RESUMO
INTRODUCTION AND OBJECTIVES: Multiple sclerosis (MS) is a disease of the central nervous system associated with immune dysfunction, demyelination, and neurodegeneration. The disease has heterogeneous clinical phenotypes such as relapsing-remitting MS (RRMS) and progressive multiple sclerosis (PMS), each with unique pathogenesis. Metabolomics research has shown promise in understanding the etiologies of MS disease. However, there is a paucity of clinical studies with follow-up metabolomics analyses. This 5-year follow-up (5YFU) cohort study aimed to investigate the metabolomics alterations over time between different courses of MS patients and healthy controls and provide insights into metabolic and physiological mechanisms of MS disease progression. METHODS: A cohort containing 108 MS patients (37 PMS and 71 RRMS) and 42 controls were followed up for a median of 5 years. Liquid chromatography-mass spectrometry (LC-MS) was applied for untargeted metabolomics profiling of serum samples of the cohort at both baseline and 5YFU. Univariate analyses with mixed-effect ANCOVA models, clustering, and pathway enrichment analyses were performed to identify patterns of metabolites and pathway changes across the time effects and patient groups. RESULTS AND CONCLUSIONS: Out of 592 identified metabolites, the PMS group exhibited the most changes, with 219 (37%) metabolites changed over time and 132 (22%) changed within the RRMS group (Bonferroni adjusted P < 0.05). Compared to the baseline, there were more significant metabolite differences detected between PMS and RRMS classes at 5YFU. Pathway enrichment analysis detected seven pathways perturbed significantly during 5YFU in MS groups compared to controls. PMS showed more pathway changes compared to the RRMS group.
Assuntos
Esclerose Múltipla Crônica Progressiva , Esclerose Múltipla , Humanos , Seguimentos , Estudos de Coortes , MetabolômicaRESUMO
BACKGROUND: Heart failure (HF) is a leading cause of cardiac morbidity among women, whose risk factors differ from those in men. We used machine-learning approaches to develop risk- prediction models for incident HF in a cohort of postmenopausal women from the Women's Health Initiative (WHI). METHODS: We used 2 machine-learning methods-Least Absolute Shrinkage and Selection Operator (LASSO) and Classification and Regression Trees (CART)-to perform variable selection on 1227 baseline WHI variables for the primary outcome of incident HF. These variables were then used to construct separate Cox proportional hazard models, and we compared these results, using receiver-operating characteristic (ROC) curve analysis, against a comparator model built using variables from the Atherosclerosis Risk in Communities (ARIC) HF prediction model. We analyzed 43,709 women who had 2222 incident HF events; median follow-up was 14.3 years. RESULTS: LASSO selected 10 predictors, and CART selected 11 predictors. The highest correlation between selected variables was 0.46. In addition to selecting well-established predictors such as age, myocardial infarction, and smoking, novel predictors included physical function, number of pregnancies, number of previous live births and age at menopause. In ROC analysis, the CART-derived model had the highest C-statistic of 0.83 (95% confidence interval [CI], 0.81-0.85), followed by LASSO 0.82 (95% CI, 0.81-0.84) and ARIC 0.73 (95% CI, 0.70-0.76). CONCLUSIONS: Machine-learning approaches can be used to develop HF risk-prediction models that can have better discrimination compared with an established HF risk model and may provide a basis for investigating novel HF predictors.