Walsh 1 1 Arthritis UK Pain Centre, Academic Rheumatology, University of Nottingham, Nottingham, 2 Department of Rheumatology, West Hertfordshire Hospitals NHS Trust, St Albans, 3 Department of Rheumatology, St Georges Healthcare NHS Trust, London, UK Background: Pain in RA remains a problem and priority for patients and physicians alike. Pain may be associated with different factors, such as mood, co-morbidities, gender and inflammation. However, despite many people with RA fulfilling FM criteria, the role of augmented pain processing is poorly understood and infrequently measured. DAS28-P is a derived index which represents the patient- reported proportion of disease activity and may reflect central pain processing in RA. This project examined characteristics associated with pain progression during early RA. We examined characteristics changing during the 3 year follow-up and their associations with pain. Methods: Data were drawn from the Early Rheumatoid Arthritis Network (ERAN), an inception cohort of 1236 early RA patients from UK and Eire. Pain levels (SF36-Bodily Pain) were examined over the first 3 years after diagnosis. Generalized Estimating Equation (GEE) analyses, adjusting for confounders, were used to examine the associations of pain with demographic and clinical characteristics at baseline and each follow-up visit. Results: Pain improved from baseline to 1 year (median (IQR) 41(22– 62) and 51 (31–72) respectively) and then remained constant after- wards. DAS28-P at baseline had median (IQR) of 0.42 (0.35–0.51) and did not change substantially during the 3 year follow-up. Initial GEE analysis showed that high DAS28P was consistently associated with worse pain throughout the follow-up (Table 1). Additional GEE analysis found that high DAS28-P significantly predicted the next year’s pain (Table 1). Conclusion: DAS28P is associated with RA pain at presentation and throughout follow-up; and it also predicts pain for the next 1 year. Disability, fatigue, current pain may also predict future pain better than some well-established measures of RA severity/prognosis. Funding: This work was supported by Pfizer (I-CRP grant WS2307457 to D.F.M. and D.A.W.). 42 TABLE 1 Variables associated with pain and pain predictors across 3 years of early RA Factor GEE analyses from baseline to year 3 Current pain Pain at next year B (95% CI) P B (95% CI) P Age 0.17 (0.09, 0.26) <0.001 0.07 (–0.01, 0.14) 0.075 Gender (female) –5.64 (–8.18, –3.10) <0.001 0.10 (–0.06, 0.26) 0.211 Smoking status 2.46 (0.16, 4.76) 0.036 0.09 (–0.04, 0.23) 0.174 ACR 1987 classification 1.56 (–0.80, 3.92) 0.196 0.05 (–0.09, 0.19) 0.479 DAS28 –3.41 (–4.29, –2.52) <0.001 0.06 (–0.05, 0.17) 0.253 HAQ –13.30 (–14.93, –11.68) <0.001 –0.28 (–0.37, –0.19) <0.001 SF36- Mental Health –0.01 (–0.07, 0.06) 0.852 –0.02 (–0.10, 0.06) 0.701 SF36- Vitality 0.03 (–0.03, 0.09) 0.284 0.11 (0.02, 0.19) 0.015 DAS28-P –30.18 (–39.46, –20.90) <0.001 –0.11 (–0.19, –0.02) 0.015 Current pain (SF36-Bodily Pain) Not applicable 0.33 (0.23, 0.44) <0.001 GEE analysis examining the association between DAS28-P and pain (current or next year’s pain) during 3 years of follow-up, with adjustment for identified confounders. GEE analysis examined longitudinal changes during follow-up and also cross-sectional associations at each time point. GEE: generalized estimating equation. i71 Downloaded from https://academic.oup.com/rheumatology/article-abstract/53/suppl_1/i71/1795216 by guest on 08 June 2020