Background: At least 14% of Parkinson disease (PD) patients develop impulse control disorders (ICDs). The pathophysiology behind these behaviors and the impact of deep brain stimulation in a real-life setting remain unclear. Objectives: We prospectively examined the impact of bilateral subthalamic nucleus deep brain stimulation (STN-DBS) on ICDs in PD patients, as well as the relationship between impaired sensorimotor gaiting and impulsivity. Methods: Patients undergoing bilateral STN-DBS were assessed for ICDs preoperatively and 1-year postoperatively using a validated questionnaire (QUIP-RS). A subset of patients completed the Balloon Analogue Risk Task (BART) and auditory prepulse inhibition (PPI) testing. Results: Analysis revealed 12 patients had an improvement in score assessing ICDs (‘good responders'; p = 0.006) while 4 had a worse or stable score (‘poor responders'; p > 0.05). Good responders further exemplified a significant decrease in hypersexual behavior (p = 0.005) and binge eating (p = 0.01). Impaired PPI responses also significantly correlated with impulsivity in BART (r = -0.72, p = 0.044). Discussion: Following bilateral STN-DBS, 75% of our cohort had a reduction in ICDs, thus suggesting deep brain stimulation effectively manages ICDs in PD. The role of impaired PPI in predisposition to ICDs in PD warrants further investigation.

Impulse control disorders (ICDs) such as pathological gambling, hypersexuality, compulsive shopping, binge eating, punding and dopamine dysregulation syndrome are recognized as ‘behavioral addictions' that develop in at least 14% of PD patients, and can lead to devastating social, familial and financial consequences [1, 2, 3]. Unfortunately, the current treatment options for ICDs are restricted to decreases in dopamine agonists (DAs) [4, 5] or the addition of psychoactive medications [2], both of which have disadvantages. Due to such a paucity of treatments, alternative therapies for ICDs such as subthalamic nucleus deep brain stimulation (STN-DBS) have significant implications for PD [6].

Why ICDs develop in PD patients remains ill defined. There is some evidence that sensorimotor gating, as measured by prepulse inhibition (PPI), may be irregular in patients with PD, schizophrenia and obsessive compulsive disorder [7, 8, 9, 10]. PPI is known to be disrupted in disorders with basal ganglia dysfunction due to a lack of midbrain dopamine [9, 10]. A common measure of sensorimotor gating is a startle response [9]. PPI can be induced when a weak sensory event preceding a strong startle-inducing stimulus reduces the magnitude of the startle response [11, 12, 13]. Abnormalities signify the inability to properly attend to or ignore environmental stimuli. Impaired PPI may lead to ICDs in certain PD patients.

There is mounting evidence that STN-DBS may also reduce ICD behavior in PD patients. Specifically, in a study by Lhommée et al. [3], STN-DBS ameliorated dopamine dysregulation syndrome in 4/4 patients, behavioral addictions in 17/17 patients and compulsive dopaminergic medication use in 9/9 patients. In their study, however, DA reduction was more dramatic than typically occurs in practice. This study aims to shed light on the role of deep brain stimulation in ICDs, as well as to examine the role of PPI in impulsivity secondary to PD.


Consecutive patients undergoing bilateral STN-DBS for treatment of PD were offered participation. Those who qualified for surgical treatment completed the Unified Parkinson's Disease Rating Scale (UPDRS) and neuropsychological testing as part of routine preoperative workup. Patients who did not improve more than 30% on CAPSIT ON/OFF medication testing were not considered acceptable candidates for surgery, as well as those who demonstrated dementia (DRS score ≤130) or significant cognitive impairment at baseline testing. Subjects who could not complete testing due to language barriers and/or dementia were excluded from participation in the study. Institutional review board approval for the study was obtained.

Questionnaire for Impulsive-Compulsive Disorders in Parkinson's Disease

After giving informed consent, participants completed an impulsivity questionnaire, the Questionnaire for Impulsive-Compulsive Disorders in Parkinson's Disease Rating Scale (QUIP-RS), which was given within 1 month preoperatively and 1 year postoperatively to track changes in behavior. The QUIP-RS was specifically developed to analyze impulsive behaviors in the Parkinson's patient population - including both the screening and follow-up of impulsivity in PD patients [14]. The questionnaire was administered in the clinic with a research associate available for assistance (e.g. help filling out questionnaires if writing was difficult due to tremors).

