Does NAD Decline With Age in Human Tissue?

In 2015 a group at the University of Minnesota put healthy volunteers in a high-field magnetic resonance scanner and read nicotinamide adenine dinucleotide out of their occipital lobes without taking a sample. The phosphorus-31 spectrum resolves NAD+ and NADH separately, so the assay returns both the pool sizes and the redox state in intact brain. Across their volunteers, NADH rose with age while NAD+, total NAD and the NAD+/NADH redox potential fell.

That study is cited constantly, and it deserves to be. It is also one of a very small number of measurements behind a claim now made routinely. So it is worth asking the plain question directly: does NAD decline with age in human tissue, how was each number obtained, and what does the decline actually consist of?

What is being measured when someone says NAD

NAD exists in four interconvertible species: NAD+, NADH, NADP+ and NADPH. The first pair carries hydride in catabolic metabolism; the second is the reductive currency for biosynthesis and antioxidant defence. An assay that reports “NAD” without saying which species and which oxidation state has not reported much.

NAD+ has a second job that matters here. It is consumed, not merely cycled, as a substrate by poly(ADP-ribose) polymerases (PARPs) in DNA repair, by sirtuins, and by the glycohydrolase CD38. Consumption cleaves the molecule and releases nicotinamide, which must be resynthesised. A tissue therefore runs a continuous cycle of breakdown and salvage, and a concentration measured at one instant is the standing balance of two rates rather than a rate itself.

Measurement is not trivial either. Trammell and Brenner noted that high-performance liquid chromatography with ultraviolet detection does not offer the specificity and sensitivity needed for robust quantification of NAD metabolites in complex samples, which is why they developed a targeted assay using HPLC coupled to mass spectrometry, and they describe ionisation suppression of particular compounds as a problem their method had to correct in an earlier published approach.

Does NAD decline with age in human tissue

Three human datasets carry most of the weight, and they sample three different compartments.

Study	Compartment	Method	Participants	Reported direction
Massudi et al., PLoS ONE, 2012	Pelvic skin	Biochemical assays on tissue samples	49 donors, newborn to 77 years	NAD+ negatively correlated with age in males (p = 0.001, r = -0.706) and females (p = 0.01, r = -0.537); PARP activity rose with age in males (p < 0.0001, r = 0.768) and correlated inversely with tissue NAD+ (p = 0.0003, r = -0.639)
Zhu et al., PNAS, 2015	Brain, in vivo	Phosphorus-31 magnetic resonance spectroscopy	Healthy volunteers	Age-dependent increase in NADH; age-dependent reductions in NAD+, total NAD and NAD+/NADH redox potential
Clement et al., Rejuvenation Research, 2019	Plasma	Liquid chromatography-tandem mass spectrometry	Healthy subjects, 20-87 years	NAD+, NADP+ and NAAD declined; NADH, NADPH, nicotinamide, N-methyl-nicotinamide and ADP-ribose increased; nicotinic acid, NMN and NAMN showed no significant change

Two details in that table are usually dropped in summary. The skin study found its associations mainly in males; lipid oxidation and sirtuin activity correlated with age in men and not in women in the same cohort. And the plasma study is not a uniform decline at all – several metabolites rose, and three did not move significantly. A sentence such as “NAD falls with age” compresses a pattern that includes increases.

Declared interests belong in the reading. Two authors of the skin study were directors of a start-up biotechnology company in the NAD field; the first author of the plasma study was affiliated with a company selling longevity interventions. Neither fact invalidates a measurement, and both were disclosed by the authors. They are reasons to want independent replication rather than reasons to discard the data.

Why NAD studies disagree with each other

The sharpest challenge to the simple story came from isotope tracing rather than from a better concentration assay. McReynolds and colleagues infused labelled precursors into young and aged mice and followed NAD synthesis across tissues by mass spectrometry. Four findings cut against the standard narrative at once.

Tissue NAD+ depletion in aged mice was modest, with a median decrease of about 30% – not the collapse implied by the phrase “NAD decline”.
Circulating NAD+ precursors were not significantly changed, and tracing showed unimpaired synthesis of nicotinamide from tryptophan.
In most tissues, turnover of the smaller aged NAD+ pool was modestly faster, so absolute NAD+ biosynthetic flux was maintained. Production was not failing.
Calorie restriction partially mitigated the age-associated decline by decreasing consumption, while acute inflammatory stress from lipopolysaccharide lowered NAD+ by impairing synthesis in young and aged animals alike.

Their conclusion is that the decline in NAD with normal ageing is relatively subtle and occurs despite maintained production, most likely because consuming enzymes are more active. A lower concentration with an unchanged flux is a different physiological statement from a failing synthetic pathway, and the two are not distinguishable by any single-timepoint measurement.

A companion review from the same group sets out the remaining problems plainly: many aspects of NAD+ metabolism remain poorly understood and human data are limited; changes in the composition of cells or tissues with age can themselves shift a bulk NAD+ reading; and the field lacks tools to resolve NAD+ content in specific cells and subcellular compartments, as well as any knowledge of the threshold at which depletion produces physiological consequences in a given tissue. Without a threshold, a 30% difference cannot be called meaningful or trivial.

Why sample handling and compartment decide the number

Three structural limits sit under every value above.

