What is the difference between NAD+, NMN and NR?

NAD+, NMN and NR are not three separate products but three points on one shared salvage pathway. NAD+ is the pathway's endpoint, a dinucleotide of about 663 Da. NMN (about 334 Da) is the direct precursor that NMNAT enzymes convert into NAD+, while NR (about 255 Da, the smallest of the three) sits one step further upstream and is converted to NMN by NRK1 and NRK2 before it ever becomes NAD+. All three are studied for the same goal, raising cellular NAD+ availability, but they differ in molecular size, in which enzymes process them, and in how much human bioavailability data exists for each.

Which precursor raises blood NAD+ most effectively in human trials?

Both oral NR and oral NMN reliably raise blood or plasma NAD+ metabolites in randomized trials. NR at 1000 mg per day for six weeks raised PBMC NAD+ about 60 percent versus placebo (PMID 29599478), while a 60 day dose ranging NMN trial raised blood NAD+ significantly at 300, 600 and 900 mg per day (PMID 36482258). Neither trial record points to a clear winner between the two precursors. Direct IV NAD+ has the weakest data of the three: the one available pharmacokinetic pilot found infused NAD+ did not raise blood NAD+ in the first two hours (PMID 31572171).

Is NMN or NR better supported by the clinical evidence?

Both have multiple independent randomized trials with a consistent short-term safety profile, so neither has a clearly stronger overall evidence base. NMN has more trials reporting a functional benefit alongside a confirmed blood NAD+ rise, improved insulin signaling and glucose disposal (PMID 33888596) and improved six-minute walk distance (PMID 36482258), though other NMN trials found no effect on VO2max (PMID 34238308) or on muscle mass and frailty in a pooled meta-analysis (PMID 40275690). NR has the one trial to reach brain NAD via MRS, in Parkinson's patients (PMID 35235774), but also a clean negative-efficacy trial on insulin sensitivity and body composition at a high 2000 mg per day dose (PMID 29992272). Reading either as definitively superior overstates what the current trial record supports.

Do NAD+, NMN or NR extend lifespan in humans?

No. There is no completed human lifespan or mortality-outcome trial for NAD+, NMN or NR, and essentially all lifespan-extension data for these compounds comes from yeast, worms, flies and mice. Human trials top out at surrogate or functional endpoints over weeks to a few months, and even those are frequently mixed to negative: Dollerup found no metabolic benefit despite raising NAD+ substrates (PMID 29992272), and a 2025 meta-analysis found no muscle or frailty benefit in adults with a mean age over 60 (PMID 40275690). A more recent nicotinamide riboside trial in long COVID patients missed its primary cognitive endpoint despite blood NAD+ rising 2.6 to 3.1 fold (PMID 41357333), the same biomarker-versus-outcome gap that runs through this evidence base. Raising a biomarker is not evidence of a lifespan effect.

What is the salvage pathway and how do NAD+, NMN and NR fit into it?

The salvage pathway is the route adult tissue uses to maintain its NAD+ pool, more active than the minor de novo route built from dietary tryptophan. Nicotinamide, the breakdown product of NAD+ itself, is recycled by the rate-limiting enzyme NAMPT into NMN, an enzyme reported to decline with age in multiple tissues. NR bypasses NAMPT entirely and is converted straight to NMN by NRK1 and NRK2, the specific reason it stays of research interest even where NAMPT activity looks impaired. NMN, however it is produced, is the shared final precursor that NMNAT1, 2 and 3 convert into NAD+ (Covarrubias et al., Nat Rev Mol Cell Biol 2021, PMID 33353981).

Why can't NAD+ itself simply cross a cell membrane the way NMN and NR can?

NAD+ is a dinucleotide, roughly 663 Da with a double phosphate charge, structurally different from NMN's single mononucleotide (about 334 Da) or NR's bare nucleoside (about 255 Da). That size and charge mean NAD+ cannot cross an intact cell membrane by simple diffusion, and no confirmed dedicated whole-NAD+ importer has been established in mammalian cells. NR, the smallest of the three, is thought to cross membranes via nucleoside transporters plus an NRK-driven intracellular trapping step, one proposed reason it is orally bioavailable while intact NAD+ largely is not. Oral NAD+ also has to survive gut enzymes and microbiota before absorption even becomes a question.

What NR doses were used in human clinical trials?

