Lithium levels in tap-water & suicide

These commentaries are based on Dr Gillman’s peer reviewed scientific papers, see Publications


There are many papers, over the last 15 years, reporting that violence and suicide is lower in areas where the natural lithium intake from municipal reticulated tap-water supplies is higher.

However, there is no credible evidence of a correlation between the very low levels of lithium ingested in tap-water and the level of lithium in people’s blood. Indeed, quite the reverse is the case, because serum lithium — like sodium — does not vary much with intake: they are both tightly regulated physiologically. Even if there was a relationship between normal low-level total dietary lithium intake, and serum levels, that would have little to do with tap-water, because most ingested lithium does not come from tap-water. Therefore, any attempt to correlate the lithium level in tap-water with the serum level, in people from different regions, with different water supplies, can be confidently predicted to be a failure. Research published so far ignores that lack of relationship and, with no rational basis, attempts to link tap-water directly to suicide: that is not only, a priori, misguided, but, worse than that, it is so utterly ill-conceived that it is a good example of the third-rate science that continues to erode confidence in the medical literature.


There have been repeated claims, for some years, that the suicide rate and violence is lower in areas where the natural lithium intake from municipal reticulated tap-water supplies is higher. There are a number of papers on this topic (1-12).

For reasons that will be mystifying to any scientist, none of the above papers have seen fit to measure the serum levels of lithium in subjects, in the different areas where they have assessed the rate of suicide, in order to establish that they are in fact different.That is the essential and obvious first step in the scientific investigation of such a question.Yet, none of these researchers have done anything except to measure the lithium level in the municipal water supply, they have not accounted for lithium from any other source.

The established facts about lithium levels detailed in this commentary indicate that speculative associative epidemiology, attempting to link tap-water with suicide and violence is, a priori, highly unlikely to be correct, and even less likely to reveal any actual cause-effect relationship.

Let us look at the evidence.

Endogenous serum lithium

The best replicated and latest data indicate that typical endogenous serum lithium levels are around 0.0003 mmol/L and that levels show low variation (~ 0.0001 -0.0005) in relation to typical dietary ingestion amounts (which are < 1 mg/day).

It may be noted that the measurement of nano-gram levels of endogenous lithium (and other trace metals) in biological tissues and fluids involves technical challenges and difficulties (13-16) and some reported measurements require replication before one can have confidence in them.

Lithium intakes above 1 mg per day probably overcome the physiological control mechanisms (see below) and lead to an increase in serum levels, exhibiting linear pharmaco-kinetics, as we see with therapeutic doses.

The most recent large samples (thousands of subjects) of endogenous serum lithium levels in humans are from Bochud et al. (17) and Seidlerova et al. (18).

The first step

The first logical step in any such research must necessarily be to demonstrate that a representative sample of people, ingesting water with a substantially different lithium concentration, actually do have serum lithium levels that are reliably and reproducibly different, compared to people in other areas where the water has a different lithium concentration. Simple.

Not to establish that before proceeding with detailed research, ignores logic and scientific methodology. Current evidence contradicts the proposition that tap-water is a determinant of total dietary lithium intake, and therefore, on the face of it (i.e. ‘a priori’) that makes a complete nonsense of all this kind of research.

Tap-water: less than one tenth of daily intake

The data indicating that tap-water constitutes less than one tenth of typical daily dietary lithium intake is clear and relevant, because these researchers have sought to link the varying amount of lithium in the municipal tap-water supply, in different districts, to suicide rates and violence etc. Our best current estimation of lithium intake, from the French ‘Total diet study’, is an average of 0.5 mg per day (19): the amount of tap-water which would have to be consumed, to reach just one 10th of that level, using the mean figure from a recent large survey of European water sources (20), would be 20 L per day (i.e. 0.002 mg/L x20 = 0.04 which is about 1/10 of the daily total of 0.5 mg). That is clearly absurd.That is a conservative estimate, for a large proportion of people in many areas where tap-water has even less lithium, the proportion of lithium ingested from tap-water would be more like 1% of the total lithium intake.

Municipal reticulated water supply (tap-water) and serum levels

The key step, missing in all the papers published so far, is to elucidate the relationship between tap-water and serum lithium. The data of Bochud et. al. (17), and Seidlerova et al. (18)has never been cited by any psychiatric publication on lithium. They both measured serum lithium concentrations & 24-hr. urinary excretion, in a couple of thousandsubjects, and also in the tap-water they consumed.

