DOMINION Study Drills Down on Impulse Control Disorders in Parkinson's Disease

Susan Jeffrey

May 13, 2010

May 13, 2010 — Results of a large cross-sectional study show that impulse control disorders (ICDs), including such behaviors as compulsive gambling, shopping, eating, and hypersexuality, occur in about 14% of all patients with Parkinson's disease (PD). That number climbs to about 17% of those patients taking a dopamine agonist, 2 to 3 times higher than those not receiving a dopamine agonist.

All 4 behaviors appeared to occur with similar frequency and at similar rates with either pramipexole or ropinirole, the dopamine agonists most commonly prescribed in this cohort. Interestingly, there was also a relationship between these disorders and levodopa treatment, particularly at higher doses.

The findings highlight the need for patients with PD to be screened for a range of ICD symptoms as part of routine clinical care, said lead study author Daniel Weintraub, MD, from the University of Pennsylvania, Philadelphia.

"Clinicians probably should have a conversation with their patients prior to initiating treatment about monitoring for these symptoms and being aware of their possible development," Dr. Weintraub told Medscape Neurology.

The findings from phase 1 of the cross-sectional DOMINION study are published in the May issue of the Archives of Neurology.

Results from phase 2 of the DOMINION study, a case-control study using subjects drawn from the phase 1 cohort, were presented recently at the American Academy of Neurology (AAN) 62nd Annual Meeting.

Here, the researchers found that ICD patients are much more functionally impaired, are more depressed and anxious, and have more obsessive-compulsive symptoms, as well as scoring higher on scales of novelty seeking and impulsiveness, compared with PD patients who had not developed such disorders.

"What's intriguing about this group of patients is that they're all exposed to the same dopamine agonist, but for some reason, some people develop specific behaviors, and not other behaviors, and the question is why that's the case," said presenter Valerie Voon, MD, from the University of Cambridge, England.

"It may be that some of these associated factors might reflect underlying mechanisms or individual susceptibilities, and we clearly need more prospective studies to identify risk factors for developing these behaviors," she concluded.

Compulsive Behaviors

Impulse control disorders have previously been reported in PD, particularly in association with dopamine agonist treatment, but also with levodopa therapy and deep-brain stimulation surgery, the study authors write. Previous studies, though, have had relatively small sample sizes, examined a limited range of ICDs, and did not examine clinical correlates with development of these disorders.

The DOMINION study included 3090 PD patients treated at 46 movement disorder centers in the United States and Canada. Each was assessed using the Massachusetts Gambling Screen for current problem or pathological gambling, the Minnesota Impulsive Disorders Interview for compulsive sexual behavior and buying, and the Diagnostic and Statistical Manual of Mental Disorders research criteria for binge eating.

Overall, an impulse control disorder was identified in 13.6% of all PD patients, and 3.9% of all patients had 2 or more disorders.

"We found that the relative frequencies of the disorders were pretty similar; there were some differences, but all 4 were relatively common," Dr. Weintraub noted. "The one that was least common was compulsive sexual behavior, and the reason for that appears to be that almost no females met criteria for that, so that only occurred in male patients."

Table. Prevalence of Specific Impulse Control Disorders in Parkinson's Disease Patients

Impulse Control Disorder Patients, %
Gambling 5.0
Compulsive sexual behavior 3.5
Compulsive buying 5.7
Binge-eating disorder 4.3


As expected, ICDs were more common in those treated with a dopamine agonist, occurring in 17.1% vs 6.9% among those not taking a dopamine agonist (odds ratio [OR], 2.72; 95% confidence interval [CI], 2.08 – 3.54; P < .001). The frequency was similar though between those taking pramipexole vs ropinirole (17.7% vs 15.5%; OR, 1.22; 95% CI, 0.94 – 1.57; P = .14).

The large sample size also allowed them to look at other variables that might be associated with developing an ICD, Dr. Weintraub said, "and some interesting things emerged from this."

Among them was that younger age (<65 years) was independently associated with having an ICD, being not married as opposed to married, residence in the United States vs living in Canada, current smoking, and having a family history of gambling problems.

"Levodopa treatment was also independently associated with having an impulse control disorder, although the strength of the association, or the odds of having an ICD, was approximately half for levodopa as opposed to dopamine agonist treatment," Dr. Weintraub noted.

The association with levodopa treatment was particularly evident at higher doses, but there was no such dose-response relationship seen with dopamine agonist therapy, he added. "It only mattered whether you were on or not on a dopamine agonist, not how much you were on, whereas for levodopa it seemed to make a difference how much you were taking, so that's an important distinction to make."