Human NR trials have used single doses from 100 to 1000 mg in the first pharmacokinetic study (PMID 27721479), 1000 mg per day for six weeks in adults aged 55 to 79 (PMID 29599478), up to 2000 mg per day for twelve weeks in a metabolic trial (PMID 29992272), and 1000 mg per day for thirty days in the NADPARK Parkinson's trial (PMID 35235774). These figures describe what researchers used under clinical supervision with pre-trial safety review. They are study data, not a human dosing recommendation, and this article is written for research and educational context only.

What NMN doses were used in human clinical trials?

NMN trials have used single doses up to 500 mg in the first safety and pharmacokinetic study (PMID 31685720), 250 mg per day for ten to twelve weeks in two separate trials (PMID 33888596 and PMID 35479740), and a dose ranging design testing 300, 600 and 900 mg per day for sixty days, with 600 mg per day flagged by the authors as an apparent sweet spot (PMID 36482258). As with NR, these numbers describe trial protocols, not a dose recommended here. No chronic human dosing beyond about ninety days has been tested for NMN in the trials reviewed in this article.

What side effects have been reported for oral NR and NMN in clinical trials?

Across the RCTs reviewed here, spanning independent trial groups in the US, Japan, China, Denmark and Norway over roughly eight years, oral NR and NMN were consistently well tolerated with no serious adverse events and no clinically significant abnormalities on routine safety labs. One NMN trial reported a single participant with mild, self-resolving abdominal pain (PMID 35479740). Trial duration was short, mostly six to twelve weeks with one dose ranging NMN trial running sixty days, and sample sizes were small to moderate, typically ten to about forty-eight participants per arm, so the trials are underpowered to detect rare adverse events and say nothing about multi-year safety.

What side effects have been reported for IV or injected NAD+?

Injected NAD+ has a distinct, less reassuring safety signal than the oral precursors. A retrospective review of real-world IV clinic clients found all 6 NAD+ recipients reported moderate to severe symptoms, abdominal cramping, nausea, vomiting, raised heart rate and chest pressure, and their infusions took about 60 percent longer to run, a slower drip rate used to manage the symptoms, versus the IV NR comparison group (Reyna et al., Front Aging 2026, PMID 41704678). The pharmacokinetic pilot study reported no adverse events during its single 6-hour infusion, but that trial only enrolled 8 dosed subjects and did not test repeated dosing (PMID 31572171). No RCT reviewed here specifically evaluated drug interactions, pregnancy, pediatric use or long-term cancer-risk signals for any of the three compounds.

Is NAD+, NMN or NR approved as a drug, or legal to sell as a supplement, in the EU and the US?

None of the three is FDA-approved or EMA-approved as a drug for any indication, and IV NAD+ clinics operate in a largely unregulated wellness space rather than a medically validated one. NR has been sold in the US and EU as a dietary supplement or novel-food ingredient without the exclusionary-clause problem NMN ran into. NMN's US status flip-flopped: the FDA ruled it excluded from the dietary-supplement definition in fall 2022, making it technically unlawful to sell as a US supplement for roughly three years, before reversing that decision in September 2025 after a citizen petition and industry lawsuit. That history is a reminder that regulatory status is not a quality or efficacy signal in either direction, and any of the three should be treated as a research-use material in the EU, not an approved medicine.

Are IV NAD+ drips the most effective way to raise NAD+ levels?

The evidence points the other way. IV NAD+ has the weakest human data of the three compounds discussed here: the one pharmacokinetic pilot found infused NAD+ did not raise blood NAD+ in the first two hours (PMID 31572171), and a separate retrospective review found IV NAD+ caused moderate to severe symptoms in all recipients versus a milder profile for IV NR (PMID 41704678). Oral NR and NMN, by contrast, have multiple RCTs showing a reliable blood NAD+ rise with a consistent short-term safety record. No trial here directly compares oral precursor and IV NAD+ efficacy head to head, but nothing in the current data supports IV being the superior route.

