This article was prepared with AI assistance and has not yet been reviewed by a qualified medical or scientific professional. It is for general information, not medical advice.

The short answer: In a randomized study published July 21, 2026, the NAD+ precursor nicotinamide riboside substantially increased blood NAD+ in older adults with mild cognitive impairment. It did not significantly improve the prespecified cognitive outcomes over 12 weeks compared with placebo. The result does not prove NAD+ biology is irrelevant to the brain; it shows why making a biomarker rise and making a person function better are separate achievements.

There is something wonderfully straightforward about a blood test: one number goes up, another goes down, and a complicated story seems to resolve into a line on a chart. Longevity science is especially vulnerable to that comfort. A molecule associated with cellular energy becomes a product; a product changes the molecule; and the change is quietly recast as rejuvenation.

A randomized 2026 trial interrupts this smooth story at precisely the point a person—not a molecule—enters the picture.

The result in one sentence

Researchers led by Christopher Martens reported a 12-week, double-blind, randomized placebo-controlled pilot trial of nicotinamide riboside (NR) in people aged 65 or older with amnestic mild cognitive impairment. Participants had measurable memory impairment without dementia. The NR group received 500 mg twice daily.

Of 52 participants randomized, 42 completed the study: 22 in the NR group and 20 in the placebo group. Measurements of the blood NAD+ metabolome were available for 19 NR participants and 17 placebo participants. The treatment approximately doubled blood NAD+ concentrations in measured participants receiving NR; the corresponding placebo measurements did not show a similar rise.

The study's principal clinical question was whether cognition improved. The researchers found no statistically significant between-group improvement in the primary cognitive measures over the intervention period. They also did not find a statistically significant between-group improvement in whole-brain cerebral blood flow or blood pressure, both measured as secondary outcomes. Some exploratory analyses suggested changes in blood flow in particular brain regions, including the hippocampus, but these findings need confirmation and do not establish a memory benefit.

The biochemical door opened. The clinical door did not—not in these measured outcomes, not over these twelve weeks.

Why NAD+ attracted so much attention

Nicotinamide adenine dinucleotide, or NAD+, is involved in fundamental energy-transfer reactions and is used by enzymes implicated in cellular stress responses and DNA repair. Scientists have studied whether altered NAD+ metabolism contributes to functional decline in aging tissues and whether precursor supplementation can modify disease-related processes.

NR is one precursor used to make NAD+. It is taken orally, unlike intravenous NAD+ administration, and it has its own absorption and metabolism. A result obtained with oral NR capsules should not be relabeled as evidence that NAD+ infusions, injections, NMN, or an unrelated longevity stack improve memory. Shared metabolic vocabulary does not make delivery methods interchangeable.

The early research proposition was plausible: improve NAD+ availability; potentially influence energy metabolism or vascular function; possibly preserve cognition. But a plausible sequence still contains several links that need to be demonstrated individually.

The two clocks in this experiment

A molecule can change within weeks. The progression of cognitive impairment can take years. That mismatch creates a genuine scientific challenge.

Twelve weeks may be adequate to detect biochemical changes and some short-term symptoms. It is unlikely, by itself, to answer whether a compound prevents progression to Alzheimer's disease years later. Yet it is also too generous to say the null clinical result means nothing: the trial looked for cognitive benefit over 12 weeks and did not find a statistically significant advantage over placebo.

The useful interpretation is narrower than both marketing extremes. NR changed the blood biomarker. A corresponding cognitive improvement was not demonstrated on the measured timescale. Whether a longer exposure, another population, or a different outcome would change the answer remains an open research question, not an established promise.

What the exploratory brain scans do—and do not—add

The researchers used MRI methods to estimate cerebral blood flow. Whole-brain results did not show a significant treatment benefit. Exploratory analyses suggested a difference in the left hippocampus, a structure involved in memory. None of the regional findings would have survived correction for multiple comparisons.

It would be a mistake to dismiss the region finding automatically. Biological interventions sometimes have local effects that are diluted in whole-organ averages. It would be equally mistaken to turn an exploratory image into evidence that memory improved. Exploratory analyses are especially vulnerable to chance findings when many regions or comparisons are examined.

The study also reported a relatively large, though non-significant, pattern for one delayed-recall subtest. A probability value near a conventional threshold does not turn a result into proof. The planned outcome, totality of tested endpoints, and confidence in replication matter more than selecting the most favorable row.

Why the study is stronger than a before-and-after testimonial

Random allocation and placebo comparison help distinguish a treatment effect from practice on memory tests, expectation and variation over time. The investigators found that both groups improved modestly on some tests, a pattern compatible with becoming familiar with the testing procedure. Without a comparison group, those changes could easily have been credited to the supplement.

The trial was still small. Its analysis included the 42 completers rather than all 52 randomized participants, adding uncertainty from missing follow-up data. That sample also offers limited power to detect modest effects and limited ability to identify uncommon harms. The investigators reported no serious adverse effects and broadly similar adverse-event profiles over 12 weeks, but that does not establish long-term safety. Capsules and placebo were provided by Niagen Biosciences (formerly ChromaDex); the published paper also lists federal research funding and states that the company had no other involvement. Readers should evaluate both design and relevant material support.

The study was conducted in people with mild cognitive impairment. Its results should not be transplanted without qualification to healthy young adults, people with diagnosed Alzheimer's dementia, or anyone receiving injected NAD+.

What should a reader watch next?

Better trials would recruit enough participants to detect clinically important differences, measure changes in daily cognition and function, track disease progression over longer periods, and test whether an increase in blood NAD+ predicts a useful response. Measurements in blood should not be assumed to reflect equal changes in relevant brain cells. Replication of the hippocampal signal would be interesting, especially if paired with functional improvement.

Until then, a doubled laboratory number should be interpreted as target engagement—not as two extra years of youth or a reversal of brain aging.

The trial's most important contribution is not that it disproves the entire NAD+ field. It is that it separates a result researchers can measure today from the patient benefit they are still trying to establish.

Study-at-a-glance

Feature	Verified study detail
Publication	Martens et al., Alzheimer's & Dementia, first online July 21, 2026
Design	Randomized, double-blind, placebo-controlled pilot
Population	Older adults with amnestic mild cognitive impairment
Size	52 randomized; 42 completed
Duration	12 weeks
Biomarker	Blood NAD+ rose approximately twofold on NR in measured samples
Primary cognitive outcome	No significant between-group improvement
Interpretation	Biochemical effect demonstrated; clinical cognitive benefit not established

Sources

• Martens et al., Alzheimer's & Dementia (2026), doi:10.1002/alz.71605 — primary human study; methods, results, limitations and funding disclosures.
• PubMed record, PMID 42478598.
• ClinicalTrials.gov, NCT03482167 — registered protocol and outcome framework.

Editorial note: This is research reporting, not a recommendation to treat memory impairment with NR or to substitute supplements for medical evaluation.