First report from Epiterna on the search for drugs that can extend human lifespan

Twitter/X Source:

Although most drugs currently approved are meant to treat specific diseases or symptoms, it has been hypothesized that some might bear a beneficial effect on lifespan in healthy older individuals, outside of their specific disease indication. Such drugs include, among others, metformin, SGLT2 inhibitors and rapamycin. The UK biobank has recorded prescription medication and mortality data for over 500’000 patients during more than 40 years. In this work we examined the impact of the top 406 prescribed medications on overall mortality rates within the general population of the UK. As expected, most drugs harbor a negative effect on lifespan, probably due to the underlying negative effect of the disease the drug is intended for. Importantly, a few drugs seem to have a beneficial effect on lifespan, including notably Sildenafil, Atorvastatin, Naproxen and Estradiol. These retrospective results warrant further investigation in randomized controlled trials.

2024.03.08.24303967v1.full.pdf (1.4 MB)


Do they plan to publish the whole dataset with the 400 tested and the 14 “winners”?

All SGLTi seem to extend lifespan according to them btw @Neo.

“ACEi (HR 1.11, CI 1.06-1.15) was associated with increased mortality.” :grimacing: (poke @desertshores, they looked at ramipril specifically) Did they also try ARBs? Telmisartan?

I’d like to see the results for lithium, GLP1RAs, and finasteride as well.


For Atorvastatin we observed a J-shaped dose-response effect, with no effect on mortality at 10mg (HR 0.96, CI 0.89-1.02), a reduction in mortality at 20mg (HR 0.87, CI 0.82-0.93), no effect at 40mg (HR 1.03, CI 0.95-1.11) and increased mortality at 80mg (HR 1.17, CI 1.03-1.33).

(ping @AnUser).


1 Like

It’s noteworthy that three out of five drugs that extend lifespan in males are immunomodulators: antibiotic, steroid, and nonsteroidal anti-inflammatory drug.

Atorvastatin appears to confer benefits for median lifespan initially, but these benefits gradually diminish over time.

1 Like

On SGLT2i:

Statins (HR 0.97, CI 0.94-1.00) and Estrogen (HR 0.76, CI 0.67-0.85), as a class, reduced mortality, so did SGLT2i (HR 0.64, CI 0.45-0.89), although with a much lower sample size.

They looked at “SGLT2 inhibitors (Canagliflozin, Empagliflozin, Dapagliflozin)”. In the UK here were the most prescribed SGLT2i from April 2021 to April 2022:

  • Empagliflozin: 2,268,049/y
  • Dapagliflozin: 1,928,883/y
  • Canagliflozin: 731,557/y

So we can assume that dapagliflozin and empagliflozin extend lifespan.


Also, among statins, only atorvastatin had a significant positive effect on mortality:

Importantly, we identified 14 drugs that increased lifespan, compared to health matched controls (Figure 2, Data Table 2), independently of current smoking, cancer diagnosis, diabetes, gender, and age at recruitment. These included notably the statin Atorvastatin (HR 0.91, CI 0.87-0.95), the PDE5 inhibitor Sildenafil (HR 0.85, CI 0.78-0.93), the anti-inflammatory drug Naproxen (HR 0.90, CI 0.85- 0.96), and the estrogen related drugs Estraderm (HR 0.67, CI 0.51-0.88), Vagifem (HR 0.73, CI 0.59-0.91), Estriol (HR 0.74, CI 0.60-0.92) and Estradiol (HR 0.75, CI 0.59-0.95). Others included, Lymecycline, Otomize, Marvelon, and 2 vaccines (Avaxim, Revaxis).

Statins as a class had a tiny positive effect: “Statins (HR 0.97, CI 0.94-1.00)”

Does it mean that some statins have a detrimental effect? We really need the whole dataset to conclude… But in the meantime, should people prefer atorvastatin to other statins?

I also find it interesting that after simvastatin failed the ITP, they decided to test for atorvastatin.

(and what are the 2 drugs not listed out of the 14?!)


That 80 mg atorvastatin increase mortality looks like reverse causation to me.
I guess this study shows which drugs have an outsized effect when prescribed to disease process relative to healthy population? SGLT2i? (Like the metformin story long ago, that those given metformin for diabetes lived longer than healthier controls - but was debunked later?).


Yes there might be confounders. For instance for viagra, I guess people who use it tend to be in a relationship, maybe a happy one, etc. They’re also probably not hospitalized.


