The largest epidemiological study related to the prevalence of ICBs in patients with movement disorders is the multicenter cross-sectional, North American DOMINION study [1]. Prevalence rate of ICBs in PD treated with DRT was 13.6% [1] compared to 0.5–1% in the general population [42]. However, a recent study with 115 PD, did not show a significant difference in the rate of tobacco or alcohol addiction and pathological gambling compared with the general population, suggesting that PD patients do not have specific profiles for ICBs. In the DOMINION study (Fig. 17.2a), the most common ICDs was compulsive shopping (5.7%) followed by pathological gambling (5.0%), binge eating disorder (4.3%), and compulsive sexual behavior (3.5%). The study showed also that most frequently a single ICB is observed in PD patients rather than multiple ICBs. Regarding DAs, their use is clearly associated with a higher risk of developing ICBs. In the DOMINION study [1], ICBs were more common in users than in patients not taking DAs (17.1% vs. 6.9%, Fig. 17.2b) confirming a previous study in which DAs significantly increased the frequency of pathological gambling, hypersexuality, and compulsive shopping in PD patients [43]. Consistently, a similar relationship between ICDs and dopaminergic agents was documented in a retrospective study documenting the development of new-onset gambling or hypersexuality in 7 out of 267 patients [44]. Notably, all of these patients were in the DA-treated arm (7/66 patients; 10.6%).

As far as RLS is concerned, DAs treatment seems to have a similar effect on impulsive behaviors. In a case–control study based on a questionnaire followed by a phone interview, the ICBs frequency in RLS patients treated with dopaminergic agents was 17% [45]. The same study showed a latency of 9.5 months between initiation of the DA treatment and the onset of impulsive behaviors. ICBs were represented by compulsive eating with a frequency of 11% followed by compulsive shopping (9%), punding (7%), pathological gambling (5%), and hypersexuality (3%); [45].

In a cross-sectional study, Voon and colleagues [46] revealed a lower prevalence of any ICBs in RLS patients (11%) with a higher risk in patients with higher Das dosage, younger age of onset of RLS, history of experimental drug use, female sex, and a family history of gambling disorders. However, in this study the patients were screened using self-administered patient questionnaires followed by a confirmatory psychiatric interview. Among the ICBs, the same authors reported that the most frequent was compulsive eating (4.3%), followed by compulsive shopping with a rate of 3.6%. Pathological gambling had a frequency of 2.1%, punding of 2.1%, and hypersexuality of 1.4% [46]. Higher ICBs frequency was observed in a recent retrospective case series of 28 RLS patients treated with the transdermal DA rotigotine. In this study, the overall prevalence of ICBs, including binge eating, hypersexuality, compulsive shopping, pathological gambling, and punding equals a rate of 21%. History taking and scoring with the Zurich Screening Questionnaire Symptoms were used for the ICBs assessing prior to and after initiation of rotigotne treatment. Of note, both patients with and without ICBs and irrespective of the therapy totalized higher scores on the Zurich Screening Questionnaire [47]. Similarly, regardless of the pharmacological treatment, RLS seem to be associated with impaired executive performances [48], reduced decision-making abilities under ambiguity but not risk [49] and deficits in short-term attention and verbal fluency [50]. These results could be an expression of intrinsic cognitive and behavioral traits of RLS patients consequently predisposing them to ICBs. Contrary to this hypothesis, patients suffering from multiple sclerosis with secondary RLS seem to be exposed to the same risk of DA-induced ICBs as patients with primary RLS [51]. Thus, rather than being a specific feature of RLS patients, the propensity to ICDs could be related to the sleep disturbance caused by RLS symptoms. Interestingly, ICBs can also be observed in a wide gamut of disorders with the same risk of DA use, such as progressive supranuclear palsy [52], multiple system atrophy [53], and pituitary adenomas [54] in addition to PD and RLS.

Another issue of undeniable importance in the risk of developing ICBs is the multiple co-medication regime. This is a crucial issue for PD patients that usually undergo a complex therapeutic scheme, but much less of an issue for RLS patients for which monotherapy is the rule. Thus, the DOMINION study showed that the combination of DA agonist and levodopa treatment induces a greater risk of ICDs [1]. The same study observed a higher risk in those patients with a higher levodopa dose, too. As far as the use of amantadine concerned, the DOMINION study demonstrated an association with ICDs [1] consistent with another study on pathological gambling in 1167 PD patients [55]. In this regard, these observations are contrary to those seen in a double-blind crossover study in which amantadine was actively administered to 17 PD patients with pathological gambling with beneficial effects [56].

There are no studies so far concerning the nature of DA agonist formulation and its association with ICBs, although some reports have hypothesized that long-acting drugs are associated with a lower risk of ICBs. However, in the above-mentioned study, transdermal rotigotine with putative long-lasting action is associated with the development of ICBs [47] similar to that observed for the other DA agonists. To date, no clear evidence is available for understanding the effect of dopaminergic drugs’ half-life on ICDs.

