NMN/NR Real-World Evidence: What Registries and Observational Data Actually Show

Why Real-World Evidence Matters for NAD+ Precursors

Real-world evidence fills the gap between controlled trial conditions and actual patient outcomes at population scale. For NMN and NR, this gap is particularly wide because randomized trials have enrolled small numbers of participants (typically 20-50) over short durations (4-12 weeks), leaving questions about long-term efficacy and safety unanswered.

Traditional RWE sources for pharmaceuticals include insurance claims databases, electronic health record (EHR) systems, and disease-specific registries. NAD+ precursors face a structural disadvantage here: NR is sold as a dietary supplement (ChromaDex's NIAGEN), and NMN's regulatory classification was challenged by the FDA in late 2022 when the agency argued NMN could not be marketed as a supplement because it was under investigation as a drug. This regulatory ambiguity means neither compound generates the claims data or prescription records that fuel traditional pharmacoepidemiology.

The absence of large observational databases does not mean zero real-world data exists. Several smaller registries and biomarker-tracking platforms have begun collecting longitudinal user data. These efforts produce hypothesis-generating signals, but they lack the methodological rigor of claims-linked RWE studies powering decisions for GLP-1 agonists or SGLT2 inhibitors.

The Yoshino Trial: Bridging RCTs and Real-World Questions

The most cited efficacy signal for NMN comes from a randomized, placebo-controlled trial published in Science. Yoshino et al. (2021) enrolled 25 postmenopausal women with prediabetes and administered 250 mg/day NMN for 10 weeks. The primary finding: NMN increased skeletal muscle insulin sensitivity by approximately 25% (measured by hyperinsulinemic-euglycemic clamp), with concurrent increases in muscle NAD+ metabolites.

This trial established biological plausibility but left critical RWE questions open. Twenty-five participants over 10 weeks cannot address durability of response, dose-response across populations, or whether insulin sensitivity gains translate to reduced diabetes incidence over years. The study's inclusion criteria (BMI 25-40, age 55-75, postmenopausal, prediabetic) also limit generalizability to the broader population purchasing NMN supplements, which skews younger and includes both sexes.

A key limitation: no follow-up registry tracked these 25 women after the trial ended. Whether their insulin sensitivity improvements persisted, whether they continued supplementation, or whether any progressed to type 2 diabetes remains unknown. This is precisely the gap that prospective RWE registries could fill.

NR Clinical Trials: Larger Numbers, Similar Limitations

Nicotinamide riboside has a broader published trial base than NMN. Martens et al. (2018) conducted a crossover trial in 24 healthy older adults (55-79 years) using 1 to 000 mg/day NR for 6 weeks and demonstrated a 60% increase in whole-blood NAD+ levels alongside trends toward reduced aortic stiffness (systolic blood pressure decreased by 5 mmHg, P=0.07). Elhassan et al. (2019) confirmed NAD+ elevation in skeletal muscle of older men (N=12) at 1 to 000 mg/day NR over 3 weeks.

The CHROMADEX-sponsored Dollerup et al. (2018) trial randomized 40 obese men to 2 to 000 mg/day NR or placebo for 12 weeks. NAD+ metabolites rose significantly, but the study found no change in insulin sensitivity, mitochondrial function, or body composition. This negative metabolic result at high doses contradicts the positive Yoshino NMN data at lower doses, raising questions about dose-response kinetics, population differences, or whether NMN and NR have meaningfully different tissue distributions.

These discrepancies are exactly what well-designed RWE could help resolve. A registry following thousands of users across different doses, formulations, and durations would provide the statistical power to detect subgroup effects that 12-to-40-person RCTs cannot.

Existing Real-World Data Collection Efforts

Several organizations have attempted to fill the RWE gap, though none yet meets the standard of a validated outcomes registry.

AgelessRx ALIVE Study. AgelessRx, a longevity-focused telehealth platform, launched the ALIVE (Aging Longitudinal Interventions for Vitality Enhancement) observational study, enrolling approximately 1,000 participants using various longevity interventions including NMN and NR. Participants self-report outcomes and submit periodic blood work. The study tracks epigenetic age (via DNA methylation clocks), metabolic biomarkers, and patient-reported quality of life. Preliminary data presented at conferences suggests epigenetic age deceleration in a subset of NMN users, but peer-reviewed publications from this registry remain limited.

