Patients with early inflammatory arthritis who fulfil the 2010 American College of Rheumatology–European League Against Rheumatism classification criteria for rheumatoid arthritis have increased mortality compared with those who do not: results from the Norfolk Arthritis Register

Abstract

BackgroundMortality is increased in rheumatoid arthritis compared with the general population. Most studies have used the 1987 American College Rheumatology criteria to define rheumatoid arthritis when investigating mortality. The aims of this study were to examine whether, in a cohort of patients with early inflammatory polyarthritis, the 2010 American College of Rheumatology–European League Against Rheumatism classification criteria for rheumatoid arthritis identify those with decreased survival; and if so to identify which components of the criteria are the strongest predictors of mortality. MethodsAdults with two or more swollen joints for 4 or more weeks were recruited to the Norfolk Arthritis Register (NOAR) between 1990 and 2009. Patients included in this analysis had symptom duration of at least 2 years and had not received any disease modifying therapy at initial assessment. Data on the components of the 2010 criteria, along with demographic details, were obtained at baseline-visit through a nurse-administered questionnaire and joint examination. Bloods samples were taken for C-reactive protein (CRP), rheumatoid factor (RF), and anti-citrullinated protein antibody (ACPA) estimation. All patients registered with NOAR are flagged with the UK Office for National Statistics (ONS) which provides mortality data. Survival analyses were performed with Cox proportional hazards models in univariate analyses, then adjusted for age and sex. A multivariable model was then developed including all components of the 2010 criteria as well as baseline smoking status, age, and gender. Findings1671 patients had complete data for analysis, with 20 488 person-years of follow-up. 1092 (65%) patients were female and there were 471 deaths reported by the ONS by Dec 31, 2011. 905 (54%) patients fulfilled the 2010 criteria at baseline, and they had a significantly increased risk of death compared with patients in NOAR who did not fulfil the 2010 criteria, both in univariate analyses and in the age-adjusted and sex-adjusted model (adjusted hazard ratio (HR) 1·39 [95% CI 1·15–1·68]). The multivariable model identified high titre RF, ACPA (or both) positivity (more than three times the upper limit of normal) and raised CRP as significant predictors of mortality (HR 1·64 [95% CI 1·31–1·97] and 1·25 [95% CI 1·02–1·52], respectively). InterpretationPatients presenting with early inflammatory polyarthritis who fulfil the 2010 criteria have significantly increased mortality compared with those who do not. The components of the 2010 criteria that seem to be important predictors of mortality are high titre RF or ACPA positivity, and abnormal CRP at baseline. FundingArthritis Research UK.