Background Depressive symptoms are intuitively related to the disease activity of systemic lupus erythematosus (SLE) but studies on these relationships have yielded inconsistent and even conflicting results. Indeed, the inter-correlation between, and co-existence of depression and anxiety may potentially engender inconsistency in addressing the relationship between the severity of depression and disease activity of SLE.

Objectives We evaluated whether lupus disease activity is independently associated with the severity of depression and anxiety in lupus patients.

Methods Adult lupus patients (age≥21) were assessed for the severity of depressive and anxiety symptoms, lupus disease activity and SLE-related disease damage by using the Hospital Anxiety and Depression Scale (HADS), SLEDAI and SLICC/ACR SDI respectively. Age- and gender-matched healthy controls (HC) were recruited for comparison. Prevalence and severity of depressive and anxiety symptoms were compared between lupus patients and HC. A score of ≥8 in the HADS-anxiety and HADS-depression subscales was conventionally considered as clinically severe anxiety and depression respectively. Independent relationships between the severity of anxiety and depressive symptoms, and SLEDAI were studied with linear regression models.

Results In total, 110 lupus patients and 110 HC were studied. While age and gender were matched between the patients with SLE and HC, there were also no statistically significant differences between both groups in body mass index and smoking. In the SLE group, the mean ± SD duration of illness, daily prednisolone dose, cumulative prednisolone dose, SLEDAI and SLICC/ACR SDI were 72.31±81.1 months, 14.31±15.6mg, 12.64±14.6gm, 5.71±5.5 units and 0.44±0.8 units respectively. The mean ± SD HADS-anxiety and HADS-depression scores were significantly higher amongst the SLE patients than those of the HC (for anxiety: 6.64±3.5 versus 4.61±3.2, p<0.001; for depression: 4.18±3.7 versus 2.73±2.5, p=0.001). Significantly more lupus patients had clinically severe anxiety (40.9% versus 21.8%, p=0.002) and depressive symptoms (15.5% versus 6.4%, p=0.025) than HC. In SLE patients, univariate analyses revealed that the number of regular medications that patients were taking (β=0.231, p=0.031), current use of cyclophosphamide (β=7.335, p<0.001), SLEDAI (β=0.173, p=0.007) and HADS-anxiety score (β=0.585, p<0.001) were significantly associated with HADS-depression score, whereas SLEDAI (β=0.155, p=0.012) and HADS-depression score (β=0.533, p<0.001) were associated with the severity of anxiety. Multiple linear regression analyses revealed that SLEDAI (β=0.160, p=0.016), calcineurin inhibitor non-use (β=-1.929, p=0.041) and past cyclophosphamide non-use (β=-1.603, p=0.039) independently predicted HADS-anxiety amongst lupus patients even after adjusting for HADS-depression. Conversely, SLEDAI (β=0.014, p=0.834) lost its significant univariate correlation with HADS-depression after controlling for HADS-anxiety in the multivariate regression model.

Conclusions Anxiety is more common in lupus patients than HC, and its severity is independently associated with more active SLE regardless of the presence and absence of concomitant depression. Lupus patients, especially in those who have active disease, should be seriously considered for proper referrals for full and formal psychiatric assessment.

Disclosure of Interest None declared