The specific aim of this application is to afford the applicant, a physicist, the opportunity to become an independent researcher applying magnetic resonance spectroscopy (MRS) in affective disorders. Two components are proposed to accomplish this: formal academic training in psychology and neuroscience; and two research projects. There is an established literature that suggests the pathophysiology of affective illness may lie in irregularities in second messenger and signal transduction pathways; in particular the phosphatidylcholine (PtdCho) and phosphatidylinositol (PtdIno) cycles. Elements of the PthCho and PtdIno cycles are detectable using proton (1 H) and phosphorous (31 P) MRS, in particular choline containing compounds (Cho), myo-Inositol containing compounds (Ino), phosphomonoesters (PME), phosphodiesters (PDE) and the nucleotide triphosphates (NTPs). Over the five year course of the study 35 subjects with Major Depressive Disorder (MD), 35 subjects with Bipolar I Disorder (BD) and 20 comparison subjects (NC) will be recruited. A priori regions of interest are the anterior cingulate gyrus and the caudate nuclei since functional neuroimaging studies have noted mood state dependent alterations in metabolic activity in these regions and preliminary MRS data from these brain regions have demonstrated mood-state, medication, and diagnosis- dependent alterations in Cho, Ino, beta-NTP and PME. In the first two years of funding subjects will be examined using proton echo planar spectroscopic imaging (PEPSI) at 1.5 T. For the final three years of funding subjects will be examined using 1H decoupled 31P MRSI at 4.0 T. The use of PEPSI at 1.5 T and 1H decoupled 31P MRSI at 4.0 T will allow for the acquisition of spectra from smaller regions of interest, such as the anterior cingulate and the caudate, than is possible with older techniques. In addition, PEPSI affords a time advantage which allows for the absolute quantification of the metabolites detected. All subjects will be examined on two occasions six weeks apart and the following hypotheses will be tested: Increased right cingulate cortex Cho will be associated with depression. A decrease in the left cingulate cortex Cho will occur as a result of antidepressant treatment. Decreased right cingulate cortex Ino will be associated with depression. Decreased caudate nucleus beta-NTP will be associated with depression. An increase in caudate nucleus PME will occur as a result of lithium treatment.