The average concentration of lithium in tap-water in the Belgium cohort (Bochud) was 0.01 mg/L (high for tap-water), & in South Africa was 0.00021 mg/L (pretty low) — a fifty-fold difference — the serum lithium levels in the two cohorts were identical. The mean 24-hour urinary lithium excretion was higher in the Belgian sample (0.055 mg vs 0.021 mg per 24 hours) and more variable.

Serum lithium levels were almost identical: Belgium 0.0022 mg/L vs. South Africa = 0.0023 mg/L. And that in relation to a fifty-fold difference in tap-water lithium concentration.

Seidlerova et al. (18) measured serum Lithium in more than one thousand subjects (from northern Belgium) with similar results — average Lithium level was 0.18 µmol/L (0.0013 mg/L), with low variance.

That demonstrates unequivocally that serum lithium is tightly regulated even when there are variations in dietary intake (at these trace-intake levels from natural dietary sources of < 1 mg/day). Lithium and sodium are both alkali metals and sodium is tightly regulated, in such a manner that increases in sodium intake do not cause much variation in the plasma level: it is hardly surprising that same is true for lithium.

Micro-doses of lithium

de Roos et al. (21) used a lithium ‘supplement’ as a marker in a diet study: their 78 subjects had baseline levels of 0.0064 mg/L on normal diet (somewhat high compared to other results herein, but in the right ball-park)and on the supplement of 1.75 mg lithium daily themean lithium level was 0.046 mg/L (steady state, 24 hrs. post-dose).

Likewise, with similar results, Donahoo et al. (22) also using a ‘supplement’ dose of 1.75 mg, in 7 subjects, found a serum level of 0.035 mg/L (9 hrs. post-dose).

Shiotsuki et al. (23) looked at serum lithium levels in 43 subjects visiting a health spa and found they were much elevated after drinking the lithium-rich spa-water, the estimated lithium intake in these people was between 10 mg and 70 mg over a few hours. Serum levels rose from 0.026 mmol/L (not a true ‘baseline’ measurement) to 0.073 mmol/L (0.5 mg/L), the therapeutic range is around 0.4 up to 1.0 mmol/L.

Nunes et al. (24) did a study over many months giving a dose of only 0.3 mg daily of lithium to patients with dementia. The information in this commentary reveals why this study was of doubtful utility: they are barely getting any more than a typical daily intake, which as we have seen above, does not increase lithium levels. Furthermore, they did not measure serum lithium levels. Therefore, just like all the silly studies on suicide and tap water, this study falls into the same trap. Future studies need to measure serum lithium levels and use doses that actually are demonstrated to elevate serum lithium above the tightly regulated ‘endogenous background level’.

The above results suggest that intakes above around 0.5 – 1 mg per day saturate the physiological control mechanisms regulating lithium and lead to a progressive increase in serum levels exhibiting near-linear pharmaco-kinetics, as we see with therapeutic doses.

At the 1000 times higher therapeutic intake levels, of ~ 1,000 mg/d, serum lithium is determined by glomerular filtration rate and alters in direct proportion to daily dose: so, if 500 mg gives a steady state level of 0.3 mmol/L then 1,000 mg will give 0.6 mmol/L.

Other results for ‘endogenous’ lithium levels — for the record, but all are smaller sample sizes compared to Bochud & Seidlerova — Miller found serum lithium levels around 0.00016 mmol/L for normal subjects dwelling in the Denver metropolitan area, and the mean 24-hr excretion rate was 0.005 mmol/day (25). Folkerd (26) healthy volunteers (n = 25), mean 0.00027 +/- 0.02 mmol/L (range 0.00013-0.00055 mmol/L). Lehmann, small sample, maximum of 0.06 mg/L (0.009 mmol/L) (27). However, older assays may be inaccurate.

Endogenous lithium levels and suicide

In Bayesian terms, the above indicates that the prior probability of there being any relationship between low levels of lithium ingestion and any human behaviour is exceedingly small. Strange and tortuous assumptions and arguments would need to be adduced for any such hypothesis to be entertained, even fleetingly, by the most simple-minded reductionist.

There have been repeated claims, over the last 15 years, that violence, and the suicide rate, is lower in areas where the natural lithium intake from municipal reticulated tap-water supplies is higher. There are now a large number of papers touching on this topic (1-12).

Note, the recent Knudsen (28) and Kessing (29) studies were negative.However, they involved a lot of time and resources which this analysis indicates were a completely pointless effort.