Balloon Analogue Risk Task and PPI

A subset of patients consented to additional testing consisting of the Balloon Analogue Risk Task (BART) and auditory PPI. Both were performed preoperatively on medications. BART is a computer game validated to test impulsivity [15, 16]. The goal of the participant is to collect as much virtual money as possible by pumping up 30 balloons at a rate of 5 cents per pump. The more pumps, the more potential money they will earn; however, if the balloon pops, they will lose the money. At any point, they may choose to stop pumping and press ‘collect $$$', whereby the money will be safe while they complete the rest of the 30 balloons. The balloon randomly pops between 1 and 128 pumps. The two measurements included for analysis were the average adjusted pump count and the total number of explosions.

In PPI, to elicit the startle response, subjects were exposed to a strong startling tone (115 dB) in the presence and absence of a prestartle ‘warning tone' (74 dB). The orbicularis oculi contractile response was concurrently recorded via EMG (Nicolet Viking Quest EMG Machine), and the response data was analyzed for maximal response per auditory stimuli. EMG activity (filter 80-2,500 Hz) was continuously recorded from the right orbicularis oculi and digitized (sampling frequency 1,000 Hz). The EMG continuous recording filters for PPI were: interview scope sweep 100 ms/div, sensitivity 0.2 mV/div, main scope sweep 10 ms/div, input sensitivity 30 μV/div. PPI was assessed by measuring the reduction of the startle response. Patients with severe head tremors that interfered with EMG recordings were excluded, as well as those with hearing impairments. Auditory PPI results were correlated with the BART impulsivity results using Pearson's correlation analysis.


The mean age of our subjects at the time of surgery was 57.63 years (range: 34-73) with an average disease duration of 12.19 years prior to surgical intervention (range: 6-20). There were 10 male participants and 6 female participants. Table 1 outlines the demographics as well as the lead locations and programmed stimulation settings of patient devices.

Table 1

Patient demographics

Patient demographics
Patient demographics


Analysis of the QUIP-RS results showed that 12 patients had an improvement in ICD score [‘good responders (GRs)'] while 4 had a worse or stable score [‘poor responders (PRs)'; fig. 1] with no obvious demographic differences (see online suppl. table 1; for all online suppl. material, see The 12 GRs had a significant decrease in ICD score [preoperative 11.75 ± 2.62 to postoperative M = 3.42 ± 1.00; t(11) = 3.36, p = 0.006], while the 4 PRs had insignificant increases in scores [preoperative 12.50 ± 5.55 to postoperative M = 18.00 ± 1.78; t(3) = -1.29, p = 0.288] (fig. 1). When examining specific impulsive behaviors in GR patients, analyses revealed a significant decrease in hypersexual behavior [t(11) = 3.55, p = 0.005] and binge eating [t(11) = 3.08, p = 0.01] following STN-DBS (online suppl. fig. 1).

Fig. 1

Significant decrease of ICD score in GRs to STN-DBS in comparison to PRs (n = 16). PreOp = Preoperatively; PostOp = postoperatively. * p < 0.05.

Fig. 1

Significant decrease of ICD score in GRs to STN-DBS in comparison to PRs (n = 16). PreOp = Preoperatively; PostOp = postoperatively. * p < 0.05.

Close modal

Treatment Differences

In an effort to determine differences between the GR and PR subsets, we examined the total daily levodopa equivalent dose and DA dose. There was a significant decrease in levodopa equivalent dose over time [t(15) = 4.62, p = 0.0003], but no difference between groups. Similarly there was no difference in total DA dose between groups (p = 0.257), lead location and/or stimulation parameters. Two of the 4 PRs had increases in DA versus 2 of 12 GRs. Of the 2 PRs who did not have DA increases, one patient had a preexisting abuse history of cocaine, amphetamines, marijuana and hallucinogens, suggesting addictive predispositions preceding PD. The other patient developed a unique sensory wearing-off phenomenon characterized by upper abdominal pain, which was only responsive to levodopa after surgery (despite continued programming and other medications). QUIP showed a de novo appearance of gambling, punding and hobbyism, and compulsive use of PD medication. In real life, he did not gamble or exhibit hobbyism.

Of note, 3 of 4 nonresponders had double monopolar settings, in comparison to 5 of 12 GRs. Only one PR had interleaved stimulation programming in contrast to 5 of 12 GRs. Monopolar stimulation was more commonly used in GR as compared to PR (82.4 vs. 17.6%, respectively).

PPI and Impulsivity

Analysis of the PPI and BART data from a separate subset of patients (n = 8) revealed a significant correlation between both overall PPI responses and total number of balloon explosions in BART (r = -0.72, p = 0.044), as well as responses to nonstartling auditory stimuli and total number of balloon explosions in BART (r = -0.77, p = 0.026; online suppl. fig. 2). There were also trends between responses to nonstartling stimuli and total number of balloon pumps in BART (r = -0.67, p = 0.069).