In the general population as well as in PD, younger age, poor socioeconomic status, maniac and depressive disorders, or alcohol or substance use have been recognized to be factors associated with the development of pathological gambling [57, 58]. Whereas on one side, DAs use is the stronger predictor of ICBs development (see the paragraph above), on the other hand, limited data are available about the other likely risk factors. In the last decade, great efforts have been made to identify intrinsic and environmental features of patients with impulsive disorders. Their recognition and prevention represent key advances in PD and RLS care limiting its subsequent socioeconomic consequences.

As stated above, both in the general population and in a large cohort of PD patient’s, ICBs were independently associated with younger age [1, 59]. ICBs were also independently correlated with family history of gambling disorders, concurrent cigarette smoking, and single marital status [1]. Despite some evidence that males are more prone to gambling disorders in the general population; in the DOMINION study, there was no significant difference in gender amongst PD patients with ICBs. In this regard, the same study showed that females are more likely to express compulsive eating and shopping while men hypersexuality [1].

Recently, in a prospective cohort study, one hundred and sixty four PD patients with no previous impulsive deficiency treated with a DAs were followed longitudinally for 4-years in order to observe new-onset ICBs. 39.1% developed new-onset ICBs with a median onset time 23 months after DRT initiation. Even though baseline demographic characteristics were similar in patients with and without ICBs, subjects with ICBs had surprisingly a greater baseline prevalence of caffeine use and higher lifetime prevalence of cigarette smoking [60]. This finding appears remarkable in the light of the putative protective role of these habits in the development of PD [61].

Psychiatric symptoms such as depression, anxiety, obsessive-compulsive symptoms, disinhibition, and appetite disturbance are also associated with ICBs [62, 63]. For instance, pathological gambling correlates with two personality traits represented by high novelty seeking and high impulsivity [59]. These results point to the role of individual genetic background of DA system in the susceptibility to ICBs (see the next paragraph). Similarly, regardless of the pharmacological treatment, patients with RLS show preferences toward risk choices, as recently demonstrated in a case-control prospective study with 89 RLS patients [19]. The propensity to make greater impulsive choices under ambiguity, assessed by the Iowa Gambling task (IGT) [64], was comparable in drug-free and treated patients but it was significantly different compared to control voluntaries.

Theoretically, genes provide the first contribution to an individual’s predisposition to impulsivity and the development of ICBs. Family and twin epidemiological studies corroborate this general concept [65, 66]. Since DA system plays a crucial role in impulsivity and in addiction, greater attention has been paid to the exploration of the genetic background involved in dopaminergic neurotransmission. The association between DA and impulsivity is remarkable for some disorders, such as attention deficit hyperactivity disorder (ADHD). This disorder is characterized clinically by impulsivity, and amphetamine, a molecule able to increase the DA cleft release, is considered the gold standard for its treatment [67].

DA receptor (DR) polymorphisms have been studied in the general population and DRD2 Taq1A for instance, has been associated with ICBs [68]. However, recent functional imaging studies have not found clear differences in D2/D3 levels between healthy subjects and pathological gamblers as have been observed in substance abuse disorders [69]. Other DRs have been linked to ICBs. Among these, DRD1 polymorphisms have been associated with pathological gambling [70] while DRD4 polymorphism has been controversially reported in impulsive personality trait and ICBs subjects [71, 72]. Of interest, a recent study explored the allelic variants of DRD2, catechol-O-methyltransferase (COMT), and DA transporter (DAT) in 48 idiopathic PD patients without ICDs and 41 with ICDs. Surprisingly, no differences were observed in the frequency of variant of DRD2, COMT, and DAT1 between the two groups suggesting that polymorphisms of dopaminergic genes do not play a relevant role in the development of ICB at least in PD [73].

Other neurotransmitters seem to play a role in deficient impulse control and impulsive personality. Serotonin (5-HT), for instance, has been extensively studied [74–76]. Functional variants in the 5-HT transporter promoter gene (5-HTTLPR) have an effect on the neural mechanisms of disorders relating to impulse control [77] or in other neuropsychiatric disorders, such as anxiety or depression [78, 79]. Serotoninergic system is largely involved in aggressiveness [74, 76] or in impulsive behaviors as demonstrated in a Hungarian population cohort of healthy volunteers showing significant association with impulsiveness scales and genotype analysis for the receptor gene HTR1A [80]. Similarly, pathological gambling and risky-choice behaviors were considered in relation to alterations in serotonergic system functioning [81]. Consistent evidence has suggested that changes in brain-derived neurotrophic factor (BDNF) brain expression and release are involved in mood, anxiety, and eating disorders [82]. Genetic studies on the functional polymorphism Val66Met in the BDNF gene in patients and control subjects are in agreement with this hypothesis [83, 84]. The BDNF Val66Met may affect emotional decision-making performance assessed by using IGT in one hundred sixty-eight healthy subjects [85]. Indeed, BDNF largely determines the development and integrity of several systems, such as the noradrenergic, dopaminergic serotonergic, glutamatergic, and cholinergic neurotransmitter systems [86]. Additionally, BDNF has been found to exert important influences upon feeding behavior and compulsive hoarding [87].

