Lithium, used widely to treat bipolar disorder, does double duty as a neuroprotective and neurotrophic agent. Low levels of lithium, found naturally in drinking water, seem to ward off psychiatric disease, but whether they do the same for dementia is controversial. In a new study of groundwater lithium concentrations and dementia incidence across the U.S., William Parker and colleagues at the University of Chicago find that, at first blush, high lithium levels do accompany a lower incidence of dementia. However, the association disappears when the researchers control for demographic factors like age, ethnicity, and access to health care. Their results appear in a letter in the May 23 JAMA Psychiatry.

  • Epidemiological data questioned if low levels of lithium in tap water reduce dementia incidence.
  • In the U.S., the link vanishes when scientists control for differences in health care access and demographics.
  • Lithium in drinking water is unlikely to affect dementia risk.

The results run contrary to a large study linking high lithium concentrations with lower dementia incidence in Denmark (Aug 2017 news), and a more recently reported inverse relationship between lithium levels and AD-related mortality in Texas (Fajardo et al., 2018). Importantly, those studies did not control for the variables the Parker group did.

In the new study, the investigators collated data on lithium in groundwater from 3,000 drinking water wells in 174 counties across the U.S., sampled between 1992 and 2003. They tallied diagnoses of dementia in the same counties using insurance claim data covering 4.2 million adults. In 32 high-lithium counties, where the mean lithium exposure reached 141 μg/L, dementia prevalence was significantly lower compared to low-lithium counties, where the mean exposure was 6 μg/L. The same was true for bipolar disorder, previously linked to groundwater lithium concentrations. The association also held for three “negative control” conditions with no known link to lithium in drinking water, namely major depressive disorder, myocardial infarction, and prostate cancer.

However, the researchers found significant differences between the high- and low-lithium counties in terms of health care resources. The high-lithium areas had fewer hospital beds, primary care physicians, and psychiatrists per capita, a disparity that could result in lower rates of diagnosis of dementia. In addition, the residents of the high-lithium counties were younger, less educated, and more were Hispanic. When the investigators adjusted for those differences, the association between lithium and dementia, or any of the other diseases, disappeared.

“This indicates the purported association of high lithium concentrations in drinking water with mental health disorders is driven by unaccounted variation in demographics, health care resources, and diagnostic practices,” the authors concluded.

At therapeutic doses, lithium potently inhibits glycogen synthase kinase-3, decreases tau phosphorylation, lowers amyloid deposition, and improves cognition in animal models of AD. In a clinical study, lithium slowed cognitive decline and reduced levels of CSF tau in people with mild cognitive impairment (Forlenza et al., 2011). But as the authors point out, people in the highest lithium counties would have to drink more than 1,000 liters of water every day to ingest the 150 mg dose that elicited those effects.

Still, the debate goes on. Studies employing microdoses of lithium, equivalent to the amount present in just a few liters of water, or novel formulations, hint at benefits of sub-therapeutic doses in people (Nunes et al., 2012) and in animal models (Nunes et al., 2015; Wilson et al., 2017Habib et al., 2017). 

How low is too low? Jun Tan and Roland Shytle, University of South Florida, Tampa, said they agreed with the authors that concentrations of lithium in drinking water were unlikely to benefit those suffering with dementia. Nonetheless, they wrote in an email to Alzforum, “These findings should in no way reduce the enthusiasm for recent research exploring the possible therapeutic benefits of lithium … at broader and safer dose ranges lower than those already well-established for the treatment of bipolar disorder.”—Pat McCaffrey


  1. This study is both interesting and important for the reasons outlined in the paper. We agree with the authors that the therapeutic doses of lithium traditionally used in medicine are “orders of magnitude higher” than what is found to occur naturally in drinking water. For example, it’s well established that chronic lithium therapy is effective in stabilizing mood and reducing the risk of suicide in the treatment of bipolar disorder. Unfortunately, the high doses of lithium needed to achieve these therapeutic effects require routine monitoring of lithium blood levels in order to minimize the risks of adverse side effects.

    However, a growing body of recent preclinical and clinical research suggests that prophylactic and/or acute therapeutic benefits of lithium may be possible for a variety of other medical conditions, including dementia, at better-tolerated doses of lithium lower than those typically used in the treatment of bipolar disorder. Our own recent preclinical studies with a novel cocrystal of lithium support this hypothesis (Habib et al., 2017; Smith et al., 2013​). 

    So while we agree that the very low concentrations of lithium in the drinking water characterized in this study are unlikely to produce therapeutic benefits for those suffering with dementia, these findings should in no way reduce the enthusiasm for recent research exploring the possible therapeutic benefits of lithium for a variety of other medical conditions at broader and safer dose ranges lower than those already well-established for the treatment of bipolar disorder. 


    . LISPRO mitigates β-amyloid and associated pathologies in Alzheimer's mice. Cell Death Dis. 2017 Jun 15;8(6):e2880. PubMed.

