White matter microstructural properties in bipolar disorder in relationship to the spatial distribution of lithium in the brain

Background: Lithium treatment is associated with an increase in magnetic resonance imaging derived measures of white matter integrity, but the relationship between the spatial distribution of brain lithium and white matter integrity is unknown. Methods: Euthymic patients with bipolar disorder receiving lithium (n = 12) and those on other medications but naive to lithium (n = 17) underwent diffusion imaging alongside matched healthy controls (n = 16). Generalised fractional anisotropy (gFA) within white matter was compared between groups using a standard space white matter atlas. Lithium-treated patients underwent novel multinuclear lithium magnetic resonance imaging (Li-7-MRI) to determine the relative lithium concentration across the brain. The relationship between Li-7-MRI signal intensity and gFA was investigated at the resolution of the Li-7-MRI sequence in native space. Results: Lithium-treated bipolar disorder and healthy control groups had higher mean white matter gFA than the bipolar disorder group treated with other medications (t = 2.5, p < 0.05; t = 2.7, p < 0.03, respectively). No differences in gFA were found between patients taking lithium and healthy controls (t = 0.02, p = 1). These effects were seen consistently across most regions in the white matter atlas. In the lithium-treated group, a significant effect of the Li-7-MRI signal in predicting the gFA (p < 0.01) was identified in voxels containing over 50% white matter. Limitations: Cross-sectional evaluation of a relatively small cohort. Conclusions: The higher gFA values observed in the lithium-treated bipolar disorder group suggests that long-term lithium is associated with greater white matter integrity. Our novel analysis supports this further, showing a positive association between white matter gFA and the spatial distribution of lithium.