Where can research-grade NAD+, NMN or NR be sourced for laboratory use in the EU, and how should it be stored?

peptidesdirect.io ships NAD+ within the EU as a research-use-only material, supplied lyophilized for laboratory reconstitution, not as a consumer supplement or a dosed product. Note that the human trials reviewed above used oral capsule or tablet forms of NR and NMN, or an IV solution for NAD+, so those results describe the compound's biology, not this product's specific formulation or storage stability. As with other lyophilized research materials in the catalog, unreconstituted vials should be kept refrigerated or frozen, protected from light and from repeated freeze-thaw cycles, and any reconstituted material degrades faster and should be used promptly and kept cold. Always follow the product's own current storage documentation, since stability data differs by molecule.

How do I verify the purity of a NAD+, NMN or NR research vial with a Certificate of Analysis?

Every batch peptidesdirect.io sells is tested by an independent third-party lab, Janoshik, with results published rather than self-reported. The shop's public CoA page lists current lab reports by product and batch, and each report links to a verify.janoshik.com test page showing the raw chromatogram and purity percentage for that specific batch. Buyers should match the batch code printed on their own vial label to the batch shown in the report before treating a listed purity figure as applicable to their unit, since purity is tested and certified per batch, not per generic product listing.

NAD+ vs NMN vs NR: Which Precursor Does the Research Actually Support?

Q: Does taking NAD+ directly, orally or by IV, actually raise NAD+ levels?

The human data for direct NAD+ is the thinnest of the three compared here. The one available pharmacokinetic pilot infused 750 mg intravenously over 6 hours and found blood NAD+ did not rise in the first two hours, suggesting substantial extracellular breakdown before the molecule reaches tissue (PMID 31572171). Because NAD+ cannot cross an intact cell membrane by simple diffusion and no confirmed whole-NAD+ transporter exists, the leading model is that even infused NAD+ is likely broken down and resynthesized through the same nicotinamide, NR and NMN salvage steps rather than absorbed intact. Oral NAD+ faces the same problem plus gut-enzyme and microbiota breakdown before absorption is even relevant.

Q: Does a rise in blood NAD+ mean tissue NAD+ also increases?

Not necessarily, and this is the central unresolved question in the field. A comprehensive review of NR and NMN human trials found both reliably raise blood or plasma NAD+ metabolites but concluded that many trials to date do not provide evidence the supplementation raised tissue, as opposed to blood, NAD+ levels (Freeberg et al., J Gerontol A Biol Sci Med Sci 2023, PMID 37068054). Animal models point to tissue NAD+, not blood NAD+, as the more likely driver of age-related decline (PMID 33353981), but that causal link is not established in humans. NADPARK is one of the few trials to measure NAD+ directly in a tissue, the brain, via MRS, and it did find a significant but variable rise (PMID 35235774).

TL;DR: NAD+ vs NMN vs NR

NAD+, NMN and NR are not three competing products. NAD+ is the endpoint of a shared salvage pathway; NMN and NR are two precursors that feed into it through a common final enzyme step. Oral NR and NMN both reliably raise blood NAD+ metabolites in randomized human trials, at doses from 250 mg to 2000 mg/day, with a consistent short-term safety record. Direct NAD+ (oral or IV) has the thinnest human data of the three; one small IV pilot found infused NAD+ did not raise blood NAD+ in the first two hours. Raising blood NAD+ does not reliably translate into functional benefit: an RCT on insulin sensitivity found no improvement despite a confirmed rise in NAD+ precursors, a meta-analysis on muscle strength and frailty found no benefit either, and a separate exercise trial showed mixed functional results without blood NAD+ having been measured at all. No human trial of any precursor has measured lifespan or mortality. Everything below is preclinical mechanism plus short human trials, not a longevity guarantee.

One Pathway, Three Entry Points

Marketing around NAD+ boosters tends to frame NAD+, NMN and NR as rival products, as if choosing between them were like choosing between two different drugs. Biochemically that framing is wrong. All three converge on the same enzymatic pathway.

Cells maintain their NAD+ pool mostly through the salvage pathway, the dominant route in adult tissue versus the minor de novo route from dietary tryptophan. Three inputs feed into one shared enzyme step:

Nicotinamide (NAM), the breakdown product of NAD+ itself, is recycled by NAMPT (nicotinamide phosphoribosyltransferase) into NMN. This is the rate-limiting step of the whole pathway, and NAMPT activity is reported to decline with age in multiple tissues.

Nicotinamide riboside (NR) is phosphorylated directly into NMN by nicotinamide riboside kinase 1 and 2 (NRK1/2), bypassing NAMPT entirely. That is the specific reason NR is of research interest even where NAMPT activity looks impaired.

