Back in March I discussed bipolar disorder’s very complex biology, which includes abnormalities in glial-neuronal interactions, immune cells and inflammatory processes throughout the brain and body, the hypothalamic-pituitary axis, a host of genes, and several neurotransmitter systems. Here I will look at what neuroimaging tells us about abnormalities in brain regions and their connections, and then at what and advanced statistical approach to analyzing fMRI data says about the biology of emotion.
In a 2014 review, Mary Phillips and Holly Swartz of the University of Pittsburgh discussed what was known from fMRI, volumetric analyzes, diffusion imaging, and resting state techniques. Most of the studies imaged patients in the euthymic state; few included manic or depressed subjects because of the many practical difficulties conducting this type of research, including the problems disturbed patients experience cooperating with often lengthy and uncomfortable scanning procedures. Experiments typically involve presenting emotionally laden images, such as faces, to subjects before or during scanning.
The overall picture is of “parallel dysfunction in bilateral prefrontal cortical (especially ventrolateral prefrontal cortical)-hippocampal-amygdala emotion processing and emotion regulation neural circuitries, together with an ‘overactive’ left-sided ventral striatal-ventrolateral and orbitofrontal cortical reward processing circuitry.”
Both “voluntary” and “automatic (implicit)” processes contribute to emotion regulation. The more voluntary, or conscious, processes involve lateral regions of the prefrontal cortex, while more medial structures, including the orbitofrontal cortex, anterior cingulate cortex, mediodorsal prefrontal cortex, and hippocampus, mediate nonconscious emotion regulation. Bipolar disorder is associated with abnormally elevated amygdala responses to emotional stimuli (whether positive or negative in valence) as well as decreased activity in the ventrolateral prefrontal cortex and in connectivity between the ventrolateral prefrontal cortex and the amygdala.
In addition, there is bias toward positive emotions: individuals with bipolar disorder show increased amygdala and medial prefrontal cortex responses to positive stimuli such as happy faces, and decreased orbitofrontal cortex-amygdala connectivity, which the authors suggest may indicate underlying attentional bias to positive emotional stimuli.
Further, abnormally elevated amygdala activity occurs during performance of non-emotional tasks, suggesting a “heightened perception of emotional salience in non-emotional contexts.”
Interestingly, people with bipolar disorder also have increased sensitivity to rewarding stimuli. The brain’s reward system includes the nucleus accumbens and ventrolateral prefrontal cortex, which are activated during arousal by emotional stimuli, and the orbitofrontal cortex, which is involved in encoding reward value. The medial prefrontal cortex regulates the nucleus accumbens's responses to potentially rewarding stimuli, thus helping to manage appetitive behaviors. Bipolar disorder is associated with abnormally increased nucleus accumbens and left ventrolateral prefrontal cortical activity in reward anticipation.
These findings suggest a picture of heightened emotional responsiveness in regions and circuits connected with the amygdala, as well as reduced activity in circuitry involving the prefrontal cortex which controls emotional responsiveness. The article does not make clear whether the reduced emotional control occurs primarily in the “voluntary” or “automatic” circuits, or both. In addition, there is bias toward positive emotions as well as heightened emotional circuit activity in response to nonemotional stimuli. These phenomena would appear to be the basis for the mood intensity and lability we see in patients with bipolar disorder, the euphoric moods, and the attribution of emotional significance to mundane events and activities. In parallel with this, the overactive reward circuitry involving the nucleus accumbens and ventrolateral prefrontal cortex may underlie the unrealistic anticipation of positive outcomes to often risky behavior in patients with bipolar disorder. (It is also worth noting that not all manic patients are euphoric—some have predominately irritable moods, and there is some evidence that euphoric and irritable mania are associated with different genetic variants.)
More recently, Philip Kragel and Kevin LaBar Duke University have discussed the use of advanced mathematical analyses of fMRI data LINK, called “multivoxel pattern analysis,” to further understand how emotions operate in the brain. These techniques use neural networks, or machine learning, to match patterns of activity on MRI scans with emotional stimuli or behavior. Among their findings are that the dimensions of valence (positive vs. negative emotions) and level of emotional arousal are inadequate descriptions of people’s emotional experience. The so-called basic emotions—happiness, sadness, disgust, fear, and anger—have been associated with particular patterns of neural activity. These authors note that “it is not clear whether these category-specific, distributed activation patterns reflect evolutionarily ingrained networks, constructive processes, or a combination of factors.” In the earlier review, Phillips and Swartz briefly discussed these approaches, noting pattern analysis of fMRI scans can distinguish bipolar disorder from major depression, as well as high versus low genetic risk for bipolar disorder in healthy adolescents.
This field is evolving rapidly. Earlier studies attempted to correlate psychiatric diagnoses, mood states, and particular emotions with differences in the size or activity of brain regions, with limited success. Now measurements of the activity of neural circuits—systems involving more than one brain region along with associated white-matter connections—have added another dimension to the picture. Pattern recognition techniques are beginning to refine the picture still further and have the potential to contribute to more precise diagnoses and predictions of treatment response for individual patients.
The picture we now have of bipolar disorder—increased activity of amygdala-centered emotion circuitry, reduced prefrontal control, increased responsiveness to positive and non-emotional stimuli, and overactive reward circuitry—is certainly incomplete, but its elements provide ways of thinking about the dimensions of the mood disturbances we see in patients with bipolar disorder.