NAD+ and Alzheimer's Disease: What 2024-2026 Clinical Research Actually Shows

Bottom line: as of 2026, no clinical trial has shown that NAD+ or any NAD+ precursor — nicotinamide riboside (NR), NMN, or IV NAD+ — treats, slows, or prevents Alzheimer's disease in humans. The preclinical case in mouse models is strong and biologically coherent. The human evidence is limited to one published Phase 1 safety study in Alzheimer's patients, a handful of small mild-cognitive-impairment pilots, and an active research pipeline that has not yet produced a Phase 3 efficacy readout.

Why is NAD+ a target in Alzheimer's disease?

Alzheimer's disease has three pathologies that intersect with NAD+ biology: amyloid-beta plaques, neurofibrillary tau tangles, and a steady metabolic failure of neurons that precedes overt symptoms by years. Each of these pathologies consumes or destroys NAD+, and each is in turn worsened by NAD+ depletion. The result is a self-reinforcing loop that the NAD+ precursor hypothesis is designed to interrupt.

The clearest NAD+ depletion mechanism in Alzheimer's is chronic activation of poly-ADP-ribose polymerase 1 (PARP1), the DNA-damage-repair enzyme that consumes NAD+ as its sole substrate. Oxidative stress, amyloid-beta toxicity, and tau-driven genomic instability all trigger sustained PARP1 activation in neurons and glia. Postmortem Alzheimer's brains show elevated PARP1 activity and reduced NAD+ in vulnerable regions including the hippocampus and entorhinal cortex (Love et al., Brain 1999, PMID 10072011; Strosznajder et al., Mol Neurobiol 2012, PMID 22302351).

When NAD+ falls, sirtuin function falls with it. SIRT1 is required for the autophagic clearance of amyloid-beta and pathological tau; SIRT3 is the principal regulator of mitochondrial quality control inside neurons. SIRT3 deficiency in the Alzheimer's brain correlates with worse mitochondrial dysfunction and accelerated cognitive decline (Yin et al., Aging Cell 2016, PMID 27613551). The hypothesis is that restoring NAD+ restores sirtuin signalingand therefore restores at least the proteostatic and mitochondrial arms of the Alzheimer's pathology.

How depleted is NAD+ in the Alzheimer's brain?

Brain NAD+ falls with age across all species studied. In Alzheimer's disease, the decline accelerates beyond what age alone predicts. Postmortem studies of Alzheimer's hippocampus and cortex report 30-50% reductions in NAD+ compared with age-matched cognitively normal controls (Liu et al., Cell Metab 2019, PMID 31257032; Hou et al. PNAS 2018, PMID 29432159, supplementary postmortem data). The depletion is most severe in regions with the highest amyloid and tau burden — exactly the regions where mitochondrial function fails first.

Compounding the loss, the activity of CD38, an enzyme that consumes NAD+ as part of immune signaling, rises with age and with neuroinflammation. Alzheimer's brains show elevated CD38 expression in microglia and astrocytes (Guerreiro et al., Nat Commun 2020), which further accelerates NAD+ consumption. Between PARP1 upregulation, CD38 upregulation, and reduced NAMPT-driven salvage capacity, the Alzheimer's brain is in a chronic state of NAD+ deficit that gets worse with disease progression.

Brain NAD+ trajectory in healthy aging vs Alzheimer's disease

The preclinical case: mouse models, C. elegans, and proteostasis

Preclinical evidence for NAD+ precursors in Alzheimer's disease is consistent, mechanistically coherent, and — by the field's own cautious standards — heavily overinterpreted by the supplement industry and lay press. The three foundational papers are worth understanding in detail because they define the upper bound of what the precursor hypothesis claims, and the gap between what they show and what human trials have not yet shown.

Hou et al. PNAS 2018 (PMID 29432159)

Hou and colleagues at the NIH National Institute on Aging used the 3xTgAD and APP/PS1 mouse models of Alzheimer's — the workhorse transgenic systems that overexpress mutant human amyloid precursor protein and presenilin. Treatment with nicotinamide riboside (12 mM in drinking water for 3 months) reduced phosphorylated tau, lowered amyloid-beta deposition, attenuated neuroinflammation, and improved contextual memory and spatial learning on standard rodent cognitive tests. The mechanism centered on reduced DNA damage and restored mitochondrial function via the NAD+ salvage pathway.

