Friday, June 27, 2014

Effort of movements in Parkinson’s disease

The very terms that we use to describe the motor symptoms of Parkinson’s disease (PD) imply a subjective scaling of time and space: bradykinesia (slowness of movement), tachyphemia (cluttering of speech), and micrographica (smallness of handwriting).  Although these symptoms are stable features of the disease, a remarkable property of PD is that under some conditions the symptoms can spontaneously improve.

In 1965, R.S. Schwab and I. Zieper, two neurologists at the Massachusetts General Hospital, described the case of a 62-year old male PD patient who exhibited severe tremor and severe rigidity and was totally dependent on his wife.  His wife would start her day by dressing him, laying out his breakfast, making his lunch, go to work, then come back in the afternoon to make his dinner and finally get him undressed and ready for bed.  One evening his wife had severe abdominal pain and had to be taken to the hospital for emergency surgery.  The next day she woke worried about her husband, and was surprised when the nurse told her that he had come to visit her.  He had dressed himself, made his own breakfast, and then took a taxi to the hospital.  At the hospital his neurologist noticed him and upon examination found that he was able to walk 50% faster than in past examinations.  “All his motor tests were improved in spite of the presence of the same amount of rigidity and tremor that had been present before.”

A second case was another elderly male with advanced stage PD with severe rigidity who was confined to a wheelchair, unable to walk alone, living on the first floor of his home in Providence, RI.  A hurricane approached the city and his wife left to get some supplies from the drugstore.  “As a result of the storm the harbor overflowed 10 feet into the street.  The patient, sitting in his wheelchair, suddenly saw the door blown in and a wall of water entered the house.  Exactly how he did it is not clear, but he managed to get out of his wheelchair and climbed the steps to safety on the second floor where he was found several hours later by his wife, the waters having subsided. She found him seated in a chair as helpless as he was before.”

While these examples are anecdotal, there are other more controlled instances in which the PD patients show marked improvements in their movements.  One example of this is in the movements that are made during sleep.  Although healthy people do not move during REM sleep, people with PD sometimes experience REM sleep behavior disorder (RBD).  Valerie Cochen De Cock and her colleagues studied movements made during sleep by PD patients and reported that the movements were “surprisingly fast, ample, coordinated and symmetrical, without obvious signs of parkinsonism”.  They found one patient singing a song with a “strong and sonorous voice, a wide smile on his face” (he used to sing before his PD), another “declaiming political speeches with a loud voice” (he used to give speeches at the town council), another “shouting and getting hold of a heavy oak table and throwing it across the room”, and another “fighting with an invisible foil, with great agility” (apparently to save his lady-love from an attacking knight).

The mechanisms with which the brain of a Parkinsonian patient produces these feats remain a complete mystery.  But these observations do hint that latent in the PD brain is the ability to make fairly normal movements.  Yet, the movements are apparently unavailable for expression except under extraordinary circumstances.  Why?

Neuroeconomics of movements

Pietro Mazzoni, Anna Hristova, and +John Krakauer studied this question by asking PD patients and healthy controls to reach with their dominant (and more affected) arm to a target.  Visual feedback for the hand was removed at reach onset, and at the end of each reach the volunteers were given feedback with regard to the speed and accuracy of their movement.  Crucially, the trial had to be repeated if the speed was outside the requested range.  The authors found that for a given reach velocity, the endpoint accuracy of the movements made by the PD patients was similar to controls.  This again illustrated the latent abilities of the patients.  However, the patients required many more attempts in order to produce a reach that was as fast as the requested speed.  That is, the patients were capable of producing movements of normal speed and accuracy, but it took them more trials to become motivated to make the fast movements.  The authors proposed that under normal conditions, the patients seem to lack the “motor motivation” that healthy people possessed in generating their movements.

I have suggested that one way to view this result is to consider the possibility that in the brain, each movement is a balance between two factors: the reward that one expects to acquire at the end of the movement, and the effort (or motor cost) that will be spent in generating that movement (Shadmehr et al., Journal of Neuroscience, 2010).  The reward that we expect to acquire represents the subjective value of the movement.  For example, if you see a dear friend, the subjective value for the steps that you are about to take toward your friend are higher than if you are walking to greet someone that you may not be so fond of.  As a result, you will walk faster toward the dear friend.  (I have often thought that to examine how my brain currently values people in my life, I should measure the speed at which I walk toward them.)

