Oral single-ingredient NR and endurance performance: benefit in trained adults is not established

Useful facts when choosing a product

• NR is not merged with nicotinic acid, nicotinamide, NMN, NAD+ or NRH.
• NR plus pterostilbene is not single NR. [S12]
• Gray used 1000 mg once; Dolopikou 500 mg once; Stocks 1000 mg/day for one week; Martens 1000 mg/day for six weeks per condition. These are study regimens, not dosing recommendations. [S01-S04]
• Martens explicitly used chloride; other study salts and counterion-free active amounts were not inferred from a brand.
• Unverified deficiency diagnosis and total B3 intake do not prove nutritional sufficiency.

What the research actually shows

# Oral single-ingredient NR and endurance performance: benefit in trained adults is not established

TASK-1021 / R01-061 | Same-chat task 3/5 | Evidence cutoff 2026-09-17 (KST)

## 30-second answer

**Evidence does not establish that oral single-ingredient nicotinamide riboside (NR) improves endurance exercise performance, particularly sport-specific time-trial performance, in trained adults.** The small thesis that directly tested a 5 km cycling time trial excluded endurance-trained athletes. In 10 paired analyses, its main text reports NR 819.5+/-239.8 seconds versus placebo 823.4+/-250.0 seconds (SD), P=0.72; the abstract reverses the condition labels. This does not establish exact zero, equivalence, or applicability to trained adults. [S01](https://etheses.bham.ac.uk/id/eprint/8925/13/Gray2019MScbyRes.pdf)

An older-men local repeated-contraction fatigue signal, measured/predicted VO2max, time to exhaustion, and metabolism during fixed-workload exercise are separate endpoints. None was recoded as a time-trial improvement. **The current efficacy judgment is D, fixed score 28; safety Caution; kind/category S / sports.** This is the supplied Chamgap rule-based index, not a probability of success or official GRADE score. [S02](https://physoc.onlinelibrary.wiley.com/doi/10.1113/JP280825) [S03](https://www.researchgate.net/publication/330914286_Acute_nicotinamide_riboside_supplementation_improves_redox_homeostasis_and_exercise_performance_in_old_individuals_a_double-blind_cross-over_study) [S04](https://www.nature.com/articles/s41467-018-03421-7) [S05](https://media.springernature.com/original/springer-static/esm/art%3A10.1038%2Fs41467-018-03421-7/MediaObjects/41467_2018_3421_MOESM1_ESM.pdf) [Calculation audit](calculation_audit.json)

## 1. Question and completed-record boundaries

The proposed question is **trained adults -> oral single-ingredient NR -> endurance exercise performance -> standardized time-trial completion time -> the actual placebo/appropriate comparator**. Proposed population and formulation are not observed study facts. Cycling, running, and swimming times are not pooled. Faster completion and longer time to exhaustion have opposite favorable directions. Botanical species/parts do not apply to this purified chemical question; salt and active amount are study-specific.

All 3136 index records and the seven supplied originals, 314/929/1068/3093/3094/3135/3136, were examined for duplication. Their completed claims concern NAD+/aging, subjective fatigue/energy, glucose metabolism, gait in frail older adults, cognition, muscle mass, and strength. No exact completed independent claim for the present sport-performance question was identified: operation is `new`. Existing extraction, searches, and source maps were reused without rewriting or regrading earlier clinical content. The current supplied technical record confirms actual subsequent publication of 3135 and 3136; unassigned statuses in original manifests are historical. [Duplicate/completion audit](duplicate_audit.json)

