Divergent immune cascades activated by metabolic versus mechanical exercise intensity: revisiting cytokine heterogeneity in “moderate-intensity exercise”

“Moderate-intensity exercise” is widely regarded as anti-inflammatory; however, meta-analyses of key cytokines have yielded highly inconsistent results. We argue that this inconsistency reflects not mere statistical noise, but a fundamental conceptual conflation: aerobic exercise intensity is anchored to metabolic rate (%VO 2 max), whereas resistance training intensity is anchored to neuromuscular force output (%1RM). These quantify orthogonal dimensions of physiological stress and are not physiologically interchangeable, yet the shared label “moderate intensity” groups them into a single exposure category. Crucially, these two exercise modalities activate inflammatory signaling pathways in opposite directions: aerobic exercise engages a myokine-mediated anti-inflammatory axis (IL-6→AMPK→IL-10), whereas resistance exercise initiates a damage–repair immune cascade (DAMPs→M1→M2 macrophage polarization), yielding fundamentally divergent acute cytokine profiles. The aggregation of incommensurable physiological stimuli is a major contributor to the extreme heterogeneity reported in current meta-analyses. The proposed “dual-axis classification” quantifies metabolic and mechanical dimensions in parallel rather than as mutually exclusive categories; for mixed-modality interventions (e.g., HIIT), both dimensions should be reported concurrently. Accordingly, we recommend that future studies adopt a dual-axis framework that reports metabolic and mechanical loads separately, implement time-resolved inflammatory sampling, and apply causal inference methods to re-stratify existing evidence by exercise modality, thereby addressing this central source of inconsistency.