Glucose metabolism in the amygdala in depression: relationship to diagnostic subtype and plasma cortisol levels.

In a previous positron emission tomography (PET) study of major depression, we demonstrated that cerebral blood flow was increased in the left amygdala in unipolar depressives with familial pure depressive disease (FPDD) relative to healthy controls [J. Neurosci. 12 (1992) 3628.]. These measures were obtained from relatively low-resolution PET images using a stereotaxic method based upon skull X-ray landmarks. The current experiments aimed to replicate and extend these results using higher-resolution glucose metabolism images and magnetic resonance imaging (MRI)-based region-of-interest (ROI) analysis. The specificity of this finding to FPDD was also investigated by assessing depressed samples with bipolar disorder (BD-D) and depression spectrum disease (DSD). Finally, the relationship between amygdala metabolism and plasma cortisol levels obtained during the scanning procedure was assessed. Glucose metabolism was measured using PET and 18F-fluorodeoxyglucose (18FDG) in healthy control (n=12), FPDD (n=12), DSD (n=9) and BD-D (n=7) samples in the amygdala and the adjacent hippocampus. The left amygdala metabolism differed across groups (P<.001), being increased in both the FPDD and BD-D groups relative to the control group. The left amygdala metabolism was positively correlated with stressed plasma cortisol levels in both the unipolar (r=.69; P<.005) and the bipolar depressives (r=0.68;.1<P<.05). In contrast, neither significant main effects of diagnosis nor significant relationships with plasma cortisol were evident in post hoc analyses of metabolism in the right amygdala or the hippocampus. Preliminary assessment of BD subjects imaged during remission suggested that amygdala metabolism is also elevated in remitted subjects who are not taking mood-stabilizing drugs, but within the normal range in subjects taking mood stabilizers. These data confirm our previous finding that neurophysiological activity is abnormally increased in FPDD, and extend it to BD-D. These abnormalities were not accounted for by spilling in of radioactivity from the adjacent hippocampus. The correlation between left amygdala metabolism and stressed plasma cortisol levels may conceivably reflect either the effect of amygdala activity on corticotropin-releasing hormone (CRH) secretion or the effect of cortisol on amygdala function.