## 2. Actual study conditions

| Study / access | Actual participants and training | Single NR, route, regimen, comparator | Shared exercise, endpoint and denominator | |---|---|---|---| | Gray, repository 2019 / original thesis PDF and figures | 11 healthy men, age 24.91+/-2.34 years(SD). At least 150 min moderate or 75 min vigorous activity/week. **Endurance-trained athletes excluded**; eligible VO2max<60; baseline 43.8+/-6.50 mL/kg/min(SD) | Oral 1000 mg once, four 250 mg capsules the preceding evening approximately 12 h before exercise. NIAGEN named, but **study-specific salt/counterion-free active amount unverified**. Microcrystalline-cellulose placebo; >=1 week crossover interval. Approximately 4 weeks/5 visits is not 4 weeks of dosing | 60 min cycling at 55% baseline VO2max, 2 min rest, then 5 km self-paced cycling TT. Two familiarizations. Eleven randomized orders; one equipment-failure exclusion; **10 paired TT analyses** | | Stocks 2021 / primary abstract, author PDF inaccessible | Eight men, age 23+/-4; VO2peak 46.5+/-4.4 mL/kg/min. Specialized training status not verified from accessible abstract | Oral 1000 mg/day NR for **1 week** versus cellulose; double-blind crossover. Salt/active equivalent/washout unverified from abstract | **60 min at 60% Wmax**; muscle biopsies before/immediately/3 h after. A fixed-duration metabolic experiment, not a TT or TTE efficacy test | | Dolopikou, online 2019 / 2020; author full-text HTML/table | Twelve young men 22.9+/-1.0 years and 12 older men 71.5+/-1.0 years(SEM). Described as recreationally trained, yet recent 2 week regular resistance/aerobic activity is also excluded. Not established endurance athletes | Oral **500 mg once**, two 250 mg capsules, tests approximately **2 h** later. Life Extension product; salt/active equivalent unverified. Identical lactose placebo. **10 days is washout**, not repeated treatment | Same dynamometer/cycling tests; no training program. Twelve paired participants per age group. Local fatigue and VO2 measures are not a TT | | Martens 2018 / original and supplement | Healthy men and women aged 55-79, not a recruited trained-sport cohort. Thirty randomized; 24 completed; **aerobic figure n = 21** | Explicit oral **NR chloride**, 500 mg twice/day=1000 mg/day, **6 weeks per condition**, placebo; no separate crossover washout. Counterion-free active equivalent unreported | Modified Balke graded treadmill VO2max and time to exhaustion. No additional common training program. Actual Supplementary Figure 2A/B; not a fixed-distance TT |

Original locations: Gray printed pp 27-39/48-49, Figure 11, p 57; Stocks abstract; Dolopikou Participants/test methods/Table 3; Martens Methods and actual Supplementary Figure 2. [S01](https://etheses.bham.ac.uk/id/eprint/8925/13/Gray2019MScbyRes.pdf) [S02](https://physoc.onlinelibrary.wiley.com/doi/10.1113/JP280825) [S03](https://www.researchgate.net/publication/330914286_Acute_nicotinamide_riboside_supplementation_improves_redox_homeostasis_and_exercise_performance_in_old_individuals_a_double-blind_cross-over_study) [S04](https://www.nature.com/articles/s41467-018-03421-7) [S05](https://media.springernature.com/original/springer-static/esm/art%3A10.1038%2Fs41467-018-03421-7/MediaObjects/41467_2018_3421_MOESM1_ESM.pdf)

## 3. Separated numeric evidence

| Outcome | NR / placebo or contrast | Analysis, time and interpretation | |---|---|---| | **Gray 5 km cycling completion time**, lower is better | **819.5+/-239.8 / 823.4+/-250.0 seconds(SD)**; paired n = 10 | After the overnight single dose and fixed exercise; paired t test **P=0.72**. Arithmetic **-3.9 seconds(NR-placebo)** uses rounded main-text means; it is not an author-reported adjusted effect or verified paired CI. Abstract condition labels are reversed; original data/correction resolution is unavailable | | Gray fixed-ride power | 136+/-11 W | Prescribed 60 min workload, **not TT mean power**. Initial VO2max is a baseline characteristic, not a treatment effect | | Dolopikou older local fatigue index | **36.3+/-1.1 / 42.6+/-1.4**, mean+/-SEM; 12 paired men | Thirty maximal concentric knee contractions at 180 degrees/s; index from last 5/first 5 torque ratio, in percent; **lower is better**. Body/table **P=0.013**, abstract **P=0.012**. Difference column **-4.9+/-8.5**, 95% CI **-10.3 to 0.5**, conflicts with arithmetic **-6.3 percentage points**; CI includes zero. Retained as conflicting | | Dolopikou young local fatigue index | 36.8+/-1.9 / 38.9+/-1.5 SEM; n = 12 | P=0.369; difference column-2.1, 95% CI-6.8 to 2.8. Not a whole-body TT benefit | | Dolopikou young directly measured VO2max | 46.9 /49.1 mL/kg/min; n = 12 | P=0.455. Dispersion decimal points in accessible HTML are unreliable; SEM was not silently reconstructed | | Dolopikou older **YMCA-predicted VO2max** | 33.0 /30.7 mL/kg/min; n = 12 | P=0.507. Not labeled direct maximal oxygen measurement. Table CI/contrast correspondence remains unresolved and is not used for TT precision | | Martens VO2max / time to exhaustion | mL/kg/min /minutes; **paired n = 21** for each | Six weeks per condition; authors report no significant aerobic-performance effect. Actual Supplementary Figure 2 shows mean+/-SD without precise condition means/paired effects/CI labels; bar heights were not converted into invented numbers. Different from a fixed-distance TT |