No sure actually. Otherwise why would atorvastatin outperform while rosuvastatin and others failed?

1 Like

Actually, they did: Association between prescription drugs and all-cause mortality risk in the UK population | medRxiv

A few interesting things:

  • Telmisartan is the only sartan with a low HR (HR all: 0.86 CI 0.56–1.32, HR males: 0.76 CI 0.46–1.25), but it didn’t reach significance, probably because of the small sample size, as it’s rarely prescribed in the UK (n=306 vs 7860 for losartan, for instance).
  • Amlodipine is neutral (HR: 0.99, CI 0.95–1.04). In males, 5 mg (HR 0.98 CI 0.91–1.04) might be better than 10 mg (HR 1.06 CI 0.97–1.16). Lercanidipine is another interesting dihydropyridine CCB: HR 0.93 CI 0.77–1.12.
  • They don’t give data for each flozin alone; they only group them (with a small total sample size n=563).
  • Indapamide is interesting: HR 0.87, CI 0.756–0.999 but p=0.048 (poke @DrFraser you’ll like this). Contrary to thiazide: HR: 1.0. Indapamide 1.5 mg SR (HR 0.83 CI 0.66–1.04) might be a bit better than indapamide 2.5 mg (HR 0.89 CI 0.75–1.05). (Chlorthalidone is rarely prescribed in the UK)
  • Finasteride: HR male: 0.94 CI 0.84–1.05.
  • Pioglitazone: HR all: 0.91 CI 0.76–1.08 (30 mg is doing great in males: HR 0.74 CI 0.57–0.97).
  • Liraglutide is the only GLP1RA listed: HR 0.81 CI 0.56–1.18.
  • Lithium carbonate (but 400 mg!): HR 2.15 CI 1.22–3.80 :scream:
  • Sirolimus and other -rolimus drugs not listed.
  • All statins:
  • Ezetimibe: HR 1.11 CI 0.97–1.27 (similar HR among males only).
  • Bempedoic acid and PCSK9i: not listed.

This is great, I saw that they said “as a class”, did they also say “all”?

You’re right, they did not, I clarified that point after looking at the raw data: First report from Epiterna on the search for drugs that can extend human lifespan - #10 by adssx

Atorvastatin has already been proven to be superior to rosuvastatin in terms of side effects, lower diabetes risk and cataract surgery:

So this more info in favor of atorvastatin. But has atorvastatin reduced all-cause mortality, which rosuvastatin has?


I find it odd that Naproxen is on the list as I thought NSAIDs increased the risk of death.
“It is widely considered a strong painkiller and is part of the category of medicines known as non-steroidal anti-inflammatory drugs (NSAIDs).”

“Naproxen and other nonsteroidal anti-inflammatory drugs (NSAIDs) can increase the risk of heart attack and stroke”

“NSAID use causes an estimated 3,300 deaths and 41,000 hospitalizations each year among older adults.”
“NSAID use is associated with an increased risk of death after one year and five years.”


Yes, there are several surprising results.

For instance, they found HR 2.15 for lithium carbonate. And yet, another Swiss team looking at the same dataset (UK Biobank) found that “Subsequent multivariate survival analyses reveal lithium to be the strongest factor in regards to increased survival effects (hazard ratio = 0.274 [0.119–0.634 CI 95%, p = 0.0023]), corresponding to 3.641 times lower (95% CI 1.577–8.407) chances of dying at a given age for lithium users compared to users of other anti-psychotic drugs.” (Lithium treatment extends human lifespan: findings from the UK Biobank 2023)


I love that Nystatin made it in the statin’s list … this is actually an anti fungal for candida (usually use for thrush or vaginal candidiasis). Anyway, not a good one for aging apparently. But also not a cholesterol medicine.


It may be because dose for longevity is much lower than for psychiatric illnesses
Doses of lithium for longevity purposes are about 20 mg / day, doses for illness are about 1800 mg / day - 90x difference

1 Like

I’m sorry. It was my mistake. I simply filtered for “statin” in my spreadsheet. I’ll edit my message.

1 Like

This one SHOCKED me. I’ve always felt so guilty whenever I took one if I had a headache or felt sick. Now I guess I don’t feel so badly about it. It also reduces pain before working out.

Well I guess I might have to ask my doctor to switch me from 5mg Rosuvastatin to 10mg Atorvastatin now.

1 Like