Nicotinamide mononucleotide (NMN), whether produced endogenously or supplied as a supplement, is the shared final precursor. NMNAT1, 2 and 3 (three tissue-localized isoforms in the nucleus, Golgi and mitochondria) add an adenylyl group from ATP to convert NMN into NAD+ (Covarrubias et al., Nat Rev Mol Cell Biol 2021, PMID 33353981).

So NR and NMN are not two destinations

They are two different on-ramps to the same NMN-to-NAD+ final step. The real research question is not "NAD+ or NMN or NR", it is which on-ramp actually gets more intact molecule into cells, and whether raising the pool matters functionally once it gets there.

Direct NAD+ is a different case structurally: a dinucleotide (roughly 663 Da, two nucleotides joined by a pyrophosphate bridge), versus NMN's single mononucleotide (about 334 Da) and NR's bare nucleoside (about 255 Da, the smallest of the three, thought to cross membranes via nucleoside transporters plus an NRK-driven intracellular trapping mechanism). NAD+'s size and double phosphate charge mean it cannot cross an intact cell membrane by simple diffusion, and no confirmed dedicated whole-NAD+ importer exists in mammalian cells. Oral NAD+ is largely broken down by gut enzymes and microbiota before absorption ever becomes relevant.

NAD+longevity

Essential cellular coenzyme that declines with age. Powers energy metabolism in every cell, activates sirtuins (longevity genes), and supports DNA repair. A cornerstone molecule in aging and longevity research.

What the Human Trials Actually Measured

Below is a run-through of the controlled human data for each precursor: what was dosed, for how long, and what changed.

Nicotinamide riboside (NR)

The first human pharmacokinetic trial of NR gave single oral doses of 100, 300 and 1000 mg to human subjects. It established that NR is orally bioavailable and produces a dose-dependent rise in the blood NAD+ metabolome, with a pilot single subject showing blood NAD+ rising up to 2.7-fold, and critically, without the flushing reaction that nicotinic acid causes (Trammell et al., Nat Commun 2016, PMID 27721479).

A 2x6-week randomized, double-blind, placebo-controlled crossover trial then gave 1000 mg/day NR to healthy adults aged 55 to 79. NAD+ in peripheral blood mononuclear cells rose about 60 percent versus placebo, and NAAD, a more sensitive synthesis biomarker, rose roughly 5-fold. The regimen was well tolerated, with no significant change in blood pressure or arterial stiffness, though those were secondary, exploratory endpoints and the trial was underpowered to detect an effect on them (Martens et al., Nat Commun 2018, PMID 29599478).

A 12-week RCT then pushed the dose to 2000 mg/day NR in 40 obese, insulin-resistant men aged 40 to 70. It was safe, with no serious adverse events, but insulin sensitivity, glucose disposal and oxidation, energy expenditure, fat metabolism and body composition were all unchanged versus placebo, a clean negative-efficacy result despite the precursor doing what it is supposed to do biochemically (Dollerup et al., Am J Clin Nutr 2018, PMID 29992272).

The most unusual NR trial to date is NADPARK, a Phase I RCT of 30 people with Parkinson's disease. NR was well tolerated and produced a significant, though variable, rise in cerebral (brain, measured by MRS) NAD levels, along with reduced CSF inflammatory markers and modest, variable clinical change, one of the only human trials to show a precursor can raise brain NAD, not only blood NAD (Brakedal et al., Cell Metab 2022, PMID 35235774).

Nicotinamide mononucleotide (NMN)

The first acute safety and pharmacokinetic study of NMN gave single oral doses up to 500 mg to 10 healthy Japanese men. It was safe, well tolerated and efficiently metabolized, though small, short and lacking chronic-dosing data (Irie et al., Endocr J 2020, PMID 31685720).

A 10-week RCT gave 250 mg/day NMN to postmenopausal, overweight or obese, prediabetic women. Insulin-stimulated glucose disposal (clamp) and skeletal-muscle insulin signaling both increased versus placebo, and muscle-remodeling genes such as PDGFRB were upregulated. Body weight did not change, and no adverse effects were reported (Yoshino et al., Science 2021, PMID 33888596).

