Monday, January 19, 2015

Objectivity? A Role For Emotion In Decision Making

One of the reasons I like listening to Public Radio is that they provide a lot of clinicopathological case material that is usually quite illustrative, and frees me from the huge hurdle of being suspected of disclosing confidential patient information on this blog.  Just a note on the historical context.  When I started out, deidentified clinical information was a mainstay of teaching.  It was presented at case conferences and in medical journals.  At some point that became a lot less likely and in my opinion that adversely affects teaching in a way that could be dangerous to the health of patient.  The best physicians depend on pattern matching to recognize diseases and many of those patterns are recalled not just from live patients but also pictures, images, and numbers that are remembered independent of any real contact with a live patient.  When an administrator did not allow me to use deidentified MRI scan images for teaching residents, those residents end up knowing a little bit less, not in terms of book learning but in terms of the experiential aspects of medicine.  The most unique technical skill that your physician has that nobody else does is access to a vast array of patterns that were experienced in medical school and post graduate training.

I was driving around today, listening to public radio when a show came on called Radiolab.  I have heard it before and it is interesting because it tried to present science in interesting ways and in many cases that involves medicine.  Today's show was all about choices and I happened to pick it up about 1/3 of the way in or roughly the 20 minute mark.  At that point Antoine Bechara, MD, PhD began discussing the case of Elliot, a young accountant.  Elliot was working for a corporation as a successful upper level manager.  He was married and had children.  He was considered to be smart, successful, and religious.  One day a small tumor was discovered in his orbitofrontal cortex and it was successfully removed.  Post surgical neuropsychological testing showed that was still in the 97th percentile in terms of IQ testing.  He returned home and then went back to work.

What occurred following the successful neurosurgery was unexpected.  He was no longer able to make even routine decisions.  As an example, when he tried to decide what pen to use to sign a contract, it took him 30 minutes to decide whether to use a black pen or a blue pens.  All possible permutations of the decision were explored and evenly considered.  In the case where many more choices were available (the program used the example of a breakfast cereal aisle in the grocery store), the decisions became more impossible.  He was so disabled by this problem that he lost his job and eventually his marriage and family.   He got involved with a con man.  He lost his savings and went back to live with his parents.  Somewhere along the way he was seen by the behavioral neurologist Antonio Damasio, MD, PhD who tested him with visual stimuli designed to elicit strong emotional responses.  These visual stimuli failed to elicit these responses in the patient.  Damage to his orbitofrontal cortex had caused this disconnection.  Disconnecting the emotional response resulted in an impairment in decision making rather than and expected improvement.  Without the feeling state he was pathologically indecisive.

Major Anatomical Connections of the Ventral Medial Prefrontal Cortex from:  Euston DR, Gruber AJ, McNaughton BL. The role of medial prefrontal cortex in memory and decision making. Neuron. 2012, figure 3 with permission.

When I heard Dr. Bechara talking I remembered an excellent presentation that he gave on how people with addictions tend to respond to the Iowa Gambling Task (GT) and how some responses in that paradigm are consistent with increased risk for substance use.   This test looks at subjects attempts to optimize a $100 bet on choices from 4 decks of cards (A, B, C, and D).  The subject is to make 100 card selections in any order from any of the decks.  A selection from deck A or B results in a $100 reward.   There are unpredictable punishments so that the subject loses $1250 in every 10 cards selected from deck A or B.  Selections from decks C and D result in a $100 reward with unpredictable punishments resulting in a loss of $250 in every 10 cards from decks C and D.  The penalties are not fixed and some of them are substantial.  Take a look at this video for an example of how it works (the initial sum used in the video is substantially higher than quoted in the research literature).  Normal subjects eventually learned that they are more likely to get punished choosing from decks A and B and they will gravitate toward decks C and D.  The Iowa group used this test paradigm and modifications to investigate aspects of decision making in the ventromedial prefrontal (VM) cortex (bilateral lesions to the gyrus rectus, mesial half of the orbital gyrus and the inferior half of the medial prefrontal surface).

Subjects with lesions in the VM do not reduce their selection of decks A or B or increase their selection of choices in decks C and D.  The impairment in decision making can be replicated over time.  In order to investigate whether any emotional process was involved, the investigators looked at skin conductance resistance (SCR) associated with the decisions.  They looked at a window of +/- 5 seconds on either side of the decision to examine anticipatory, reward and punishment SCRs.  Normal subjects develop anticipatory SCR and they are more pronounced before selections from the disadvantageous decks (A and B).  Even the 20% of normal subjects who are self professed risk takers develop anticipatory SCRs but they are lower in magnitude when selecting for the disadvantageous versus advantageous decks.   VM subjects had no anticipatory SCR suggesting that these patients had a compromised ability to change their somatic state (skin conductance) in anticipation of an imagined scenario in an uncertain condition.

