Effects of swinging exercise on immune biomarkers: a systematic review and meta-analysis with machine learning-based identification of responder profiles

TL;DR

Swinging exercises reduce T-cells but boost B-cell immunity, with machine learning identifying key immune markers that predict individual responses.

Contribution

Combines meta-analysis and machine learning to identify immune responder profiles to swinging exercise.

Findings

Swinging exercise decreases T-cell markers but increases B-cell and cardiorenal markers.

CD4+/CD8+ ratio, IgA, and IgG are top predictors of immune response to swinging exercise.

Inflammatory markers show no significant change with swinging exercise.

Abstract

This study integrated meta-analysis and machine learning to elucidate the effects of swinging exercise on key immune biomarkers and identify distinct responder profiles.

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we systematically searched PubMed, Web of Science, the Cochrane Library, Google Scholar, and CNKI databases through February 2025.

Fourteen studies involving 440 participants were included for meta-analysis, examining T-cell subsets (CD3+, CD4+, CD8+, and CD4+/CD8+ ratio), B-cell immunoglobulins (IgA, IgG, and IgM), inflammatory markers (TNF-α, IL-6, and IFN-γ), and cardiorenal indices [creatine kinase (CK), lactate dehydrogenase (LDH), and blood urea nitrogen (BUN)]. Random-effects models revealed a significant decrease in T-cell markers (SMD = −1.24, 95% CI: −1.58 to −0.90) but a concurrent increase in B-cell markers (SMD = 0.86, 95% CI: 0.42–1.30) and cardiorenal markers (SMD = 0.94, 95% CI: 0.55–1.33). The effect of swinging exercise on inflammatory markers is not significantly different (p > 0.05). Meta-regression showed no significant moderating effects of age, exercise intensity, or duration (all p > 0.05). Machine learning analysis [random forest, K-means clustering, and principal component analysis (PCA)] of individual participant data (211 exercisers) identified the CD4+/CD8+ ratio (feature importance = 0.24), IgA (0.19), and IgG (0.18) as the top discriminators between responders and non-responders. Responders exhibited a balanced immune profile characterized by higher CD4+/CD8+ ratios and elevated immunoglobulin levels.

Swinging exercise induces a dual immune response: transient T-cell suppression coupled with enhanced humoral immunity. The inter-individual variability highlights the need for personalized training regimens based on immune monitoring. We recommend integrating immune profiling into athletic programming to optimize health and performance outcomes. The observed increase in markers of muscle damage and metabolic stress (CK, LDH, and BUN) confirms the substantial physiological stimulus provided by these sports.

Figures