Selegiline Revalidated as a Late-Life Longevity Intervention

A single-author meta-analysis from Johns Hopkins pooled 22 rodent lifespan experiments (spanning 27 years, 6 countries, 4 species) on the FDA-approved MAO-B inhibitor L-deprenyl (selegiline). The drug produced a statistically significant, moderate increase in average lifespan (standardized mean difference 0.68, p = 0.0002). Effects were larger at higher doses and when treatment started in older animals. Critically, the benefit was not an artifact of unusually short-lived control animals. However, a re-analysis of the one dog study showed the survival benefit vanished once age at enrollment was accounted for, underscoring that translation to humans remains unproven.

L-deprenyl, sold as selegiline, has spent three decades as a bit-player in longevity science: an FDA-approved Parkinson’s and depression drug that a handful of labs kept quietly showing could make old rodents live longer. The problem was that the field never trusted the result. The studies were old, scattered across continents and decades, and geroscience has learned the hard way that “lifespan extension” often evaporates on closer inspection.

This new analysis, published in the high-impact journal Ageing Research Reviews, is the most rigorous attempt yet to settle the question. Michael Bene of Johns Hopkins gathered every credible rodent survival experiment on the drug and ran them through a modern statistical meta-analysis. The verdict: the effect is real, reproducible, and of moderate size. Across mice, rats and hamsters, treated animals reliably out-survived their controls, with an effect size (SMD 0.68) that sits comfortably between the noise of a null result and the strong signals seen with heavyweight interventions like caloric restriction and rapamycin.

The Big Idea is not just “a drug works.” It is that L-deprenyl clears the bars that usually trip up longevity claims. There was no detectable publication bias. And, most importantly, the effect was not driven by the field’s most notorious confounder: sickly, short-lived control animals that make any intervention look good by comparison. When Bene tested for this, he found the opposite pattern - the benefit was, if anything, larger in studies where controls lived longer.

Two findings sharpen the translational angle. Effects grew stronger with higher doses and, strikingly, with older age at treatment onset - hinting that this is a drug you might start late in life rather than in youth. That fits the geroscience dream of a pill for people who are already old.

The cautionary counterweight is the dogs. A 1997 study had reported that L-deprenyl prolonged life in elderly dogs. Bene re-analyzed the raw survival data and found that once you adjust for the fact that treated and untreated dogs entered the study at very different ages, the “benefit” was no longer statistically significant (hazard ratio 0.38, but confidence interval crossing 1.0). The trend still favored the drug, and the study was badly underpowered - but it is a reminder that a robust rodent signal is not a human promise. No human longevity trial has ever been run. This paper makes the strongest case yet that one should be.

Impact Evaluation:


8 Likes

Was it anywhere apparent the best starting age for humans? (or translated in human years/age). I could only see in old age and that can be anywhere from 60-100.

2 Likes

Thanks for this. I’ve been taking generic Selegiline since my early fifties ~ 25 years. My dose today is 5 mg every other day. Unfortunately I have just taken it orally - there are no dosing instructions on the packet. I shall try putting it under my tongue as I do with Modafinil.

2 Likes

What’s the paper title or the link?

2 Likes

Fixed it: L-deprenyl extends lifespan across mammalian species: A meta-analysis of 22 longevity experiments - PubMed

More on the paper:

Study Design Specifications

  • Type: Random-Effects Meta-Analysis of 22 longevity experiments (15 included in final quantitative synthesis).
  • Subjects:
    • Species: Mouse, Rat, Hamster.
    • Sex: Heavily skewed Male (N=1220 combined) vs. Female (N=263 combined).
    • Re-analysis: Domestic Dogs (Beagles) from Ruehl et al., 1997.

Mechanistic Deep Dive

The analysis highlights three primary potential mechanisms:

  1. Dopaminergic Preservation: L-deprenyl protects dopaminergic neurons, which are uniquely sensitive to aging. This may involve upregulation of antioxidant enzymes (SOD, Catalase) in the striatum.
  2. Catecholaminergic Activity Enhancer (CAE): The “Knoll Hypothesis” suggests the drug acts as a synthetic enhancer of catecholamine release. However, the meta-regression favoring higher doses contradicts the micro-dose CAE theory.
  3. Hormetic Stress/Immunomodulation: The drug may modulate interferons, TNF-alpha, and interleukins, suggesting a systemic immunomodulatory effect.