Sorrentino et al. Nature 2017 (PMID 29211722)

Sorrentino and colleagues at EPFL used a multi-model approach (C. elegans, transgenic mouse Alzheimer's models) to show that NAD+ repletion restored mitochondrial unfolded protein response (UPRmt) signaling, reduced amyloid-beta proteotoxicity, and improved organismal healthspan. The proposed mechanism was sirtuin-dependent activation of mitophagy and the UPRmt — a more sophisticated proteostatic model than simple substrate replenishment. The Sorrentino result is one of the most-cited preclinical NAD+ papers because it ties NAD+ to proteostasis through a defined molecular pathway rather than a generic "energy" argument.

Fang et al. and the mitophagy pathway

Vilhelm Bohr's group at the NIH (Fang et al., Nat Neurosci 2019, PMID 30742114) added a third mechanistic pillar: NAD+ precursors and NAD+-elevating compounds restore neuronal mitophagy in mouse models of Alzheimer's, clearing damaged mitochondria that would otherwise accumulate and amplify oxidative stress. Restoration of mitophagy reduced both amyloid-beta plaque burden and tau hyperphosphorylation in their hands. The mitophagy framework is the cleanest single-arrow explanation for why NAD+ precursors might help any neurodegenerative disease with mitochondrial dysfunction.

Human clinical trials of NAD+ precursors in Alzheimer's

The published human evidence in Alzheimer's is far thinner than the preclinical literature suggests. As of 2026, there is one published Phase 1 safety trial in Alzheimer's patients specifically, several small Phase 2 trials in mild cognitive impairment (a population at risk for Alzheimer's), and active recruitment for follow-up studies. No Phase 3 efficacy trial has reported results.

Wang et al. NR safety trial in Alzheimer's (NCT03482167)

The University of Washington and VA Puget Sound Health Care System registered and conducted the first NR Phase 1 trial in Alzheimer's disease (ClinicalTrials.gov NCT03482167). The trial tested 1000 mg/day oral nicotinamide riboside in mild-to-moderate Alzheimer's patients with co-primary objectives of safety, tolerability, and peripheral NAD+ elevation measured in peripheral blood mononuclear cells (PBMCs). Secondary outcomes included exploratory cognitive measures and plasma biomarker shifts.

The published and abstract-disclosed findings report acceptable tolerability, PBMC NAD+ elevation consistent with prior healthy-adult NR pharmacokinetic studies (Martens et al. Nat Commun 2018, PMID 29453352; Conze et al. Sci Rep 2019, PMID 31506573), and no signal for hepatotoxicity or hematologic abnormality at the 1 g/day dose. Cognitive endpoints were exploratory only; the trial was not powered to detect cognitive benefit. The trial's value lies in establishing that NR can be administered to Alzheimer's patients without acute safety concerns — not in demonstrating efficacy.

Mild cognitive impairment trials (NR and NMN)

Mild cognitive impairment (MCI) is the prodromal stage from which most Alzheimer's dementia eventually emerges, and several small Phase 1-2 trials have tested NR or NMN in MCI cohorts as a proxy for Alzheimer's prevention. Reported outcomes are mixed: peripheral NAD+ elevation is consistent and reproducible; cognitive endpoints on standard batteries (MMSE, MoCA, ADAS-Cog) have not shown robust improvement versus placebo in published readouts. Phase 2 trials with biomarker progression endpoints (plasma p-tau 217, neurofilament light chain, amyloid PET) are needed to clarify whether NR or NMN modifies the trajectory of MCI toward Alzheimer's.

The MIND-AD trial and the "MIND-AD NAD" misnomer

Online discussions sometimes conflate the MIND-AD multi-domain lifestyle intervention trial with NAD+ research. MIND-AD (and the related FINGER 2.0 family of trials) tests combinations of the MIND diet, structured exercise, cognitive training, and cardiovascular risk management in prodromal Alzheimer's. NAD+ precursors are not an MIND-AD intervention arm in any published protocol. The broader FINGER 2.0 program does include nutritional supplement components in some cohorts, but the active comparator is not NR or NMN as a standalone NAD+ trial. Treat any claim that "MIND-AD tests NAD+" with skepticism — the literature does not support it as of 2026.