The ‘Kapusta’ paper

An examination of the paper by Kapusta et al. (5), published in the British J of Psychiatry, a prestigious journal, and a follow up effort (4), will illustrate the kind of nonsensical material that is being published, even in supposedly first division journals. Kapusta’s paper seeks to make a correlation between population suicide rates and the differing concentrations of lithium in tap-water between different districts in Austria. The majority of the sample reported were exposed to low levels of lithium in the tap-water in their districts, and all within a relatively narrow range of concentrations (0.002-0.01 mg/L): since the established range in fresh water is 0.000005 to 0.3 mg/L, that constitutes a small window into the possible range.

The most recent best estimate of dietary lithium intake is, as above, around 0.5 mg per day (19, 30, 31). Intake from tap-water must represent less than one tenth of total daily intake — if typical tap-water is 0.002mg/L then drinking 2L per day is barely 1% of the total 0.5 mg estimated total daily intake. Lithium in food probably contributes ten times as much as tap-water. Not only that, but many of the foods (especially sea-food) and vegetables, and particularly bottled water, in peoples’ diet have come from different districts and different countries: which indicates that the locality in which one lives has nothing to do with lithium intake.

The consumption of bottled spring-water supports a multi-billion-dollar industry and bottled waters have, on average, a five times higher lithium concentration: viz. median level of bottled waters is 0.010 mg/L vs. 002 mg/L for tap water (20). In Europe, the EU average consumption, for 2015, was 110 L per capita per year, Austrian data indicates around 90 l per year.

Source: European Federation of Bottled Waters.

Bearing in mind that children, and the financially less-advantaged, and geriatrics, are unlikely to drink mineral water (all this is so silly that it is difficult not to be facetious about it), we can estimate the typical adult is drinking about 150 L per year, or 3 L per week. That is equivalent to 15 L of ordinary tap water in terms of lithium content — based on the assumption from the data above, that the average difference between mineral water and tap water is five-fold.

That indicates a substantial proportion of Kapusta’s sample (or any other sample) will have been ingesting more lithium from bottled spring-water than from tap-water, certainly enough to make complete and utter nonsense of their data.

Lastly, Kapusta did not measure serum levels of lithium of a sample of people in areas exposed to different levels of lithium in their tap-water in order to demonstrate a correlation between the tap-water level and the serum level: that is the essential and simple requirement needed to establish a relationship between those two variables. Not to do that represents an epic fail — if there is no relationship between the lithium level in tap water in the different areas, and the average lithium level in the people living in those areas, then the principle behind such research is invalidated (cf. Bochud et al.), never mind the issue of lithium from other sources.

Remember, two huge studies of large samples ingesting water with a fifty-fold concentration difference found identical serum levels: they therefore concluded that, at those levels of intake, serum levels are tightly regulated and vary little with intake.

Without the correlation between serum levels and lithium concentration in tap-water (which the currently available data indicates is simply not present) it is obvious the paper is of no value. It is just silly and unscientific, I feel a sorry for them wasting their time. What is more of a concern is the ‘expert referees’ who decided to waste money by giving them a research grant, and also the journal’s referees who missed these obvious and major mistakes, which should have caused the paper to be rejected.

Lastly, in regard to not having accounted for the consumption of lithium either in bottled water, or in the diet, Kapusta et al. blithely state: ‘For obvious reasons, data for both of these factors are not available at aggregate levels; hence we were unable to consider these factors.’ That statement is glaringly incorrect: TDS1 had already been published (30) and estimates of bottled water sales/consumption are readily available. How facile is that? to so blithely dismiss a failure to address a key assumption at the heart of the paper. Perhaps we should not be surprised, since I have just noticed (in the small-print) that this gentleman comes from the Department of psychoanalysis at Vienna! Now then, which well-known fraud sprouted from there a hundred or so years ago?


Research attempting to link lithium in tap-water with suicide, violence etc. represents a wildly misguided misallocation of research resources.

Such studies do not deserve to be published, they devalue science, and they exemplify the truth of Ioannidis proposition: ‘Why most published research findings are false’ (32-34).

I hope the data and reasoning above persuade my readers that Kapusta et al. (and all the other silly papers seeking to establish such links) are quite obviously incorrect, and that all such similar efforts to correlate lithium in tap-water with suicide, violence, or in-growing toenails, are fundamentally and fatally flawed.

Please, stop wasting your time and our time.

As ever, caveat lector.


Before leaving this subject, it may be noted that the above criticisms do not preclude the possibility that at some point, at a very low level of lithium intake, the above evidence suggests at around 1 mg per day, the body’s ability to regulate lithium may be overwhelmed, and levels start to go up. At that point, one might suppose that levels would be proportional to dose, as they are with therapeutic levels, even though those are 500 – 1000 times higher.