Novelty

  • Late-Life Efficacy: This is the most distinct finding. Unlike interventions that require early adoption, L-deprenyl efficacy scaled positively with older age at initiation (p < 0.0001).
  • Dog Study Debunking: The authors obtained raw data from the famous Ruehl et al. (1997) dog study. When adjusted for the fact that the treated dogs were slightly younger at enrollment, the statistical significance of survival vanished (p = 0.1534), although the Hazard Ratio remained favorable at 0.38.

The Translational Protocol

1. Human Equivalent Dose (HED)
The meta-analysis identifies the effective rodent dose range as 0.25 to 0.5 mg/kg.

  • Source Conversion: The study explicitly states these doses “are roughly equivalent to a 1.5–3 mg dose for a 70 kg human”.
  • Comparison: Standard tablet size for Parkinson’s is 1.25 mg to 5 mg. This aligns perfectly with the effective longevity window described in the text.
  • Reference: FDA Guidance on Dose Conversion

2. Pharmacokinetics (PK/PD)

  • Bioavailability: Oral bioavailability is impacted by a “substantial first pass” effect.
  • Metabolites: Selegiline metabolizes into L-methamphetamine and L-amphetamine. While L-isomers are less centrally potent than D-isomers, they are vasoactive.
    • Biohack: Transdermal patches (6-12 mg) or Sublingual administration bypass the gut/liver. This reduces metabolite toxicity while maintaining MAO-B inhibition.
  • Reference: PubChem: Selegiline Pharmacokinetics

3. Safety & Toxicity Check

  • The Cheese Effect (Tyramine): At doses < 10 mg/day, Selegiline is selective for MAO-B. At higher doses (>10mg), it loses selectivity, risking hypertensive crisis if aged cheese/wine is consumed.
  • Interactions: Strict contraindication with SSRIs, SNRIs, and Tricyclics due to serotonin syndrome risk.
  • Reference: DrugBank: Selegiline Interactions & Toxicity

4. Biomarker Verification Panel

  • Efficacy Marker: Phenylethylamine (PEA). Selegiline prevents the breakdown of PEA. Elevated urinary or plasma PEA confirms MAO-B inhibition.
  • Efficacy Marker: Prolactin. Dopamine inhibits Prolactin. Successful dopaminergic modulation should result in low-normal Prolactin levels.
  • Safety Monitoring: Blood Pressure and Liver Panel (ALT/AST).

5. Feasibility & ROI

  • Sourcing: Prescription only (USA) (FDA Approved).
  • Cost: Generic Selegiline (5mg) is inexpensive, approx $14-$20/month via discount pharmacies.
  • Cost-Benefit: High. The study concludes L-deprenyl is a “promising candidate” with “abundant safety data”.

Part 4: The Strategic FAQ

1. If the dog study failed significance, is the translational potential dead?
No. The Hazard Ratio in the dog study was 0.38. This means the treated dogs had a 62% reduction in mortality risk at any given time, but the study only had ~15 dogs per group. It was a statistical power failure; an N of 35 per group would be required to detect the effect with 80% power.

2. Will I test positive for Methamphetamine on a drug test?
Yes. Selegiline metabolizes into L-methamphetamine. Standard urine screens do not distinguish between the L-isomer (medicine) and D-isomer (abuse). You must request a Chiral Chromatography confirmation test.

3. Why does the study suggest “Starting Late” is better?
The data showed a significant p-value (<0.0001) for age at initiation. The authors speculate that treatment in later life exhibits “particularly beneficial effects”, likely because it counteracts age-related increases in MAO-B activity.

4. Can I take this if I am on SSRIs?
No. The combination carries a high risk of Serotonin Syndrome.

5. What is the difference between Selegiline (L-deprenyl) and Rasagiline?
Rasagiline is a newer MAO-B inhibitor that does not metabolize into amphetamines. However, the vast majority of longevity data is specific to L-deprenyl.

