When I first started using lithium carbonate (LiCO3) I described it as comparable to ‘insulin’ that is, a substance that the body needed but was not capable of producing in effective quantities itself. This study seems to confirm that or at least hints at it:
The study, published Jan. 6 in the journal Molecular Psychiatry, revealed differences in the white matter of patients’ brains and in the cerebellum, an area of the brain not previously linked with the disorder. Interestingly, the cerebellar differences were not present in patients taking lithium, the most commonly used treatment for bipolar disorder.
http://medicalxpress.com/news/2015-01-mri-scan-sensitive-metabolic-reveals.html
http://medicalxpress.com/news/2015-01-mri-scan-sensitive-metabolic-reveals.html
Stannous Flouride's Book of Truthiness 