A 12-week RCT at 250 mg/day NMN in 30 healthy adults found a significant rise in whole-blood NAD+ versus placebo with no adverse effects, one of the cleanest "oral NMN reliably raises blood NAD+" datasets available (Okabe et al., Front Nutr 2022, PMID 35479740).

A 60-day, 4-arm, dose-ranging RCT (placebo, 300, 600, 900 mg/day) in healthy middle-aged adults found blood NAD+ significantly increased in all NMN arms by day 30 and day 60 (p less than or equal to 0.001). Six-minute walk distance improved most at 600 to 900 mg/day, and biological-age markers worsened on placebo while staying stable on NMN, with no adverse events at any dose; the authors flagged 600 mg/day as an apparent sweet spot (Yi et al., Geroscience 2023, PMID 36482258).

A 6-week, 4-arm RCT (placebo, 300, 600, 1200 mg/day) in 48 amateur runners found no significant difference in VO2max at any dose, but ventilatory threshold and oxygen-utilization efficiency improved dose-dependently at the medium and high doses. Blood NAD+ itself was not reported in this trial, a real limitation given it only measured functional exercise readouts (Liao et al., J Int Soc Sports Nutr 2021, PMID 34238308).

Raising the biomarker is not the same as the functional outcome

A comprehensive human-trials review of NR and NMN summarized the field bluntly: both are safe, well tolerated, and reliably raise blood or plasma NAD+ metabolite levels, but "many clinical trials to date do not provide evidence that oral supplementation of these NAD+ precursors raised tissue NAD+ levels," as opposed to blood, and only limited, inconsistent data show clinically relevant functional benefits (Freeberg et al., J Gerontol A Biol Sci Med Sci 2023, PMID 37068054). That blood-versus-tissue gap is the central unresolved question in this entire field, and it is exactly why a trial can raise NAD+ precursor levels and still report nothing functional changed in the same paper, as Dollerup (PMID 29992272) did.

A meta-analysis pooling NMN and NR trials in adults with a mean age over 60 found no significant effect of NMN on skeletal muscle index, handgrip strength, gait speed, or the 5-times chair-stand test. NR showed mixed results: improved walking distance only in peripheral artery disease patients, and worse physical-function scores in patients with mild cognitive impairment. The authors concluded current evidence does not support NMN or NR for preserving muscle mass, function, or frailty in older adults, a direct challenge to the "raise NAD+, get a muscle benefit" assumption behind most marketing for these compounds (Prokopidis et al., J Cachexia Sarcopenia Muscle 2025, PMID 40275690).

Direct NAD+

The one available human pharmacokinetic pilot for direct NAD+ infused 750 mg intravenously over 6 hours into 8 healthy men, with 3 saline controls, and had no clinical efficacy endpoints. The key finding deflates the "just infuse it" assumption: infused NAD+ did not raise blood NAD+ during the first 2 hours, suggesting substantial extracellular or first-pass breakdown before it reaches tissue (Grant et al., Front Aging Neurosci 2019, PMID 31572171). If IV infusion of intact NAD+ does not straightforwardly raise blood NAD+, an oral NAD+ product, which must additionally survive gut enzymes and extracellular first-pass breakdown, faces an even steeper problem.

MOTS-clongevity

Mitochondrial-derived signaling peptide (16 amino acids) that mimics the effects of exercise at the cellular level. Activates AMPK, improves glucose uptake, and enhances fat metabolism - a key tool in metabolic and longevity research.

Epitalonlongevity

Tetrapeptide (Ala-Glu-Asp-Gly) that activates telomerase, the enzyme responsible for maintaining telomere length. One of the most studied peptides in longevity research, developed by Prof. Khavinson at the St. Petersburg Institute of Bioregulation.

Three Misconceptions Worth Retiring

"NMN has its own dedicated transporter, so it enters cells better than NR." This claim traces to a single 2019 mouse-intestine study proposing that the protein Slc12a8 is a sodium-dependent NMN transporter (Grozio et al., Nat Metab 2019, PMID 31131364). The same year, a direct rebuttal from Charles Brenner's NAD+ biochemistry group re-analyzed the transport data and found the methodology unsound (Schmidt and Brenner, Nat Metab 2019, PMID 32694648). The original authors defended their data in reply, so the claim remains scientifically disputed, not confirmed. The more conservative, widely cited model is that extracellular NMN is first dephosphorylated to NR, by the enzyme CD73, before it enters cells at all, meaning NMN may largely travel in through the NR route rather than a private shortcut of its own.