Indeed, humans and other animals tend to move faster toward things that they value more.  This was first illustrated by Okihide Hikosaka and his colleagues in saccadic eye movements of monkeys.  In these experiments, thirsty monkeys were trained to move their eyes to a location in exchange for a reward (juice).  In some blocks of trials, the juice volume was a little larger, and in some blocks the volume was a little smaller.  The peak velocity of the saccadic eye movements in blocks in which there was more juice at stake was larger.  That is, the monkey’s eye movements were faster when the subjective value of the movement was higher.

In the real world we do not make saccadic eye movements in exchange for juice.  Rather, we move our eyes to place the part of the visual scene that we are interested in examining on our fovea.  Do we make faster saccades to things that we value more?  In humans, this idea was first illustrated by my former student +Minnan Xu-Wilson.  She asked people to make a saccadic eye movement to spots of light, but after the saccade was completed she ‘rewarded’ them by showing them a picture of a face, an object, or simply a noisy picture.  She found that saccades that were made in anticipation of viewing a face were faster.

These experiments illustrate that one of the factors that influences the speed by which we move, that is the vigor of our movements, is the subjective value of the reward that we expect to attain at the end of the movement.  The higher this expected value, the faster the movement.

The second factor is the subjective cost of the effort that is required to make the movement.  If the subjective value of the reward associated with two potential movements is the same, people pick the movement that requires less effort.  

Now suppose that we have to move a given distance.  How does the brain decide on the speed of the movement?  The faster the speed with which we move to cover that distance, the greater force we have to produce.  If effort is related to force (perhaps because of metabolic cost of generating force), then the subjective cost of effort will be higher for the faster movements as compared to the slower movements that cover the same distance.  So if we move slower, we will produce smaller forces with our muscles and have a lower subjective cost of effort. 

However, the slow movement will bring us to our goal later.  Time discounts reward.  That is, it is better to arrive at a valuable state sooner rather than later.  So the subjective value of the movement drops if we arrive later at the destination, making it better to move fast so we get to our goal sooner. 

In summary, the subjective cost of effort makes it better to move slow so we produce smaller forces, but passage of time makes reward less valuable.  These two factors compete and the movement that the brain produces appears to be one that is the best possible given these two competing factors.  That is, the speed at which we move is one that produces the smallest possible effort (encouraging us to move slow), while at the same time maximizing the subject value of the reward we hope to attain (encouraging us to move fast).

Dopamine disorders alter the neuroeconomics of movements

In Parkinson’s disease, some of the neurons in the substantia nigra, a nucleus in the basal ganglia, gradually degenerate and die.  These neurons provide dopamine to much of the brain, and in particular the striatum, another region of the basal ganglia.  Dopamine appears to play a critical role in regulating the two factors that control movements: subjective value of reward and subject cost of effort.

In the course of the last two decades, John Salamone and his colleagues have been investigating the effects that loss of dopamine has on behavior of rats.  When rats are offered a choice between pressing a lever a few times to obtain good food, vs. eating a less preferred food for which they do not have to press levers, they choose to spend the effort and press the lever to get the preferred food, but only if the lever pressing requires modest effort.  But when a drug is injected into their basal ganglia that acts as an antagonist to dopamine, the rats become less willing to press the lever and forego the better food, settling for the less effortful choice.  On the other hand, if a drug is injected that enhances action of dopamine, the animal becomes more willing to press the lever, even if it has to press it many times in order to earn the better food.  

Therefore, it appears that when dopamine’s actions are disrupted, the balance between subjective value of reward and cost of effort shifts.  Loss of dopamine shifts the balance by increasing cost of effort and decreasing value of reward, whereas increase of dopamine shifts the balance by decreasing cost of effort and increasing the subjective value of reward.  

In this framework, loss of dopamine in PD shifts the neuroeconomics of movements towards ones that have smaller effort costs, which include movements that are slow. This speculation would not explain why certain movements of the patients are better during REM sleep, but does provide a framework for understanding the paradoxically fast and able movements that they exhibit under extraordinary circumstances: perhaps under these conditions, a greater proportion of available dopamine is engaged, increasing the expected reward for the movement, countering the effort costs.