Brain structural and functional changes in patients with major depressive disorder: a literature review

Gray matter changes

VBM is widely used in the study of abnormal brain anatomy. This technique uses the statistical parameter map method to measure the volume and density of each voxel corresponding to gray matter and white matter. Changes were quantitatively calculated to assess changes in gray matter and white matter.

Meta-analyses indicate that hippocampal volume reduction is the most common brain anatomical change in patients with depression (Arnone et al., 2012; Cole et al., 2011), and, especially, the atrophy is most pronounced in the cornu ammonis, dentate gyrus, and subiculum (Roddy et al., 2019). As an essential part of the limbic system, the hippocampus plays a vital role in memory processing and emotional management. Moreover, studies (Buddeke et al., 2017; Den Heijer et al., 2011) found that depressive symptoms and hippocampal atrophy are mutually reinforcing and aggravating. Also, the volume of the cingulate cortex, another part of the limbic system associated with memory and mood formation, was smaller in depressed patients than in healthy controls (Rodriguez-Cano et al., 2014; Wise et al., 2017).

Frontal atrophy is also one of the critical changes in depression (Grieve et al., 2013). Studies have shown that the medial prefrontal cortex, frontal cortex, dorsolateral prefrontal cortex atrophy is particularly significant (Bludau et al., 2016; Van Eijndhoven et al., 2013; Zhao et al., 2014). The frontal cortex plays an important role in emotional cognition and working memory (Bludau et al., 2014).

In addition to the frontal lobe, the volume of the bilateral putamen and left thalamus in patients with depression is also smaller in contrast to healthy controls (Lu et al., 2016). These gray matter nuclei are related to memory, information transmission, and emotional management. Also, the degree of atrophy of the amygdala is positively correlated with the severity of depressive symptoms in patients (Zhang et al., 2016). And meta-analysis showed that patients with depression who had a larger gray matter before treatment were also better treated with medication (Fonseka, MacQueen & Kennedy, 2018).

White matter changes

Abnormal white matter is also widespread in patients with depression (Liao et al., 2013). DTI shows the location and direction of the white matter bundle. Tract-based spatial statistics (TBSS) can quantitatively measure the fraction anisotropy (FA) of the nerve white matter fibers, and compare the white matter bundle skeletons of different subjects to locate the microstructure abnormalities of the brain white matter accurately (Smith et al., 2006).

Meta-analysis (Jiang et al., 2017a) indicated that the FA values in the corpus callosum, white matter in the right cerebellar hemisphere and bilateral superior longitudinal plasma of depressed patients were significantly lower than those in the healthy control group, and it was indicated that the abnormality of the corpus callosum was particularly prominent (Han et al., 2014). And Studies (Cole et al., 2012; De Diego-Adelino et al., 2014) showed that the extent of the FA decline in the corpus callosum, and bilateral upper longitudinal was positively correlated with the severity of depressive symptoms and duration of onset. And patients with suicide attempt history had lower FA in the dorsomedial prefrontal cortex than those without suicide attempt history and healthy controls (Olvet et al., 2014). Not only that, but the lower FA value in the ventral medial prefrontal area is more pronounced in patients with refractory depression (De Diego-Adelino et al., 2014). The decrease in the FA value of the superior frontal gyrus, superior longitudinal fasciculus, and corpus callosum can even predict the depression in the elderly (Reppermund et al., 2014). A study (Henderson et al., 2013) of adolescents with depression found that patients with more severe depressive symptoms had a greater FA reduction in sagittal stratum, anterior thalamic radiation, genu of the corpus callosum and anterior cingulate near the precuneus. At present, there are still many inconsistencies in the study on the abnormal white matter fiber bundles in depression. In the future, multi-site large sample studies can be carried out to verify the above research results.

Cerebrovascular changes

On the other hand, compared with the healthy control group, elderly patients with depression have more severe cerebral vascular lesions such as white matter hyperintensities (WMH), subcortical lacunar, microinfarction, and microangiopathy (Wang et al., 2014). As early as 1997, Alexopoulos (Alexopoulos et al., 1997) proposed the “vascular depression hypothesis”, which believes that cerebrovascular disease and its subsequent white matter changes are an essential part of the pathogenesis of late-onset depression. A recent meta-analysis (Van Agtmaal et al., 2017) indicates that white matter hyperintensities are significantly associated with the incidence of depression. These white matter lesions are considered to be significant predictors of late-onset depression (Park et al., 2015), of which subcortical white matter lesions are strictly related to the severity of depressive symptoms and cognitive impairment (Taylor, Aizenstein & Alexopoulos, 2013). Besides, patients with depression with severe changes in leukoencephalopathy have worse symptoms and cognitive function after antidepressant treatment (Sheline et al., 2010).

Regional brain activity changes

Depending on the purpose and the experimental design of the study, the fMRI study can be divided into two types: resting-state and task-based. The former mainly examines the spontaneous nerve activity of the subject in a calm and awake state, while the latter mainly explores the activity state of the brain when the individual is subjected to emotional stimulation or completing specific tasks.

