Parkinson's patients are bad liars, here may be why

Neurological diseases can be strange in that they often have additional personality effects. If someone gets a cold, they sneeze a bunch but are basically the same person they were before the cold. In contrast, meningitis can include mental status and personality changes in its early stages -- including irritability and sleepiness. When a disease involves the brain, it can change who we are in addition to making us sick.

In this vein, I found this paper in the journal Brain particularly interesting. Abe et al. report that Parkinson's patients tell fewer lies than controls in a task where they were prompted to do so. In addition, the resting metabolic rate in areas of the brain associated with deception in healthy patients is lower in Parkinson's patients.

The authors argue that -- in addition to the movement difficulties associated with the disease -- Parkinson's may erode other parts of the brain resulting in the lying "deficit." (Whether you consider that a deficit or not is a matter of personal preference.)

I really didn't know any of this about Parkinson's. Most of my experience with Parkinson's in school has been about disordered movement and how to manage it. We were also taught that patients can develop dementia in some cases. But I had never heard about personality changes in the disease.

However, Abe et al. note numerous anecdotal reports, dating from the recognition of it as a disease, that patients with Parkinson's are remarkably industrious, inflexible, and honest. They also note that Parkinson's patients have measurably lower metabolism in certain areas of the prefrontal cortex and that this brain region has been implicated in deception:

Certain personality traits have long been noted as being characteristic of Parkinson's disease patients. In 1913, Carl Camp wrote 'It would seem that paralysis agitans affected mostly those persons whose lives had been devoted to hard work ... The people who take their work to bed with them and who never come under the inhibiting influences of tobacco or alcohol are the kind that are most frequently affected. In this respect, the disease may be almost regarded as a badge of respectable endeavor'. Since the publication of Camp's report, many researchers have investigated the association of Parkinson's disease with personality or behavioural traits, and have consistently shown that Parkinson's disease patients have characteristic personality traits such as industriousness, seriousness and inflexibility.

Parkinson's disease patients have also been described as 'honest', in the sense that they tend not to tell lies. Although the possibility that honest people are particularly vulnerable to this disease cannot be ruled out, insidious neuropathological changes in the course of the illness might underlie this specific trait. In relation to this idea, a previous study reported that the personality change in Parkinson's disease patients was primarily the result of the disease rather than aging, and some researchers have suggested the possibility that the personality traits are associated with Parkinson's disease-specific brain damage. However, it may not be the case that Parkinson's disease patients choose not to tell lies, but rather that they actually have difficulty lying due to cognitive deficits resulting from pathological changes in certain brain regions.

One potentially critical contender for the role of mediator in complex cognitive processes such as deception is the prefrontal cortex, a structure known to support executive function. In particular, it is widely assumed that the lateral (especially dorsolateral) prefrontal cortex supports cognitive processes requiring executive function such as response inhibition and cognitive control. Some clinical studies have already implicated the prefrontal cortex as being responsible for executive dysfunction in Parkinson's disease patients. Impairment in the prefrontal executive system can prevent people exhibiting flexible and goal-directed behaviours, which are regarded as essential features of human deceptive behaviour. In support of the clinical findings mentioned above, recent neuroimaging studies involving healthy individuals have provided substantial evidence that the prefrontal cortex is consistently active during the making of deceptive responses relative to honest responses. (Emphasis mine. Citations removed.)

These observations prompted the authors to hypothesize that prefrontal dysfunction in Parkinson's account for the tendency to not lie.

To test this hypothesis, they perform two key experiments.

First, they want to verify that Parkinson's patients actually don't lie. To do this, they compared the performance of patients with relatively early Parkinson's to age- and sex-matched controls. The participants were also matched according to performance on a variety of cognitive tests, to help prevent cognitive ability from being a confound in their data. They showed the participants a series of pictures of common living and inanimate objects. (To confirm that both groups remembered the objects -- and thus did not have a deficit in memory -- they were tested on them afterward. Also, consistent with an absence of a memory deficit, the participants could identify whether they were animate or inanimate.)

Then in the testing phase, the participants were then shown videos of 4 actors. An actor would hold up a picture and ask "Have you seen this?" Prior to the test, the participants were instructed to lie to one of the actors, but tell the truth to the other 3. The experimenters measured the ability to lie successfully. This was quantified as the percentage of times the participant correctly told the truth minus the percentage of times the participant correctly told a lie -- i.e. lied to the actor they were supposed to.

One would assume that the resulting difference would be close to zero. Assuming the participants can remember 1) the right actor to lie to and 2) whether they had seen the picture or not, they should be equally successful at doing so in the Lie or Truth condition. This is exactly what they see in the healthy controls. By contrast, the Parkinson's patients show reduction in the percentage of times they lied correctly.

