Skip to content

AI–IPL Functional Connectivity Identified as Potential Biomarker of Bipolar Disorder

Key findings

  • On functional MRI, impaired connectivity between the anterior insula (AI) and the inferior parietal lobule (IPL) of the frontoparietal executive control network distinguished patients with bipolar depression from those with both unipolar depression and healthy control subjects
  • Altered AI–IPL functional connectivity was related to greater impairments in the behavioral dimensions of perceived emotion control and reward sensitivity
  • AI–IPL functional connectivity may be a biomarker of bipolar disorder, and the IPL has potential to be a target for focal interventions

Bipolar depression is often misdiagnosed as unipolar depression, which delays proper treatment. Worse still, inappropriately prescribing an antidepressant can prompt a manic episode. Researchers led by Kristen K. Ellard, PhD, a psychologist with the Dauten Family Center for Bipolar Treatment Innovation, and Joan A. Camprodon, MD, PhD, director of the Division of Neuropsychiatry at Massachusetts General Hospital, published preliminary evidence in Biological Psychiatry: Cognitive Neuroscience and Neuroimaging that functional magnetic resonance imaging (fMRI) can distinguish between the two types of depression, a finding that may improve the accuracy of clinical diagnosis.

The study participants included 35 patients with unipolar depression, 24 patients with bipolar depression and 39 healthy control subjects. They completed the Affective Control Scale, which is a measure of perceived emotion control, and the Behavioral Inhibition System/Behavioral Activation System, which measures behavioral inhibition and activation and affective responses to impending reward and punishment.

All subjects underwent resting-state functional MRI using the same scanner and protocols. The researchers examined correlations between signals in the anterior insula (AI) and signals in prespecified regions of the frontoparietal executive control network, default mode network and salience network. Those functional networks have been linked to deficits that underlie mood disorders, including increased self-processing (rumination) in depression, decreased behavioral inhibition in mania and emotion and attention dysregulation in all mood disorders.

The researchers took two approaches to the MRI data, examining them according to syndrome (unipolar depression, bipolar depression or no depression) and results on the psychological scales that the patients completed.

The first finding was that patients with bipolar depression exhibited significantly altered bilateral dorsal AI functional connectivity to the left inferior parietal lobule (IPL), a key node of the executive control network. That finding differentiated bipolar depression subjects from both patients with unipolar depression and healthy control subjects.

With regard to the psychological test results, impaired perceived emotion control and increased behavioral drive toward rewarding experiences were significantly associated with weaker functional connectivity between the right AI and left IPL. That was true across groups regardless of diagnosis.

The second set of findings, the authors explain, may be tied to patients’ inability to engage regulatory control in response to physiological shifts in autonomic arousal. Patients with bipolar depression might not adaptively respond to shifts in physiological states due to a disruption in the coordination between interoceptive awareness and regulatory control.

AI–IPL Functional Connectivity as a Biomarker

The researchers used receiver-operating characteristics analyses to explore how well AI functional connectivity could discriminate between bipolar and unipolar depression. They found that right dorsal AI–left IPL functional connectivity significantly distinguished patients with bipolar depression from the two other groups of participants (area under the curve = 0.67, SE = 0.059, P = .01, 95% confidence interval, 0.57–0.79). AI–IPL functional connectivity had 71% sensitivity, 65% specificity, a negative predictive value of 87% and a positive predictive value of 40%.

The researchers conclude that impaired AI–IPL functional connectivity appears to be a useful biomarker of bipolar disorder, and the IPL might be a suitable target for focal interventions, such as neuromodulation.

Learn more about the Dauten Family Center for Bipolar Treatment Innovation

Learn more about the Division of Neurotherapeutics

Related topics


Previous research has shown that diagnostic codes routinely collected in electronic health records can help predict domestic abuse an average of two years in advance. Could EHR systems also be used to predict suicidal behavior?


To examine how cannabis use affects the brain, most studies use positron emission tomography, which exposes subjects to radiation. Mass General researchers examined whether functional near-infrared spectroscopy is a feasible alternative for such research.