Evidence of diagnostic specificity in the neural correlates of facial affect processing in bipolar disorder and schizophrenia: a meta-analysis of functional imaging studies

Psychol Med. 2013 Mar;43(3):553-69. doi: 10.1017/S0033291712001432. Epub 2012 Jul 9.


Background: Schizophrenia (SZ) and bipolar disorder (BD) may overlap in etiology and phenomenology but differ with regard to emotional processing. We used facial affect as a probe for emotional processing to determine whether there are diagnosis-related differences between SZ and BD in the function of the underlying neural circuitry.

Method: Functional magnetic resonance imaging (fMRI) studies published up to 30 April 2012 investigating facial affect processing in patients with SZ or BD were identified through computerized and manual literature searches. Activation foci from 29 studies encompassing 483 healthy individuals, 268 patients with SZ and 267 patients with BD were subjected to voxel-based quantitative meta-analysis using activation likelihood estimation (ALE).

Results: Compared to healthy individuals, when emotional facial stimuli were contrasted to neutral stimuli, patients with BD showed overactivation within the parahippocampus/amygdala and thalamus and reduced engagement within the ventrolateral prefrontal cortex (PFC) whereas patients with SZ showed underactivation throughout the entire facial affect processing network and increased activation in visual processing regions within the cuneus. Patients with BD showed greater thalamic engagement compared to patients with SZ; in the reverse comparison, patients with SZ showed greater engagement in posterior associative visual cortices.

Conclusions: During facial affect processing, patients with BD show overactivation in subcortical regions and underactivation in prefrontal regions of the facial affect processing network, consistent with the notion of reduced emotional regulation. By contrast, overactivation within visual processing regions coupled with reduced engagement of facial affect processing regions points to abnormal visual integration as the core underlying deficit in SZ.

Publication types

  • Meta-Analysis

MeSH terms

  • Adult
  • Affect
  • Bipolar Disorder / diagnosis
  • Bipolar Disorder / physiopathology*
  • Brain Mapping / statistics & numerical data*
  • Emotions*
  • Facial Expression*
  • Female
  • Humans
  • Likelihood Functions
  • Limbic System / physiopathology
  • Magnetic Resonance Imaging
  • Male
  • Prefrontal Cortex / physiopathology
  • Pulvinar / physiopathology
  • Schizophrenia / diagnosis
  • Schizophrenia / physiopathology*
  • Sensitivity and Specificity
  • Visual Cortex / physiopathology
  • Visual Perception / physiology