The effect of anti-inflammatory treatment on depressive symptoms in spondyloarthritis: does the type of drug matter?

Abstract

To investigate the effect of pharmacological treatment of SpA on depressive symptoms and explore whether this effect differs between drug classes. Data from the observational Assessment of SpondyloArthritis international Society Health Index Validation Study were used. Patients were assessed at baseline and after initiation of NSAIDs/conventional synthetic DMARDs (csDMARDs)/TNF inhibitors (TNFis). Depressive symptoms were assessed with the Hospital Anxiety and Depression Scale depression subscale [HADS-D; 0-21 (best-worst)]. Covariables included demographics and disease characteristics, including disease activity [Ankylosing Spondylitis Disease Activity Score (ASDAS)/BASDAI]. The change in HADS-D from baseline was compared between treatments (NSAIDs/csDMARDs/TNFis) with analysis of variance and multivariable regression analysis. A total of 304 patients were included; 102/45/157 initiated NSAIDs/csDMARDs/TNFis and 260 (85%) / 44 (15%) had axial/peripheral SpA. At baseline, the mean HADS-D was 6.9 (s.d. 4.2); 126 (42%) were possibly depressed (HADS-D ≥8) and 66 (22%) were probably depressed (HADS-D ≥11). At follow-up, depressive symptoms significantly improved in all treatment groups. In multivariable regression without disease activity measures, initiating TNFis compared with NSAIDs was associated with greater improvement in depressive symptoms [β = -1.27 (95% CI -2.23, -0.32)] and lower odds of possible depression at follow-up [odds ratio 0.47 (95% CI 0.23, 0.94)]. This association was attenuated after additional adjustment for disease activity (ASDAS/BASDAI) but not CRP. csDMARDs did not differ from NSAIDs regarding their effect on HADS-D. Between-drug class results were confirmed in axial SpA (axSpA), although less clear in peripheral SpA. Treatment of active SpA also improves depressive symptoms. Especially in axSpA, TNFis have a greater effect than NSAIDs, which is mainly explained by a stronger effect on disease activity. We found no evidence for a direct link between CRP-mediated inflammation and depressive symptoms in SpA.