You take an NAD+ booster because you're tired. A new network meta-analysis suggests that is the one thing none of them reliably fixed.
The study, published in Frontiers in Nutrition by Gao and colleagues at Beijing Sport University, compared NAD+ precursors head-to-head across different kinds of fatigue. It found that these supplements do different jobs, and that none of them clearly made people feel less fatigued.
The Study: What Was Actually Done
The team ran a systematic review and network meta-analysis following PRISMA-NMA guidelines and registered it in PROSPERO (CRD420251237193). A network meta-analysis lets researchers rank treatments even when few trials compared them directly.
- Included: 27 studies, 25 randomized controlled trials and 2 quasi-experimental studies
- Participants: 1,599 adults in total
- Compounds: nicotinamide riboside (NR), nicotinamide mononucleotide (NMN), NADH and nicotinamide (NAM)
- Outcome domains: perceived fatigue, cognitive performance, exercise performance (endurance and strength), quality of life (QoL) and sleep
Results are reported as standardized mean differences (SMD), where larger numbers mean larger effects, and compounds were ranked with SUCRA scores. The authors report no commercial or financial conflicts of interest.
The Ranking, Domain by Domain
The authors argue the answer to "which booster works?" depends on what you want it to do.
Endurance: NMN takes it, and the effect scales with dose
NMN showed a dose-dependent improvement in endurance:
- Medium dose: SMD 0.66 (95% CI 0.09–1.23)
- High dose: SMD 1.02 (95% CI 0.43–1.60)
By conventional benchmarks an SMD near 1.0 is a large effect, which makes this the headline result. The confidence intervals are wide, though, especially at the medium dose, where the lower bound sits close to zero.
Cognition: NADH leads, with a major caveat
NADH produced the strongest cognitive improvement (SMD 1.50, 95% CI 0.82–2.19), mostly in cohorts with cognitive impairment. The authors state the finding was sensitive to influential studies. In practice, removing one or a few trials could shrink or change the result. It's a promising signal, not a settled one, and it applies mainly to people who already had cognitive impairment, not to healthy adults seeking a mental edge.
Quality of life: modest gains, mainly from NMN
QoL improved moderately, driven mainly by NMN:
- High dose: SMD 0.56 (95% CI 0.14–0.98)
- Low dose: SMD 0.45 (95% CI 0.11–0.78)
The two doses look similar here, so this benefit doesn't show the same clean dose-response as endurance.
Strength: unchanged
Neuromuscular strength did not change. Whatever NMN did for endurance, it didn't make people stronger.
The Surprise: Fatigue and Sleep Didn't Budge
The key finding is a negative one. No precursor significantly improved perceived fatigue or sleep quality.
NAD+ boosters are widely marketed as energy supplements, and NAD+-dependent energy dysfunction is considered a key mechanism in fatigue. Yet in pooled data, people didn't report feeling less fatigued, even though measured endurance improved with NMN.
The authors frame fatigue as multidimensional, with cognitive, physical and subjective components. A supplement can shift one component without moving the others. The data suggest NAD+ precursors may change what the body can do on an endurance task without changing how tired a person feels day to day.
The abstract doesn't explain why, so any explanation is speculation. Subjective fatigue scales may be noisy. The trials may have enrolled people who weren't very fatigued. Many studies were small. Or the supplements may simply not affect subjective tiredness.
What About NR?
NR is probably the most-discussed precursor, but the abstract doesn't report a standout win for it in any domain. That is not the same as showing NR is useless. The abstract's specific conclusions point to NMN for endurance and QoL and NADH for cognition. How NR and NAM compared in finer detail would need to be checked in the full paper.
Limitations You Should Know
- Heterogeneous, mostly small trials. Pooling different populations, doses, durations and outcome measures adds uncertainty.
- Dose categories were defined by the authors. The abstract doesn't give exact milligram cut-offs for "low," "medium" and "high," so those labels shouldn't be read as specific doses. For general context only, human NMN trials commonly fall roughly in the 250–1,200 mg/day range.
- NADH's cognitive result is fragile and concentrated in cognitively impaired groups.
- Mixed designs. Two of the 27 studies were quasi-experimental, not randomized.
- Indirect comparisons. A network meta-analysis ranks compounds partly by inference, so the ranking is not equivalent to a large head-to-head trial.
Practical Takeaways
Dose: The best evidence is a dose-response for NMN on endurance, with the larger effect at the higher category. The paper's abstract doesn't specify milligram thresholds, so check the full text or the individual trials before choosing a dose, and don't assume more is safe.
Timing: The abstract doesn't address time of day or relation to meals or workouts, so there's no evidence-based timing advice here.
Who it might suit:
- Adults interested in endurance performance, where NMN had the clearest signal
- People wanting a modest quality-of-life lift, with realistic expectations
- Clinicians or patients exploring NADH in cognitive impairment, as a research-stage option
Who should be skeptical or skip it:
- Anyone buying an NAD+ booster mainly to feel less tired or sleep better, since no precursor significantly improved either
- Anyone hoping for strength gains, since strength was unchanged
- Pregnant or breastfeeding people, and anyone with a medical condition or on medication, who should talk to a clinician first (general caution, not a finding of this study)
Bottom Line
This analysis supports a narrower story than the marketing. NMN has the best case for improving endurance, with a modest QoL bonus. NADH has a tentative cognitive signal in impaired populations. None of the precursors made a measurable difference to how fatigued people felt or how well they slept. The authors call for a precision approach that matches the compound, dose and population to the outcome you actually want.
