Disability in multiple sclerosis is associated with age and inflammatory, metabolic and oxidative/nitrosative stress biomarkers: results of multivariate and machine learning procedures

The aim of this study was to evaluate the immune-inflammatory, metabolic, and nitro-oxidative stress (IM&NO) biomarkers as predictors of disability in multiple sclerosis (MS) patients. A total of 122 patients with MS were included; their disability was evaluated using the Expanded Disability Status Scale (EDSS) and IM&NO biomarkers were evaluated in peripheral blood samples. Patients with EDSS >= 3 were older and showed higher homocysteine, uric acid, advanced oxidized protein products (AOPP) and low-density lipoprotein (LDL)-cholesterol and higher rate of metabolic syndrome (MetS), while high-density lipoprotein (HDL)-cholesterol was lower than in patients with EDSS <3; 84.6% of all patients were correctly classified in these EDSS subgroups. We found that 36.3% of the variance in EDSS score was explained by age, Th17/T regulatory (Treg) and LDL/HDL ratios and homocysteine (all positively related) and body mass index (BMI) (inversely related). After adjusting for MS treatment modalities, the effects of the LDL/HDL and zTh17/Treg ratios, homocysteine and age on disability remained, whilst BMI was no longer significant. Moreover, carbonyl proteins were associated with increased disability. In conclusion, the results showed that an inflammatory Th17 profile coupled with age and increased carbonyl proteins were the most important variables associated with high disability followed at a distance by homocysteine, MetS and LDL/HDL ratio. These data underscore that IM&NO pathways play a key role in increased disability in MS patient and may be possible new targets for the treatment of these patients. Moreover, a panel of these laboratory biomarkers may be used to predict the disability in MS.