Jinfiniti NAD+ Testing Platform. Jinfiniti offers an intracellular NAD+ test used by thousands of supplement consumers. Their aggregated (anonymized) data provides population-level NAD+ distributions by age, sex, and supplement use. While not a clinical outcomes registry, this represents a form of biomarker-level RWE. Their published data suggests baseline NAD+ levels decline approximately 1.5% per year after age 40, and NMN/NR supplementation at 500+ mg/day restores levels to those typical of individuals 10-20 years younger.

Bryan Johnson's Blueprint Protocol. While an N=1 case rather than a registry, the extensively documented self-experimentation by Bryan Johnson (including NMN as part of a multi-intervention longevity protocol) has generated continuous longitudinal biomarker data spanning multiple years. His published data shows sustained NAD+ elevation alongside improvements in vascular age, epigenetic pace of aging, and VO2 max. The obvious limitation: you cannot disentangle NMN effects from dozens of concurrent interventions.

Regulatory Barriers to Traditional RWE Generation

The structural reasons for sparse NMN/NR real-world evidence deserve explicit mention. RWE in standard pharmacoepidemiology depends on three data layers that NAD+ precursors largely lack.

Prescription records. NR and NMN are purchased over-the-counter (or were, before the FDA's 2022 NMN determination). They do not appear in pharmacy claims databases. No ICD-10 code captures "NAD+ deficiency" as a billable diagnosis. Physicians rarely document supplement use in structured EHR fields.

Insurance claims linkage. Because no insurer covers NMN or NR, there is no claims trail connecting supplement exposure to downstream outcomes like hospitalizations, cardiovascular events, or new diabetes diagnoses. This eliminates the possibility of retrospective cohort studies using administrative databases.

Standardized dosing. Unlike prescription drugs with fixed formulations, NMN/NR products vary enormously in purity, bioavailability (standard capsules vs. liposomal vs. sublingual), and actual content vs. label claims. A 2024 analysis by the Natural Products Association found that some NMN products contained <50% of labeled dose. This heterogeneity makes it difficult to attribute outcomes to "NMN" as a single exposure in observational data.

Biomarker RWE vs. Hard-Outcome RWE

A critical distinction exists between the two forms of evidence that NMN/NR registries might generate. Biomarker-level RWE, such as documenting NAD+ levels, epigenetic clocks, or glucose homeostasis metrics across user populations, is achievable now. Hard-outcome RWE, such as linking NMN/NR use to reduced cardiovascular events, cancer incidence, or all-cause mortality, requires either very large registries (tens of thousands) or very long follow-up (a decade or more), or both.

Current data sits firmly in the biomarker category. Conze et al. (2019) published a safety analysis of NR (NIAGEN) pooling data from multiple trials and showing no serious adverse events at doses up to 1 to 000 mg/day, with the most common side effects being mild GI discomfort (nausea, bloating) in approximately 10% of users. This safety profile, confirmed across ~300 human exposures in clinical trials, represents the closest thing to aggregated RWE currently available for NAD+ precursors.

The longevity field faces a fundamental measurement problem here. If the primary claim for NMN/NR is healthspan extension, proving this requires outcomes that take decades to manifest. Surrogate endpoints (epigenetic age, NAD+ levels, insulin sensitivity) can be tracked in registries now, but their validation as predictors of lifespan or disease-free survival remains incomplete.

Comparisons to RWE for Established Longevity Interventions

Context helps. Metformin, often called the most studied longevity drug, has massive RWE from decades of diabetic patient populations. The UKPDS (1998) and subsequent observational analyses of millions of diabetic patients provide hard-outcome data (reduced cardiovascular mortality, possible cancer risk reduction) that NMN/NR may never achieve without similarly large exposed populations.

Rapamycin, another longevity candidate, similarly lacks RWE outside transplant populations. The difference: rapamycin has 30+ years of post-marketing surveillance in immunosuppressed patients generating safety data, while NMN/NR have perhaps 5 years of meaningful human supplementation at scale.