6. Does the “High Dose” recommendation in the study risk the “Cheese Effect”?
The study supports doses up to 0.5 mg/kg in rats. The text states this equates to 1.5–3 mg in humans. This is safely under the 10 mg threshold where MAO selectivity is typically lost.

7. Why is the data for females so poor?
The authors note that less than 25% of the total sample size was female. In the subgroup analysis, the effect size for females was 0.26 (non-significant) vs 0.84 for males (significant). It is unknown if this is a biological lack of effect or simply a lack of statistical power.

8. Is the oral tablet the best delivery method?
The study noted that an oral study using 1 mg/kg failed to show benefits, likely due to the “substantial first pass” effect. Sublingual or transdermal methods bypass this first-pass metabolism.

9. Can I combine this with Rapamycin?
Data is absent. However, Rapamycin mimics nutrient scarcity, while Selegiline promotes dopaminergic drive. No direct contraindications exist in literature, but monitor blood pressure closely.

10. What specific product was used in these studies?
The meta-analysis covered 23 experiments using L-deprenyl (Selegiline). The text highlights that it is the “same” L-deprenyl approved for Parkinson’s and depression.

3 Likes

Who do people recommend as a source of Selegiline. My rapamycin source does not do Selegiline.

1 Like

All the Indian pharmacies often discussed here have it in 5mg tabs for dirt cheap.

2 Likes

@Cole , please add the following output constraints to your prompts, so that it removes all the random formatting characters ($, etc.) from the output:

Output Constraints:

  • Use Markdown formatting.
  • Do not use LaTeX, python code, or special characters that break simple text parsers or reveal formatting codes, etc…
1 Like

My quality provider/mfgr prompt gives these two companies/products:

[Selegiline]

  • Selgin by Intas Pharmaceuticals Ltd. - 5mg, 10mg
  • Elegelin by Sun Pharmaceutical Industries Ltd. - 5mg
1 Like

Knoll’s command didn’t abandoned researches of selegilin. And BPAP (a derivative of selegiline, which is not MAO-B inhibitor at all, but 160 times more potent catecholamin enhanser, which is very interesting).
The idea of Knoll was that amphetamine-like catecholamin action is to release a lot of products of neuromediators even without signal. But CAE, catecholamin enhancer effect is to enlarge the release when we have signal, but do nothing if there is no signal at all. So, stimulation only when we have a stimul.
There are several recent studies of deprenyl, and BPAP. And Rasagiline, which is MAO-B inhibitor, but not CAE. So only deprenyl is MAO-Bi AND CAE.

Analysis of my AI follows:

User question

MAO-B itself may also affect longevity. It seems to reduce inflammation, and maybe there are other mechanisms too. In that case, BPAP may be better as an enhancer, but worse than deprenyl, because the MAO-inhibition logic is absent. That could explain why its longevity effect drops off. So deprenyl may turn out to be the best of both worlds.


Assistant answer

Yes, I would formulate it exactly like that: your model is very plausible.

BPAP may be “better” as a pure enhancer, but worse as a geroprotector, because the MAO-B component is cut off. And deprenyl/selegiline is a hybrid that has both layers:

  1. MAO-B inhibition / PEA-dopamine metabolism / anti-inflammatory-neuroprotective axis
  2. CAE / TAAR1 / impulse-dependent monoamine release enhancer axis

And this is precisely why it may win on lifespan, even if BPAP is pharmacologically more “elegant” in the enhancer sense.

1. Two different logics

Logic A: MAO-B inhibition

This is not just “less dopamine breakdown.”

With aging, MAO-B is connected to several things:

  • age-related increase in MAO-B activity;
  • more oxidative monoamine metabolism;
  • Hâ‚‚Oâ‚‚/aldehyde byproducts;
  • astrocyte/glial axis;
  • neuroinflammation;
  • dopamine/PEA tone;
  • possible effects on mood, motivation, and cognitive energy.

