Cognitive symptoms and associated disease and non-disease related factors in patients with systemic lupus erythematosus: a longitudinal study

Abstract

BACKGROUND/PURPOSE: Cognitive impairment is commonly reported in patients with systemic lupus erythematosus (SLE). Its associated disease and non-disease related factors have been inconsistently reported. Objective: To examine cognitive symptoms and its relation to disease related factors including disease activity, antiphospholipid antibody, previous neuropsychiatric history (NPSLE) and non-disease related factors such as anxiety and depression over time. METHODS: Cognitive symptoms inventory (CSI) was used to measure perceived cognitive impairment serially at 3 time-points 12 months apart. Disease activity was measured by SLEDAI. Depressive and anxiety symptoms were measured by HADS-D and HADS-A respectively. RESULTS: 304 SLE patients were recruited at baseline (T0) among whom 144 had first re-evaluation (T1) and 34 had second re-evaluation (T2) at 12-month interval. Majority (73.5%, 25/34) of patients had stable CSI whereas 5.9% (2/34) of patients had persistently worsened CSI over 24 months. At T0, multivariate analysis revealed that higher CSI was associated with history of NPSLE (p = 0.005) and psychiatric disease (p = 0.04), higher HADS-A (p < 0.001) and HADS-D (p < 0.001) scores. CSI of active patients (SLEDAI > 6) was not different from inactive patients and did not change despite regression of disease activity in 12 months. There was no difference in CSI between T0 and T1 regardless of history of NPSLE, psychiatric history, change in depressive status at T1 (HADS-D > 11 as cutoff) but CSI was significantly different in patients who demonstrated change in anxiety status at T1 (HADS-A > 11 as cutoff) (p = 0.03). Multivariate linear regression analysis revealed change in HADS-A as the only significant predictive factor of change in CSI over time (β = 0.774, 95% CI 0.43 – 1.12, p < 0.001). CONCLUSION: 5.9% of unselected SLE patients reported persistent cognitive symptoms. Patients with history of NPSLE and psychiatric illness, high anxiety and depressive symptoms had worse CSI than those without these conditions. CSI was sensitive, but modestly, to change in anxiety symptoms over time.