Comparative Effects of Different Exercise Modalities on Circulating Adipokines and Inflammatory Cytokines in Adults...

To compare the effects of different exercise modalities on adipokines and inflammatory markers in patients with Type 2 diabetes mellitus (T2DM), and to explore population-specific response patterns and associations with exercise training characteristics.

Web of Science, PubMed/MEDLINE, Embase, the Cochrane Library and EBSCOhost were systematically searched from inception to 22 September 2025. Randomized controlled trials (RCTs) involving structured exercise training interventions in adults with T2DM were included. Bayesian network meta-analysis was performed to estimate pooled effects and 95% credible intervals (CrIs), followed by secondary Bayesian cluster analyses to explore population-specific patterns.

A total of 60 RCTs involving 2741 patients with T2DM were included. Compared with non-exercise controls (CON), exercise interventions significantly reduced circulating visfatin levels (Hedges' g = -0.60, 95% confidence interval [CI] [-1.25, 0.04], prediction interval [PI] [-2.38, 1.17], I ² = 27%, p = 0.04). Aerobic exercise (AE) significantly increased adiponectin and reduced leptin (mean difference [MD] = -2.53, 95% CrI [-4.08, -0.82]), tumour necrosis factor-α (TNF-α, MD = -1.78, 95% CrI [-2.50, -1.09]) and interleukin-6 (IL-6, MD = -1.54, 95% CrI [-2.82, -0.46]). High-intensity interval training (HIIT), classified as an aerobic interval-based modality distinct from continuous AE, significantly reduced TNF-α (MD = -2.04, 95% CrI [-3.62, -0.50]), whereas resistance training (RT) was associated with statistically credible reductions in both TNF-α and IL-6. AE ranked highest for adiponectin and leptin, whereas RT showed the highest descriptive ranking for inflammatory cytokines; however, the IL-6 findings should be interpreted cautiously because this network showed local inconsistency. In secondary Bayesian and subgroup analyses, RT was associated with comparatively larger reductions in TNF-α (MD = -2.23, 95% CrI [-4.08, -0.45]) and IL-6 (MD = -10.71, 95% CrI [-17.14, -4.48]) among younger adults (< 60 years) and overweight individuals, with the largest reductions in IL-6 (MD = -15.94, 95% CrI [-25.85, -6.26]) observed in older adults (≥ 60 years). AE was associated with comparatively larger reductions in TNF-α (MD = -1.92, 95% CrI [-2.70, -1.00]) and IL-6 (MD = -1.63, 95% CrI [-3.34, -0.37]) among individuals with obesity. Higher AE training frequency was associated with greater reductions in circulating IL-6 (β = -0.44, p < 0.05). Within AE interventions, sessions lasting 45-59 min and intervention durations ≤ 8 weeks were associated with larger improvements in adiponectin (g = 1.42) and leptin (g = -1.36), whereas RT programmes with session durations < 60 min and weekly training volumes > 150 min were associated with larger reductions in TNF-α (g = -0.97).

AE and RT demonstrated different association patterns with adipokine- and inflammation-related outcomes in adults with T2DM. AE was more consistently associated with favourable adipokine-related changes, whereas RT showed comparatively larger associations with some inflammation-related outcomes. Additional exploratory analyses suggested that these patterns may vary across age and weight subgroups, and that training frequency and session duration may also be associated with selected outcomes. However, the IL-6 findings should be interpreted with particular caution because the corresponding network showed local inconsistency. Therefore, these results should be regarded as hypothesis-generating rather than as a basis for modality prioritization or formal exercise recommendations.