[S01](https://etheses.bham.ac.uk/id/eprint/8925/13/Gray2019MScbyRes.pdf) [S03](https://www.researchgate.net/publication/330914286_Acute_nicotinamide_riboside_supplementation_improves_redox_homeostasis_and_exercise_performance_in_old_individuals_a_double-blind_cross-over_study) [S04](https://www.nature.com/articles/s41467-018-03421-7) [S05](https://media.springernature.com/original/springer-static/esm/art%3A10.1038%2Fs41467-018-03421-7/MediaObjects/41467_2018_3421_MOESM1_ESM.pdf) [Structured extraction](study_extraction.json)

**Conflicts:** Gray's Results, Discussion and Figure 11 direction agree, so the main-text assignment is the conditional working value while the opposite abstract assignment is retained. This is not a substitute for an author correction. P=0.85 compares first versus second experimental visits, not NR versus placebo. Nonsignificance also does not establish that familiarization/order effects were completely eliminated. The positive older-men local-fatigue report is retained, but unresolved correspondence among means, difference, CI and P prevents treating it as a clean effect estimate or a TT benefit. [Correction/conflict log](correction_log.json)

## 4. Primary-outcome hierarchy, design and funding

Gray states two hypotheses, fat oxidation and TT, without verified agreement to a registered single primary endpoint and prospective analysis plan. A third researcher generated the computer-randomized order and A/B capsule codes, disclosed after analyses. Thus allocation-concealment implementation was documented, not declared absent. One equipment-failure exclusion is reported; 10 pairs are not 20 independent participants. An accepted thesis is not represented as a peer-reviewed journal article. [S01](https://etheses.bham.ac.uk/id/eprint/8925/13/Gray2019MScbyRes.pdf)

Dolopikou has multiple redox/performance aims without an identified unique prospective registered primary hierarchy. The completed 3136 strength judgment was not reopened; the new extraction concerns repeated-contraction fatigue and VO2. Martens mainly addresses safety/NAD, with exploratory physiological performance. NCT02921659 is confirmed in the paper, but the official registry page yielded only a shell, so full original registration agreement is unverified. The Stocks abstract does not establish complete primary-outcome, concealment, funding or adverse-event reporting. [S02](https://physoc.onlinelibrary.wiley.com/doi/10.1113/JP280825) [S03](https://www.researchgate.net/publication/330914286_Acute_nicotinamide_riboside_supplementation_improves_redox_homeostasis_and_exercise_performance_in_old_individuals_a_double-blind_cross-over_study) [S04](https://www.nature.com/articles/s41467-018-03421-7)

Commercial product use and the repository funding field do not fully establish Gray's financial independence. Dolopikou's no-conflict declaration does not prove complete funding independence. Martens' public support plus ChromaDex contribution/product record was reused. Product identity, funding verification and bias assessment remain separate. [S01](https://etheses.bham.ac.uk/id/eprint/8925/13/Gray2019MScbyRes.pdf) [S03](https://www.researchgate.net/publication/330914286_Acute_nicotinamide_riboside_supplementation_improves_redox_homeostasis_and_exercise_performance_in_old_individuals_a_double-blind_cross-over_study) [S04](https://www.nature.com/articles/s41467-018-03421-7)

