Neuromyelitis optica spectrum disorder (NMOSD) is a rare inflammatory demyelinating disease of the central nervous system (CNS) that primarily affects the optic nerves and spinal cord, often resulting in substantial and frequently permanent neurological deficits and disabilities, including blindness and paralysis (Wingerchuk and Lucchinetti, 2022). NMOSD has been reported worldwide with a prevalence of 0.7 to 10.0 per 100,000 (Wingerchuk and Lucchinetti, 2022). East Asians appear to have a higher prevalence of NMOSD (around 3.5 per 100,000) as compared to the Whites and other Asian racial groups, and the prevalence among the Chinese population is 3.31 per 100,000 (Hor et al., 2020). Pathogenic aquaporin-4 immunoglobulin G (AQP4-IgG) is identified in the serum of over 80% of NMOSD patients (Lucchinetti et al., 2014). AQP4-IgG in peripheral serum crosses the impaired blood–brain barrier (BBB) into the CNS, binds to AQP4 on astrocytic end-feet, and triggers the classical complement cascade ultimately causing oligodendrocyte damage and demyelinating injury (Jarius et al., 2020). Seronegative NMOSD represents a heterogeneous group of disorders with mechanisms that remain poorly understood, and is recognized as a relatively benign disease course (Yeo et al., 2019; Dauby et al., 2022). More than 90% of patients experience relapses, and each relapse can lead to progressive accumulation of disability (Wingerchuk and Lucchinetti, 2022). Among untreated AQP4-IgG positive patients, nearly one-quarter require gait assistance within 5 years of disease onset, more than 40% experience blindness in at least one eye, and the mortality rate approaches 10%, posing a significant disease burden on both patients and their families (Kitley et al., 2012; Ajmera et al., 2018). Therefore, early assessment of disability status and relapse risk at the time of NMOSD diagnosis is crucial for long-term disease management and evaluation of pharmacological treatment outcomes.

Lipids are fundamental components of cellular membranes and play crucial roles in energy supply, signal transduction, and other physiological functions. In the CNS, the myelin sheath is a lipid-rich multilamellar structure formed by oligodendrocytes (Stadelmann et al., 2019). Tian et al. (2020) conducted the first nationwide survey covering all 1,665 tertiary hospitals and revealed that approximately 12.8–13.9% of Chinese NMOSD patients are comorbid with hyperlipidemia. An animal study revealed that genetic deletion of ApoA-I in mice resulted in lymphocyte activation and autoantibody production, suggesting ApoA-I maybe a critical regulator of immune homeostasis (Wilhelm et al., 2009). A research suggested that dyslipidemia with low HDL-C is associated with disease activity of NMOSD (Cho et al., 2020). HDL modulates immune homeostasis through fine-tuning cholesterol bioavailability in lipid rafts, thereby influencing the activation states of B cells and T cells (Norata et al., 2012). Given the correlation between lipid profiles and autoimmune diseases, as well as the link between lipid metabolism and immune homeostasis, we hypothesized that lipid profiles may correlate with neurological impairment and relapse in NMOSD. Therefore, the present study aims to investigate the associations of lipid profiles with disability and relapse, and to evaluate the potential predictive value of lipid parameters for neurological deficit and prognosis of AQP4-IgG positive NMOSD.

In this retrospective study, we investigated correlations between lipid parameters and neurological deficits as well as relapses. The results found that lower ApoA-I and HDL-C levels were associated with greater neurological impairment and higher relapse risk of NMOSD. Therefore, ApoA-I and HDL-C may serve as potential biomarkers for acute-phase disability assessment and prognosis evaluation.

Emerging evidence has been reported to evaluate the degree of neurological impairment and relapse in NMOSD. Although serum GFAP has been proposed as a reliable indicator of disease severity and activity in AQP4-IgG positive patients (Schindler et al., 2021; Kim et al., 2022), its clinical application remains limited due to low specificity and the burden of high cost. Thus, more practical and economic biomarkers in disease assessment of NMOSD are needed. Latest study found that high ApoA-I and HDLwere associated with relapse and attack of NMOSD (Li et al., 2025). The divergence between studies likely arised from distinct lipid measures: prior work analyzed lipoprotein concentrations (HDL/LDL), while our study specifically evaluated cholesterol carried by these lipoproteins (HDL-C/LDL-C), which may represent different pathophysiological aspects. Furthermore, proteomic analysis in this study revealed that the differentially expressed proteins (DEP)were associated with inflammatory response and lipid metabolism (Li et al., 2025). This result was consistent with the correlations observed between ApoA-I, HDL-C, and CRP in our study. Ding et al. demonstrated a positive association between elevated LDL-C and relapse risk (Ding et al., 2024). The study findings also indicated that cholesterol metabolism was significantly associated with relapse in NMOSD. However, this study found an association between elevated LDL-C and increased relapse risk, contrasting with our results which indicated reduced HDL-C was linked to higher relapse risk. This discrepancy may arise from differences in the enrolled populations: this study cohort was recruited from Southern China, whereas our participants were from Northern China, suggesting potential regional variations. Prior studies suggested that environmental factors may contribute to the pathogenesis of NMOSD, although conclusive evidence remains elusive (Wingerchuk and Lucchinetti, 2022). Previous studies also demonstrated associations between the lipid profile with disease activity and prognosis in NMOSD, with HDL-C being significantly lower during attacks than in remission, and triglycerides showing positive correlations with neurological disability severity and poor prognosis (Wu et al., 2019; Cho et al., 2020). ApoA-I was associated with neuroaxonal injury and cerebral perfusion of multiple sclerosis (MS) (Jakimovski et al., 2020; McComb et al., 2020). HDL-C exerts protective effects against BBB injury and also associated with cerebral perfusion of MS (Fellows et al., 2015; Jakimovski et al., 2020). The above studys revealed that dyslipidemia is closely related to the inflammation demyelinat on of CNS. Therefore, ApoA-I and HDL-C may be protective against the severity of neurological disability and relapse in NMOSD. Furthermore, we found that ApoA-I and HDL-C exhibit an inverse correlation with CRP via univariable linear regression. Given that this was an exploratory study rather than an in-depth investigation of mechanism, we did not adjust for potential confounders, and this finding is consistent with Li et al.’s proteomic discovery that DEP were enriched in chronic inflammatory response pathways (Li et al., 2025).

