Endurance Exercise, Mitochondrial Peptides, and Metabolism

Endurance exercise may trigger a mitochondrial signal that resistance exercise does not reproduce. In a small controlled human trial, 45 minutes of cycling at about 70% of VO₂max significantly raised circulating humanin, a mitochondria-derived peptide. A resistance workout did not raise humanin or MOTS-c. The result does not prove that these peptides cause exercise's metabolic benefits, but it offers a specific, testable explanation for one reason aerobic work may matter.

The stakes are larger than an interesting blood marker. A 2024 meta-analysis found that circulating MOTS-c declined across worsening type 2 diabetes categories, with reported study-level values falling from 235 pg/mL in normoglycemia to 186 pg/mL in well-controlled diabetes and 157 pg/mL in poorly controlled diabetes. That pattern makes mitochondria-derived peptides worth watching. It does not make them a diagnostic test, a treatment target ready for consumers, or a reason to buy a sketchy peptide vial from the internet.

What the exercise study actually found

Von Walden and colleagues' 2021 controlled trial assigned 10 participants per arm to either endurance or resistance exercise. The endurance session was 45 minutes of cycling at approximately 70% VO₂max. The resistance session used four sets of seven-repetition-maximum leg press and knee extension.

After cycling, circulating humanin rose significantly. MOTS-c moved upward only as a non-significant trend. After resistance exercise, neither peptide increased. The study therefore supports a narrow conclusion: under these specific conditions, acute endurance exercise increased humanin, while the tested resistance protocol did not increase either measured peptide.

That distinction matters because the workouts imposed different physiological demands. Muscle biopsies showed that endurance exercise, but not resistance exercise, activated PGC-1α-ex1b and 45S pre-rRNA transcription. The authors' findings fit a mechanism in which sustained oxidative phosphorylation demand helps prompt peptide release. Mechanical loading alone did not produce the same response in this experiment.

Still, a mechanism-shaped result is not a completed mechanism. The trial had only 10 people in each arm, and the supplied findings do not provide a dose-response curve, confidence interval, or evidence for easier aerobic sessions. We do not know whether a brisk walk, 20 minutes of cycling, intervals, or a longer ride produces the same humanin response. We also cannot treat the non-significant MOTS-c result as proof that MOTS-c fails to respond; a study this small can miss a real effect.

Why the diabetes association draws attention

Zhou and colleagues' 2024 systematic review and meta-analysis pooled seven studies, 11 comparison groups, and 661 participants examining humanin and MOTS-c across metabolic states. Its most consistent observation involved type 2 diabetes: MOTS-c was lower with progressively worse glycemic status, and humanin showed a similar suppression pattern in metabolic disease conditions.

This is an association, not a timeline. Cross-sectional data cannot show whether chronic hyperglycemia suppresses peptide production, whether mitochondrial dysfunction lowers peptide levels before diabetes develops, or whether both follow another metabolic process. No prospective cohort in this evidence base tracked peptide levels as people moved from metabolic health to diabetes.

The obesity literature provides a useful warning against tidy stories. Studies in the review pointed in conflicting directions: some found lower MOTS-c in obesity, while one reported higher levels in obese females. Those results should not be averaged into a consumer-facing claim that obesity either raises or lowers MOTS-c. The inconsistency has no settled explanation here, and the assays themselves create a major problem.

Why the assay problem is more than a laboratory footnote

Across the studies in the meta-analysis, absolute MOTS-c values spanned roughly three orders of magnitude. Some reports used concentrations in the low-to-mid hundreds of pg/mL; at least one used arbitrary units. That spread suggests non-standardized assays, differences in calibration, or differences in antibody specificity. No standardized reference material or published inter-assay validation study appears in this evidence base.

Statistical pooling cannot repair incompatible measurements. Standardized mean differences can compare within-study separation, but they do not make one laboratory's absolute MOTS-c reading interchangeable with another's. A commercial MOTS-c result therefore has no established clinical interpretation. It cannot tell you whether your mitochondria need rescuing, whether your training plan works, or whether you are headed toward diabetes.

Does regular endurance training change resting levels?

Alser and colleagues' 2024 athlete comparison directly compared resting MOTS-c and humanin profiles in professional endurance athletes and sedentary controls. The groups differed, which is consistent with an adaptation to chronic endurance training.

"Consistent with" carries the appropriate weight. This was a cross-sectional comparison, not a trial that followed previously sedentary adults through a training program. Endurance athletes may differ before training begins in ways that affect peptide secretion. Training could cause the difference, but self-selection remains a live alternative explanation.

The same uncertainty appears in the 2021 controlled trial. Resting peptide levels did not correlate with VO₂max, leg strength, or mitochondrial DNA copy number within that sample. That finding does not erase the athlete study, but it argues against assuming that higher current fitness automatically means higher resting MDP levels. Training history, genetic background, recent activity, and measurement variation may all matter.

Population differences could change the signal

A separate 2021 randomized exercise study in Hispanic and non-Hispanic White breast cancer survivors followed a 16-week combined aerobic and resistance intervention and found ethnicity-specific MOTS-c responses. That result came from a distinct population and a combined program, so it cannot confirm or overturn the acute cycling trial. It does show that a universal exercise-to-MOTS-c rule would be premature.

MOTS-c is encoded in mitochondrial DNA, making mitochondrial haplogroup variation a plausible contributor to different responses across ancestry groups. Most studies have not stratified participants by genetic ancestry or haplogroup. This leaves an important confound in comparisons across studies and populations.

What this means for training

If your goal is to raise mitochondria-derived peptides specifically, the only directly tested human stimulus here is sustained cycling: about 45 minutes at about 70% VO₂max. That session significantly raised humanin. It did not significantly raise MOTS-c, and resistance training at four sets of seven-repetition-maximum leg press and knee extension raised neither peptide.

That should not turn resistance training into the villain of a mitochondrial morality play. The trial tested one resistance protocol and one biomarker question. Resistance exercise remains useful for strength, muscle mass, bone health, function, and metabolic health through pathways this literature did not measure. It simply has not shown the same MDP response under the studied conditions.

Nor does higher humanin after a ride prove that humanin improved glucose control. In this evidence base, claims that MOTS-c or humanin improve glucose regulation, preserve muscle, or protect tissues come from cell and animal work, not human trials of administered peptides. No human randomized trial has tested exogenous MOTS-c or humanin against HbA1c, insulin resistance, body composition, cardiovascular events, or other metabolic outcomes.

For now, the practical recommendation is ordinary but evidence-based: include regular moderate-to-vigorous aerobic exercise because its health benefits are established well beyond this peptide literature. Treat MDPs as an emerging molecular explanation, not a personalized treatment target.

The verdict

Humanin has the clearest acute human signal: in one small controlled trial, moderate-to-vigorous cycling raised it significantly while the tested resistance workout did not. MOTS-c has a compelling association with worsening type 2 diabetes, but its acute response to endurance exercise was not statistically significant, its obesity data conflict, and its assays lack standardization.

The exercise–MDP–metabolism hypothesis is biologically plausible and worth following. It still lacks the studies that would make it clinically actionable: larger randomized exercise trials, longitudinal training interventions, standardized assays, ancestry-aware analyses, and human trials that connect peptide changes to metabolic outcomes. Until then, exercise is the intervention. The peptides are the clue.