Melbourne-led research has developed a visually predictive model that can be used to screen for major depressive disorders in young people.
Using the world’s largest mega-analysis (by sample size) of resting state functional connectivity in youth MDD – comprising 27 global datasets – the team showed there was significant disruption to a young person’s attentional systems in MDD-related hyperconnectivity compared with healthy people.
Regions of the brain containing each pair of significant connections were then assigned to their respective networks in a sample of 440 youths with MDD and 370 healthy people aged between 12 and 25 years.
Lead author, Dr Andrew Zalesky, said this revealed that increased connectivity between the dorsal attention network (DAN) and several other networks, including the salience/ventral attentional network (VAN), somatomotor network (SMN) and central executive network (CEN), was evident in MDD.
“Hyperconnectivity within the VAN was also observed, but conversely, the MDD group also demonstrated significantly reduced connectivity involving many core, default mode network (DMN) regions, including the medial prefrontal cortex (mPFC), rostral anterior cingulate cortex (ACC), subgenual cingulate cortex (SGC), posterior cingulate cortex (PCC) and superior frontal gyrus,” he said.
The orbitofrontal cortex and the superior parietal, temporal pole, and somatosensory regions as well as the insula and thalamus also demonstrated hypoconnectivity.
“In addition, a significant sex-by-diagnosis interaction was evident,” Dr Zalesky said.
“Connectivity in the visual, cuneus, somatosensory, premotor, and dorsal mPFC as well as the anterior thalamus was lower in female MDD compared with male MDD individual. On the contrary, male MDD participants demonstrated lower connectivity lateralized to right visual, somatosensory, posterior parietal and retro splenial cortex relative to female counterparts.”
No significant effects were detected for age-by-diagnosis interaction.
“Together, these findings indicate that greater symptom severity was associated predominantly with attentional and sensory network hyperconnectivity in youth MDD,” he said.
“Among all hub regions, the highest consistency was observed for those of the DMN across all analyses.”
At the network level, the DMN similarly demonstrated widespread hypoconnectivity with multiple networks spanning the DAN, VAN, and CEN.
“These findings indicate that DMN- and attentional network-centred hypoconnectivity also relate to symptom severity,” Dr Zalesky noted.
Based on the observed patterns of functional connectivity, the team developed a computer model that could accurately predict individual-level diagnostic status and symptom severity for MDD.
“Collectively, our findings indicated that youth MDD was associated with robust changes in functional connectivity anchored to core components of the default mode and attentional networks. These regions tended to have a higher level of hubness and demonstrated significant predictive accuracy on independent machine-learning analyses,” Dr Zalesky said.
“This suggests a potential trajectory of disruption that may begin in youth and remains constrained to a common set of processes and brain systems across the illness course.
“Importantly, the topological properties of these hub regions may represent opportunities for non-invasive, neuromodulatory intervention refinement of potential stimulation targets, capitalizing on the high degree of neural plasticity of the adolescent brain.”