Amygdala volumes in mood disorders - Meta-analysis of magnetic resonance volumetry studies

Background: The amygdala plays an important role in the regulation of emotions and has been implicated in the pathophysiology of mood disorders. Studies of amygdala volumes in mood disorders have been conflicting, with findings of increased, decreased and unchanged amygdala volumes in patients relative to controls. We present the largest meta-analysis of amygdala volumes in mood disorders and the first one to investigate modifying effects of clinical, demographic and methodological variables. Methods: We reviewed 40 magnetic resonance imaging studies investigating amygdala volumes in patients with unipolar or bipolar disorders. For meta-analysis we used standardized differences in means (SDM) and random effect models. In the search for sources of heterogeneity, we subdivided the studies based oil diagnosis, setting, age, medication status, sex, duration of illness, slice thickness, interrater reliability of tracing and anatomical definitions used. Results: The volumes of the left and right amygdala in bipolar (N=215) or unipolar (N=409) patients were comparable to controls. Bipolar children and adolescents had significantly smaller left amygdala volumes relative to controls (SDM=-0.34, 95%CI=-0.65; -0.04, z=-2.20, p=0.03), whereas bipolar adults showed a trend for left amygdala volume increases (SDM=0.46, 95% CI=-0.03; 0.96, 7= 1.83, p=0.07). Unipolar inpatients had significantly larger left (SDM=0.35, 95%CI=0.03; 0.67, z=-2.17, p=0.03) amygdala volumes than controls, with no significant amygdala volume changes in unipolar Outpatients. Limitations: Heterogeneity of included studies. Conclusions: The absence of overall differences in amygdala volumes, in the presence of significant and sometimes mirror changes in patient subgroups, demonstrates marked heterogeneity among mood disorders. Amygdala volume abnormalities may not be associated with mood disorders per se, but rather may underlie only some dimensions of illness or represent artifacts of medication or comorbid conditions. (C) 2008 Elsevier B.V. All rights reserved.