IV NAD+ infusion clinics and the absence of Alzheimer's evidence

Direct intravenous NAD+ infusion is offered commercially in the United States and elsewhere, sometimes with marketing claims targeting Alzheimer's, dementia, or "brain fog." There is no randomized trial showing that IV NAD+ therapyimproves cognition, slows Alzheimer's progression, or alters disease biomarkers. The pharmacokinetics of IV NAD+ are poorly characterized and the proportion of an IV NAD+ dose that crosses the blood-brain barrier in humans has not been measured directly. Commercial IV NAD+ protocols for cognitive indications are off-label and unsupported by clinical-grade evidence.

Timeline of NAD+ in Alzheimer's research

Mechanistic detail: PARP, SIRT3, and mitophagy in the Alzheimer's brain

Three mechanisms account for nearly all of the proposed therapeutic rationale for NAD+ precursors in Alzheimer's. Each is biologically defensible, each is supported by mouse data, and none has been demonstrated to be the rate-limiting step in human disease.

PARP overactivation

PARP1 detects DNA single- and double-strand breaks and consumes NAD+ to assemble poly-ADP-ribose chains that recruit the repair machinery. In Alzheimer's disease, the source of DNA damage is not transient — amyloid-beta toxicity, tau aggregation, and chronic neuroinflammation impose continuous oxidative damage on neurons. PARP1 stays activated, NAD+ stays drawn down, and the salvage pathway cannot refill the nuclear pool fast enough. The downstream consequence is sirtuin failure, because sirtuins also require NAD+ as substrate. The PARP-sirtuin trade-off is a recurring theme across neurodegeneration; it is most clearly demonstrated in Alzheimer's.

SIRT3 and mitochondrial quality control

SIRT3 is the mitochondrial sirtuin and the principal deacetylase for mitochondrial proteins. It controls antioxidant defense (via SOD2 deacetylation), substrate selection (via PDH regulation), and mitochondrial unfolded protein response signaling. SIRT3 protein levels are reduced in Alzheimer's hippocampus (Yin 2016, PMID 27613551), and SIRT3 knockout exacerbates amyloid pathology in transgenic mouse models. Restoring NAD+ should restore SIRT3 function and therefore mitochondrial quality control — that is the proposed mechanism by which NR rescued mitochondrial function in the Hou 2018 and Fang 2019 mouse studies.

Mitophagy collapse

Damaged mitochondria in Alzheimer's neurons are not cleared efficiently. Mitophagy — the selective autophagic degradation of damaged mitochondria — fails in the Alzheimer's brain, and the accumulated damaged organelles produce oxidative stress that accelerates the disease. NAD+ precursors restore mitophagy in mouse models of Alzheimer's (Fang et al. 2019, PMID 30742114) through a sirtuin-dependent pathway. Whether mitophagy is the rate-limiting bottleneck in human Alzheimer's is unresolved.

Which NAD+ precursor — and does any of them reach the brain?

The pharmacology of oral NR, oral NMN, oral nicotinamide, and IV NAD+ in humans diverges sharply once the dose is past the gut. None of them crosses the blood-brain barrier intact in meaningful quantities. The brain NAD+ rise documented after oral NR in Parkinson's patients (Brakedal et al., Cell Metabolism 2022, PMID 35235774, in NADPARK) is mediated indirectly: peripheral conversion to nicotinamide, which crosses the BBB readily, and brain NAMPT salvagerebuilds central NAD+. For Alzheimer's, this matters because the proposed therapeutic mechanism depends on raising NAD+ specifically in vulnerable neurons — hippocampal pyramidal cells, entorhinal cortex layer II projection neurons, basal forebrain cholinergic neurons — where PARP1 hyperactivity is driving the local depletion.

