Prevalence, involved domains, and predictor of cognitive dysfunction in systemic lupus erythematosus

Abstract

Cognitive Dysfunction (CD) can occur in Systemic Lupus Erythematosus (SLE) before the occurrence of Neuropsychiatric Lupus Erythematosus (NPSLE). Given the reversibility and fluctuation of SLE-related CD, the research for possible predictors is of great significance for early detection and intervention. We sought to determine the prevalence, involved domains, and possible predictors of CD in SLE patients. We conducted a retrospective cross-sectional study at Nanfang Hospital from 2018 to 2019. A total of 78 SLE patients were recruited. The Montreal Cognitive Assessment (MoCA) scale was used to screen cognitive function. Demographic, clinical, and laboratory characteristics were collected. The serum anti-methyl-d-aspartate receptor (anti-NMDAR) antibody and S100β were measured by enzyme-linked immunosorbent assay (ELISA). Multivariate logistic regression analysis and ROC curve were used to assess the predictor of SLE-related CD. Of 78 recruited patients,53 (67.9%) had CD. It mainly involved delayed recall, abstract generalization, verbal repetition, and fluency. The disease activity index (SLEDAI) was not associated with SLE-related CD (p > 0.05). Multivariate logistic regression showed that an increase in each year of education there was a decrease in the likelihood of CD (OR 0.261, CI 0.080-0.857, p = 0.027) whereas with each unit increase in serum anti-NMDAR antibody there was an increased likelihood of SLE-related CD (OR 1.568, CI 1.073-2.292, p = 0.020). The prevalence of SLE-related CD was 67.9% in our study and SLE-related CD was not associated with disease activity. Serum anti-NMDAR antibody can be used as a predictor for SLE-related CD.