Prophylactic mood stabilization: What is the evidence for lithium exposure in drinking water?

Abstract

Large-scale population registers and databases have great potential for use in hypothesis-generating epidemiological research. Utilising such data, cross-sectional and ecological studies are relatively inexpensive and time-efficient observational studies that can identify correlations between exposures and outcomes. In the field of lithium research, this has generated interest in exploring the effect of exposure to micro-levels of lithium levels in drinking water. The studies carried out so far have predominantly examined associations between lithium exposure and rates of suicide and violence and only a paucity looked at associations between lithium exposure and mood. This is interesting considering that, for therapeutic levels of lithium in bipolar disorder, the evidence first supported its efficacy on mood and it only recently demonstrated an independent effect of lithium on suicidality, which may be specific and not secondary to the improvement of affective symptoms.1 Ecological studies have reported associations between geographical differences in lithium levels in the drinking supply and rates of suicide, suggesting that exposure to microlevels of lithium may have a prophylactic effect.2 One important limitation of such population studies, which rely on anonymised datasets, is their inability to link individual exposures to outcomes. This is overcome when all national datasets and registers use a common identifier, as is done in Denmark through the use of unique personal identification numbers, thereby allowing exposures to be linked to outcomes at an individual level. Kessing et al. have utilised this added strength of Danish population datasets to explore the link between exposure to microlevels of lithium in drinking water and incidence of bipolar disorder.3 The study estimates individual average lithium exposure by combining longitudinal data on areas of residence with data from drinking water lithium levels in the various municipalities. Comparing average lithium exposure in individuals with a hospital diagnosis of bipolar disorder to matched controls over a 13 year period, it found no association between higher long-term lithium exposure and reduced incidence of bipolar disorder.