## 5. Nutrition and safety

This is an additional-supplementation question, not treatment of diagnosed deficiency. Gray replicated a preceding-day food diary, required an overnight fast, and restricted exercise/caffeine/alcohol/other supplements for 24 h. Dolopikou used the same 3-day diet and reported no other medications/nutritional supplements. Neither establishes quantified total dietary B3/NR or a diagnostic deficiency test; nutritional sufficiency was not assumed proven. Unverified diet/vitamin co-use for Stocks and Martens is retained in the structured extraction. Lower older-person NAD-related redox values are not an NR-deficiency diagnosis; Gray's planned PBMC NAD results were not reported. Specific laboratory-test interference is not established by these records. [S01](https://etheses.bham.ac.uk/id/eprint/8925/13/Gray2019MScbyRes.pdf) [S02](https://physoc.onlinelibrary.wiley.com/doi/10.1113/JP280825) [S03](https://www.researchgate.net/publication/330914286_Acute_nicotinamide_riboside_supplementation_improves_redox_homeostasis_and_exercise_performance_in_old_individuals_a_double-blind_cross-over_study) [S04](https://www.nature.com/articles/s41467-018-03421-7)

**Safety: Caution.** The TT source lacks an analyzable adverse-event table, and adjacent performance studies are short or involve other populations. The existing primary extraction from the MCI trial is reused for safety only: over 10 weeks, 18 events in 7 of 10 NR participants versus 21 events in 7 of 10 placebo participants. Events and affected people are different denominators. One NR participant had a temporary dose reduction for severe nausea/heartburn; this does not estimate causal risk or incidence in athletes. The prior record's severity/seriousness terminology limitation is preserved rather than asserting that all events were mild. [S06](https://www.brennerlab.net/wp-content/uploads/2023/12/Orr23.pdf) (Exact current input, previous TASK-1020, S06 extraction reused.)

Short-term tolerability is not extrapolated to long-term high-dose use, pregnancy/lactation, children, liver/kidney disease, or multiple-drug use. The 500 mg/1000 mg regimens are study conditions, not personal dosing instructions. Missing active-amount or batch-specific chemical information was not filled by inferring a salt from the brand.

## 6. Counterevidence, adjacent findings and current screening

The older-men positive local-fatigue result is retained as counterevidence. Its small male population, acute dosing, training qualifiers and numeric inconsistencies prevent a conclusion about endurance records in trained adults. Martens' nonsignificant TTE/VO2 and Stocks' fixed-workload metabolic findings were not counted as repeated TT refutations; a small benefit is not excluded. [S02](https://physoc.onlinelibrary.wiley.com/doi/10.1113/JP280825) [S03](https://www.researchgate.net/publication/330914286_Acute_nicotinamide_riboside_supplementation_improves_redox_homeostasis_and_exercise_performance_in_old_individuals_a_double-blind_cross-over_study) [S04](https://www.nature.com/articles/s41467-018-03421-7)

Heikkinen 2026 reanalyzes the existing Lapatto NR twin cohort for epigenetic outcomes alongside **separate HIIT cohorts**. VO2 changes in different exercise cohorts were not attributed to an NR+exercise placebo-controlled intervention. Lin 2025 studies 6 week NR/walking in sedentary older hypertensive adults with daytime SBP primary. For the 2026 Friedreich ataxia publication, only the primary indexed title/identifiers were verified; the clinical-population boundary was retained without filling inaccessible dose or numeric results. [S07](https://onlinelibrary.wiley.com/doi/10.1111/acel.70638) [S09](https://link.springer.com/article/10.1007/s11357-025-01815-2) [S10](https://pubmed.ncbi.nlm.nih.gov/42009009/)

Zone2Training NCT07344636 is a potentially relevant registered lead, but the official shell/API did not expose current result or primary-outcome fields. We do not assert that results are absent, or import current recruitment/sample details from mirrors as primary-verified facts. Accessible official results are a revision trigger. The NMN runner trial and NR+pterostilbene study are excluded for molecular/combination mismatch. Existing Werner, cancer-survivor, transplant-survivor, frailty, muscle-mass and gait records retain their independent endpoint boundaries. [S08](https://clinicaltrials.gov/study/NCT07344636) [S11](https://link.springer.com/article/10.1186/s12970-021-00442-4) [S12](https://insight.jci.org/articles/view/158314) [S13-S19 source map](sources.json)

## 7. Final rule application and classification

**Axes: P / R1 / I1 / E0 / B2 / CX.** Clinical claim type B; `no_human_study=false`; big-hard-endpoint RCT, split-coprimary and failed-primary/positive-secondary flags are false. An unverified hierarchy was not treated as a known primary failure.