Brain delivery routes for major NAD+ precursors

The practical implication for Alzheimer's drug development is that the precursor route is unlikely to be the breakthrough delivery method. CNS-penetrant NAD+ precursors, brain-targeted prodrugs, or small molecules that activate the brain's own salvage pathway (NAMPT activators, CD38 inhibitors) are more biologically defensible targets for a disease-modifying therapy. See our review of CD38 inhibitors as a NAD+ longevity target and NAMPT as the rate-limiting enzyme for the alternative-target landscape.

How does the NAD+ field compare with other Alzheimer's disease-modifying programs?

Alzheimer's disease-modifying therapeutics have a long history of mechanistic promise followed by clinical disappointment. The table below puts NAD+ precursors in context against the major disease-modification programs of the last fifteen years.

Three observations are worth making about that table. First, the lecanemab/donanemab readouts establish that disease modification is achievable but modest — a 27% slowing in cognitive decline on a composite measure over 18 months is the current best-case outcome for any AD therapy. Second, every previous intervention with compelling mouse-model data has failed or produced equivocal readouts in Phase 3. Third, the NAD+ precursor program is at least five years behind the amyloid-clearance programs in clinical development — the field has not even attempted a powered Phase 2 cognition trial in Alzheimer's, let alone Phase 3.

Evidence-tier summary: NAD+ in Alzheimer's by claim

The evidence base for NAD+ in Alzheimer's breaks down differently depending on which claim is being evaluated. The following grading follows the four-tier system documented at our methodology page: strong, moderate, emerging, and preclinical.

• NAD+ is depleted in the Alzheimer's brain. Moderate evidence — multiple postmortem cohorts and converging mechanism (PARP/CD38 upregulation, reduced salvage capacity).
• NAD+ precursors raise peripheral NAD+ in Alzheimer's patients. Strong evidence — replicated PBMC NAD+ elevation in Wang Phase 1 and adjacent healthy-aging trials.
• NAD+ precursors raise brain NAD+ in Alzheimer's patients. Emerging evidence — extrapolated from Parkinson's 31P-MRS data (Brakedal 2022); no published direct measurement in Alzheimer's patients yet.
• NAD+ precursors reduce amyloid or tau in mouse Alzheimer's models. Preclinical only — robust and replicated, not translated to humans.
• NAD+ precursors improve cognition in Alzheimer's patients. No evidence — no powered trial has tested this endpoint.
• NAD+ precursors slow Alzheimer's disease progression. No evidence — no longitudinal trial has tested this endpoint.

What we still do not know

Five questions are open in 2026, and each represents a study that the field needs before NAD+ precursors can be considered for Alzheimer's care.

1. Do NAD+ precursors raise NAD+ in the Alzheimer's brain specifically? The 31P-MRS data in NADPARK established brain target engagement in Parkinson's patients. The same measurement has not been published in Alzheimer's patients, where the BBB and salvage dynamics may differ.
2. Does brain NAD+ elevation reduce neurodegeneration biomarkers in humans? Plasma p-tau 217, neurofilament light chain (NfL), and GFAP are now established biomarkers of Alzheimer's progression. A Phase 2 trial with these as primary endpoints is the obvious next step.
3. What is the right dose, route, and duration? NADPARK used 1 g/day NR for 30 days. Alzheimer's disease-modification trials would need years of treatment at doses that may be higher than current safety data covers.
4. Do NAD+ precursors interact with anti-amyloid therapy? Lecanemab and donanemab are now standard-of-care for early Alzheimer's. The interaction between NAD+ precursors and these antibodies has not been studied. NR alters sirtuin and PARP activity; both pathways are downstream of the inflammatory cascade that anti-amyloid therapy modulates.
5. Is the right target NAD+ replenishment or PARP inhibition? If PARP overactivation drives the NAD+ deficit, a PARP1 inhibitor might be more directly therapeutic than a precursor. PARP inhibitors are FDA-approved oncology drugs and have been proposed for repositioning in neurodegeneration — none has reached an Alzheimer's efficacy trial.

What this means for patients, caregivers, and clinicians

Three takeaways are worth internalizing if you or a family member has been diagnosed with mild cognitive impairment or Alzheimer's disease and is considering NAD+ supplementation.