Reduction in ICDs

We examined how STN-DBS affects impulsivity in PD as well as the possible influence of abnormal sensorimotor gating. We demonstrated that 75% of patients (12/16) had improvement in ICDs following surgery. This prospective study is the first to examine the effect of STN-DBS on ICDs prospectively in routine clinical settings.

Our findings are in agreement with the finding that STN-DBS coupled with aggressive DA reduction results in ICD reduction [3]. Other previous reports regarding the impact of STN-DBS on impulsivity in PD have been inconsistent [17, 18, 19]. One retrospective study showed a decrease in pathological gambling following bilateral STN-DBS, and while a general trend supporting this finding was found in our study, the change did not reach statistical significance [20]. Conversely, several case reports have documented the development of ICDs de novo following surgery [21, 22, 23].

In our study, there was a variable response observed to STN-DBS. In 2 PRs, the DA dose increased postoperatively. Why the 2 GRs with the DA increase did not see the same rise in ICDs suggests that it is not DAs alone that result in postoperative ICDs. When examining stimulation parameters in GRs versus PRs, there are many variables that may account for the differences in response. First, inadvertent stimulation of limbic and associated territories may account for increases in impulsivity. In contrast to delineation of striatal territories, the borderlines between the functionally segregated territories in the subthalamic nucleus overlap [24, 25]. With this model, one can argue overflow of stimulation into unintended areas of the subthalamic nucleus causing unwanted side effects.

Also 3 PR patients were stimulated with 1 or 2 medially located active contacts. Increased charge needed to control motor symptoms may have elicited impulsivity based on spreading of the current to the medial and anterior limbic areas.

Further, there were more double monopolar settings in PRs as opposed to more interleaving in GRs. Interleaved stimulation allows for two programs to be used in an alternating fashion on the same lead and has been shown to reduce side effects [26]. Theoretically, interleaving shapes the current to reduce spread, whereas double monopolar stimulation spreads the current. This finding suggests that interleaved stimulation might reduce ICDs in PD.

Mechanisms of Impulsivity

We also showed a significant negative correlation between impaired PPI and impulsivity on BART, suggesting an important disruption in sensorimotor information gating. Disorders with impaired PPI including PD all generate abnormalities in cortico-striato-pallido-pontine and cortico-striato-pallido-thalamic circuitry, which notably modulate PPI [8, 9, 10]. Therefore, neurodegeneration associated with PD is likely to affect both information gating and reward processing. PPI is known to be improved by DAs [9]. Furthermore, lower PPI is associated with increased distractibility [9], and impairments have been linked to increased impulsivity and risk-taking behaviors [27]. Our study adds to the literature by showing that reduced PPI correlates with impulsivity in PD patients. However, despite the novelty of these present findings, we acknowledge small sample size and possible epiphenomenon. Future studies of PPI and impulsivity in PD, and more so how STN-DBS can impact PPI, are needed.

In conclusion, this study provides prospective data indicating STN-DBS improves ICDs in the majority of PD patients. Due to the lack of treatment options for ICDs in PD patients, these results have encouraging clinical implications.

Ms. Lucy Gee received funding from NIH grant 5 T35 HL 071483. Dr. Julie Pilitsis is a consultant for Medtronic, St. Jude and Boston Scientific, and receives grant support from Medtronic, Boston Scientific, St. Jude and NIH 1R01CA166379. Dr. Adolfo Ramirez-Zamora is a consultant for Teva Neuroscience and received clinical trial support from Boston Scientific. Dr. Eric Molho is a consultant for Lundbeck, US World Meds and Merz, and has received speaking honoraria from US World Meds. He receives grant support from the Cure Huntington Disease Initiative, Kyowa, Teva, US World Meds, Auspex, Acadia, Merz and Boston Scientific.

