Causal Networks of the Brain in Depression

Research: Going beyond functional brain networks toward effective connectivity.

Posted Oct 30, 2017

What is Effective Connectivity?

Rolls and colleagues (2017) describe their work on "effective connectivity" in analyzing brain activity in depressed patients. Effective connectivity goes beyond standard functional connectivity (which is itself becoming very sophisticated) by using analytic methods to determine not only correlations among different brain areas, but also the direction of influence of one brain region on another. Using an approach known as dynamic causal modeling (DCM) to work out effective connectivity allows researchers to say more about cause-and-effect relationships among different brain regions, adding a needed layer of complexity onto existing approaches which can only look at correlation in activity.

When the activity of two brain regions is correlated, it does not mean that one necessarily influences the other or that if you change one (e.g. with a therapeutic intervention), you will get the desired effect. Correlated areas may simply both be active as a result of an underlying process, for example, and may not directly influence one another. Effective connectivity measures allow us to look at how brain regions may be directly influencing one another, though the work is early on.


In the research described here, the study authors compare brain activity in people with clinical depression with non-depressed controls. In reviewing the prior extensive literature examining functional brain networks in depression, they describe the importance of the following areas of function in major depressive disorders, which are a growing concern in the U.S. and globally:

  • Altered processing of reward and incentives, leading to impaired motivation, loss of interest (apathy) and lack of capacity to enjoy experiences (anhedonia);
  • Difficulty regulating worry and anxiety, leading to difficulty from excessive anxiety and over-reactivity to negative information;
  • Impaired flexibility in cognition and behavior, contributing to rigid negative thinking, pessimism and difficulty changing course;
  • Altered sensory and social information processing, contributing to errors in perception and interpretation of experiences supporting negative emotional states (dysphoria);
  • Impaired memory and concentration, affecting executive function including self-referential processing necessary for normal performance;
  • Changes in physiological regulation, associated with changes in sex drive, appetite, sleep, and in other body systems.

All of these key functional areas are partially regulated by brain areas which have been shown in prior correlational research to be different between depressed people and people without depression. Such areas of the brain (here's a good tool for understanding brain anatomy and function) include areas involved with:

  • Processing emotions including the amygdala, areas of the anterior cingulate cortex and pallidum;
  • Self-referential processing including the medial prefrontal cortex, areas of the posterior cingulate and precuneus;
  • Memory including the hippocampus and related areas;
  • Visual processing including the fusiform gyrus, lingual gyrus, lateral temporal cortex;
  • Attention and concentration including the dorsolateral prefrontal cortex [GHB: a standard target for treating depression with transcranial magnetic stimulation TMS], anterior cingulate, thalamus and insula.

Conducting analysis of effective connectivity allowed the authors of the present study to begin to extend our understanding of the nature of the relationships among this different brain regions. They note being particularly interested in areas associated with reward, and non-reward and punishment, in using effective connectivity to look at the neural basis of depression.

In addition to advancing knowledge, understanding causal relationships—related to direction of influence between one brain region and another—may provide novel therapeutic approaches for targeted interventions and allow us to refine existing interventions. Understanding the causal relationships among different brain regions may allow us to better understand other psychiatric conditions and their treatments, as well as normal brain function and pathways for performance enhancement.

Researching Effective Connectivity In Depression

The researchers recruited 336 patients diagnosed with major depression and 350 controls without depression, diagnosed according to standard criteria. Measures used included the Hamilton Depression Rating Scale and the Beck Depression Inventory, along with demographic and basic information about illness and treatment course. Subjects were imaged with MRI and the data analyzed using  dynamic causal modelling, modified to derive causal relationships from whole-brain network analysis by analyzing changes in signal across relatively short time scales for a large number of regions of interest in a relatively large group of experimental subjects.

Extending our current understanding of the neural basis of depression, the study showed the following differences in effective connectivity between depressed subjects and healthy controls:

  • Reduced drive of the medial orbitofrontal cortex by temporal regions of the brain. The medial orbitofrontal cortex is involved with subjective pleasure- and reward-based motivation, and weaker input into this area of the brain may related to feelings of unhappiness and inability to enjoy oneself which all hallmarks of depression, due to decreased ability to generate positive emotional states.
  • Increased input from temporal cortical areas to the precuneus and associated areas. The precuneus is an area brain involved with self-representation, and increased drive of this area may help to explain the strong negative sense of self and low-self esteem often associated with depression. Related findings here include effects on negative ruminations about oneself from the angular gyrus, and increased negative emotional processing of visual cues due to effects on the inferior temporal cortex.
  • Increased overall activity in the lateral orbitofrontal cortex related to reduced input from other areas in the frontal gyrus. The lateral orbitofrontal cortex is related to increased processing of non-reward and punishment, and activity here helps account for a bias toward  guilt, blame, which is connected with increased input from brain areas addressing language and sense of self.
  • Increased input to the hippocampus from temporal regions. Given the role of the hippocampus in memory, this suggests a possible mechanism for preferential attention to negative memories in depression.

Additional Considerations

In addition to elucidating greater detail about the neural features of depression, this study takes an important step in using dynamic modeling to map out directional influence among brain regions. We begin to get a clearer picture of not only which brain systems are involved in what symptoms of depression, but also how they are inter-related functionally and causally as we map out how the brain actually works. This knowledge may also be useful in sorting out different biological biotypes of depression by adding causal analysis to correlational studies.

As this and related approaches allow us to better understand the causal relationships underpinning the brain activity which gives rise to experience in health and illness, we will develop both a deeper understanding of how the mind works as well as tools for intervening which presumably will have greater predictability, utility and tolerability.


Rolls E.T., Cheng W., Gilson M., Qiu J., Hu Z., Ruan H., Li Y., Huang C.-C., Yang A.C., Tsai S.-J., Zhang X., Zhuang K., Lin C.-P., Deco G., Xie P. & Feng J., Effective connectivity in depression, Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, Volume 2, Number 7, October 2017, doi: 10.1016/j.bpsc.2017.10.004.