Last month, Anna Fels wrote in the Sunday NYTimes suggesting that lithium be added to the drinking water because trace amounts of lithium are associated with lower rates of mental illness, violence and suicide in particular. What she didn’t say was that communities with naturally-occurring lithium in their drinking water enjoy greater longevity as well. Though the evidence is still thin, there is a credible mechanism, through inhibition of a chemical signal called GSK-3β. The dosages we’re talking about are ½ to 2 mg per day, which is less than 1% of the dosage typically prescribed for bipolar disease.
Lithium is not a drug, but a chemical element. It is right above sodium in the periodic table, and the theory is that the body can’t tell the two apart. Sodium is an essential electrolyte, and the body uses it everywhere. Nerve signals are propagated as waves of substitution of sodium for potassium. The theory is that trace amounts of lithium in the body replace sodium, and move a little faster across membranes, because the ions are smaller and lighter.
The only study looking at all-cause mortality compared different regions of Japan. Drinking water was analyzed and mortality rates were compared in 18 different towns, 1.2 million people in all.
The top graph shows about 10% decrease in mortality in regions of Japan that had naturally high concentrations of lithium. Large dots correspond to counties with large populations. Note the log scale on the X axis; the high counties had about 30 times as much lithium in their drinking water compared to the low counties. 10% decrease in mortality corresponds to about an extra year of life, extrapolating boldly.
The bottom graph is a survival curve for lab worms with lithium added to their medium, compared to control worms without.
The most abundant and compelling data is about suicide and violent crime rates in towns with varying amounts of natural lithium in the water. Twenty different studies in different parts of the world are summarized here. This same review covers four studies in which lithium was used to treat Alzheimer’s Disease, and all four found benefits, measured by cognitve performance.
Lithium in the diet
There is general agreement that animal grazers don’t concentrate lithium, and the best dietary sources are plants. Which plants are best? The answer seems to depend on lithium in the soil at the site where the plants happen to be grown. We can’t characterize some foods as “good sources of lithium” except according to the location in which they were grown. So it is impractical to figure out how much lithium you’re getting in your diet. A blood test might be helpful, with the qualification that lithium passes quickly through the body (swept through with sodium), so the test is sensitive to what you happen to have eaten in the last 24 hours.
How does it work?
Of course, we don’t know how lithium works, but speculation all centers around a powerful and ubiquitous (if little-known) chemical signal called GSK-3β. Lithium suppresses the action of GSK-3β. All that I know about GSK-3 (alpha and beta forms) I learned from this article by James P Watson (not to be confused with the Watson who got the Nobel for DNA).
The alpha and beta forms of GSK-3 are distinct proteins, derived from separate genes. The nomenclature would lead you to believe that they are different forms of the same thing, but the nomenclature is deceptive. I’m not going to tell you what GSK stands for, except that the last word is kinase, because the name reflects only the particular circumstances in which GSK happened to have been discovered (in 1980).
If “energy is the currency of the body,” then phosphate groups are the dollar bills. Kinases are enzymes that activate other chemicals by attaching a phosphate group to them. The GSK-3s are kinase amplifiers. They specialize in finding placs where there is already a single phosphate, then adding more to fully activate the substrate.
GSK-3β has been implicated in the formation of amyloid plaques in the brain, which is one theory for the cause of Alzheimer’s Disease. This was the motivation for trials of lithium against AD, which have shown early signs of success.
The two forms of GSK-3 are involved in many different processes, turning on dozens of genes and turning off dozens more. The actions are varied and complex, and GSK-3 cannot be characterized as helpful or harmful. Watson refers to GSK-3α as the “mainly good guy” and GSK-3β as the “mainly bad guy”. GSK-3β expression increases with age, and it may play a predominantly pro-aging role. Suppressing it a bit seems to do some good. GSK-3α does not modulate consistently with age, either up or down. However, when the GSK-3α gene is knocked down in mice, the mice are not only viable but have lower fat mass and increased insulin sensitivity—harbingers of extended life. Watson characterizes GSK-3α nevertheless as “mainly good” based on its effect dampening three aging targets: mTOR, Wnt and P53. P53 controls apoptosis=cell suicide which, I have argued, is on a hair trigger as we age. Dialing down apoptosis non-selectively has general benefits for preserving muscle and nerve cells, but it also increases cancer risk. TOR is “Target Of Rapamycin”. Recall that rapamycin is a drug that made headlines three years ago when it was found to extend mouse life span, though fed to the mice late in life. Rapamycin is a powerful immune suppressor, and may be too dangerous a drug for general human use, but other drugs that dampen TOR signaling may be promising.
Speculation on the future
I am convinced that both forms of GSK-3 are key players in aging and other metabolic functions. However, they may not be good targets for anti-aging intervention because they play on both teams (promoting pro- and anti-aging pathways). The fact that they need “priming”—a first phosphorylation by a more specific kinase—indicates that they are not an upstream source of aging. We can find better targets.
Trace doses of lithium might prove to be useful for modestly extending life and protecting against Alzheimer’s. I’d like to see studies in mice, and larger epidemiological studies, and then clinical trials if warranted. There has been only one epidemiological study. It should not be difficult to find some part of the world where people don’t tend to move very often and then correlate local concentrations of lithium in drinking water with local variations in age at death.
For early adopters
If you want to experiment with lithium, I think that doses up to 1mg/day are safe. Just for context, this is a tiny quantity. One ounce of lithium is a lifetime supply. Larger doses are not better. They are toxic.
Lithium is not a standard ingredient in multi-mineral pills. Nor is it a standard blood test.
Lithium carbonate is available as a prescription drug, in pills that cannot easily be divided into 100 doses. If you felt comfortable with the procedure, I suppose you could dissolve a 500mg pill in a pint of water and take it by the teaspoon. There are about 100 tsps in a pint. Lithium carbonate is about ⅙ lithium. You can also get lithium salts from chemical supply houses.
Google to find on-line sources of supplements with appropriate low doses of organic salts of lithium, like lithium arginate and lithium orotate. They are not better or more bio-available than the simple carbonate, just more expensive.
Thanks for Michael Eaves for pointing me to sources for this article.