Thus, it remains entirely possible that in areas with an unusually high lithium level, and some (especially bottled mineral) drinking water gets up to a few milligrams per litre, enough lithium may be ingested by some people to affect physiology, and even illness processes.

However, the kind of research above is not going to throw light on that. Someone needs to do endogenous serum lithium levels in populations where there are very large variations in the lithium concentrations in water (cf. Bochud), and look at how ‘micro-dose’ supplements of a few mg per day, affect serum lithium levels (see above).

Finally, remember that we still do not understand what the relationship is between serum lithium, and the level in various nervous tissues. Improved scanning techniques may allow accurate estimations in humans at these very low concentrations, and may inform us about the degree to which lithium is concentrated in different tissues and brain regions; cf. thyroid tissue — where it is concentrated quite considerably — caveat, most of these studies are old and un-replicated! Time for some real science to be done? Or am I getting over-optimistic and silly in my old age.


1.         Liaugaudaite, V, Mickuviene, N, Raskauskiene, N, Naginiene, R, et al., Lithium levels in the public drinking water supply and risk of suicide: a pilot study. J. Trace Elem. Med. Biol., 2017.

2.         Vita, A, De Peri, L, and Sacchetti, E, Lithium in drinking water and suicide prevention: a review of the evidence. Int. Clin. Psychopharmacol., 2015. 30(1): p. 1-5.

3.         Helbich, M, Leitner, M, and Kapusta, ND, Lithium in drinking water and suicide mortality: interplay with lithium prescriptions. Br J Psychiatry, 2015. 207(1): p. 64-71.

4.         Bluml, V, Regier, MD, HLAVIN, G, ROCKETT, IR, et al., Lithium in the public water supply and suicide mortality in Texas. J. Psychiatr. Res., 2013. 47(3): p. 407-411.

5.         Kapusta, ND, Mossaheb, N, Etzersdorfer, E, Hlavin, G, et al., Lithium in drinking water and suicide mortality. Br J Psychiatry, 2011. 198(5): p. 346-50.

6.         Kabacs, N, Memon, A, Obinwa, T, Stochl, J, et al., Lithium in drinking water and suicide rates across the East of England. Br J Psychiatry, 2011. 198(5): p. 406-7.

7.         Ohgami, H, Terao, T, Shiotsuki, I, Ishii, N, et al., Lithium levels in drinking water and risk of suicide. Br J Psychiatry, 2009. 194(5): p. 464-5; discussion 446.

8.         Pompili, M, Vichi, M, Dinelli, E, Pycha, R, et al., Relationships of local lithium concentrations in drinking water to regional suicide rates in Italy. World J Biol Psychiatry, 2015: p. 1-8.

9.         Kapusta, ND and Konig, D, Naturally occurring low-dose lithium in drinking water. J Clin Psychiatry, 2015. 76(3): p. e373-4.

10.        Giotakos, O, Tsouvelas, G, Nisianakis, P, Giakalou, V, et al., A negative association between lithium in drinking water and the incidences of homicides, in Greece. Biol. Trace Elem. Res., 2015. 164(2): p. 165-8.

11.        Dawson, EB, The relationship of tap water and physiological levels of lithium to mental hospital admission and homicide in Texas. Lithium in Biology and Medicine, ed. GN Schrauzer and KF Klippel. 1991, Weinheim: : VCH Verlag.

12.        Dawson, EB, Moore, TD, and McGanity, WJ, Relationship of lithium metabolism to mental hospital admission and homicide. Dis. Nerv. Syst., 1972. 33(8): p. 546-56.

13.        Subramanian, KS, Determination of metals in biofluids and tissues: sample preparation methods for atomic spectroscopic techniques. Spectrochimica Acta Part B: Atomic Spectroscopy, 1996. 51(3): p. 291-319.

14.        Dell’Osso, L, Del Grande, C, Gesi, C, Carmassi, C, et al., A new look at an old drug: neuroprotective effects and therapeutic potentials of lithium salts. Neuropsychiatr Dis Treat, 2016. 12: p. 1687-703.

15.        Clarke, WB, Guscott, R, Downing, RG, and Lindstrom, RM, Endogenous lithium and boron red cell-plasma ratios: normal subjects versus bipolar patients not on lithium therapy. Biol. Trace Elem. Res., 2004. 97(2): p. 105-16.

16.        Zhao, J, Gao, P, Wu, S, and Zhu, D, Determination of trace lithium in human urine by electrothermal atomic absorption spectrometry using nitric acid as a chemical modifier to eliminate the interference of chloride. Anal Sci, 2009. 25(5): p. 639-43.

