Does NAD+ Really Decline With Age? What a 2026 Human Study Found

NAD+ is a coenzyme involved in redox reactions, cellular energy metabolism, and signaling. Its biological importance is not in dispute. The problem begins when a complex, compartmentalized system becomes a single marketing sentence: age lowers NAD+, so a supplement must restore it. ⁵

A useful NAD+ claim needs at least four details: the sample or tissue, the measurement method, the population, and the outcome. Whole blood, plasma, skeletal muscle, and a specific cell compartment are not interchangeable readouts. A result in one should not silently become a claim about all of them.

The broader NAD+ ingredient guide explains the related label identities. This article asks a narrower question: what the new whole-blood study changes about the familiar age-decline story.

The study was published in Nature Metabolism in May 2026. It brought together seven independent human cohorts totaling 303 participants. Six cohorts contributed age or lifestyle analyses. A seventh, 24-participant NR subset from a twin study, helped show that the assay could detect a targeted whole-blood NAD+ change. ¹

The investigators used a validated combination of liquid chromatography and high-resolution mass spectrometry. In plain language, the method separates molecules in a sample and then measures them by mass. The validation work also tested analytical variation and sample handling rather than assuming every stored blood sample behaves the same way. ^{1, 2}

That handling work mattered. Freezing and repeated freeze-thaw cycles reduced measured whole-blood NAD+ under the tested conditions. Differences between studies can therefore reflect not only biology, but also how a sample was collected, stored, and processed. ¹

The researchers found no statistically significant whole-blood NAD+ difference between adults younger than 30 and adults older than 60 in the dedicated cross-sectional cohort. They also found no clear age association in other cohorts spanning wider adult age ranges. ¹

Whole-blood NAD+ was not higher in elite athletes than in the comparison groups, and it did not show a consistent change after the resistance-training, protein-rich diet, or multimodal interventions included in the study. These are results from the interventions and populations examined, not proof that every form of exercise or diet has no relationship with NAD+ biology. ^{1, 2}

In contrast, whole-blood NAD+ increased in the NR supplementation cohort. The important interpretation is methodological: the assay was capable of detecting a change when a targeted intervention altered this blood marker. That analysis was used as an assay-sensitivity check, not as a trial designed to establish an anti-aging benefit. ¹

NAD+ metabolism is compartmentalized. Different tissues, cell types, and subcellular spaces maintain distinct NAD-related pools. A convenient blood measurement cannot automatically stand in for all of them. ⁵

This helps explain why apparently conflicting human findings do not necessarily cancel each other out. A 2022 cross-sectional study of skeletal muscle reported lower muscle NAD+ in older adults and associations with activity, mitochondrial measures, and muscle function. That was a muscle study with its own population and design, not a whole-blood measurement. ³

Neither study proves a universal rule. The newer paper narrows what can be claimed about whole-blood NAD+. The muscle paper suggests that tissue-specific patterns can differ. The sample type belongs in the conclusion, not only in the methods section.

A biomarker can document a measurable biological response. That is useful evidence. It does not by itself establish improved energy, cognition, muscle function, disease prevention, biological-age reversal, or longer life.

A 2025 review of NAD+ precursor research concluded that consistent age-related decline has been shown in only a limited number of human studies, that human tissue data remain sparse, and that clinical trials of precursor supplementation have so far shown limited efficacy. This is a more cautious picture than the front-label sequence of “decline, boost, benefit.” ⁴

For a supplement claim to move beyond biomarker evidence, the studied ingredient, amount, formulation, population, duration, and clinical outcome all need to match. A rise in whole-blood NAD+ is not a substitute for that chain of evidence.

The paper does not show that NR or NMN “does not work.” It also does not demonstrate that either ingredient slows aging. Its NR cohort showed that whole-blood NAD+ can change after precursor use, while the age and lifestyle cohorts challenge that same marker as a simple readout of aging.

The NR observation should not be transferred to NMN or direct NAD+ products as if the names were interchangeable. NAD+ and NADH are coenzyme forms in NAD metabolism. Nicotinic acid and nicotinamide are forms of niacin (vitamin B3), while NR and NMN are NAD-related precursors. These are different label ingredients with different evidence and amount declarations. ^{4, 6}

For the ingredient-level details, use the separate nicotinamide riboside guide, NMN guide, and NMN versus NR comparison. Keeping those questions separate prevents a study about one blood measurement from becoming a verdict on every NAD-related product.

Start with the exact ingredient name and the quantitative amount attached to it per listed serving. U.S. Supplement Facts panels generally disclose the serving size and amount per serving for each dietary ingredient outside a proprietary blend. A front label that says only “NAD+ support” does not answer those questions. ⁷

Next, match the label to the cited evidence. Was the same molecule, amount, formulation, duration, population, and endpoint studied? Liposomal, sublingual, sustained-release, and “bioavailable” wording should not be treated as automatic multipliers unless appropriate comparative evidence supports the finished formulation.

Important strengths include the validated assay, explicit sample-handling experiments, multiple independent cohorts, and a clear NR response that demonstrated analytical sensitivity. The researchers also made source data available where participant privacy allowed. ¹

The main age analyses did not follow the same people across decades. Cohorts also differed in age range, health status, fasting and sampling conditions, and intervention design. The largest single cohort contained 70 participants, so a non-significant result does not prove that every possible difference is exactly zero. ¹

Participants came from European cohorts in the Netherlands, Spain, and Finland, and ethnicity was not recorded. The study measured whole blood rather than total-body NAD+ or a panel of tissues. It also could not compare every NAD-related metabolite reliably across every cohort. ¹

The paper lists European Union, national, foundation, and institutional funding. One reused cohort, MEJNES2019, also received support from Nestlé Health Science alongside public grants. That is cohort-specific context, not evidence that the company funded the entire new analysis. The authors declared no competing interests. ¹