    . Improving lithium therapeutics by crystal engineering of novel ionic cocrystals. Mol Pharm. 2013 Dec 2;10(12):4728-38. Epub 2013 Nov 18 PubMed.

  2. This work is very interesting and brings counterpoints to Kessing and colleagues (2017). Despite the well-designed statistical analysis, some points caught my attention. One is why the authors choose as negative controls major depressive disorder, myocardial infarction, and prostate cancer. They may be unrelated to groundwater lithium, but there are many works in the literature associating the use of lithium to treat these diseases and we do not know how the lifetime consumption of groundwater lithium would affect that.

    Also, Alzheimer’s disease patients could be separated from dementia patients, since the described effects of microdose lithium treatment is associated with AD, and there are many other diseases that lead to dementia. Another remarkable point is the significantly lower number of physicians, especially psychiatrists, in the high lithium cities. Is this a reflection of the lower incidence of these diseases?

    I would have liked to see some explanation about how all the variations and co-variations could interfere in the final results. A useful strategy would be measuring lithium levels in the blood taken from the high- and low-level lithium-treated populations using inductively coupled plasma mass spectrometry to detect lithium trace concentrations.

    Finally, in the last sentence the authors state that “the high-lithium group was exposed to a mean of 141.3 μg/L in their water supply. This means that a patient would need to drink more than 1,000 L of water a day to ingest the lowest reported effective therapeutic lithium dose of 150 mg.” The cited reference states that 300 μg/day of the metal lithium was used, equivalent to 1.5 mg of lithium carbonate. In this way, ingestion of 2 L of high lithium water could give the approximated amount of lithium used in the work of Nunes and colleagues (2013). This is a realistic chance.  


    . Association of Lithium in Drinking Water With the Incidence of Dementia. JAMA Psychiatry. 2017 Oct 1;74(10):1005-1010. PubMed.

    . Microdose lithium treatment stabilized cognitive impairment in patients with Alzheimer�s disease. Curr Alzheimer Res. 2012 Jun 29; PubMed.

  3. This is a very interesting research letter. Past papers on the subject have described a putative inverse association with dementia as well as extreme manifestations of possible psychiatric illness, such as suicide and homicide, and other, subtler psychiatric symptomatology; but this is a complicated subject with many variables, and sporadic cases of dementia may have a long prodromal phase. The Parker research letter showed raw data with differences in disease rates in association with variances in lithium levels in groundwater, including dementia and bipolar disorder, between the low-lithium and high-lithium groups. The authors attributed the differences in rates for dementia and bipolar disorder to demographic factors, including age, ethnicity, and education levels; and access to health care, with major depressive disorder rates, along with myocardial infarction and prostate cancer rates, used as negative controls to adjust for possible differences in health care access between the low-lithium and high-lithium groups. The authors’ conclusion that groundwater lithium does not confer significant benefit for bipolar disorder or dementia was reached after statistical adjustments described above were performed on the raw data.

    It should be noted that lithium is used clinically as adjuvant therapy for major depressive disorder, and the practice has been well characterized (Geier, 2012) at dosing levels that can be similar to levels used for bipolar disorder. Lithium can be used effectively as monotherapy for major depressive disorder (see Sadock and Sadock, Kaplan & Sadock's Concise Textbook of Clinical Psychiatry. Wolters Kluwer/Lippincott Williams & Wilkins, 2008: 515), but the side effect profile of lithium can make SSRIs a better option. The inclusion of major depressive disorder in the Parker research as a negative control for calculating adjustments, when lithium levels in the groundwater could have influenced the rates of major depressive disorder in the study population, may invite concerns that the disease rate variances that were measured and calculated based upon the negative controls between the low-lithium and high-lithium groups were possibly impacted by lithium levels in the groundwater; and this may have potentially contributed to an overadjustment of the data.


    . Lithium as augmentation for major depressive disorder. Mental Health Clinician: July 2012, Vol. 2, No. 1, pp. 15-17.

  4. This study makes use of the relatively wide variation of groundwater lithium in the U.S. to demonstrate no correlation between levels and incidence of dementia or bipolar disorder. Their last paragraph makes the clear statement that although Li+ levels in drinking water vary, they do not come close to acknowledged therapeutic doses and likely have little to no biological effect, at least with respect to psychiatric or neurodegenerative disorders. Indeed, they might almost be considered on a homeopathic spectrum. While lithium is commonly prescribed for BPD, the jury is still out as to whether such therapeutic levels, which have recognized side effects, are protective against neurodegenerative disease.

  5. The conclusions in this comprehensive study fit with the consensus of comments on the Danish study, that the amount of lithium in drinking water was orders of magnitude below the effective concentration for treating bipolar disorder, inhibiting known lithium target enzymes, or observing effects in animal models of Alzheimer’s disease. I wonder if this suggests something else in the drinking water in Denmark has a protective effect?

