Explorative research project · 2020–2022

The study that could not answer its original question.

We wanted to know whether longer-term NMN use would be reflected in a change in epigenetically estimated age. The overall dataset showed no discernible average effect. At the same time, methodological limitations and confounding factors identified later prevented a robust answer to the question of efficacy.

Go to the result Sources and review status

The research question

Can NMN change measured epigenetic age?

The starting point was a simple but demanding question: after several months of NMN use, would two epiAge measurements show a change beyond the expected progression over time?

The test available at the time estimated epigenetic age from DNA methylation patterns in a saliva sample. Such a value is a biomarker: a computational approximation of biological processes — not the ageing process itself and not a direct measure of health, lifespan or rejuvenation.

What was actually done

An explorative before-and-after observation — not a clinical efficacy study.

Participants were self-selected customers. They purchased a discounted study bundle and provided two test results together with information on use and accompanying circumstances.

Baseline measurement

First saliva sample, epiAge result and a questionnaire on previous NMN use, dosage, other supplements, medical conditions and subjective observations.

NMN use

The bundle included two containers of 50 g NMN each. The daily dose and route of administration were not standardised in advance but recorded afterwards.

Second measurement

The second test was to take place after three months, or after six months if preferred. Interruptions, other supplements and subjective changes were recorded again.

Statistical analysis

Predicted and measured epigenetic age were compared in a weighted linear model. An uncertainty of approximately 2,8 years was taken into account for the epiAge measurement.

Methodological classification

Observational · self-selected sample · no control group · no randomisation · no blinding · no standardised daily dose

The result

On average, there was no deviation from the expected course.

An initial analysis in May 2022 included 18 complete cases and yielded a = 1,001 ± 0,049. The final internal analysis in November 2022 had 26 complete cases available.

The weighted linear model yielded a = 0,999 ± 0,024. In this model, a factor of 1 corresponds to no average difference between the predicted and measured second epiAge value.

The available data provided no evidence that NMN reduced measured epigenetic age on average.

A confounding factor identified later

COVID was a methodological problem — but not the only limitation.

Data collection took place during the COVID-19 pandemic. The original questionnaires did not systematically record either SARS-CoV-2 infections or vaccinations. The first internal statistical analysis in May 2022 therefore already noted that their possible influence needed to be investigated.

Later research showed that infections and associated changes in the immune system can affect estimates from epigenetic clocks. Short-term changes in individual clocks were also observed after mRNA vaccination — though not consistently in the direction of ‘acceleration’. Such shifts may reflect, among other things, changes in immune-cell composition.

For our small dataset, this meant that a possible NMN effect, an infection, an immune response, changes in cell composition and normal measurement variability could not reliably be disentangled retrospectively.

Limits of the evidence

Why the study could not provide a robust answer to its original question.

01

Small, self-selected sample

Only 26 cases were fully evaluable. Participants were customers who had chosen to take part in the project.
02

No control group

Without a placebo or comparison group, changes over time cannot be unambiguously attributed to NMN use.
03

Variable conditions of use

Dosage, route of administration, duration of use, interruptions and other supplements were not standardised.
04

Unrecorded pandemic-related events

Infections and vaccinations were not variables systematically collected in the original design.
05

A measurement is not ageing itself

An epigenetic clock condenses methylation patterns into an estimate. Short-term biological changes can affect the value without proving a corresponding change in the organism as a whole.
06

Analysis not fully reproduced from raw data

The internal final reports were reviewed. However, the calculation could not be independently reproduced from the complete anonymised raw-data table because that table was not found in the project archive available today.

History and funding

A research question gave rise to the Anti-Aging-Shop.

At the time, NMN and, in particular, two epiAge tests per person were so expensive that a larger study could not be offered entirely free of charge using private funds. The AFEGA Anti-Aging-Shop developed out of the attempt to fund the research project and offer participants access at a reduced price.

Two research contracts totalling 50.000 Euro were awarded for the preparation and conduct of the study.

In 2020 alone, revenue from the AFEGA Anti-Aging-Shop made it possible to provide more than 60.000 Euro for research and product development by the AFEGA Research Group. The shop was therefore not merely intended as a source of funding; it actually enabled a substantial amount of research work.

What remains scientifically

No evidence of efficacy. No disproof. But real lessons learned.

A biomarker is not the process

A precise number can convey more certainty than the study design and biological circumstances actually support.

Confounding factors must be recorded from the outset

What is not collected systematically often cannot be reliably controlled for later — even if its importance only becomes apparent after the study has begun.

Unconfirmed expectations are part of research

Scientific integrity also means fully disclosing a disappointing result or one that cannot be interpreted robustly.

The study did not become meaningless to us. It became an example of how real-world conditions can prevent scientific questions from being answered — and why precisely this failure should be documented.

Sources and review status

What this project record is based on.

Questionnaires and test reports containing personal data were not published. Only anonymised, aggregated information was used for this page.

Internal primary documents

Historical participation and recruitment page, 2021
Questionnaire, Part 1 and Part 2
Shipping and participation information for the study bundle
Statistical analyses dated 13 May, 4 June and 19 November 2022
Research contract, Part I and Part II

Review status: documents inspected. The complete anonymised raw-data table for the 26 cases was not available.

Scientific assessment

The project text at the time described the undertaking as the first European human study of NMN. This historical priority claim has not been independently verified and is therefore not repeated here as a fact.