Background Co-morbid depression and anxiety occur in the context of rheumatoid arthritis (RA). Their characteristics, including associations with RA features, have not been examined previously in an early RA inception cohort with longitudinal follow up data.

Objectives To examine the frequency of anxiety and depression in patients with early RA, over time and to explore associations with epidemiological, socioeconomic and disease-related features.

Methods The Scottish Early Rheumatoid Arthritis (SERA) inception cohort recruited newly diagnosed RA patients (fulfilling ACR-EULAR 2010 criteria) followed-up thereafter every 6 months. Pre-specified clinical, laboratory and psychosocial features, including anxiety and depression scores (measured by the hospital anxiety and depression scale; score range: 0–21 for each one), were recorded at baseline and at follow-up. Non-parameteric tests and logistic regression models were used to examine the nature and magnitude of associations.

Results Data from 848 RA patients were available. Frequency of anxiety and depression at baseline was 19.0% and 12.2% with a reduction at month 12% to 13.4% and 8.2% (p=0.004 and p=0.01, respectively). Anxiety and depression scores were correlated with DAS28 at all time points, including baseline (all p<0.0001). Change in DAS28 (final-baseline) was correlated with change in depression and anxiety scores at month 6 (p<0.0001, r=0.265 and p=0.001, r=0.230) and 12 (p<0.0001, r=0.288 and p<0.0001, r=0.217). In univariate analyses, anxiety and depression scores were associated with various features, at different time points (table 1). CRP was highly associated with depression but not anxiety scores at all time points, with change in CRP correlating with change in depression scores (month6; p<0.0001, r=0.185 and month12; p<0.0001, r=0.302). Multivariable analysis indicated that anxiety score at baseline was associated with female gender (p=0.01, Beta=0.133), younger age (p=0.007, Beta=−0.181) and patient global assessment score (PGA) (p<0.0001, Beta=0.201), and at month 6 and 12 with low BMI (month 6, p=0.024, Beta=−0.091; month 12, p=0.002, Beta=−0.139), PGA (month 6, p<0.0001, Beta=0.222; month 12, p<0.0001, Beta=0.248) and baseline anxiety scores (month 6, p<0.0001, Beta=0.623; month 12, p<0.0001, Beta=0.586). For depression scores, multivariable analysis indicated association at baseline with PGA (p<0.0001, Beta=0.286) and at month 6 and 12 with PGA (month 6, p<0.0001, Beta=0.306; month 12, p<0.0001, Beta=0.320), CRP levels (month 6, p=0.006, Beta=0.150; month 12, p=0.002, Beta=0.171) and baseline depression (month 6, p<0.0001, Beta=0.422; month 12, p<0.0001, Beta=0.356) and anxiety scores (month 6, p<0.0001, Beta=0.189; month 12, p=0.008, Beta=0.170).

Conclusions Depression and anxiety are significant comorbidities at the time of RA diagnosis. While there are also associations with socioeconomic and other variables, the close relationship between CRP and depression provides further support to the already compelling data linking inflammation and depression. Changes in the anxiety and depression scores, in tandem with disease activity over time, requires further investigation.

Disclosure of Interest None declared