The mechanisms underlying the association between lipid metabolism and autoimmune diseases remain to be fully elucidated. Lipids, as fundamental structural components of cell membranes, not only establish selective barriers but also play pivotal roles in cellular signaling—either directly as signaling molecules or indirectly by modulating membrane fluidity (Mutlu et al., 2021). Dysregulated lipid metabolism can further aggravate autoimmune inflammation in addition to driving the development of atherosclerosis. Extracellular lipids not only provide metabolic energy and nutrient sources for B cells but also serve as key signaling mediators in their adaptive immune responses. ApoA-I primarily exerts its anti-inflammatory effects by modulating the functions of immune cells, including monocytes/macrophages, dendritic cells, neutrophils, and lymphocytes (Tao et al., 2024). A proteomic analysis of peripheral serum from patients with NMOSD revealed that ApoB and ApoC4 were up-regulated in the NMOSD group, and Gene Ontology enrichment analysis indicated that the DEP were involved in processes of inflammation response and lipids regulation, such as complement activation, chronic inflammation response, and the integrity of the BBB (Li et al., 2025). This quantitative proteomics analysis suggests that apolipoproteins may play a critical role in the pathological mechanisms of NMOSD. Another proteomic analysis of serum from AQP4-IgG positive NMOSD patients founded that DEP in the relapse and attack severity groups were mainly enriched in cytokine–cytokine receptor interaction, the IL-17 signaling pathway and the chemokine signaling pathway (Wei et al., 2024). A meta analysis of whole-genome sequences revealed that two independent signals were identified within the major histocompatibility complex (MHC) region associated with AQP4-IgG, one of which may be mediated by structural variation in the complement component 4 gene (Estrada et al., 2018). The study demonstrated that HDL-C correlated negatively with inflammatory cytokines, such as IL-6, IL-8, IL-10, in MS (Rádiková et al., 2020). Thus, lipid metabolism may exert pathogenic effects in NMOSD through multiple mechanisms, including cytokine pathways, complement activation, and compromised blood–brain barrier integrity.

The study possesses several strengths. First, it was designed with stringent inclusion and exclusion criteria to enhance data reliability. Given the pathological heterogeneity between AQP4-IgG positive and negative neuromyelitis optica spectrum disorder (NMOSD), only AQP4-IgG–positive patients were enrolled to ensure cohort homogeneity. Second, thyroperoxidase antibodies and other autoantibodies are detectable in up to half of AQP4-IgG-seropositive NMOSD patients, with approximately one-third exhibiting comorbid autoimmune disorders such as thyroiditis, systemic lupus erythematosus, or Sjögren’s syndrome (Wingerchuk and Lucchinetti, 2022). The study revealed that autoimmune thyroid disorders may contribute causally to NMOSD susceptibility (Wang et al., 2022). Therefore, we excluded patients with comorbid autoimmune disorders and adjusted for thyroid hormone levels in the analysis. Additionally, patients with potential confounding factors that could influence lipid metabolism, including the use of lipid-lowering agents, corticosteroid therapy (Martinez et al., 2024), or comorbidities such as hypertension, diabetes mellitus, or nephrotic syndrome were also excluded.

This study has several limitations that should be acknowledged. First, it was a single-center retrospective analysis with a relatively short follow-up duration, which may not fully capture long-term disease progression or lipid profile fluctuations. Lipid levels can be affected by external factors, including dietary habits and medication use. Given the retrospective nature of the study, it was hard to control all potential confounding variables. Second, as a retrospective study, cytokine and complement profiles were not assessed in the majority of early-phase patients; thus, correlation analyses between serum lipid profiles and cytokine levels/complement components could not be performed. Third, the relationship between ApoA-I, HDL-C with AQP4-IgG titers could not be specified owing to the different time at which the AQP4-IgG was sampled. In addition, AQP4-IgG titers were measured in some patients after corticosteroids therapy. Fourth, stringent inclusion/exclusion criteria (e.g., inclusion of AQP4-IgG seropositive patients, exclusion of those with comorbid autoimmune disorders and use of lipid-lowering agents) resulted in a relatively small sample size, diminishing statistical power. Therefore, future prospective, multicenter studies with larger cohorts and extended follow-up periods are necessary to confirm and expand upon these findings. Meanwhile, the correlations between serum lipid profiles and cytokine levels, complement components and AQP4-IgG titers warrant further investigation.

In conclusion, this study demonstrates that low ApoA-I and HDL-C were associated with increased neurological impariment and a higher risk of relapse in NMOSD patients with AQP4-IgG. Peripheral ApoA-I and HDL-C could serve as a potential, simple and accessible biomarkers in the early evaluation of disease severity and relapse risk in NMOSD, which may aid in optimizing acute-phase treatment strategies and facilitate risk stratification for long-term disease management.