Since the best treatment for ICDs is prevention, patients with movement disorders and their caregivers should be warned about the risk of developing behavioral disorders. During follow-up visits, signs of ICBs should be actively searched for and some clinical tools have been recently developed for this purpose. The Questionnaire for Impulsive–Compulsive Disorders in Parkinson’s Disease (QUIP), for instance, is a brief, self-report screening measure able to screen ICB positive patients which can be used as an additional evaluation tool [88].

Before starting a dopaminergic treatment, and above all one which involves a DA agonist, a premorbid history of substance abuse or impulsive-compulsive behaviors should be taken from patients and family members. A sustained monitoring of ICBs is mandatory in the long term, since the range latency onset encompasses a period of 1–84 months with a mean of 23 months as observed in PD [60, 89]. In a large cohort of RLS patients, a mean treatment duration of 9.5 months was observed before the development of ICBs [45] with a comparable timing in a small case series [90]. Physicians should discuss ICDs with their patients as early as possible before starting any DRT. Factors that have been associated with ICDs should be assessed in order to screen for individuals with a higher risk. In these patients, levodopa might be the first choice rather than a DA agonist since the probability of developing ICDs is lower [1]. Of note, patients with ICBs tend to hide their problem from the family most likely due to a lack of insight regarding the consequence of these behaviors, or for guilt.

Depending on the social impact of ICBs, non-pharmacological or DRT adjustment should be considered. One approach may be crucial and consist in limiting the economic consequences of the ICBs, mainly with regard to pathological gambling or compulsive shopping. Thus, by blocking access to a bank account or to a credit card, there is a good chance of controlling these kinds of ICBs. Consequently, a daily budget for patients, if she/he agrees, may be helpful and the spouse or the family could provide and manage it. Changes in pharmacotherapy should be considered primarily. Several studies suggest that a decrease or, if necessary, withdrawal of the DA agonists is efficacious [53, 89, 91]. Parkinsonian patients showing ICBs; however, seem to have a tendency to develop the so-called DA agonist withdrawal syndrome (DAWS). This is a disabling complication of DA agonist use [92], strongly linked to impulsive disorders, characterized by symptoms resembling those of other drug withdrawal syndromes, such as anxiety, panic attacks, agoraphobia, depression, dysphoria, diaphoresis, fatigue, pain, orthostatic hypotension, and drug cravings. Of note, this syndrome responds only to DA repletion and not to levodopa or other antiparkinsonian medications [93]. Of interest, a case of DAWS has been recently described in a patient with RLS [94]. With respect to DA agonists, several studies have not found any significant differences amongst these (e.g., ropinirole, pramipexole, or rotigone) and their association with ICBs [43, 89, 95].

Nowadays, no obviously effective drugs are available once ICBs have developed. Further, the role of amantadine has not yet been established since contrasting results were obtained by two different studies [1, 56]. Naltrexone, an antagonist at mu and kappa opioid receptors, has been shown to be efficacious in treating substance addictions including alcohol and opioid dependence in general population [28]. Its efficacy was tested also in a case series of three PD patients with pathological gambling with positive outcomes [96].

Behavioral therapies, such as cognitive-behavioral, aversive, and motivational therapies have all shown efficacy in ICBs as evidenced in several trials in general population [97, 98]. In PD, a randomized controlled trial comparing up to 12 sessions of a cognitive-behavioral therapy to a waiting list control condition with standard medical care showed that behavioral therapy is effective in reducing the severity of ICBs [99].

The development of ICBs may have a serious impact on daily living and quality of life of patients and their families. Knowledge of predisposing factors, such as higher DA dosage, young age of PD or RLS onset, history of drug addiction, and a family history of gambling disorders is crucial for management of ICBs. Patients should be warned of potential behavioral side effects when starting with dopaminergic agents.

Further studies are essential to understand the pathophysiology of ICBs in movement disorder patients. Moreover, more effort is needed to identify risk factors for ICBs. Research in this field could be of vital importance for additional treatment options that are nowadays unfortunately limited.

In the future, genetic assessment of the DA or 5-HT system background of single individuals could represent a valid opportunity to identify patients with a strong possibility of developing ICBs, thereby excluding them from DA therapy or recommending that they undergo other non-pharmacological therapies. Indeed, DA agonist suspension could further complicate patient care through the development of DAWS. This side effect represents a considerable challenge for clinicians treating RLS or PD patients.