The NAD+ Salvage Pathway: How Cells Recycle Energy

NAD+ is not a nutrient you consume, and it is not built from scratch on any meaningful scale. It is a molecule your cells break and rebuild, thousands of times over, through a short recycling loop called the salvage pathway. Almost everything interesting about NAD+ — why it falls with age, why precursor supplements exist, why NMN and NR are the two molecules everyone argues about — comes from the structure of that loop.

What NAD+ is for

NAD+ (nicotinamide adenine dinucleotide) is a coenzyme — a molecule other enzymes require in order to work. It participates in two categories of reaction, and the distinction between them is the crux of the whole topic:

• Redox reactions, where NAD+ accepts an electron pair, becomes NADH, hands it off in the mitochondrial electron transport chain, and returns to NAD+. Here the molecule is a shuttle. It is not consumed — it cycles, and one NAD+ molecule can carry electrons indefinitely.
• Consuming reactions, where an enzyme cleaves NAD+ and uses part of it as a substrate. The PARP enzymes do this during DNA repair. The sirtuins do it during deacetylation. CD38, an immune-associated enzyme, does it in large quantities. In these reactions NAD+ is destroyed, and what is left over is nicotinamide.

The first category explains why NAD+ is essential. The second explains why it needs continuous replacement, and why demand rises with damage, inflammation and age: every strand break repaired costs NAD+, and so does every sirtuin-mediated signal. Sirtuin biology covers what that group does with it.

The salvage pathway, step by step

The body does synthesise NAD+ from tryptophan, de novo, but the route is long, energetically expensive, and contributes little in adult tissue. Most NAD+ is recycled from the nicotinamide left over by the consuming reactions above:

1. An NAD+-consuming enzyme cleaves NAD+ and releases nicotinamide (NAM).
2. NAMPT converts NAM to nicotinamide mononucleotide (NMN). This is the rate-limiting step of the entire pathway.
3. NMNAT — three isoforms, in the nucleus, cytosol and mitochondria — converts NMN to NAD+.
4. NAD+ is available again, and the loop closes.

This recycling is the dominant source of NAD+ in adult cells (Revollo et al., J Biol Chem 2004). Two design features are worth noticing. The loop is closed, so the pool is largely conserved and the cell is not dependent on dietary supply. And it has a single bottleneck at NAMPT, which means the whole system's throughput is set by how much of one enzyme is active — an efficient design when supply is stable and a fragile one when the enzyme declines.

The NMNAT isoforms matter more than they look. Because they sit in different compartments, NAD+ pools are compartmentalised — mitochondrial, nuclear and cytosolic pools are separately regulated and do not equilibrate freely. A whole-blood NAD+ measurement averages across pools that may be moving in different directions.

Why levels fall with age

NAD+ declines across the lifespan — the commonly cited figure is roughly a 50% fall between age 20 and 60 — and the fall is not one process but four acting together:

• NAMPT activity declines, so the bottleneck narrows and recycling throughput falls.
• Consumption rises. CD38 expression increases with age and with chronic inflammation, and CD38 is a substantial NAD+ consumer. This appears to be a major driver of the age-related decline rather than a minor contributor (Camacho-Pereira et al., Cell Metab 2016).
• DNA damage accumulates, so PARP activity — and PARP's NAD+ consumption — rises.
• Chronic disease and inflammation raise demand across the board.

That breakdown has a consequence people miss. If the dominant problem is excess consumption rather than reduced production, adding precursor is pouring water into a leaking bucket. Precursors address supply only — nothing about NMN reduces CD38 activity or repairs the DNA damage driving PARP.

Where precursors enter the loop

Every oral NAD+ product is a molecule that enters the salvage pathway at some point, and the only meaningful question is where:

• NMN (nicotinamide mononucleotide) enters downstream of NAMPT, which is the theoretical argument for it — it bypasses the rate-limiting step directly (Rajman et al., Cell Metab 2018). Whether it is absorbed intact or dephosphorylated to NR at the gut wall and reassembled inside the cell is still argued over, and if the latter is true the theoretical advantage largely disappears.
• NR (nicotinamide riboside) is converted to NMN by NR kinase, then continues to NAD+. It also bypasses NAMPT, by a different door.
• Nicotinamide (NAM) itself is cheap and abundant, but it re-enters at the top of the bottleneck and at high concentrations inhibits sirtuins — which is the opposite of the usual goal.
• Nicotinic acid (niacin) uses a separate route entirely, the Preiss-Handler pathway, and raises NAD+ efficiently. Its problem is the flushing reaction, which is dose-limiting for most people.

Human studies show reliably that both NMN and NR raise measurable NAD+ (Trammell et al., Nat Commun 2016; Yoshino et al., on NAD+ intermediates). That is the honest extent of the strong claim.

What the evidence does and does not support

The biochemistry is solid and the clinical outcome literature is thin, and those two facts need holding at the same time. Raising a biomarker is not the same as changing an outcome. Human trials of NAD+ precursors have been mostly small, mostly short, and have generally reported raised NAD+ with modest or inconsistent effects on the functional endpoints that would matter — muscle function, insulin sensitivity, physical performance.

Other questions are unresolved and worth stating. Compartmentalisation means a rise in blood NAD+ may not reflect muscle or brain. The optimum is unknown — nobody has established that pushing an older person toward young-adult levels is beneficial rather than merely different. And a molecule supporting DNA repair and cell survival has plausible reasons for caution as well as enthusiasm, which is why long-term safety data would be valuable and does not yet exist at scale. Is NAD+ worth it works through the cost-benefit.

What else moves the same variable

Precursors are the supply-side lever. The demand and regulation side has interventions with their own evidence base:

• Exercise raises NAMPT expression in skeletal muscle. It is the best-evidenced NAD+ intervention available and it is free — a structured resistance programme or zone 2 work both count.
• Energy restriction and fasting shift the NAD+/NADH ratio and activate the same downstream sensors, overlapping heavily with AMPK signalling.
• Reducing chronic inflammation lowers CD38-driven consumption, which addresses the leak rather than the supply.
• Sleep, because NAD+ metabolism is under circadian control and the clock genes and NAMPT regulate each other reciprocally.
• Intravenous NAD+ bypasses the pathway entirely and produces high acute levels. What that achieves beyond the infusion window is not well characterised.

Bottom line

NAD+ is recycled rather than synthesised, through a short loop in which NAMPT is the rate-limiting step and nicotinamide is the recovered raw material. Levels fall with age because that step slows while consumption by CD38 and PARP rises — a supply problem and a demand problem at once. NMN and NR both enter the loop past the bottleneck and both reliably raise measurable NAD+ in humans; what that translates into functionally is where the evidence thins out considerably. Exercise, sleep and inflammation control act on the same system with better evidence and no cost. Treat precursors as a plausible addition to those, not a replacement for them, and discuss any supplement with a physician alongside the rest of your care.

Educational content, not medical advice. Laboratory interpretation and any treatment decision are made by a licensed physician after individual evaluation. Individual results vary.