"A rising blood NAD+ number proves the supplement is working where it counts." As the Freeberg review above notes, most human trials confirm a blood or plasma rise without confirming a parallel rise in tissue (muscle, brain, liver) NAD+. Animal models point to tissue NAD+, not blood NAD+, as the driver of age-related decline (Covarrubias et al., PMID 33353981), but that causal link has not been established in human trials. NADPARK (PMID 35235774) is one of the very few trials to measure brain NAD directly via MRS, and it did find a rise, but a "significant but variable" one with only mild clinical change.

"IV or injectable NAD+ drips are the premium, most effective option." Ironically this route has the weakest human evidence of the three. Only one or two small pilot studies exist, one directly showing infused NAD+ failing to raise blood NAD+ in the first two hours, and separate tolerability data point to a worse side-effect profile for infused NAD+ than oral NR or NMN show in their trial records.

EU vs US: a case study in why regulatory status is not a quality signal

Neither NAD+, NMN, nor NR is FDA-approved or EMA-approved as a drug for any indication. NR has been sold in the US and EU as a dietary supplement or novel-food ingredient (marketed as Niagen and similar) without the exclusionary-clause problem NMN ran into. NMN's US status flip-flopped: in fall 2022 the FDA ruled NMN excluded from the dietary-supplement definition, on the theory it had first been authorized for drug investigation before being marketed as a supplement, making it technically unlawful to sell in the US for roughly three years. After a 2023 citizen petition and an industry lawsuit, the FDA reversed course in September 2025, ruling NMN was documented as marketed as a US supplement as early as 2017, predating the IND authorization. NMN is again lawful as a US supplement, but the episode makes a point worth keeping: none of these three compounds has ever completed a drug-approval pathway in the US or the EU, and IV "NAD+ clinics" operate in a largely unregulated wellness space, not a medically validated one.

Safety, Honestly

Across every human RCT cited here for oral NR and NMN, tolerability was consistent: no serious adverse events at studied doses, and no clinically significant abnormalities on routine safety labs, across an eight-year span of independent trial groups in the US, Japan, China, Denmark and Norway. A few honest caveats belong next to that reassurance.

Trial duration is short, most run 6 to 12 weeks, and the longest chronic human oral dosing found here is around 90 days at 600 to 900 mg/day NMN. Multi-year safety is not established, only inferred from short-term tolerability plus the longer safety history of the niacin and nicotinamide family generally. Sample sizes are also small to moderate, typically 10 to about 48 participants per arm, underpowered to detect rare adverse events.

Direct IV or injected NAD+ has a distinct, less reassuring signal. The Grant pharmacokinetic pilot above showed no early blood-NAD+ rise, and separate small retrospective data describe moderate-to-severe gastrointestinal and cardiovascular-type symptoms (flushing, chest tightness, nausea) requiring slow infusion rates, a distinct tolerability profile the oral NR and NMN trials above do not show.

No trial here specifically targeted drug interactions, pregnancy, pediatric use, or long-term cancer-risk signals. NAD+ and NAMPT biology intersect with cell-proliferation pathways, a theoretical caution some reviews raise for populations with active or recent cancer, though no RCT reviewed here tested that directly.

NAD+ in the Wider Longevity and Mitochondrial Research Context

NAD+ metabolism rarely sits in isolation in longevity research. Two other catalog compounds touch adjacent biology, with the same honesty about evidence maturity applied here.

MOTS-c is a mitochondrial-derived peptide studied mainly as an AMPK activator in animal models of metabolic and exercise physiology, adjacent to the mitochondrial energy axis NAD+ also feeds, with human outcome data still thin to nonexistent. Epitalon is studied in the geroprotector literature, mostly Russian animal and cell-culture work on telomerase activity, a similarly preclinical-dominant evidence base. Neither should be read as more validated than the NAD+ precursor data above, both are promising mechanistically and thin on human outcome data.

For more on NAD+ and brain aging specifically, see our companion article: NAD+ 2026: Aging, Brain Health and What the NMN Human Studies Really Show.

Longevity & Anti-Aginglongevity

Mitochondrial function, NAD+ metabolism, telomere maintenance

NAD+ precursor and salvage-pathway research