P means actual functional performance; R1 indicates no verified independent replication of the target question; I1 follows supplied case 29 for unconfirmed funding independence. E0 codes the observed nonsignificant adjacent TT result, not exact zero. B2 follows the supplied small-study criterion(total 11,<200) plus unverified registered-primary/analysis agreement(case 44). Acute dosing is not additionally penalized as inadequate chronic follow-up. Condition-label conflict is separately disclosed. CX means no verified paired treatment-difference CI, not C1 exclusion of benefit. Funding and CI issues were not counted again as bias defects.

Executing the supplied unchanged `derive_grade.py` E0 branch yields **D**. None of H/R2/I2/E+/C1/B0/big-hard-endpoint flags is a strength: **0 strengths -> supplied D anchor 28**. No override or invented rule was used. Separate document quality is **A**, meaning same-author checks of sources, numbers, boundaries, bilingual parity and format; not independent external review or journal certification. [Calculation and original clauses](calculation_audit.json) [Classification](classification_audit.json)

## 8. Search scope, uncertainty, revision and completion

Actual search date was 2026-09-17 KST. Web-engine queries combined English/Korean NR names with time trial/time-trial, endurance, trained, athletes, cycling, exhaustion, deficiency, years, authors, DOIs and registry IDs, while reusing the prior NR source map. University repositories, publishers, author-provided originals, supplementary PDFs and official PubMed/ClinicalTrials.gov identification pages were accessed. Queries are in [search_log.json](search_log.json). No exhaustive Embase/Cochrane/Web of Science search, fabricated PRISMA counts, or inaccessible primary numeric results are claimed. Correction/retraction searches did not yield a verified resolution; this is not a guarantee that no correction/retraction exists.

Remaining information states include conflicting condition labels/CI, inaccessible registrations or full texts, and missing nutrition, salt/active-amount and long-term safety information. **The completed narrow conclusion is that oral single-NR improvement of sport-specific endurance records in trained adults is not established.** A prespecified same-population/same-intervention TT trial, original paired data or correction, official posted results, interpretable effect CIs, or relevant long-term safety could trigger revision while preserving the eventual published ID. These limitations do not defer this manuscript or assign new clinical-value/grade decisions to the recipient.

`result_status=completed_with_uncertainty`; `publication_status=needs_id_assignment`; `content_verification_status=completed_with_declared_scope`. **The new ID, slug, URL and first_published_at remain unassigned.** This does not undo 3135/3136 publication. No server connection, deployment or next TASK was performed. No additional image was necessary to understand this evidence, so no clinical illustration was created.

Why this is classified as D (28)

The supplied unchanged calculator gives P/R1/I1/E0/B2/CX -> D; zero strengths -> fixed anchor 28. E0 codes the nonsignificant TT comparison, not exact zero effect.

Counterpoint. A positive local-fatigue signal and a possible small benefit remain, but do not establish endurance improvement in trained adults.

Rejudgment record. Same-author source, numeric and rule check; not independent consensus — P/R1/I1/E0/B2/CX; supplied unchanged calculator; zero-strength D fixed anchor 28

Review performed and remaining limitations

In 12 older men the local fatigue result favors NR at P=0.013 (abstract 0.012), but the difference column -4.9 and 95% CI -10.3 to 0.5 do not reconcile with the -6.3 percentage-point arithmetic mean difference. The signal is retained without extrapolation to time trials. [S03]

Search scope and limitations. search_log.json: 37 targeted queries; prior current-input source map reused; no exhaustive subscription-database review

Evidence Table

Receipt — 19 References

Evidence access cutoff: 2026-09-17. Each citation states its actual access level (full-text transcription, abstract, index, or other) and remaining limitations. Bibliographic checking does not certify the study results or treatment efficacy.

What this document does and does not do

Chamgap is an information source. It reports what research has and has not confirmed; it does not tell readers what to take or buy. That decision belongs to readers and, when needed, medical or legal professionals. This verdict reflects literature available up to the search date and may change as new research appears. Nothing here is medical advice.