s / Osteoarthritis and Cartilage 21 (2013) S63–S312 S252 Figure 1. Presentation of the mean values of the KOOS subscales for different grades of radiographic changes in the right knee joint (n1⁄4302). Normative values are age matched values from Paradowski et al. 2006. 480 IMPACT OF PREDICTABLE VERSUS UNPREDICTABLE INTERMITTENT PAIN ON SOCIAL ROLE PARTICIPATION IN SUBJECTS WITH KNEE OSTEOARTHRITIS T. Kendzerska y,z, M. Gignac y,x, I. Stanaitis z, G. Hawker y,z. yUniv. of Toronto, Toronto, ON, Canada; zWomen's Coll. Hosp., Toronto, ON, Canada; x Toronto Western Res. Inst., Toronto, ON, Canada Purpose: Focus group discussions in individuals with hip/knee osteoarthritis (OA) identified two types of OA pain a constant background pain and a less frequent, but more intense and often unpredictable, intermittent pain with the latter having the greatest impact on social role participation. These findings led to the creation of the OARSIOMERACT measure of Intermittent and Constant OA Pain (ICOAP). The current study sought to validate focus group findings with respect to the influence of OA pain predictability on participation restrictions. Methods: In an established community cohort aged 50+ years with hip/ knee OA, we assessed demographic characteristics, OA pain (ICOAP Knee) and disability (KOOS-PS), and participation restrictions. ICOAP is comprised of two subscales: a 5-item scale assesses constant pain and a 6-item scale assesses intermittent pain, or ‘pain that comes and goes’. Subscale scores are created by summing item scores and transforming to 0-100; higher scores indicate greater pain. Those with intermittent pain were asked to report the frequency with which the pain occurs ‘without warning’ (i.e., unpredictably) and ‘after a trigger’ (i.e., predictably), from 0 (never) to 4 (very often). To assess social role participation, participants were asked the degree to which they had given up or limited time spent in important roles due to their hip/knee arthritis, from 1, not at all, to 5, a great deal. Logistic regressionwas used to examine the effect of the frequent unpredictable and frequent predictable intermittent knee pain (often/very often yes/no) on participation restrictions (roles restricted quite a bit/a great deal yes/ no), controlling for age, gender, and OA severity (ICOAP subscale scores; KOOS-PS score). Specifically, we assessed for interactions between ICOAP intermittent scores and each of frequent unpredictable and predictable knee pain on social role restrictions. Results: 265 cohort participants with complete datawere included in our analyses. Theirmean agewas 65 years (SD 10) and 75%were female. 69.6% reported intermittent knee pain only, 22.5% constant knee pain only, and 7.9%bothpain types.Median (IQR) ICOAP intermittent and constant scores were 37.5 (20.8-45.8) and 0 (0), respectively. Of the 186 subjects who reported intermittent pain, 14.6% reported frequent unpredictable pain and 10.9% reported frequent predictable pain. 40% and 36.6% had limited time spent in ‘important roles’ ‘somewhat’ or ‘quite a bit’. Controlling for age, sex, ICOAPsubscale andKOOS-PS scores,we found a significant interaction between ICOAP intermittent pain severity and frequency of unpredictable pain on participation restrictions (p1⁄40.03), such that for individuals with similar levels of intermittent pain severity, the impact on social role participationwasgreater for thosewithmore frequent versus less frequent unpredictable intermittent pain. No interaction was found between intermittent pain severity and frequent predictable pain. Conclusion: Our results confirm findings from qualitative research that, controlling for other factors, social role restrictions are greatest among those with intermittent knee pain that frequently occurs warning influences. Further studies in larger cohorts, and with greater variability in pain types, is warranted to confirm our findings and, if confirmed, to elucidate potential explanations. 481 EFFICACY AND TOLERABILITY OF CELECOXIB AND NAPROXEN VS PLACEBO IN HISPANIC PATIENTS WITH KNEE OSTEOARTHRITIS M.N. Essex, R. Behar, M.A. O'Connell, P. Bhadra Brown. Pfizer Inc., New