The researchers also looked at the question about whether biases in this paradigm were conscious or not.  The experiment in this case used the same decks A, B, C, and D but the task was broken up into 4 different zones.  The subject was asked about their explicit knowledge of what was happening in the game after every 10 cards.  The 4 periods included:

1.  Pre-punishment period before encountering punishment.
2.  Pre-hunch period as punishment was being encountered by the subjects till had no ideas about the game.
3.  Hunch period where guesses about favorable decks begin to appear.
4.  Conceptual period when they have a clear idea about the advantageous versus disadvantageous decks.

In normal subjects the SCRs were absent pre-punishment but began to build and was sustained.  Although 30% of controls never got to the conceptual periods they all had SCRs and played the game correctly.  50% of VM subjects got to the conceptual stage in that they could explicitly state the deck types.  That did not result in them correcting their choices.  In real life this means the patient with frontal lobe damage has an awareness of what is right and what is wrong but the correct choice is not made.  The authors use the example of a person with a substance use disorders balancing the choice between taking a drug as an immediate reward and the long term reward of a stable home, family and work life they choose the drug.  The GT has been used to study the issue of substance users and impairment in decisions is noted.  

The wiring and impact of various signaling systems on the vmPFC is complex.  It is hard to imagine methods that would allow the isolation and correlation of any of these systems suggested in the clinical vignette about the patient with the brain tumor.  The neuroanatomy is also complex.  Many of us were taught to consider the supraorbital area of frontal cortex to be typical frontal cortex,  It turns out that the most medial gyri that represent the vmPFC and are more appropriately considered limbic cortex.   Looking at a recent post on the involvement of the nucleus accumbens in decision making now provides two avenues for advancing decisions - emotions and reward pushing these decisions forward.

Apart from psychiatric disorders and addictions, these brain systems have profound implications for everyday life and the illusion of free will.  Many of the biases in everyday life that many of us would deny that we have, may be the product of the reward and/or emotional valence assigned to that string of associations through these mechanisms.  Many of these biases are unconscious.  I think there is widespread confusion that emotions compromise objectivity (as in rational decision making).  One of the main outcomes of these studies is that emotions are necessary make a decision and do not necessarily compromise the rational aspects of that decision.  The other approach I see written about is the idea that there is a reptilian brain lying deep inside the human brain and this has a characteristic response pattern (anger/rage).  It was popular to talk about reptilian brains when I first learned neuroanatomy, but a lot less was known about the integration of the human brain at that time.

George Dawson, MD, DFAPA


1:  Radiolab:  Choice

2: Bechara A, Damasio H, Damasio AR. Role of the amygdala in decision-making. Ann N Y Acad Sci. 2003 Apr;985:356-69. Review. PubMed PMID: 12724171.  From a special ediction of the journal called: THE AMYGDALA IN BRAIN FUNCTION: Basic and Clinical Approaches

3: Bechara A, Damasio H, Damasio AR. Emotion, decision making and the orbitofrontal cortex. Cereb Cortex. 2000 Mar;10(3):295-307. Review. PubMed PMID: 10731224.  This is from a special edition of this journal called:  The Mysterious Orbitofrontal Cortex

4: Euston DR, Gruber AJ, McNaughton BL. The role of medial prefrontal cortex in memory and decision making.  Neuron. 2012 Dec 20;76(6):1057-70. doi: 10.1016/j.neuron.2012.12.002. Review. PubMed PMID: 23259943; PubMed Central PMCID: PMC3562704.

Supplementary 1:

Figure 3 above was reprinted from Neuron, Vol. 76 edition number 6, Euston DR, Gruber AJ, McNaughton BL. The role of medial prefrontal cortex in memory and decision making, Copyright (2012), with permission from Elsevier.  License # 3542200221086 License date Jan 04, 2015 per the Copyright Clearance Center.

Supplementary 2:

I use the following human neuroanatomy text by Paxinos and Mai with the accompanying Atlas of the Human Brain by J├╝rgen K. Mai, Joseph Assheuer, and George Paxinos.  It was recommended to me by Lennart Heimer after I took one of his courses in brain dissection at Washington University.  I requested permission from the publisher to use some of these figures for teaching purposes and the fees were astronomical.  So the text and atlas are primarily useful to clarify your own thinking rather than preparing presentations.

No comments:

Post a Comment