Background/Purpose: Depression (DEP) is a common psychiatric comorbidity of rheumatoid arthritis (RA), yet little is known of its temporal effect on disease activity and symptoms.  The study aim was to evaluate how the lifetime prevalence of DEP influenced longitudinal changes in the different domains of RA disease activity.

Methods: RA patients with self-reported DEP data and ≥ 1 follow-up visit were identified from a national observational cohort of > 34,000 individuals (The Consortium of Rheumatology Researchers of North America; CORRONA).  Linear mixed models estimated the association between the lifetime prevalence of DEP and changes in disease activity over 2-years.  Outcomes were the clinical disease activity index (CDAI), tender and swollen joint counts (TJC and SJC), patient and physician global assessment (PGA and EGA), patient-reported pain, health assessment question (HAQ), C-reactive protein (CRP), and erythrocyte sedimentation rate (ESR).  To account for potential confounders, patients with and without prior DEP were matched (1:1) using a propensity score.  Propensity score models incorporated covariates selected a priori and were the following: baseline disease severity, gender, race, ethnicity, age, health insurance, marital status, employment, alcohol use, smoking, exercise, BMI, and composite comorbidity.  The matching was performed separately for each outcome measure.  Model based estimates of disease activity were generated and trends over time between the two groups tested.

Results: The rates of change by DEP status were significantly different for the CDAI, TJC, PGA, EGA, pain, and HAQ.  Patients without prior DEP had greater decreases in disease activity.  Baseline, 1-year, and 2-year CDAI estimates in those without prior DEP were: 15.7 [14.7-16.6], 11.4 [10.5-12.4], and 7.2 [6.1-8.3]; compared to patients with prior DEP: 15.8 [14.8-16.7], 12.9 [11.9-13.8], and 10.0 [8.9-11.1].  The component measures displayed similar trends.  Also, this effect was greater for the patient-reported outcomes.  Baseline, 1-year, and 2-year PGA estimates in those with no prior DEP were: 3.7 [3.5-3.8], 3.1 [2.9-3.2], and 2.5 [2.3-2.7]; and in patients with prior DEP: 3.7 [3.5-3.8], 3.6 [3.4-3.7], and 3.4 [3.2-3.6]. In contrast, baseline, 1-year, and 2-year EGA estimates in patients without prior DEP were: 2.7 [2.6-2.9], 2.0 [1.8-2.1], and 1.2 [1.0-1.4]; and in those with prior DEP: 2.7 [2.5-2.9], 2.1 [1.9-2.3], and 1.5 [1.3-1.7].  Similar results were obtained for the TJC, a measure assessed via the patient and physician, when compared to the SJC, which is evaluated solely by clinicians.            

Conclusion: The results suggest that the presence of prior DEP in RA impacts prospective changes in metrics reported by the patient: pain, functional status, and global assessment; and to some extent measures reported by providers, but not the number of swollen joints or acute phase reactants.