Pharmacological and non-pharmacological treatments for impulsive-compulsive behaviors in Parkinson's disease

Rationale: Impulsive-compulsive behaviors (ICBs) in Parkinson's disease are associated with psychiatric comorbidities, reduced quality of life, caregiver burden and serious psychosocial consequences. These behaviors greatly impact patients and their families, and pose a challenge for clinicians. Objectives: To assess and compare the effects of pharmacological and non-pharmacological treatments for ICBs in people with Parkinson's disease, and to assess whether the effects differ according to ICB subtype. Search methods: We searched the Cochrane Movement Disorders Group Specialized Register, the Cochrane Central Register of Controlled Trials, MEDLINE, Embase, PEDro, and major trials registries, along with handsearching of abstracts, to 13 June 2025. Eligibility criteria: We included parallel-group and cross-over randomized controlled trials (RCTs) in people with Parkinson's disease with ICBs comparing pharmacological or non-pharmacological treatment with placebo or no treatment. Outcomes: Our critical outcomes were change in the frequency and severity of ICBs from baseline, and adverse effects or events. Our important outcomes were change in: quality of life, neuropsychiatric symptoms (depression, anxiety, apathy, anhedonia), impulsivity, cognition, motor and non-motor symptoms associated with ICBs in Parkinson's disease. Risk of bias: We used the revised Cochrane risk of bias tool for randomized trials (RoB 2) for outcomes reported in the summary of findings tables. Synthesis methods: We included four studies, each evaluating a different intervention. Due to this heterogeneity, we could not conduct a meta-analysis and instead provided a narrative summary of the studies and their findings. For continuous outcomes, we calculated mean differences (MDs), while for binary outcomes, we calculated risk ratios (RRs), both with 95% confidence intervals (CIs). We used the GRADE approach to assess the certainty of evidence for each outcome. Included studies: We included four RCTs: three parallel-group and one cross-over design, with a total of 151 participants. Sample sizes ranged from 17 to 50 per study; participants' mean age ranged from 57.6 to 61.2 years. Women comprised 23.5% to 32% of participants. Three studies compared pharmacological interventions - amantadine, naltrexone, and clonidine (one study each) - to placebo. One study compared a non-pharmacological intervention (cognitive behavioral therapy, CBT) to a wait-list control. Non-profit organizations supported three studies; one did not report funding information. The studies were conducted in Europe and North America and were published between 2010 and 2023. Follow-up durations ranged from one to six months. All studies evaluated critical outcome measures of interest. Only one study assessed any of the review's important outcomes of interest. Synthesis of results: Amantadine versus placebo The evidence is very uncertain about the effect of amantadine versus placebo on adverse events (RR 4.50, 95% CI 0.25 to 81.76; 17 participants). The one study exploring this comparison did not assess any other review outcomes. Naltrexone versus placebo Compared to placebo, naltrexone may result in little to no difference in change in severity of ICBs (; low-certainty evidence). The evidence is very uncertain about the effect of naltrexone on adverse events, including dizziness (RR 4.00, 95% CI 0.48 to 33.22; 48 participants), nausea (RR 15.00, 95% CI 0.90 to 248.78; 48 participants), headache (RR 1.25, 95% CI 0.38 to 4.10; 48 participants), and blood pressure changes (RR 1.67, 95% CI 0.72 to 3.86; 48 participants). [Figure: see text] Clonidine versus placebo The evidence from one study (with 39 participants assessed for all outcomes) is very uncertain about the effect of clonidine on change in frequency of ICBs (RR 0.89, 95% CI 0.54 to 1.47), change in severity of ICBs (), and adverse events, including sleepiness (RR 0.53, 95% CI 0.05 to 5.34), falls (RR 3.15, 95% CI 0.14 to 72.88), nausea (RR 3.15, 95% CI 0.14 to 72.88), orthostatic hypotension (RR 1.58, 95% CI 0.30 to 8.43), restless legs syndrome (RR 3.15, 95% CI 0.14 to 72.88), fatigue (RR 0.70, 95% CI 0.13 to 3.75) and asymptomatic bradycardia (RR 3.15, 95% CI 0.14 to 72.88). The evidence is also very uncertain about the effect of clonidine versus placebo on any changes in participants' quality of life (), depression () and anxiety (). [Figure: see text] [Figure: see text] [Figure: see text] [Figure: see text] CBT versus wait-list control The evidence suggests that, compared to wait-list control, CBT may result in little to no difference in change in frequency (RR 0.79, 95% CI 0.50 to 1.26; 42 participants; low-certainty evidence) and severity of ICBs (; 31 participants; low-certainty evidence). [Figure: see text] The variety of interventions and the small sample sizes precluded data synthesis, including meta-analysis, as well as subgroup and sensitivity analyses. We judged all four studies to have 'some concerns' or a high risk of bias overall. Authors' conclusions: The RCTs reviewed here provide only uncertain evidence regarding the effectiveness of both pharmacological and non-pharmacological treatments for ICBs. This uncertainty is due to limited evidence, small sample sizes, short follow-up periods, and an uncertain balance between side effects and efficacy. Future RCTs should include appropriate critical and important outcomes and explore different ICBs separately, with adequate sample sizes and follow-up durations. Funding: This Cochrane review had no dedicated funding. Registration: Protocol (2023): doi.org/10.1002/14651858.CD015046.