Identifying different cognitive phenotypes and their relationship with disability in neuromyelitis optica spectrum disorder

BACKGROUND: The existence, frequency, and features of cognitive impairment (CI) in patients with neuromyelitis optica spectrum disorder (NMOSD) are still debated. A precise classification and characterization of cognitive phenotypes in patients with NMOSD are lacking. METHODS: A total of 66 patients with NMOSD and 22 healthy controls (HCs) underwent a neuropsychological assessment. Latent profile analysis (LPA) on cognitive test z scores was used to identify cognitive phenotypes, and ANOVA was used to define the clinical features of each phenotype. Univariate and multivariate analyses were used to explore the predictors of severe CI, and a corresponding nomogram was created to visualize the predictive model. RESULTS: LPA results suggested four distinct meaningful cognitive phenotypes in NMOSD: preserved cognition (n = 20, 30.3%), mild-attention (n = 21, 31.8%), mild-multidomain (n = 18, 27.3%), and severe-multidomain (n = 7, 10.6%). Patients with the last three phenotypes were perceived to have CI, which accounts for 67.6% of patients with NMOSD. Patients with NMOSD and worse cognitive function were older (p < 0.001) and had lower educational levels (p < 0.001), later clinical onset (p = 0.01), worse Expanded Disability Status Scale scores (p = 0.001), and poorer lower-limb motor function (Timed 25-Foot Walk, p = 0.029; 12-item Multiple Sclerosis Walking Scale [MSWS-12], p < 0.001). Deterioration of Nine-Hole Peg Test (odds ratio, OR: 1.115 [1, 1.243], p = 0.05) and MSWS-12 (OR: 1.069 [1.003, 1.139], p = 0.04) were the independent risk factors for severe cognitive dysfunction. Finally, a nomogram was built based on the entire cohort and the above factors to serve as a useful tool for clinicians to evaluate the risk of severe cognitive dysfunction. CONCLUSIONS: We introduced a classification scheme for CI and highlighted that the deterioration of upper- and lower-limb motor disability potentially predicts cognitive phenotypes in NMOSD.