Tier 2 — strong

Anti-inflammatory effects of nicotinamide mononucleotide (NMN) in human skeletal muscle after BFR-exercise

Yang DL, Chao KC, Yang HT, Chen KH, Dewi L, Condello G, Ye M, Nicholls A, Liao YC, Huang CY, Kuo CH

Journal of the International Society of Sports Nutrition 2026 Volume 23, issue 1, article 2632284

Bibliography

DOI: 10.1080/15502783.2026.2632284
PubMed: PMID 41705654
PubMed Central: PMC12918316
Funding: The present study is sponsored by National Science and Technology Council, Taiwan, Grant number: 108-2410-H-845-022-MY3.
Competing interests: No potential conflict of interest was reported by the author(s).

Study snapshot

Design	Randomised, placebo-controlled, counterbalanced crossover trial. Muscle biopsies at baseline, 0 h and 24 h post-exercise in both conditions.
Model	Untrained young men (22.8 ± 1.5 y) with prior weight-training experience; vastus lateralis muscle biopsies analysed for mRNA (RT-PCR) and immunofluorescence (TOM20, MPO, p16INK4a, COX4, DAPI).
Sample	n=11 (15 enrolled, 4 withdrew for scheduling).
Intervention	Oral gelatin capsules containing 300 mg NMN (AbinoNutra) four times daily (total 1200 mg/d) or 300 mg cornstarch placebo, for 7 consecutive days before and including the BFR-exercise day. Three-week washout between arms. Pre-exercise stimulus: three cycles of 5-min thigh-cuff occlusion at 180 mmHg, then 4 × 8 back squats at 70% 1-RM.
Duration	Seven days of supplementation per arm; 24 h post-exercise recovery window; three-week washout between crossover conditions.
Endpoints	TNF-α mRNA in muscle (RT-PCR); IL-10 mRNA in muscle; p21 mRNA in muscle; Nucleated cell infiltration in necrotic regions (DAPI); Mitochondrial content in muscle tissue (TOM20 immunofluorescence); Rating of perceived exertion (RPE)

What the study showed, in plain terms

Blood-flow-restriction resistance exercise (BFR-exercise) deliberately deprives the working muscle of oxygen during a set of squats, then floods it with blood when the cuff is released. That surge triggers a stronger inflammatory response than a normal squat set, which in the days that follow drives muscle repair, satellite-cell activation and hypertrophy. The question this trial asked was straightforward: if a person takes NMN, does that helpful inflammatory response still happen, or does NMN dampen it?

Eleven young untrained men each did the same BFR-exercise session twice, three weeks apart — once after seven days of 1200 mg/day NMN and once after seven days of an identical-looking placebo. The order was randomised and the participants did not know which capsules they were receiving. The researchers took muscle biopsies from the thigh at three time points around each session and measured inflammatory signals, muscle repair signals and mitochondrial content.

NMN clearly suppressed the inflammatory response. The rise in TNF-α mRNA (the early "clear-out the damaged tissue" signal) was cut nearly in half, and the rise in IL-10 mRNA (the "wind down the inflammation" signal that follows) was flattened almost completely. NMN also delayed the rise in p21, a marker that muscle stem cells use as they commit to becoming new muscle fibres, and it delayed the clearance of the immune cells that had entered the damaged regions of muscle. A separate and unexpected finding was that BFR-exercise increased the mitochondrial content of the muscle tissue by 171% within 24 hours on placebo — but this gain did not happen when the participants were on NMN. Imaging suggested the extra mitochondria were being physically donated by infiltrating neutrophils into the damaged muscle fibres, and NMN appeared to interrupt that transfer.

Key findings

TNF-α mRNA in muscle rose 187% at 0 h post-exercise on placebo (p<0.01, d=1.58) and only 106% on NMN (p<0.01 vs baseline, and significantly lower than placebo at the same time point).
IL-10 mRNA rose 67% at 0 h on placebo (p<0.05, d=0.86); the rise was suppressed to a non-significant change on NMN.
p21 mRNA (myogenic terminal differentiation marker) rose 143% at 0 h and 338% at 24 h on placebo; the 24 h rise was reduced to 257% on NMN.
Nucleated cell infiltration in necrotic muscle regions returned to baseline by 24 h on placebo but stayed elevated at +142% at 24 h on NMN (p<0.05), consistent with a delayed resolution of inflammation.
Mitochondrial content in muscle tissue (TOM20 staining) rose 171% at 24 h post-exercise on placebo (p=0.02, d=2.60); this increase was completely abolished on NMN.
Immunofluorescence with TOM20, MPO and COX4 showed infiltrating neutrophils carrying substantially more mitochondria than the myofiber cytoplasm they were engaging with, with the densest mitochondrial signal at the necrotic disruption sites — consistent with a directional, immune-cell-mediated donation of mitochondria to damaged fibres within 24 hours.
NMN did not alter the participants' rated perceived exertion during the exercise session (8.1 AU on NMN vs 8.5 AU on placebo).

What this study can and cannot tell us

The sample is small (n=11) and homogeneous: young untrained men with prior weight-training experience. There is no direct read-out for older adults, women, trained athletes, or anyone whose baseline inflammatory tone is different for reasons of age, disease or medication.

The dosing period was short — seven days of 1200 mg/day. Chronic NMN users in real-world use may be dosing for months or years, and the trial cannot speak to whether a longer course would produce the same suppression or whether the muscle would adapt.

The primary read-outs are muscle-biopsy mRNA and immunofluorescence at single time points (0 h and 24 h). The trial did not follow the participants through a full training block and did not measure downstream hypertrophy, strength or performance. Whether the observed dampening of inflammation and blocking of the acute mitochondrial gain translates into a real training deficit — or, alternatively, a favourable metabolic shift — is not answered here and is called out explicitly by the authors as their central open question.

The mitochondrial-transfer mechanism is inferred from co-staining and diffusion patterns in fixed tissue, not from a live-cell tracking experiment. It is consistent with recent in vitro literature on myeloid mitochondrial transfer, but the causal chain — NMN suppresses TNF-α, which reduces neutrophil infiltration, which cuts off mitochondrial donation — is a working model rather than a proven pathway.

The trial was funded by Taiwan's National Science and Technology Council with no reported commercial conflicts. The NMN was AbinoNutra branded product; the manufacturer had no reported role in the trial.