  6. As the authors write in their discussion, “Therapeutic lithium doses are orders of magnitude larger than groundwater lithium concentrations, making a true causal relationship between groundwater lithium and mental health biologically dubious.”

    We made the same point on Alzforum about the previous study (Kessing et al, 2017). Quote from online:

    “Assuming that individuals in the study drank up to three liters of water per day, the maximum intake would be about 90 µg per day from water sources. However, the intake of lithium from food sources was reported to be 0.6 to 3.1 milligrams per day in the United States in 1985 (Schrauzer, 2002). Therefore, the contribution of water-sourced lithium to the total daily intake of lithium would be marginal. This observation alone argues that the protective or hazardous effects associated with lithium intake from water sources are unlikely to be mediated by the pharmacological effect of lithium itself, but rather, mediated by some other agent in the water that co-enriches with lithium or by some other regional environmental factor.”

    It seems like our caveat has been substantiated by this current study, and is common sense.​


    . Association of Lithium in Drinking Water With the Incidence of Dementia. JAMA Psychiatry. 2017 Oct 1;74(10):1005-1010. PubMed.

    . Lithium: occurrence, dietary intakes, nutritional essentiality. J Am Coll Nutr. 2002 Feb;21(1):14-21. PubMed.

  7. Studies which have shown beneficial effects of lithium in preclinical models of Alzheimer's disease have been conducted using doses of lithium leading to plasma concentrations far higher than would be expected from continuous ingestion of water (containing trace levels of lithium). It is highly unlikely that even continuous exposure of these trace levels of lithium over many years would lead to the same pharmacological effects of lithium when present at plasma levels similar to those detected in patients administered regular doses of lithium. Our work in preclinical models of familial Alzheimer’s disease has demonstrated a novel mechanism by which lithium can improve cognitive function and that is through enhancing brain clearance of β-amyloid via increasing blood-brain barrier efflux and CSF bulk flow of this protein. Our studies were performed using doses of lithium which lead to plasma concentrations similar to those observed in patients administered lithium for bipolar disorder. Any effect of lithium on the clearance of β-amyloid from the brain would be unexpected at the low plasma levels that would be obtained from exposure to trace levels of lithium detected in water.

  8. Following up on comments by Drs. Wallace and Buck: It seems to me that adjusting for health care resources and using major depression as a negative control group may have led to inappropriate overadjustment, masking a real effect. If lithium in the water has the kind of widespread positive effects on psychological health implied by previous work in Texas and elsewhere, it would not only lead to reduced risk of depression, but also a lower drive for health care resources—both for psychiatric diagnoses and because of reduced malingering and more optimism and confidence leading people to be less likely to visit the hospital for minor complaints. An effect on bipolar, dementia, and depression all at once might be mediated by BDNF signaling, for instance. 


    . Lithium: occurrence, dietary intakes, nutritional essentiality. J Am Coll Nutr. 2002 Feb;21(1):14-21. PubMed.

    . Lithium levels in drinking water and risk of suicide. Br J Psychiatry. 2009 May;194(5):464-5; discussion 446. PubMed.

    . Long-term lithium treatment increases intracellular and extracellular brain-derived neurotrophic factor (BDNF) in cortical and hippocampal neurons at subtherapeutic concentrations. Bipolar Disord. 2016 Dec;18(8):692-695. Epub 2016 Nov 24 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Does Lithium in Tap Water Keep Dementia at Bay?

Paper Citations

  1. . Examining the Relationship between Trace Lithium in Drinking Water and the Rising Rates of Age-Adjusted Alzheimer's Disease Mortality in Texas. J Alzheimers Dis. 2018;61(1):425-434. PubMed.
  2. . Disease-modifying properties of long-term lithium treatment for amnestic mild cognitive impairment: randomised controlled trial. Br J Psychiatry. 2011 May;198(5):351-6. PubMed.
  3. . Microdose lithium treatment stabilized cognitive impairment in patients with Alzheimer�s disease. Curr Alzheimer Res. 2012 Jun 29; PubMed.
  4. . Chronic Microdose Lithium Treatment Prevented Memory Loss and Neurohistopathological Changes in a Transgenic Mouse Model of Alzheimer's Disease. PLoS One. 2015;10(11):e0142267. Epub 2015 Nov 25 PubMed.
  5. . BACE1 inhibition by microdose lithium formulation NP03 rescues memory loss and early stage amyloid neuropathology. Transl Psychiatry. 2017 Aug 1;7(8):e1190. PubMed.
  6. . LISPRO mitigates β-amyloid and associated pathologies in Alzheimer's mice. Cell Death Dis. 2017 Jun 15;8(6):e2880. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Association Between Groundwater Lithium and the Diagnosis of Bipolar Disorder and Dementia in the United States. JAMA Psychiatry. Published online May 23, 2018.