In Bene’s meta-analysis, the mechanistic hypotheses for deprenyl include: MAO-B inhibition opposing the age-related rise in MAO-B, SOD/catalase, and immunomodulation through interferons, TNF-α, interleukins, BDNF, and NK cells. So this is not only “dopamine stimulation.”

There is also a newer independent line: deprenyl reduced the inflammatory profile in an acute SIV infection model, and the authors explicitly described broad anti-inflammatory effects of deprenyl. This is not longevity evidence, but it supports the idea that the MAO-B/MAOI axis may be immunometabolic, not merely psychopharmacological.

Logic B: CAE/enhancer

This is the Knoll line:

the compound does not crudely dump monoamines, but enhances impulse-propagation-mediated release — meaning release during a normal nerve impulse.

BPAP is stronger in this logic: in older structure-activity work, BPAP enhanced impulse-mediated catecholamine/serotonin release, and in the shuttle-box model it was roughly 130 times more potent than deprenyl.

But that does not mean “better for longevity.” It means: better as a probe/tool of enhancer regulation.

2. Why BPAP may have underperformed in longevity

Because it probably has only one layer:

BPAP = enhancer without MAO-B inhibition.

Whereas deprenyl is:

deprenyl = enhancer + MAO-B inhibition + metabolite/PEA/dopamine axis + possible anti-inflammatory/antioxidant/immunomodulatory effects.

So “the best of both worlds” is a good formulation.

If deprenyl’s lifespan effect is composed of several components, then BPAP, even if stronger as a CAE, may fail to reproduce the whole phenotype.

The recent picture supports this:

  • BPAP is mechanistically strong as an enhancer;
  • but in aged Long-Evans rats in 2023, it did not significantly improve survival or cognition;
  • deprenyl, meanwhile, shows an overall rodent lifespan signal in meta-analysis.

So mechanistic purity does not guarantee a lifespan effect.

3. Why rasagiline also loses here

Rasagiline is the reverse situation.

It preserves:

MAO-B inhibition

but, according to the Harsing/Miklya school, it loses:

the enhancer / TAAR1 / impulse-dependent release component.

The 2023 paper directly concludes: selegiline acts as a MAO-B inhibitor and dopaminergic enhancer drug, whereas rasagiline is devoid of enhancer activity and may act as a TAAR1 antagonist.

So we get a very clean matrix:

Drug MAO-B CAE/enhancer Longevity meaning
BPAP no strong pure enhancer, but may lack the MAO-B/anti-inflammatory layer
Rasagiline yes no / possibly TAAR1 antagonism pure MAO-B, but not a Knoll-style enhancer
Selegiline / deprenyl yes yes hybrid: potentially the complete geroprotective package

4. What may be the “magic” of deprenyl

I would break it down like this:

1. MAO-B/PEA axis

MAO-B destroys β-phenylethylamine very quickly. Deprenyl sharply increases PEA availability. And PEA itself is linked to trace amine / TAAR1 / motivational tone.

So MAO-B inhibition may enhance endogenous enhancer tone, not merely slow dopamine breakdown.

2. Dopamine preservation

With aging, the dopaminergic system deteriorates. Deprenyl may:

  • reduce dopamine catabolism;
  • reduce toxic byproducts;
  • support nigrostriatal function;
  • improve drive/movement/exploration.

In animals, this may translate into lifespan through activity, feeding, stress response, behavior, and thermoregulation.

3. Glia/neuroinflammation

MAO-B is especially interesting in the glial/astrocytic aging axis. If part of brain aging is astrocytic MAO-B / GABA / inflammation / oxidative metabolism, then a pure enhancer like BPAP is not necessarily expected to fix that.

4. Immune modulation

The Kitani line explicitly discusses interferons, TNF-α, interleukins, BDNF, and NK cells. This may be a peripheral lifespan component that BPAP does not cover.

5. CAE as a “functional amplifier”

And the enhancer component adds on top:

  • better impulse-dependent dopamine/noradrenaline/serotonin release;
  • without crude amphetamine-like flooding;
  • possibly supporting youthful monoaminergic responsiveness.

