Purpose Cognitive problems are often reported by patients with systemic lupus erythematosus (SLE). It is known that factors associated with cognitive dysfunction, such as anxiety and depression, negatively affect HRQoL. However, the direct impact of cognitive dysfunction on HRQoL in patients with SLE is largely unknown. We therefore evaluated the prevalence of cognitive and studied its impact on HRQoL in patients with SLE.

Methods Patients with the clinical diagnosis of SLE referred with NP symptoms to the Leiden NPSLE clinic between 2007–2019 were included in this study. All patients were evaluated in a multidisciplinary way, including a 1-hour neuropsychological assessment by an experienced clinical neuropsychologist. In a consensus meeting, NP symptoms were attributed to SLE (NPSLE, either inflammatory, ischemic or a combination of both) or to other causes. Four cognitive domains were determined with tests from the cognitive assessment: global cognitive function (GCF, score 0–30), learning and memory (L&M), executive function and complex attention (EF&CA), psychomotor speed (PS) (all T-scores). HRQoL was assessed using the SF-36 and summarized with the mental and physical component scores (MCS/PCS). The associations between cognition and NPSLE phenotype and cognition and HRQoL were assessed with multiple regression analyses and linear mixed models corrected for confounding.

Results In total, 357 patients were included, of which 86% was female and the mean age was 44 years. Approximately half of patients (169/357) had a follow-up visit after a median duration of 11 months. Impairment of the cognitive domain GCF was present in 8% of patients. Impairment of the other cognitive domains was present in approximately ±50% of patients. Most severe impairment (all domains) was seen in patients with a combined NPSLE phenotype, followed by an inflammatory phenotype. A diffuse cognitive impairment (a combination of the cognitive domains L&M, EF&CA and PS) was most common and this pattern was more frequently seen in in patients with an inflammatory NPSLE phenotype.

A weak association between cognition and HRQoL was found both cross-sectionally and longitudinally in all cognitive domains. In general, one standard deviation (SD) lower scores on the cognitive domains were associated with an at most 1/5 SD lower HRQoL (see Table 1).

Conclusion In conclusion, objective cognitive impairment is common in SLE patients with NP symptoms, especially in those with an combined an inflammatory NPSLE phenotype. However, the impact of cognitive status on quality of life appears limited.