The TAME (Targeting Aging with Metformin) trial, expected to report results by 2027, will establish a framework for measuring multi-morbidity delay as a longevity endpoint. If TAME succeeds methodologically, its endpoint design could serve as a template for future NMN/NR outcome registries.

What Clinicians Should Tell Patients About Current Evidence

The honest clinical summary: NMN and NR reliably raise NAD+ levels in human blood and tissues. Short-term RCTs (4-12 weeks) show metabolic biomarker improvements in specific populations. No real-world evidence from validated registries or health-system databases connects these biomarker changes to clinical outcomes that patients actually care about, such as fewer heart attacks, less cognitive decline, or longer life.

This does not mean the compounds are ineffective. It means the evidence base has not matured to the point where outcome claims can be made. Dr. Charles Brenner, who discovered NR's function as an NAD+ precursor, has stated: "NAD+ precursors are not going to be proven to extend human lifespan in our lifetimes. The question is whether we can demonstrate disease-modifying effects in adequately powered trials."

For patients currently supplementing, reasonable guidance includes: document your baseline (NAD+ level, metabolic panel, epigenetic age if accessible), use a product with third-party purity verification, and recognize that you are participating in an informal observational experiment rather than following evidence-based medicine in the traditional sense.

Emerging Registry Initiatives and Future Directions

Several developments may generate meaningful RWE within the next 3-5 years.

The Hevolution Foundation, backed by Saudi Arabia's Public Investment Fund with a $1 billion endowment, has funded multiple aging-focused clinical trials and registries. Their portfolio includes NAD+ biology studies that may produce longitudinal outcome data at scales previously impossible for supplement research.

Decentralized clinical trials using wearables, home blood collection, and digital health platforms could enable quasi-RWE generation for NMN/NR by continuously monitoring thousands of supplement users. Companies like TruDiagnostic (epigenetic testing) and InsideTracker (biomarker panels) already have large user bases that could be leveraged for retrospective analyses correlating supplement use with biomarker trajectories.

The UK Biobank and similar population cohorts have begun incorporating supplement-use questionnaires. If NAD+ precursor use becomes prevalent enough (estimated at 2-5% of health-optimizing adults in 2025), future Biobank analyses could provide the first population-scale associations between NMN/NR use and health outcomes, though confounding by healthy-user bias would remain a significant limitation.

The minimum viable RWE study for NMN/NR would require approximately 5,000 participants, 2+ years of follow-up, validated exposure ascertainment (pill counts, blood NAD+ confirmation), and pre-specified composite endpoints combining metabolic, cardiovascular, and functional outcomes. No such study is currently enrolled or funded, though the NIH Office of Dietary Supplements has signaled interest in supporting better post-market surveillance of popular longevity supplements.

The Healthy-User Bias Problem

Any observational study of supplement users faces a fundamental confound. People who spend $50-150/month on NMN or NR tend to exercise more, eat better, sleep more deliberately, and engage with preventive healthcare at higher rates than the general population. This healthy-user bias can make supplements appear effective when the real driver of better outcomes is the entire behavioral cluster.

Addressing this requires either active comparator designs (NMN users vs. users of another supplement like CoQ10, matched on health-consciousness markers) or detailed covariate adjustment using validated instruments for health behaviors. Neither approach has been applied to NAD+ precursor observational data at meaningful scale.

The practical implication: even if future registries show NMN/NR users have better cardiovascular outcomes than non-users, causality cannot be inferred without randomization or sophisticated instrumental variable analyses. Real-world evidence for NAD+ precursors will remain hypothesis-generating rather than confirmatory until adequately powered RCTs catch up.

Current best practice for clinicians monitoring patients on NAD+ precursors: track fasting glucose, HOMA-IR, lipid panel, and blood pressure at baseline and 12-week intervals, with optional NAD+ and epigenetic age testing for motivated patients willing to pay out of pocket. Report unexpected adverse events to the FDA MedWatch system to contribute to the passive surveillance database that currently represents the only nationwide safety-signal detection system for these compounds.