5. But there is an important trap

One cannot conclude from this:

therefore, the more MAO-B inhibition + the more CAE, the better.

Rather:

there is a window.

In the meta-analysis, two negative studies were not included in the quantitative pooling, and the author specifically notes that they may have been negative because of early initiation and/or high systemic dose: Gallagher used 0.5 mg/kg SC starting at 3 months; Carrillo used 1 mg/kg SC.

So deprenyl as “the best of both worlds” only works in a narrow sense:

low/moderate dose, no crude stimulation, not lifelong high-dose exposure from youth, and no serotonergic/MAOI interaction risks.

6. My current hierarchy after all these data

For the longevity/brain-aging hypothesis:

1st place: selegiline/deprenyl

Because:

  • it has MAO-B inhibition;
  • it has CAE/enhancer activity;
  • it has a rodent lifespan meta-analysis;
  • it is a clinical drug;
  • it has human safety history.

2nd place: rasagiline

Because:

  • clinically clean;
  • has MAO-B inhibition;
  • has a Parkinson’s delayed-start story;
  • but lacks the Knoll-style enhancer layer.

3rd place: BPAP

Because:

  • the enhancer mechanism is elegant;
  • potency is high;
  • but it lacks MAO-B inhibition;
  • has no clinical pathway;
  • the recent aged-rat result is sobering;
  • human safety is unclear.

Bottom line

Yes: deprenyl may be better than BPAP precisely because it is not “pure.”

BPAP is a beautiful tool for CAE.
Rasagiline is a clean MAO-B inhibitor.
But deprenyl is a mixed drug that simultaneously:

  • inhibits MAO-B;
  • supports PEA/dopamine;
  • provides an enhancer effect;
  • may reduce neuroinflammatory/oxidative/immunosenescence burden.

And if the lifespan effect is the sum of these axes, then deprenyl really does look like “the best of both worlds,” not like an outdated compromise.

3 Likes

And I found maybe the 3rd mechanism of action of Selegiline.

Selegiline may have a third, MAO-B-independent neuroprotective mechanism

A recent study identified protein disulfide isomerase (PDI) as a new direct target of selegiline:

https://link.springer.com/article/10.1186/s43556-025-00255-w

The authors found that selegiline covalently binds PDI, particularly near its catalytic site, reduces its enzymatic activity, and inhibits PDI-driven mitochondrial outer membrane permeabilization and cytochrome-c release. In a dopaminergic cell model, selegiline also reduced apoptosis induced by ER stress.

This potentially gives selegiline three distinct neuroprotective/geroprotective mechanisms:

  1. MAO-B inhibition
    Reduces dopamine and phenethylamine degradation, opposes the age-related increase in MAO-B activity, and may reduce MAO-B-associated oxidative stress and neuroinflammation.

  2. CAE/TAAR1 enhancer activity
    Enhances impulse-dependent monoamine release rather than causing non-physiological transmitter dumping. This is the Knoll “catecholaminergic activity enhancer” mechanism.

  3. PDI/ER–mitochondrial apoptosis modulation
    May protect stressed neurons by limiting PDI-mediated mitochondrial membrane permeabilization, cytochrome-c release, and apoptosis.

This also makes the comparison with related compounds more interesting:

  • BPAP is probably a stronger and cleaner CAE/enhancer, but it does not inhibit MAO-B and therefore lacks the MAO-B/PEA/anti-inflammatory component.
  • Rasagiline is a potent MAO-B inhibitor, but according to the Knoll/Miklya research line it lacks selegiline’s enhancer activity and may even antagonize TAAR1-related effects.
  • Selegiline may therefore combine both systems—and possibly a third proteostasis/ER-stress mechanism—making it a broader geroprotective candidate than either BPAP or rasagiline alone.

The major caveat is concentration: the PDI experiments used micromolar-to-millimolar selegiline concentrations, far above typical human plasma levels. Therefore, the PDI mechanism is currently a compelling mechanistic hypothesis, not proof that low clinical doses modify PDI in the human brain.

2 Likes