What is still not clear is what the optimal management strategies are for these patients, he added. Sometimes, reducing dopamine agonist doses or switching to a different dopamine agonist can help, but in some cases, he notes, the only ethical decision is to withdraw therapy, for example, if the patient is on the verge of bankruptcy or a danger to himself or herself because of compulsive behaviors.

Dr. Weintraub has received a grant from the Michael J. Fox Foundation to study whether adjunctive treatment with naltrexone, a drug that has been shown effective in treating pathological gambling in the general population, might be efficacious in reducing ICDs in patients with PD receiving dopamine agonists.

Case-Control Results

In her presentation at the AAN meeting, Dr. Voon reviewed results of a case-control sample taken from the cross-sectional cohort. Of 420 patients with ICDs, 282 agreed to take part in the case-control study and were matched to 282 patients without ICDs for age, sex, and dopamine agonist treatment.

They were then subdivided into those expressing compulsive gambling, shopping, hypersexuality, binge eating, or mixed behaviors, with between 40 and 60 patients per group for comparison.

Using a variety of scales, including the Geriatric Depression Scale, the State Trait Anxiety Inventory, the Obsessive Compulsive Inventory, novelty seeking (Temperament and Character Inventory), delay discounting (that is, the ability to pass up an immediate reward for a much larger but deferred reward), Barratt Impulsivity Scale, and PD clinical measures, including the United Parkinson's Disease Rating Scale, they found that, overall, patients with an ICD had greater functional impairment, were more depressed and anxious, and had more obsessive-compulsive symptoms than PD patients without an ICD. They also had higher novelty-seeking scores and higher impulsivity scores, Dr. Voon noted, "and these have been shown to some extent in much smaller studies."

As had been seen in the cross-sectional component of DOMINION, "we see there is a clear association with higher levodopa doses in the ICD patients," she said, although no such association was seen with the dopamine agonist dose.

They then compared clinical features among each ICD subgroup — pathologic gambling, shopping, hypersexuality, and eating — with the PD controls. Compared with PD patients without ICDs, patients with pathologic gambling and shopping problems were more depressed and more anxious and showed greater obsessive-compulsive symptoms, greater discounting of delayed rewards, and greater novelty seeking.

This latter finding is "intriguing," Dr. Voon noted, because novelty seeking in the overall population of PD patients has been shown to be decreased, possibly related to neurodegeneration affecting the dopaminergic projections of the striatum, whereas higher novelty seeking is a clear associated factor with substance use disorders in the general population.

The size of this cohort allowed them to look more closely at subfactors of the novelty seeking scale, such as exploratory excitability — trying new things for fun or thrills — vs stoic rigidity; disorderliness vs regimentation — being more rule-bound vs being more flexible with rules and regulations — and impulsiveness.

"What we found was that the gamblers and shoppers overall had higher novelty seeking, but we clearly did not see that in patients with hypersexuality and eating behaviors," Dr. Voon said. "We see that it's actually driven by elevations in impulsiveness and also this disorderliness, so they're less bound by rules and regulations, but we didn't necessarily see this in terms of exploratory or extravagance behaviors."

On the Barratt Impulsivity Scale, they also found differences between the subgroups on rapid decisions and the delayed discounting task measuring tendency to immediate vs delayed gratification. Compulsive gamblers and shoppers were much more likely to make rapid decisions as were binge eaters, but not the hypersexuality patients. Compulsive gamblers and shoppers were also more sensitive to immediate gratification compared with the other 2 groups, she noted.

Interestingly, the compulsive shoppers and hypersexuality patients were taking higher levodopa doses, whereas the other 2 groups were not, but again there was no interaction with dopamine agonist dose.

"Differences between behavioral subgroups suggest state differences in behavioral correlates that may reflect differences in underlying mechanisms and individual susceptibilities," the study authors conclude.

The DOMINION study was funded by Boehringer Ingelheim and designed jointly by Boehringer Ingelheim and a scientific advisory board including 6 of the 9 coauthors. Dr. Weintraub reports he has received consulting fees, honoraria, or grant support from Boehringer Ingeleheim, BrainCells, Merck Serono, Novartis, Ovation, and Wyeth. Dr. Voon has disclosed no relevant financial relationships. Disclosures for the coauthors appear in the Archives of Neurology publication.

Arch Neurol. 2010;67:589-595.

American Academy of Neurology (AAN) 62nd Annual Meeting: Abstract S58.002. Presented April 15, 2010.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.