Lithium in tap water is not associated with risk for bipolar disorder, dementia

Within the past year, two separate studies by researchers in Canada and Denmark have suggested that higher levels of lithium in drinking water are associated with lower rates of Alzheimer’s and dementia. But according to new research from the University of Chicago, the rates of mental health disorders instead seem to be influenced by regional demographics and available healthcare resources.

Robert Gibbons, PhD, director of UChicago’s Center for Health Statistics, and an interdisciplinary team of students from his Statistical Applications course completed a class project on the effects of lithium levels in drinking water and published their findings in the May 2018 issue of JAMA Psychiatry.

“I think that the idea that lithium in the drinking water could have positive mood effects has been floating around for a while,” says Nina Gao, Ph.D., a student at the UChicago Pritzker School of Medicine and an author on the paper. “It’s an interesting correlation.”

Though intrigued by these findings, Gibbons and his team were skeptical. “Given the clinical doses of lithium, it didn’t seem plausible to me,” said Gao.

Lithium is a naturally occurring trace metal typically found in soil and other mineral formations, but its levels can vary widely across the U.S. depending on climate and geology. Lithium is also used in the treatment of mood disorders, such as bipolar disorder, with doses typically in the range of 600-1200 mg/day.

Using data from the U.S. Geological Survey, the authors found lithium concentrations from more than 3,000 drinking water wells. In counties with high levels of lithium in the water supply, the average exposure was about 141.3 micrograms/liter, which is dramatically lower than the doses prescribed by physicians. If fact, a resident of one of these counties would have to drink 1,000 liters of water per day to even come close to the lowest reported effective dose of lithium.

The idea that lithium in the drinking water could have positive mood effects has been floating around for a while.

The team of students used statistical tests learned in Gibbons’s course to study the relationship between lithium levels in the water supply and the rates of diagnoses of bipolar disorder and dementia. Traditional analyses similar to those used in the Danish study showed that high lithium concentrations in the water were associated with lower rates of both disease conditions.

“[Our results] also showed associations with pancreatic cancer and depression and other things that really shouldn’t have anything to do with lithium, but are probably proxies of other sorts of confounds that happen to co-occur in those areas,” Gibbons said.

Indeed, compared to those with high lithium levels, U.S. counties with low lithium levels were much more populated and had more primary care physicians, psychiatrists, and hospital beds. Also, residents were older, more educated, and had higher household incomes.

After adjusting for these factors in more statistically rigorous analyses, there was no longer an association between drinking water lithium levels and the incidence of bipolar disorder or dementia, or any of the three negative control conditions (depression, prostate cancer, or myocardial infarction).

“You really have to think carefully about it,” says Rebecca Gorges, a Ph.D. student at UChicago’s Harris School of Public Policy and member of the study team. In counties with limited healthcare resources, the rate of many diseases will be lower, simply because you have too few doctors available to provide diagnoses. It turns out that those counties also have, on average, high lithium levels in their drinking water, giving the false impression of a relationship between lithium exposure and health outcomes.