OP0067 MORE THAN SIX-FOLD INCREASED MORTALITY RISK IN PATIENTS WITH INCIDENT RHEUMATOID ARTHRITIS AND DEPRESSION IN A LARGE COHORT WITH 10-YEAR FOLLOW-UP

Abstract

BackgroundThe occurrence of depression is increased in patients with rheumatoid arthritis (RA) compared with the background population. (1) Recently, we described that in RA the most frequent indication for filling antidepressant prescriptions is depression and the frequency of filling coincides with the occurrence of depression reported in the scientific literature. (2) In patients with stroke, depression defined as filling of antidepressants or a diagnosis with depression is associated with increased mortality risk and the risk is similar for both definitions of depression. (3)ObjectivesWe used the first filling of antidepressants as proxy for depression with the objective to describe the mortality risk associated with depression in patients with incident RA.MethodsWe included patients diagnosed with incident RA (index date) from the nationwide DANBIO register (4) from January 1, 2008 to September 30, 2018. Participants were identified by unique personal registration numbers. Included patients were without a recorded filling of methotrexate (Anatomic Therapeutic Chemical code L01BA01) and antidepressants (N06A) in the Danish National Prescription Register or recorded hospital contacts with RA and depression (International Classification of Diseases (version 10) codes M05, M06, F32) in the Danish National Patient Register, three years prior to the index date. From the index date, we defined depression as first filling of antidepressants and collected death dates from the Danish Civil Registration System. The participants were followed until December 31, 2018 and all-cause mortality estimated in two dynamic risk periods: the period from the index date until first filling of antidepressants (if it occurred) and the period after filling of antidepressants. We calculated hazard rate ratios (HRR) by modelling filling of antidepressants as time-varying exposure for total follow-up and adjusted for potential confounders defined a priori: age, sex, comorbidity, cohabitation, employment status, highest attained education, and income. Cumulative mortality was described by Kaplan-Meier curves. Results were reported with 95% confidence intervals (CI).ResultsAmong 11,071 RA patients followed for 56,993 person-years, 1,095 (10%) filled prescriptions for antidepressants. The median age at diagnosis was 61 years, 66% were female, and 64% diagnosed with seropositive RA. Adjusted HRR was highest in the age group <55 years but also increased between 55-70 years, >70 years, among females and males, and in patients diagnosed with seropositive and seronegative RA (Table 1). The cumulative mortality is seen in Figure 1.Table 1.StrataHRR (95% CI)CrudeAdjustedAge, years<558.40 (4.20-16.80)6.66 (2.80-15.85)55-703.27 (2.35-4.54)3.30 (2.27-4.80)>702.97 (2.36-3.75)2.94 (2.26-3.83)SexFemale3.72 (2.95-4.70)2.91 (2.22-3.81)Male3.10 (2.32-4.15)3.70 (2.66-5.14)RA diagnosisSeropostive (M05)3.73 (2.99-4.65)3.45 (2.66-4.47)Seronegative (M06)2.85 (2.07-3.91)3.08 (2.17-4.37)Figure 1.ConclusionDepression, defined as first filling of antidepressants, was associated with more than six-fold increased mortality risk in patients with incident RA.