To be fair, this was on the whole not a large difference -- the difference in correct percentages between the Truth and Lie conditions is a little over 10%. But if you think how difficult the task is (i.e. not), the small difference is not really that surprising. In fact, what is surprising is that the two percentages are different at all.

To establish whether this difficulty in lying has any neurological basis, the authors use PET scanning to determine the resting metabolic rate of Parkison's patient's brains. PET scanning uses a version of glucose that has been tagged with a radioactive fluorine -- 18-F-fluorodeoxyglucose (FDG). FDG cannot be degraded into energy. However, because it is pretty close to glucose, metabolic activity in the brain -- which draws on the energy glucose provides -- tends to concentrate the FDG in areas of metabolic requirement. The radioactive labeling allows us to measure where these regions are and provides some measure of relative metabolic activity.

When the authors performed PET on the brains of Parkinson's patients, they found that the activity in several regions were reduced -- not surprising because Parkinson's disorders many brain systems. This made them wonder where the reduction of brain metabolism correlated with the decreased tendency to lie. By correlating activity changes and the results of the deception task, the authors hoped to assign which region's underactivity resulted in the lying deficit.

Those results are below (Figure 2 from the paper):

Row A shows the regions that show reduced metabolism in Parkinson's patients relative to controls. Row B shows those regions that correlated with reduced lying. Row C shows the negative correlation between metabolism and the difference between correct Truth percentage and correct Lie percentage. These correlations are significant for two regions of the prefrontal cortex. Thus, the authors identified two regions where a deficit in lying correlates with a reduction in brain metabolism.

What can we take away from this paper? Are there any caveats to these conclusions?

I think it is a remarkable paper because they relate hypo-performance on a particular cognitive test to hypo-metabolism in a particular patient population. One issue that I have with imaging studies in general is that they tend to attribute the function to particular brain regions based on showing that activity increases during a particular task -- ergo, that brain region must be involved in the task. However, lots of brain regions are activated during any given task; not all of them are required for that task. It is difficult to define a brain region's function in the absence of a lesion study that shows deficits in that task -- i.e. no brain region, cannot perform the task. That is, in essence, what they have done here. Many studies show that the prefrontal cortex is active during lying, but none that I remember (although I am not a specialist in this) show less lying in a group of individuals with a functional deficit in the prefrontal.

That being said, I do have some caveats. First, I am a dubious about their selection to use resting metabolic rate PET as opposed to PET when the people were performing the deception task. Granted, numerous other studies have shown prefrontal activation during lying. Granted, weird compensatory changes can occur in patient's brains when they perform these tasks, changes that mask diminutions in activity that would seen in the resting state. I still would like to see the results of that experiment.

Further, I am not convinced (yet) that this is a deficit in lying per se rather than a deficit in set shifting. They mention above that Parkinson's patients are also reported to be inflexible. This is actually a common description of most people with executive dysfunction. Patients with executive function have difficulties changing between tasks, even though they can perform each individual task just fine. One way to interpret the Parkinson's patients' difficulty in lying is not that they had trouble lying, they just had trouble changing from telling the truth to lying. They had trouble changing their behavior to perform a novel task.

The authors counter that they looked at trials where the patients was asked to tell the Truth that immediately followed trials where the patients was asked to Lie. If a subject has a problem with set shifting, we would expect performance to be bad on the first trial after the shift from Lies to Truth, but less so on subsequent Truth trials. They found no difference between Truth trials immediately after a Lie, and those later on.

Color me unconvinced. First, I would like to see analysis showing stable performance after the second Truth trial in a row. Second -- although this would be difficult if only a quarter of the trials are Lies -- I would like to see performance on repeated lying stay relatively stable. Basically, I think that there is a performance deficit because of set shifting difficulties that would be observed if the patients were allowed to do many trials of the same type in a row.

Finally, let's make a little philosophical point: we can't really know why the Parkinson's patients aren't lying. It may be -- and the authors provide evidence for this view -- that they have difficulty lying because of something wrong with their brains. Alternatively, it may be that changes in their personality from the disease make them no longer want to lie. They can, but choose not to. I find this argument a tad far-fetched, if only because lying in the context of this study had no tangible consequences. However, it is tough to determine their motivation for telling the truth from just this experiment.

Still, in spite of those caveats, I think this is a very interesting study that definitely adds both to our understanding of Parkinson's and our understanding of the neurology of lying. Great work.

Abe, N., Fujii, T., Hirayama, K., Takeda, A., Hosokai, Y., Ishioka, T., Nishio, Y., Suzuki, K., Itoyama, Y., Takahashi, S., Fukuda, H., & Mori, E. (2009). Do parkinsonian patients have trouble telling lies? The neurobiological basis of deceptive behaviour Brain DOI: 10.1093/brain/awp052