The study by Ohgami et al ^{Reference Ohgami, Terao, Shiotsuki, Ishii and Iwata1} raises serious ethical issues related to the interpretation of research findings and, as a consequence, their possible application. While not denying that the findings are interesting and have caused a stir in the lay press and on the internet, we question the methodology and the possible implications if the results are taken seriously.

First, sociological reasons for suicide are important, and changing rates of suicide in many countries are linked to changes such as migration, poverty, relationships and economic issues. The finding that when gender was included in the analysis there was a difference in the significance levels between men and women (with the results being less significant in women) is one such example. Adding lithium to tap water is not going to change these demographic and social factors that contribute to suicide rates, and not having accounted for at least some of these is a major limitation of the study. Second, although we agree with Young ^{Reference Young2} in his commentary that more research is needed to prove or disprove this tantalising idea, it is also important to assess what the impact of different levels of tap-water lithium is going to be on thyroid function, pregnant women and on the unborn fetus. It is also important to assess whether tap-water levels of lithium directly correlate with serum lithium levels in the respective populations. The levels of lithium in body fluids in normal healthy controls have varied from 0.01 to 0.09 meq/1 in one study, ^{Reference Jathar, Pendharkar, Pandey, Raut, Doongaji and Bharucha3} but there are no data about serum lithium levels among individuals attempting suicide. Maybe assessment of serum lithium levels among those with suicidal behaviour can be a place to start. More data are also needed on the role of low-dose lithium in individuals without mood disorders who are at risk of suicide.

Finally, several foods (particularly spices) are known to have relatively high levels of lithium as reported by a study in India several years ago. ^{Reference Jathar, Pendharkar, Pandey, Raut, Doongaji and Bharucha3} This study reported levels as high as 12 μg/g of lithium in tobacco and high levels in crude salt, rock salt and several spices. Maybe, until such time that we are certain about lithium's role in decreasing suicidality in non-psychiatric populations, it might be worth conducting randomised controlled trials with these foods in individuals with suicidal behaviour to see whether low doses of lithium really help.

Let us not throw the lithium out with the tap water yet!