Leeman R, Potenza M: Impulse control disorders in Parkinson's disease: clinical characteristics and implications. Neuropsychiatry 2011;2:133-137.
Weintraub D, Nirenberg MJ: Impulse control and related disorders in Parkinson's disease. Neurodegener Dis 2013;11:63-71.
Lhommée E, Klinger H, Thobois S, Schmitt E, Ardouin C, Bichon A, Krack P: Subthalamic stimulation in Parkinson's disease: restoring balance of motivated behaviours. Brain 2012;135:1463-1477.
Emad E: Mechanisms of deep brain stimulation for obsessive compulsive disorder: effects upon cells and circuits. Front Integr Neurosci 2012;6:1-14.
Frank M, Samanta J, Moustafa A, Sherman S: Hold your horses: impulsivity, deep brain stimulation, and medication in Parkinson's disease. J Neurol Neurosur Psychiatry 2007;78:517-519.
Rodriguez-Oroz M, López-Azcárate J, Garcia-Garcia D: Involvement of the subthalamic nucleus in impulse control disorders associated with Parkinson's disease. Brain 2011;134:36-49.
Powell SB, Weber M, Geyer MA: Genetic models of sensorimotor gating: relevance to neuropsychiatric disorders. Curr Top Behav Neurosci 2012;12:251-318.
Zoetmulder M, Beirnat HB, Nikolic M, Korbo L, Jennum PJ: Sensorimotor gating deficits in multiple system atrophy: comparison with Parkinson's disease and idiopathic REM sleep behavior disorder. Parkinsonism Relat Disord 2014;20:297-302.
Braff DL, Geyer MA, Swerdlow NR: Human studies of prepulse inhibition of startle: normal subjects, patient groups, and pharmacological studies. Psychopharmacology 2001;156:234-258.
Lukhanina E, Berezetskaya N, Karaban I: Paired-pulse inhibition in the auditory cortex in Parkinson's disease and its dependence on clinical characteristics of patients. Parkinsons Dis 2010;2011:342151.
Blumenthal TD, Gescheider GA: Modification of the acoustic startle response by a tactile prepulse: effects of stimulus onset asynchrony and prepulse intensity. Psychophysiology 1987;28:296-306.
Graham FK: Control of reflex blink excitability; in Thompson RF, Hicks LH, Shvyrkov VB (eds): Neural Mechanisms of Goal-Directed Behavior and Learning. New York, Academic Press, 1980, pp 511-519.
Hoffman HS, Ison JR: Reflex modification in the domain of startle. I. Some empirical findings and their implications for how the nervous system processes sensory input. Psychol Rev 1980;87:175.
Weintraub D, Stewart S, Shea JA, et al: Validation of the Questionnaire for Impulsive-Compulsive Disorders in Parkinson's disease (QUIP). Mov Disord 2009;24:1461-1467.
Dalley J, Everitt B, Robbins T: Impulsivity, compulsivity and top down cognitive control. Neuron 2011;69:680-694.
Lejuez CW, Read JP, Kahler CW, Richards JB, Ramsey SE, Stuart GL, Brown RA: Evaluation of a behavioral measure of risk taking: the Balloon Analogue Risk Task (BART). J Exp Psychol Appl 2002;8:75.
Claassen DO, van den Wildenberg WP, Ridderinkhof R, Jessup CK, Harrison MB, Wooten K, Wylie SA: The risky business of dopamine agonist in Parkinson's disease and impulse control disorders. Behav Neurosci 2011;125:492-500.
Rao H, Mamikonyan E, Detre JA, Siderowf AD, Stern MB, Potenza MN, Weintraub D: Decreased ventral striatal activity with impulse control disorders in Parkinson's disease. Mov Disord 2010;25:1660-1669.
Wu K, Politis M, Piccini P: Parkinson disease and impulse control disorders: a review of clinical features, pathophysiology and management. Postgrad Med J 2009;85:590-596.
Ardouin C, Voon V, Worbe Y, Abouazar N, Czernecki V, Hosseini H, Krack P: Pathological gambling in Parkinson's disease improves on chronic subthalamic nucleus stimulation. Mov Disord 2006;21:1941-1946.
Lu C, Bharmal A, Suchowersky O: Gambling and Parkinson disease. Arch Neurol 2006;63:298.
Smeding HM, Speelman JD, Koning-Haanstra M, Schuurman PR, Nijssen P, Van Laar T, Schmand B: Neuropsychological effects of bilateral STN stimulation in Parkinson disease: a controlled study. Neurology 2006;66:1830-1836.
Smeding HM, Goudriaan AE, Foncke EM, Schuurman PR, Speelman JD, Schmand B: Pathological gambling after bilateral subthalamic nucleus stimulation in Parkinson disease. J Neurol Neurosurg Psychiatry 2007;78:517-519.
Mathai A, Smith Y: The corticostriatal and corticosubthalamic pathways: two entries, one target. So what? Front Syst Neurosci 2011;5:64.
Haynes W, Haber S: The organization of prefrontal-subthalamic inputs in primates provides an anatomical substrate for both functional specificity and integration: implications for basal ganglia models and deep brain stimulation. J Neurosci 2013;33:4804-4814.
Ramirez-Zamora A, Kahn M, Campbell J, DeLaCruz P, Pilitsis J: Interleaved programming of subthalamic deep brain stimulation to avoid adverse effects and preserve motor benefit in Parkinson's disease. J Neurol 2015;262:578-584.
Swann AC, Lijffijt M, Lane SD, Steinberg JL, Acas MD, Cox B, Moeller FG: Pre-attentive information processing and impulsivity in bipolar disorder. J Psychiatr Res 2013;47:1917-1924.