Chapter 5. NAD+, Redox and Sirtuins

A coenzyme with two jobs

Nicotinamide adenine dinucleotide exists in two forms. NAD⁺ accepts electrons from fuels; NADH carries them to the respiratory chain. The cell needs both, and the ratio between them, the redox state, decides which way many reactions run. That is the first job of NAD⁺: as a carrier, it is not used up, only shuttled between its two forms.

The second job was discovered later. Some enzymes do not just borrow NAD⁺; they consume it, splitting it to make a chemical signal or to modify a protein. The best known are the sirtuins.

Evidence

Shin-ichiro Imai, Leonard Guarente and colleagues found that the yeast protein Sir2, which is needed for gene silencing and limits ageing in yeast, removes acetyl groups from histones only when NAD⁺ is present, and consumes it in the process (Imai et al. 2000). The mammalian relative SIRT1 also depends on NAD⁺. In mouse muscle, Carles Cantó, Johan Auwerx and colleagues showed that the energy sensor AMPK, activated by fasting and exercise, raises NAD⁺ levels and thereby SIRT1 activity (Cantó et al. 2009). In mouse liver, fasting increases the activity of SIRT1 on the transcriptional coactivator PGC-1α, which helps switch on glucose production (Rodgers et al. 2005).

Because sirtuins read NAD⁺, they can act as sensors of the cell’s energy state. That is the link between fasting, NAD⁺ and the popular idea of “longevity genes”.

Two ratios, not one

There is not one NAD⁺/NADH ratio in a cell but several. The cytoplasm and the mitochondria hold separate pools, and most NAD is bound to proteins. To estimate the free ratio, Dermot Williamson, Patricia Lund and Hans Krebs used enzyme reactions that are close to equilibrium: if the concentrations of lactate and pyruvate are known, the free NAD⁺/NADH ratio in the cytoplasm follows from the equilibrium constant of lactate dehydrogenase, and likewise in the mitochondria from β-hydroxybutyrate and acetoacetate.

Evidence

In well-fed rat liver, the free NAD⁺/NADH ratio was of the order of 700 in the cytoplasm and of the order of 8 in the mitochondria (Williamson et al. 1967). The mitochondrial couple is far more reduced, because it is where fuels hand over their electrons.

Calculation

The redox potential of the NAD couple follows from the Nernst equation, $E = E^{\circ\prime} + \frac{RT}{2F} \ln \frac{[\text{NAD}^+]}{[\text{NADH}]}$ , with $E^{\circ\prime} = -0.32$ V at pH 7. The free energy released when a pair of electrons from NADH reduces oxygen ( $E^{\circ\prime} = +0.816$ V) is $\Delta G = -2F\,\Delta E$ .

from math import log

R, T, F = 8.314, 310.0, 96485.0
E0_NAD, E0_O2 = -0.320, 0.816               # standard redox potentials at pH 7, volts

def E_nad(ratio_nad_to_nadh):
    return E0_NAD + (R * T / (2 * F)) * log(ratio_nad_to_nadh)

print("compartment / state                  NAD+/NADH   E (V)   free energy per NADH -> O2 (kJ/mol)")
for label, ratio in (("cytosol (free couple, typical)", 700), ("mitochondria, typical", 8),
                     ("mitochondria, more reduced", 4), ("mitochondria, more oxidised", 16)):
    E = E_nad(ratio)
    dG = -2 * F * (E0_O2 - E) / 1000
    print(f"{label:36s} {ratio:9d}   {E:6.3f}   {dG:8.1f}")
# compartment / state                  NAD+/NADH   E (V)   free energy per NADH -> O2 (kJ/mol)
# cytosol (free couple, typical)             700   -0.233     -202.3
# mitochondria, typical                        8   -0.292     -213.9
# mitochondria, more reduced                   4   -0.301     -215.6
# mitochondria, more oxidised                 16   -0.283     -212.1

Two things are visible. A hundredfold difference in the ratio moves the potential by only about 0.06 V, because of the logarithm. And the total energy per pair of electrons, about 210 kJ/mol, is enough in principle for about three and a half ATP at −60 kJ/mol each; the observed 2.5 reflects the losses of real machinery. In Veech’s argument of Chapter 4, a more reduced mitochondrial NAD couple (a lower ratio) means a slightly more negative potential at the top of the chain and slightly more energy to pass down it.

What fasting and ketones do to the ratio

Here the popular story gets complicated, because the same molecule pushes the ratio in different directions in different organs.

In the liver, making β-hydroxybutyrate from acetoacetate uses NADH and produces NAD⁺. Ketogenesis therefore tends to oxidise the liver’s mitochondrial NAD couple.

In the brain and heart, burning β-hydroxybutyrate does the reverse: the first step, back to acetoacetate, produces NADH inside the mitochondria and makes the couple more reduced. That is precisely what Veech described.

In cells under energy stress, as during fasting or exercise, AMPK and other signals raise total NAD⁺ in some tissues (above). Total NAD⁺ and the free NAD⁺/NADH ratio are different quantities, and measurements of one are often reported as if they were the other.

