Health Assessment Questionnaire at One Year Predicts All‐Cause Mortality in Patients with Early Rheumatoid Arthritis

OBJECTIVES Higher self-reported disability (high HAQ) has been associated with hospitalizations and mortality in established RA but associations in early RA (ERA) are unknown. METHODS ERA patients (symptoms <1 year) enrolled in the Canadian Early Arthritis Cohort (CATCH) who initiated DMARDs and had completed HAQ data at baseline and 1 year follow up were eligible. Discrete-time proportional hazards models were used to estimate crude and multi-adjusted associations between HAQ at baseline and 1 year, respectively, with all-cause mortality in each year of follow up. RESULTS 1724 patients with ERA; mean age was 55 years and 72% female were included. Over 10 years, 62 deaths (2.4%) were recorded. Deceased patients had higher HAQ scores at baseline 1.2 (SD 0.7) and at 1 year 0.9 (SD 0.7) vs non-deceased 1.0 (SD 0.7) and 0.5 (SD 0.6), p<0.001. DAS28 was higher in deceased vs non-deceased at baseline [5.4 (SD 1.3) vs 4.9 (SD 1.4)] and at 1 year [3.6 (SD 1.4) vs. 2.8 (SD 1.4)], p<0.001. Age, male sex, lower education, smoking, more comorbidities, higher baseline disease activity score (DAS) and glucocorticoid use were associated with mortality. Contrary to HAQ at baseline, the association between all-cause mortality and HAQ at 1 year remained significant even after adjusting for confounders. HAQ baseline unadjusted hazard OR was 1.46 (CI 1.02-2.09) and adjusted 1.25 (CI 0.81-1.94) vs. HAQ at 1-year unadjusted hazard OR was 2.58 (CI 1.78-3.72) and adjusted 1.75 (CI 1.10-2.77). CONCLUSION Higher HAQ and also DAS at 1 year was significantly associated with all-cause mortality in a large ERA cohort.