Compartment. Plasma NAD metabolites are extracellular. The Clement study explicitly frames its result as the plasma NAD+ metabolome, and its authors describe age-related alterations in that extracellular pool. Treating a plasma value as a readout of intracellular hepatic or neuronal NAD is an inference, not a measurement.

Averaging. A whole-tissue extract averages cytosolic, mitochondrial and nuclear pools that are separately regulated, and averages across every cell type present. An age-related shift in which cells make up a biopsy will move that average without any cell changing.

Analytical chemistry. NAD+ and NADH differ in stability and are sensitive to extraction conditions, and the reduced forms are easily lost. This is why the targeted mass-spectrometry approach replaced ultraviolet detection, and why cross-study comparison of absolute concentrations, measured in different laboratories with different extraction protocols, remains unreliable.

CD38, apigenin and where reference standards enter

If consumption rather than production drives the decline, the consuming enzymes become the interesting targets. Escande and colleagues, reporting on CD38 – which their group had previously identified as the primary NAD+ase in mammals – characterised two flavonoids as CD38 inhibitors: quercetin and apigenin. In their experiments, pharmacological inhibition of CD38 raised intracellular NAD+; treating cell cultures with apigenin decreased global protein acetylation and the acetylation of p53 and RelA-p65; and administering apigenin to obese mice increased NAD+, decreased global protein acetylation and altered several measures of glucose and lipid homeostasis.

That work is in cell culture and in mice. It establishes a molecular target and a chemical tool, not a human outcome, and the authors present it as identifying CD38 as a pharmacological target rather than as a finished intervention.

The chemistry matters for anyone repeating such experiments. Apigenin is 5,7-dihydroxy-2-(4-hydroxyphenyl)-4H-chromen-4-one, PubChem CID 5280443, molecular formula C15H10O5, molecular weight 270.24 g/mol, CAS 520-36-5. It is one of a large family of closely related flavones that differ by a single hydroxyl or methoxy group and co-occur in plant extracts, so an inhibition constant is only interpretable when the material tested is a characterised analytical reference standard with documented identity and purity rather than a botanical fraction of nominal composition.

Frequently asked questions

How is NAD measured in human tissue?

Three ways, each with a different compartment. Biochemical assays on biopsy material give whole-tissue content, as in the 2012 human skin study. Targeted liquid chromatography-tandem mass spectrometry resolves individual metabolites in plasma or tissue extracts. Phosphorus-31 magnetic resonance spectroscopy measures NAD+ and NADH non-invasively in intact organs, so far principally the brain.

Can NAD be measured in the living human brain?

Yes. A 2015 study in the Proceedings of the National Academy of Sciences used phosphorus-31 magnetic resonance spectroscopy at high field to quantify intracellular NAD+ and NADH contents and the NAD+/NADH redox state in intact human brain. It reported an age-dependent increase in NADH alongside reductions in NAD+, total NAD and redox potential in healthy volunteers.

What is the difference between NAD concentration and NAD flux?

Concentration is the standing pool at one instant; flux is the rate at which the pool is synthesised and consumed. A pool can shrink while flux is unchanged if consumption and synthesis both increase. Isotope tracing in aged mice found roughly 30% lower tissue NAD+ with absolute biosynthetic flux maintained – a distinction invisible to concentration assays.

How much does NAD decline with age in mice?

In the isotope-tracing study across multiple tissues of aged mice, the median decrease in tissue NAD+ was about 30%. The authors describe the decline as relatively subtle, note that circulating precursors were not significantly changed, and attribute the lower concentration to more active NAD+-consuming enzymes rather than to failing synthesis.

Why do NAD studies disagree with each other?

Because they measure different compartments with different instruments. Plasma, skin biopsy and in vivo brain spectroscopy are not interchangeable readouts, extraction protocols differ between laboratories, and bulk tissue values average subcellular pools and cell types. The field also lacks a defined threshold below which depletion has physiological consequences, so the same number supports opposite interpretations.

Is plasma NAD a good marker of tissue NAD?

It has not been established as one. The plasma NAD metabolome is an extracellular pool, and the study that characterised it across ages found some metabolites falling, others rising and several unchanged. Inferring intracellular NAD status in liver, muscle or brain from a plasma value is an assumption, and no published calibration supports it.

What does CD38 have to do with NAD decline?

CD38 is an NAD+-consuming glycohydrolase, reported by Chini and colleagues as the primary NAD+ase in mammals. Because isotope tracing points to increased consumption rather than impaired synthesis as the driver of age-related decline, consumers such as CD38 and PARP became the mechanistic focus. Apigenin and quercetin were characterised as CD38 inhibitors in that context.

References

Research use only. Nonsensia supplies analytical reference standards for laboratory and research applications. This article is published for scientific and educational purposes. It is not medical advice, it does not describe any use in humans, and nothing in it should be read as a recommendation to administer any substance to a person or animal.

Filed under: Longevity Science

This article is part of our guide to The Biology of Ageing: Hallmarks, Chemistry and What the Evidence Supports.

The compound discussed in this article is available as an analytical reference standard: Apigenin ≥99% – Analytical Reference Standard, 1000 mg | CAS 520-36-5.

NAD+ Decline With Age: The Evidence and Its Limits