Increased DNA repair/genomic stability: A possible longevity mechanism of lithium

With a handful of people on this site reporting that they are taking lithium for longevity (and also quite a few people on Reddit taking it for biohacking or longevity purposes), it’s something I’ve started looking into myself.

I wanted to call attention to some papers supporting a novel longevity mechanism of lithium: genomic stability. The first two papers show directly an effect of lithium on genomic stability, although they don’t provide evidence that this occurs through the lithium target GSK3β. The next three papers show that GSK3β inhibition may increase genomic stability, and although they don’t use lithium, they provide indirect evidence that lithium will do the same. In the third part of this post I’ll try to provide some evidence that subclinical lithium doses can effectively inhibit GSK3β.

Onto the papers…

  1. Direct Effects of Lithium on Genomic Stability

Screen Shot 2024-01-02 at 5.35.40 PM
In primary rat retinal neurocytes, serum deprivation led to an increased number of phosphorylated H2AX-positive cells, while pretreatment with 1mM LiCl (0.16mM lithium) reduced the number of number of γH2AX-positive cells. As γH2AX is generally considered to represent unrepaired double-strand breaks/DSBs, this data suggests that lithium increases the efficiency of DSB repair.

The primary pathway of DSB repair is nonhomologous end-joining/NHEJ, and in the serum deprived cells, lithium gave a 3x efficiency increase in the NHEJ assay.

Lithium led to a 3x increase in LIG4 protein levels (this is a DNA ligase which helps to reconnect the separated DNA strands in the NHEJ pathway) and this was proposed as the mechanism leading to increased DNA stability.

A second study showed that a week of pretreatment with 3mM lithium in HT-22 cells (mouse hippocampal cell line) could reduce comet tail formation in response to radiation, which is a powerful inducer of DSBs. These comet tails are formed in response to DNA damage, with the ‘mean tail moment’ reflecting the degree of damage. In this same cell line, lithium enhanced repair efficiency in the NHEJ assay, and they found this NHEJ enhancement to require phosphorylation of Thr2609 on DNA-PK.

They then showed that similar lithium pretreatment reduced the persistence of γH2AX foci, in either HT-22 cells or primary neurons exposed to radiation. Additionally, in the primary neurons, lithium reduced the number of γH2AX foci per cell. Similar activity was demonstrated in-vivo, as 40mg/kg ip for 7 days reduced the number of γH2AX-positive hippocampal cells in irradiated mice. This study found no effect of lithium on homologous recombination, an important DSB repair pathway in dividing cells.

  1. Direct Role of GSK3β in Genomic Stability

We present evidence to suggest that the stimulation of A2AR markedly facilitated DNA repair through the TRAX/DISC1/GSK3β complex in a rat neuronal cell line (PC12), primary mouse neurons, and human medium spiny neurons derived from iPSCs. A2AR stimulation led to the inhibition of GSK3β, thus dissociating the TRAX/DISC1/GSK3β complex and facilitating the non-homologous end-joining pathway (NHEJ) by enhancing the activation of a DNA-dependent protein kinase via phosphorylation at Thr2609. Similarly, pharmacological inhibition of GSK3β by SB216763 also facilitated the TRAX-mediated repair of oxidative DNA damage

Specifically, this study showed that inhibition of GSK3β accelerated double strand-break (DSB) repair efficiency in irradiated mouse hippocampal neurons, as assessed by the neutral comet assay. This coincided with attenuation of IR-induced γ-H2AX foci, a well characterized in situ marker of DSBs. To confirm the effect of GSK3 activity on the efficacy of DSB repair, we further demonstrated that biochemical or genetic inhibition of GSK3 activity resulted in enhanced capacity in nonhomologous end-joining–mediated repair of DSBs in hippocampal neurons. Importantly, none of these effects were observed in malignant glioma cells

Here, we showed that TRAX participates in the ATM/H2AX-mediated DNA repair machinery by interacting with ATM and stabilizing the MRN complex at double-strand breaks.

