BACKGROUND: Dopaminergic drugs are the primary risk factor for Impulse Control Behaviours (ICB) in Parkinson's disease (PD), others being early-onset disease and gender.

OBJECTIVE: This report further explores ICB symptom relationships with motor and mood phenotypes, the complex relationship with dopaminergic medications, and hypothesizes a model with potential clinical implications.

METHODS: Data from 500 PD patients were analyzed. Hypersexuality, gambling and shopping behaviour were assessed using selected questions from the Minnesota Impulsive Disorders Interview questionnaire. Local questions assessed hobbyism. Motor characteristics considered were akinetic-rigid/gait disturbance (PIGD) and 'non-PIGD' phenotypes, motor severity, motor progression, and presence/absence of motor fluctuations. Other variables included anxiety, depression, current levodopa and agonist use, age, gender and cognition.

RESULTS: Overall, ICB symptom frequency was 17.8%. There was no relationship between PIGD/non-PIGD motor phenotypes and ICB symptoms. Those with ICB symptoms had higher total combined levodopa/agonist equivalent intake, but not current agonist-only equivalent intake. ICB symptoms were reported by 23.1% of those taking combined levodopa and agonist compared to 19.2% on agonist monotherapy and 11.6% levodopa monotherapy. Compared with non-ICB patients, patients with ICB symptoms were more likely to show an anxious mood phenotype, reported more motor fluctuations, and were younger.

CONCLUSIONS: Both PIGD and non-PIGD phenotypes are equally affected. Dose-related risk applies to total anti-parkinsonian medication and not just current agonist-only. Anxious mood phenotypes may carry increased risk. A role of anxiety, either as a marker of risk, indirect causal factor, or maintaining factor is incorporated into a preliminary model. We discuss implications for clinical management.