Exercise is prescribed for obesity, type 2 diabetes, and hypertension, but the circulating molecules that carry its benefits — and whether intensity changes them — have been hard to pin down. This study ran sprint-interval exercise (SIE) against moderate-intensity exercise (MIE) in the same protocols, profiled plasma proteins and metabolites at rest, immediately after, and three hours after, then cross-referenced the results against disease outcomes in 53,026 UK Biobank participants. The short version: intensity does not just change how much exercise happens. It changes what the body releases.

How it worked

  • Acute sessions in young, active, metabolically healthy males: MIE was 90 minutes of continuous cycling at 90–100% of the first lactate threshold; SIE was 6 × 30-second all-out cycling bouts at 0.075 kg·kg⁻¹ body mass with 4-minute rests
  • Plasma was sampled at rest, 0 h, and 3 h; Olink Explore quantified 2,884 proteins, plus untargeted polar metabolomics
  • Two further cohorts (one detecting 2,787 proteins) and in vitro work: azide-tagging of newly secreted proteins in human muscle cells, iPSC-derived skeletal muscle and brown adipocytes, electrically stimulated C2C12 myotubes, and adipocytes bathed in post-exercise plasma
  • Predicted tissue of origin came from GTEx tissue-enriched expression (at least 4× expression versus other organs); predicted targets came from CellTalkDB ligand–receptor pairs
  • UK Biobank: 53,026 participants and 1,066 incident or prevalent diseases

What they found

  • SIE changed 714 proteins immediately after exercise — nearly a quarter of everything measured. 280 proteins differed between SIE and MIE at 0 h; only 2 did by 3 h
  • Both protocols moved the metabolome, on different clocks. SIE peaked early (203 metabolites at 0 h, 199 at 3 h, including oleic acid, 7-hydroxylauric acid, eicosadienoic acid); MIE was delayed (31 metabolites at 0 h rising to 183 at 3 h)
  • Muscle fibers and adipocytes were the most intensity-sensitive tissues, with broad changes in what they secrete and express. Predicted signalling reached the brain, immune cells, adrenals, intestine, and kidney
  • Against UK Biobank there were 46,993 protein–disease associations across 741 exercise-regulated proteins. Most sat in the risk direction — consistent with the acute inflammatory and stress program that exercise triggers by design. Filtering to the protective direction left 143 proteins across 14 disease chapters, and nearly all of the strongest candidates were sprint-regulated: 32 of 33 prioritized proteins were changed by SIE

Why it matters

The paper’s framing is that intensity is a parameter rather than a detail. The two protocols produce overlapping but distinguishable signatures, and the markers most associated with cardiometabolic protection skewed toward the sprint protocol — the one that takes a fraction of the time. The proposed mechanism is cumulative: single bouts are transient, and the benefit is thought to come from repeated transient exposure.

The authors are explicit about the limits. The cohorts are small and male-biased (19 of the original 28 participants were usable in the acute analysis), organ of origin and destination could not be determined for the metabolites, and the disease associations are observational rather than causal. Whether the pattern holds in women, older adults, or untrained populations is unresolved.