A study (Hamilton et al., 2012) indicated, compared to the control group, MDD patients showed increased resting activity in the pulvinar, which is an essential nucleus in the thalamus and is thought to be functionally synchronized with nodes in the salient network, such as the amygdala, the insular lobes, and the anterior cingulate gyrus. Thus, it may enhance the response to negative emotional information in the salience network. Also, a meta-analysis (Kuhn & Gallinat, 2013) showed, at resting state, increased activity in the ventral medial prefrontal cortex, the left ventral striatum, and left thalamus and decreased activity in the left postcentral gyrus, left fusiform gyrus and left insula relative to controls in patients with depression.

On the other hand, due to the differences in experimental task design and sample selection methods, although many task-based fMRI experiments have studied the abnormalities of brain activity patterns in patients with depression during cognitive and emotional processing, consistent conclusions are still lacking (Muller et al., 2017). A meta-analysis (Miller et al., 2015) of adolescent depression found that hyperactivity of the anterior cingulate gyrus and thalamus may lead to depression patients being highly sensitive to emotional stimuli, while anhedonia may be caused by hypoactivation of the cuneus and posterior insula during reward processing. Interestingly, in depressed patients, the amygdala showed a “dual-separation” pattern of hyperactivity in response to negative stimuli (Tao et al., 2012) and decreased response to positive stimuli (Suslow et al., 2010). Likewise, in a small sample reward study (Takamura et al., 2017), MDD patients did not experience increased striatal activity in response to reward stimuli as healthy controls did.

Young et al.’s small sample studies (Young et al., 2017a; Young et al., 2017b) found that after real-time functional magnetic resonance imaging neurofeedback training, the amygdala activity of depressed patients could be relatively restored to normal, and patients’ depressive symptoms were reduced and their ability to recall positive memories was improved. A study (Holmes & Pizzagalli, 2008) indicates that the decline in cognitive ability in patients with depression is caused by the distribution of excessive neural pathway resources in negative consciousness and rumination. A small sample study by Liao et al. (2012) showed that patients with depression had a bias in their perception of pleasure and neutral stimuli, which was associated with abnormal activity in the bilateral amygdala and the right dorsolateral prefrontal cortex.

Brain network functional connectivity changes

Functional connectivity is defined as the temporal correlation of multiple brain regions. Each brain region, defined as a node, is connected, eventually forming a brain network with highly complex and concentration. These brain networks play an important role in cognitive and emotional processing (Park & Friston, 2013). In 2011, Menon (Menon, 2011) proposed the “triple network model” theory, which concluded that abnormal functional connectivity of the default mode network (DMN), the central execution network (CEN), and the salience network (SN) (Fig. 2) is closely related to various mental illnesses, including depression. Not only that, by examining abnormalities in brain network functional connectivity in 711 depressed patients, Drysdale et al. (2017) defined four neurophysiological subtypes of depression and to some extent, successfully predicted their rTMS treatment effect.

The most common methods for studying brain function connectivity include seed-based correlation analysis (SCA) and independent component analysis (ICA). SCA predetermines “seed” (a region of interest) based on previous assumptions, and calculates the correlation with the other voxels or specific other regions of the brain by their BOLD signal fluctuations. In contrast, ICA uses all available data in the fMRI image and decomposes it several independent components. An increase in functional connectivity represents increased synchronization between the two regions.

Salience network

The SN consists of the insular cortex, dorsal anterior cingulate cortex, temporal pole and amygdala and is responsible for detecting and filtering stimuli, as well as in recruiting relevant functional networks (Menon & Uddin, 2010). Completing a variety of complex functions, including communication, social behavior, and self-awareness.

Abnormal salience network connectivity is considered to be one of the crucial links in the pathogenesis of depression, especially in the insula and amygdala. Salience network, especially the right anterior insula, is thought to be critical in the transition from the central execution network’s dominant “execution state”. to the default state-preferred “default state” (Goulden et al., 2014). A study (Connolly et al., 2013) found elevated connectivity between the subgenual anterior cingulate cortex and insula, which may result in enhanced functional connectivity between the default mode network and the salience network, thus hindering the above transition (Fig. 3).

Another important anomaly node in the salience network is the amygdala. In adults and adolescents with depression and children at high risk of depression, the functional connectivity between the amygdala and the hippocampus is found to be decreased (Cullen et al., 2014; Luking et al., 2011; Zeng et al., 2012), while hyperactivity in the amygdala, as found in brain activity studies (Liao et al., 2012; Tao et al., 2012), is considered to be a compensation mechanism for this weak functional connectivity. Also, increased functional connectivity between the amygdala and subgenual anterior cingulate cortex is thought to be associated with long-term negative emotions in patients (Davey et al., 2015). Moreover, functional connectivity between the amygdala and the brainstem and precuneus are reduced in depressed patients compared with controls (Zhang & Li, 2012).