  1. Potency of Lithium’s GSK3β inhibition/Intracellular Lithium Concentrations Achieved with sub-mM Serum Levels

One of lithium’s primary targets is GSK3β, where lithium’s Ki is 2mM. It’s also worth knowing that GSK3β is constitutively active, so its activity is regulated by inhibitory phosphorylations and/or trapping within protein complexes.

As lithium appears to inhibit GSK3β via competition with the magnesium co-factor binding site, it’s been argued that lithium’s in-vivo GSK3β IC50 is 0.8mM or less.

Let’s say someone taking daily lithium for longevity has serum levels of 0.05mM, is this a dose that will actually inhibit GSK3β? We can look to a study in rats, which showed that with chronic dosing, lithium concentrates intracellularly. After 3 weeks of daily administration, intracellular levels in the brain were over 4x higher than CSF or serum levels. [ref]

Screen Shot 2024-01-02 at 10.24.32 PM

Let’s assume then that with chronic dosing, a serum level of 0.05mM corresponds to concentration of 0.2mM within brain cells. This would correspond to 20% inhibition of GSK3β, assuming a GSK3β IC50 of 0.8mM. [ref]

The actual level of GSK3β inhibition produced by 0.2mM lithium might be greater than 20% though, see for example

The lithium-induced increase in serine9-phosphorylation of GSK3β may be critical because it provides a mechanism whereby a moderate direct inhibitory effect of lithium achieved by therapeutically relevant lithium concentrations is amplified to achieve more pronounced inhibition of GSK3β by the therapeutic level of lithium [ref]

Moreover, the act of serum deprivation in the above study rapidly leads to loss of GSK3β-ser9 phosphorylation, and if you recall the first paper at the top of this post (lithium both reduces γH2AX-positive cells and enhances NHEJ assay efficiency in serum deprived media), this further supports the idea that lithium activates DNA repair pathways via GSK3β.


Interesting post. It will take me a bit of time to digest it all.

Just an FYI for people, there have been some discussions in the past focused on lithium for longevity:

Here: A primer on Lithium, Lithium + Rapamycin, etc. potentially more stable and safer pharmacokinetics

Here: Lithium Supplementation

Here: Lithium Orotate

Here: Rubidium increases lifespan in a way similar to lithium


Thank you for this post. However, in the end the question to be anawered is whether to dose with lithium and what dosage to take. I concluded that targeting 50 micro molar was a good idea and after testing have settled on 2mg per day to hit this.

I will read the referenced papers with interest.


Oddly enough higher levels of lithium inhibit some of the citrate transporters:


Was about to start 5 mg orotate i.e. the lower level of the dose mainly taken by biohackers and, longevity enthusiasts when I noted your 2 mg dosing. Is there anything you could clarify on the latter apart from your crunching of data and your own lithium level - could 2 mg be much better thsn 5 mg? I note concern about inhibition of citrate transporters.

1 Like

I could not find any research on this. Hence i tried various doses and tested the outcomes. 1mg was too low and 5mg too high. Other people may not be the same.

1 Like

Could you say in what way the outcome of 5 mg made you decide it was too high?

I did a lithium serum test after a few weeks of shifting to 5mg per day (I can tell you how many weeks if you ask me to check, but it was enough for a steady state) and the serum concentration was 0.1 mmol/L. 2 mg per day was under 0.05mmol/L. Lithium is something I want to keep really low, but around 0.05mmol/L. 1mg was also under 0.05mmoll/L. Although you cannot assume a linear relationship between supplementation and serum levels I would think at low doses it would be quite likely. Perhaps 3mg/d would be better, but I am happy with my 2.

I really don’t want to inhibit citrate transporters although the mechanism that gets citrate into most cells may not be part of the SLC13 family. There is, however, a very small amount of SLC13 expression in lots of cells. Hence I am erring on the side of safety.