1. No NAD+ precursor is approved for Alzheimer's treatment in any jurisdiction. Cholinesterase inhibitors (donepezil, rivastigmine, galantamine), memantine, and the newer anti-amyloid antibodies (lecanemab, donanemab) remain the pillars of standard care. NR, NMN, and IV NAD+ exist in the supplement and wellness markets, but their supplement status is not a substitute for the disease-modifying evidence that approval would require.
2. Trial participation is the most defensible route to NAD+ exposure in this disease. If the biology interests you, the strongest move is to ask the treating neurologist about active Alzheimer's or MCI NAD+ trials. Trial participation includes monitoring, structured outcome measurement, and contributes to the evidence base. Self-supplementation does none of those things and may interfere with disease tracking.
3. Long-term safety in Alzheimer's populations is not established. The Wang Phase 1 trial established short-term safety. Long-term safety beyond 12-24 months in a cognitively vulnerable population, interactions with anti-amyloid antibody therapy, and effects on disease trajectory have not been studied. Decisions to chronically supplement should involve the clinician managing the patient's broader pharmacotherapy.

Adjacent NAD+ evidence in neurodegeneration

Reading the Alzheimer's NAD+ literature in isolation overstates the case. Placed alongside the broader neurodegenerative-NAD+ evidence base, the pattern is consistent: encouraging preclinical biology, biomarker-level target engagement in Phase 1 humans, and no Phase 3 efficacy in any indication.

• Parkinson's disease. NADPARK and NR-SAFE established brain NAD+ engagement and 12-month tolerability; N-DOSE is dose-finding. No Phase 3 efficacy data.
• Ataxia telangiectasia. Small open-label NR studies report cerebellar NAD+ elevation and modest ataxia score improvement; sample sizes are small and trials lack placebo control.
• Mild cognitive impairment. Multiple small NR and NMN trials show peripheral NAD+ elevation; cognitive endpoints have not consistently improved.
• Friedreich's ataxia. Mitochondrial dysfunction is central; NR pilots report acceptable tolerability; no efficacy data.
• Long COVID and post-viral fatigue. Mitochondrial dysfunction is a leading hypothesis; NAD+ precursor trials are early.

The Alzheimer's NAD+ story sits inside this broader picture rather than standing alone. Across the neurology literature, NAD+ precursors look like a class with consistent target engagement, mild biomarker effects, and no demonstrated disease modification.

Bottom line

Three things are simultaneously true about NAD+ and Alzheimer's disease in 2026.

First, the biology is real. NAD+ is depleted in the Alzheimer's brain. PARP1 overactivation, sirtuin failure, and mitophagy collapse are documented features of the disease, and each is a credible target for NAD+-based therapy. Mouse models — 3xTgAD, APP/PS1, and C. elegans amyloid lines — respond to NR or NMN with reduced pathology and improved cognition across at least three independent laboratories.

Second, the human evidence is thin. One published Phase 1 safety trial of NR in Alzheimer's patients (Wang et al., NCT03482167) established short-term tolerability and PBMC NAD+ elevation, not efficacy. Mild-cognitive-impairment Phase 2 trials have produced mixed cognitive readouts. Brain target engagement in Alzheimer's patients has not been directly measured by 31P-MRS in published work. No Phase 3 trial of any NAD+ precursor has reported in Alzheimer's, and none is currently registered in the late-phase pipeline.

Third, the historical base rate is unforgiving. Roughly 99% of Alzheimer's drug candidates fail in clinical trials. The two successes of the last decade — lecanemab and donanemab — produced modest effect sizes after extensive Phase 2 biomarker work that NAD+ precursors have not yet done. Treating NR or NMN as an evidence-based Alzheimer's therapy in 2026 is not what the clinical literature supports.

For patients today, the most useful action is to ask a neurologist about active Alzheimer's or MCI clinical trials and to keep standard-of-care therapy at the center of the treatment plan. For the field, the next milestones are a Phase 2 NR or NMN trial in MCI or prodromal Alzheimer's with plasma p-tau 217 and neurofilament light chain as biomarker endpoints, and direct 31P-MRS measurement of brain NAD+ engagement in Alzheimer's patients. Until that work is done, NAD+ precursors in Alzheimer's remain a promising mechanism in search of a clinical proof.