Evidence

In rats fed a ketogenic diet, Marwa Elamin, Susan Masino and colleagues reported a higher NAD⁺/NADH ratio and more total NAD⁺ in the hippocampus (Elamin et al. 2017). That is consistent with fasting signals raising NAD⁺ in brain tissue; it does not settle the free mitochondrial ratio, which such measurements of whole tissue cannot resolve.

The safest summary is that fasting and ketones change NAD metabolism in tissue-specific ways, that sirtuins are among the enzymes that read the change, and that “ketosis raises NAD⁺” is too simple to be true everywhere.

Supplements of NAD+ precursors

NAD⁺ cannot be taken directly in a useful way, but its precursors, such as nicotinamide riboside (NR) and nicotinamide mononucleotide (NMN), can. Because NAD⁺ in tissues declines with age in animals, these precursors are widely sold.

Evidence

In a randomised crossover trial in healthy middle-aged and older adults, six weeks of NR was well tolerated and raised NAD⁺ in blood cells by about 60 per cent; effects on blood pressure and arterial stiffness were suggestive but not conclusive (Martens et al. 2018). In older men, three weeks of NR raised NAD⁺-related metabolites in muscle without improving mitochondrial function (Elhassan et al. 2019). In a ten-week trial in women with prediabetes, NMN improved insulin sensitivity in muscle but did not change body weight or body composition (Yoshino et al. 2021).

The trials are small and short, and their positive findings are on markers rather than on illness, function or lifespan. None tested attention. This book gives no supplement doses and recommends no product; the disclosure in the preface explains why the reader should know that the author’s group runs a website on this topic.

Hypothesis · not established

Not established. Fasting may raise NAD⁺-dependent signalling in brain tissue in ways that matter for attention or for the long-term health of neurons. The hypothesis would be supported by human studies that measure brain NAD⁺ (for example by magnetic resonance spectroscopy) before and after defined fasts and relate the change to cognitive measures; it would be weakened if brain NAD⁺ did not change during fasts of up to 48 hours.

A map of the mechanisms

The first five chapters can now be put side by side, with the time at which each mechanism is known to act and the kind of evidence behind it.

Mechanism	Becomes large after	Best evidence	Human evidence at 16 h?
Liver glycogen falls, gluconeogenesis rises	Hours	Human (NMR)	Yes
Ketones as brain fuel	2–3 days and more	Human (arteriovenous, PET)	Small share (~10%)
β-Hydroxybutyrate as a signal	1–2 mM, a day or more	Cell and mouse	No
Autophagy	1–2 days in mice	Yeast, mouse	Markers only
ATP free energy (Veech)	Several mM ketones	Perfused rat heart	Not reproduced
NAD⁺ and sirtuins	Tissue-dependent	Mouse	Not measured

The table is the honest starting point for Chapter 6: most of the mechanisms that make fasting famous are known to act at depths of fasting that a 16:8 schedule does not reach.

References

Cantó, Carles, Zachary Gerhart-Hines, Jerome N. Feige, et al. 2009. “AMPK Regulates Energy Expenditure by Modulating NAD+ Metabolism and SIRT1 Activity.” Nature 458 (7241): 1056–60.

Elamin, Marwa, David N. Ruskin, Susan A. Masino, and Paola Sacchetti. 2017. “Ketone-Based Metabolic Therapy: Is Increased NAD+ a Primary Mechanism?” Frontiers in Molecular Neuroscience 10: 377.

Elhassan, Yasir S., Katarina Kluckova, Rachel S. Fletcher, et al. 2019. “Nicotinamide Riboside Augments the Aged Human Skeletal Muscle NAD+ Metabolome and Induces Transcriptomic and Anti-Inflammatory Signatures.” Cell Reports 28 (7): 1717–28.

Imai, Shin-ichiro, Christopher M. Armstrong, Matt Kaeberlein, and Leonard Guarente. 2000. “Transcriptional Silencing and Longevity Protein Sir2 Is an NAD-Dependent Histone Deacetylase.” Nature 403 (6771): 795–800.

Martens, Christopher R., Blair A. Denman, Melissa R. Mazzo, et al. 2018. “Chronic Nicotinamide Riboside Supplementation Is Well-Tolerated and Elevates NAD+ in Healthy Middle-Aged and Older Adults.” Nature Communications 9: 1286.

Rodgers, Joseph T., Carlos Lerin, Wilhelm Haas, Steven P. Gygi, Bruce M. Spiegelman, and Pere Puigserver. 2005. “Nutrient Control of Glucose Homeostasis Through a Complex of PGC-1

\alpha

and SIRT1.” Nature 434 (7029): 113–18.

Williamson, D. H., P. Lund, and H. A. Krebs. 1967. “The Redox State of Free Nicotinamide-Adenine Dinucleotide in the Cytoplasm and Mitochondria of Rat Liver.” Biochemical Journal 103 (2): 514–27.

Yoshino, Mihoko, Jun Yoshino, Brandon D. Kayser, et al. 2021. “Nicotinamide Mononucleotide Increases Muscle Insulin Sensitivity in Prediabetic Women.” Science 372 (6547): 1224–29.

Fan, Y. (2026). Fasting, Ketones and the Metabolic Basis of Attention: Autophagy, β-Hydroxybutyrate and the Bioenergetics of Intermittent Fasting [断食、酮体与注意力的代谢基础] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23684203 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).