17.        Bochud, M, Staessen, JA, Woodiwiss, A, Norton, G, et al., Context dependency of serum and urinary lithium: implications for measurement of proximal sodium reabsorption. Hypertension, 2007. 49(5): p. e34.

18.        Seidlerova, J, Staessen, JA, Maillard, M, Nawrot, T, et al., Association between arterial properties and renal sodium handling in a general population. Hypertension, 2006. 48(4): p. 609-15.

19.        Kalonji, E, Sirot, V, Noel, L, Guerin, T, et al., Nutritional Risk Assessment of Eleven Minerals and Trace Elements: Prevalence of Inadequate and Excessive Intakes from the Second French Total Diet Study. European Journal of Nutrition & Food Safety, 2015.

5(4): p. 281-296.

20.        Demetriades, A, Reimann, C, and Birke, M, European Ground Water Geochemistry Using Bottled Water as a Sampling Medium in Clean Soil and Safe Water. 2012, Springer Netherlands: Dordrecht. p. 115-139.

21.        de Roos, NM, de Vries, JH, and Katan, MB, Serum lithium as a compliance marker for food and supplement intake. The American journal of clinical nutrition, 2001. 73(1): p. 75-79.

22.        Donahoo, W, Bessesen, D, Higbee, D, Lei, S, et al., Serum lithium concentration can be used to assess dietary compliance in adults. The Journal of nutrition, 2004. 134(11): p. 3133-3136.

23.        Shiotsuki, I, Terao, T, Ogami, H, Ishii, N, et al., Drinking spring water and lithium absorption: a preliminary study. German Journal of Psychiatry, 2008. 11: p. 103-106.

24.        Nunes, MA, Viel, TA, and Buck, HS, Microdose lithium treatment stabilized cognitive impairment in patients with Alzheimers disease. Curr Alzheimer Res, 2012.

25.        Miller, NL, Durr, JA, and Alfrey, AC, Measurement of endogenous lithium levels in serum and urine by electrothermal atomic absorption spectrometry: a method with potential clinical applications. Anal. Biochem., 1989. 182(2): p. 245-9.

26.        Folkerd, E, Singer, DR, Cappuccio, FP, Markandu, ND, et al., Clearance of endogenous lithium in humans: altered dietary salt intake and comparison with exogenous lithium clearance. Am. J. Physiol., 1995. 268(4 Pt 2): p. F718-22.

27.        Lehmann, K, Endogenous lithium levels. Pharmacopsychiatry, 1994. 27(3): p. 130-2.

28.        Knudsen, NN, Schullehner, J, Hansen, B, Jørgensen, LF, et al., Lithium in Drinking Water and Incidence of Suicide: A Nationwide Individual-Level Cohort Study with 22 Years of Follow-Up. International Journal of Environmental Research and Public Health, 2017. 14(6): p. 627.

29.        Kessing, LV, Vradi, E, and Andersen, PK, Nationwide and population-based prescription patterns in bipolar disorder. Bipolar Disord, 2016. 18(2): p. 174-82.

30.        Leblanc, J-C, Dietary exposure estimates of 18 elements from the 1st French Total Diet Study. Food additives and contaminants 22.7: 624-641. 2005.

31.        Anon, Second French Total Diet Study (TDS 2): Report 1 Inorganic contaminants, minerals, persistent organic pollutants, mycotoxins and phytoestrogens. 2012: p. 

32.        Ioannidis, J, Lies, Damned Lies, and Medical Science. Atlantic, 2010. November 17th.

33.        Ioannidis, JP, Why most published research findings are false. PLoS Med, 2005. 2(8): p. e124.

34.        Ioannidis, JPA, The Reproducibility Wars: Successful, Unsuccessful, Uninterpretable, Exact, Conceptual, Triangulated, Contested Replication. Clin. Chem., 2017. 63(5): p. 943-945.

Consider Donating to PsychoTropical

PsychoTropical is funded solely through generous donations, which has enabled extensive development and improvement of all associated activities. Many people who follow the advice on the website will save enormously on doctors, treatment costs, hospitalization, etc. which in some cases will amount to many thousands of dollars, even tens of thousands — never mind all the reduction in suffering and the resultant destruction of family, work, social, and leisure capability. A donation of $100, or $500, is little compared to those savings. Some less-advantaged people feel that the little they can give is so small it won’t make a difference – but five dollars monthly helps: so, do not think that a little donation is not useful.

– Dr Ken Gillman

Dr Ken Gillman