Depression has a consistent mark in the brain even when symptoms are absent

By Pooja Toshniwal PahariaReviewed by Danielle Ellis, B.Sc.Sep 6 2024

Persistent salience network growth linked to depression, regardless of emotional state

A recent study in Nature used precision functional mapping to understand better, at the brain level, the neurobiological mechanisms behind distinct depressive symptoms and mood changes over time. 

Background

Depression is a neuropsychiatric illness that causes impairment globally. Decades of neuroimaging research show minor changes in brain structure and connectivity. However, researchers have a limited understanding of the processes that cause mood-state shifts. Neuroimaging studies are scarce.

The emerging field of precision functional mapping uses high-resolution data to uncover distinct changes in brain network topology and connectivity between healthy individuals and those with depression. However, researchers have not used this method for depression.

About the study

The study evaluated whether depression changes the functional network structure in the brain and the processes that govern mood shifts over time.

Researchers mapped functional brain networks in six highly sampled patients suffering from unipolar severe depression. These individuals underwent 622 minutes of multi-echo functional magnetic resonance imaging (fMRI) over 22 sessions. Their HDRS17 score was 16, indicating medium to severe depression. The study involved 37 healthy individuals who had 328 minutes of fMRI across 12 sessions.

The researchers validated the results using samples of depressed individuals provided by Stanford University and Weill Cornell Medicine. To further understand the consequences of depression on functional brain architecture, researchers examined large MRI datasets. The datasets included 812 controls, 120 healthy individuals, and 299 individuals with treatment-resistant depression.

Linear support vector machine classifiers separated depression patients from healthy controls based on functional network size. The classifiers used data from 37 healthy controls and 141 depressed individuals collected from several scanners. Researchers studied whether ectopic intrusions or boundary shifts caused salience network changes. They also evaluated the effects of depression on uni- or hetero-modal sensorimotor networks.

Researchers quantified the salient network's encroachment on each functional network. They measured node connections between invaded areas compared to healthy ones. To evaluate the stability of the salience network, they used divided halves of each individual's fMRI dataset. They examined changes in salience network size among those who took rapid-acting antidepressants. These individuals received repeated transcranial magnetic stimulation (rTMS) for six weeks or an intense one-week session.

The researchers investigated whether changes in salience network size predict depression in the long term. They analyzed the Adolescent Brain Cognitive Development (ABCD) study data of children without major depression aged 10 to 12 who experienced depressive symptoms between 13 and 14 years. They investigated SIMD data to determine whether variations in functional connection strength between salience network nodes correlate with or predict fluctuations in symptom severity over time.

Results

The frontostriatal salience network was found to be twice as large in depressed individuals. This network, which processes rewards and integrates autonomic signals and reactions with internal objectives and environmental demands, was found to cover 73% more cortical surface area than the average in those with depression.

The team reproduced the results three times in a sample of depressed patients who underwent fMRI (n=135), as well as in extensive datasets of 299 depressed individuals and 932 controls. They found comparable results in validation testing using different network parcellation strategies. This demonstrated the robustness of the findings.

Network boundary movements were the primary drivers of salience network growth. The salience network encroached on the frontoparietal, cingulo-opercular, and default mode networks in diverse ways. The enlargement was more common in brain areas with less intracortical myelin. This allowed for higher synaptic plasticity. The size of the frontostriatal salience network remained unaltered after antidepressant therapy.

Salience network growth was consistent throughout time, regardless of emotional state. The enlargement was also prominent in youngsters before late-onset depressive symptoms during adolescence.

Longitudinal examinations of individuals scanned a maximum of 62 times in 18 months revealed mood-dependent alterations in striatal connections with the salience network's insular and anterior cingulate nodes. This network monitored changes in anxiety and anhedonia. It predicted anhedonic symptom onset during future visits. Support vector machine classifiers distinguished depressive patients from healthy controls with 78% accuracy.

Conclusion

The study showed that frontostriatal salience network enlargement is a consistent trait in individuals with severe depressive illness, but it is not a predictor of depression episodes. Symptom intensity or chronicity is not related to salience network enlargement. The study identified a characteristic-like network structure in the brain that may increase depression risk.

Researchers also identified mood-state-based connectivity alterations in frontostriatal network circuits that estimate the onset and reduction of depression over time. Future studies should test the specificity of the results for different psychopathologies and also assess their clinical value.