Effects of Physical Exercise on Neurofilament Light Chain and Glial Fibrillary Acidic Protein Level in Patients with Multiple Sclerosis: A Systematic Review and Bayesian Network Meta-Analysis.

The prognosis of people with multiple sclerosis (MS) has improved substantially in recent decades due to advances in diagnosis and treatment. Due to the unpredictable course and heterogenous treatment response in MS, there is a clear need for biomarkers that reflect disease activity in the clinical follow-up of these patients. We conducted a systematic review with Bayesian network meta-analysis with the aim of analyzing the effects of physical exercise on neurofilaments (NfL) and glial fibrillary acidic protein (GFAP) levels in patients with MS. A systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, starting with a PICO (patient/population, intervention, comparison, and outcome) question: what are the clinical effects of physical exercise (with independence of the type) on NfL and/or GFAP levels in patients with MS compared with other interventions or no intervention whatsoever? A systematically comprehensive literature search was conducted from January to March 2024 to identify original studies that answered the PICO question, using the main data sources. The quality of the studies included was assessed using the Quality Index of Downs & Black. For studies included in the systematic review that followed a randomized controlled trial (RCT) design, the methodological quality of each paper was assessed using the Physiotherapy Evidence Database (PEDro) Scale. Risk of bias was also explored by two independent reviewers. Finally, all articles were classified according to the levels of evidence and grades of recommendation for diagnosis studies established by the Oxford Center for Evidence-Based Medicine. For continuous outcome measures with enough comparisons and a methodological quality greater than or equal to good according to the PEDro scale, a Bayesian network meta-analysis (NMA) was applied. The statistical analyses were performed in R (version 4.1.3, R Core Team 2023) using the "BUGSnet" and "gemtc" packages. Bayesian NMA can be used to obtain a posterior probability distribution of all the relative treatment effects, which allows us to quantify the uncertainty of parameter estimates and to rank all the treatments in the network. Eight studies were included in this systematic review and six articles in the NMA, and they were appraised for quality. The characteristics of the included studies, types of training and described protocols, methodological quality, risk of bias, and clinical effects on the studied biomarkers were outlined. Qualitative synthesis, effects of different exercise modalities in NfL with the Bayesian NMA, selection of the final model and model assessment, and ranking of interventions are also shown. Our findings indicated that moderate-intensity exercise is more likely to reduce NfL concentration compared to high-intensity exercise, and, in turn, high-intensity exercise is more likely to reduce NfL concentration than low-intensity exercise. However, the effects of high-intensity exercise on GFAP levels were inconclusive.