Brain anatomy of major depression II. Focus on amygdala

Abstract

Here, we briefly summarize the most consistent structural magnetic resonance imaging (MRI) studies on amygdala in major depression and debate the effects of clinical variables on amygdalar morphology.

Major depressive disorder (MDD) has been associated with morphological changes of medial temporal lobe's structures, particularly of the amygdala, possibly being part of an altered limbic–thalamic–cortical circuitry (Zou et al. 2010; Bellani et al. 2010). This structure is part of the ventral limbic system and is functionally connected to the prefrontal cortex, cingulate gyrus and hypothalamus. It is a key component for affective modulation (such as negative emotions and fear), memory encoding and social behaviour (Baxter & Murray, 2002). Several magnetic resonance imaging (MRI) studies have found reduced amygdala volumes in patients suffering from depression (Sheline et al. 1998, 1999; Campbell et al. 2004; Hickie et al. 2007), specifically in children (Rosso et al. 2005), unmedicated (Caetano et al. 2004; Tang et al. 2007; Kronenberg et al. 2009), multiple episode (Bremner et al. 2000; Caetano et al. 2004; Hastings et al. 2004; Monkul et al. 2007), psychotic and female patients (Sheline et al. 1999; Hastings et al. 2004; Tang et al. 2007; Keller et al. 2008; Lorenzetti et al. 2009). In this regard, chronic or recurrent MDD patients are persistently exposed to stress-induced glucocorticoids, which may have neurotoxic effects, potentially leading to amygdala shrinkage (Hamidi et al. 2004). Interestingly, slight volume reductions of amygdalar grey matter have been shown over time without significant gross abnormalities (Frodl et al. 2008a, b), suggesting the presence of subtle microstructural processes occurring during a depression episode and after recovery. However, other structural investigations have shown preserved volumes (Mervaala et al. 2000; Munn et al. 2007; MacMaster et al. 2008), mainly in current non-suicidal patients (Monkul et al. 2007), in non-psychotic depressed patients (Keller et al. 2008) or in recovered patients (van Eijndhoven et al. 2009; Lorenzetti et al. 2010). Moreover, enlarged amygdalar volumes have also been reported (van Elst et al. 2000), particularly in subjects using antidepressants (Frodl et al. 2003; Weniger et al. 2006) and in those with severe illness or at the early stage of the disease (Frodl et al. 2003, 2008a; Lange & Irle, 2004; Lorenzetti et al. 2010; Weniger et al. 2006).

In summary, although there is some evidence that amygdalar size is reduced in MDD patients, particularly in those with recurrent episodes (Hamilton et al. 2008; Lorenzetti et al. 2009), preserved and increased volumes have also been reported. The heterogeneity of the results summarized here (see Table 1) may in part be due to socio-demographical and clinical differences of the samples (age of onset, single or multiple episodes, familiar history of MDD, medication, psychotic symptoms and phases of the illness) (Hajek et al. 2009). Furthermore, the proximity of the amygdala to the head of the hippocampus makes the anatomical delineation of this structure difficult. Indeed, as reported in the meta-analysis by Campbell et al. (2004), MRI studies considering the amygdala–hippocampus complex revealed no significant differences between depressed and control subjects. In order to better clarify the role of amygdala for the pathophysiology of MDD, future MRI studies should explore amygdalar morphology in large sample of drug-naïve patients at their first episode of depression in comparison with matched healthy individuals, longitudinally following them after recovery.

Table 1. Cross-sectional and follow-up studies investigating amygdalar volumetry in adult patients with MDD compared with healthy control subjects

*Follow-up studies: AD, patients on antidepressants at the time of the MRI scanning; HC, healthy control subjects; HR, high risk; L, left; R, right; ROI, region of interest; T, twins; VBM, voxel-based-morphometry; M, male; F, female.