In the broader sense I do think my protocol works. However, because it has lots of parts and lithium is only one I have only a limited time to do lithium testing. When I started with Lithium it had some clear positive effects, but I am going by the published research on this which points to a target of 50 microMolar.


I’m less bullish on low dose lithium after I came across this 2021 study on mice

Plasma levels over 400 micromolar. I aim to be below 50. There will be a citrate transport inhibition effect.

I find that Lithium supplementation at 1 mg daily improves my mood. Unfortunately, I can’t quantify feeling happier.

1 Like

Would’ve been nice to see them run even lower doses. 5mg orotate/day, gives 0.0027mg elemental lithium per kg bodyweight per day (assuming a 80kg human).

At the lowest dose tested for longevity, those mice are getting 0.02g LiCl per kg of diet. Let’s say a male C57BL/6J mouse weighs 35g and eats 6g food/day. The food consumption is an approximation based on similarly sized mice of a different strain and experimental condition, but it’s probably close enough.

  • On the 0.02g LiCl/kg diet, this comes to 0.56mg elemental lithium per kg bodyweight per day.

  • For the 0.5g LiCl/kg diet, this comes to 14mg elemental lithium per kg bodyweight per day.

  • For the 1g lithium carbonate/kg diet, this comes to 16mg elemental lithium per kg bodyweight per day.

The lowest of these doses (0.02g LiCl/kg diet), using the standard conversion, gives a human equivalent dose of 0.046mg elemental lithium per kg bodyweight per day. It would require 85mg of lithium orotate daily for a 80kg human to supplement this much.

Also, the study showed that mice are developing kidney disease at blood lithium concentrations which are well-tolerated by humans, so the mice may be more sensitive to its toxic effects.

Part of the toxicity of higher lithium doses is probably oxidative stress, and concentrations of 800 micromolar or higher have been shown to increase mitochondrial ROS production and activate the DNA damage response in proliferating primary human chondrocytes.

This increase ROS production might directly activate NRF2 pathway’s redox sensors (that would be my guess at least), and NRF2 expression is necessary for lithium’s lifespan increase in drosophila. So even if lithium is increasing DSB repair efficiency, this may be outweighed by unchecked oxidative stress caused by global NRF2 knockdown in presence of increased ROS production.


My mistake. I don’t understand how to convert rodent doses into humans and didn’t realize the doses were so high. Hard to differentiate in the studies whether they used prescription level doses vs supplemental doses sometimes.

The study reveals the serum levels which are in the range of a mental health dose. Far too high for longevity and likely to inhibit the citrate transporter.

1 Like

C57Bl/6J mice fed with 1.05–2.79 g/kg LiCL in the diet showed lithium plasma levels between 0.4 and 0.8 mM/l.

The highest LiCl dose they used for the lifespan studies was 0.5g/kg diet, which should produce plasma levels <0.4mM. It’s too bad they don’t provide plasma levels at the lower dose of 0.02g/kg.

True. I think citrate transporter inhibition is probably the cause of the kidney issues.

1 Like

I believe, based on the literature I have read, that a small or microdose is optimal. This is one supplement where it is easy to take too much and get negative effects. Better to err on the small side for dosing for this one.

1 Like

Does anyone in the US do bloodwork for lithium? If so what test?

For instance, there seems to be a “cheap” one at Quest, but then there is another expensive one that they call Lithium (RBC)

@John_Hemming can I ask what is your testing protocol is? Do you take the last dose the day before?

I take Lithium in the morning. When I am testing for serum Lithium I take lithium on that day as well. I don’t think it metabolises in any way that affects the test result. Hence it is merely a question as to a steady state in terms of input and excretion (I assume mainly through urine).

Perhaps at say noon having taken the supplement at 9am it might be slightly higher than average. I would not think, however, it would make that much difference. Not that I have tested this.

1 Like

Thx. As in you test an RBC version or a normal (us cost seems to be 25 dollars vs 150 dollars for the RBC version…)