Dopamine Agonists vs Levodopa in Impulse Control Disorders

We read with interest Weintraub and colleagues' cross-sectional study of 4 impulse control disorders (ICDs) in Parkinson disease.¹ The point prevalence for ICDs was 13.6%. Impulse control disorders were more common in dopamine agonist (DA)–treated patients (17.1%) than all other cases (6.8%; odds ratio, 2.72; P < .001). Known associations with ICDs were found, eg, men, younger onset age, being unmarried, and current smoking. Two findings were new and require detailed consideration.

First, ICDs were more likely in US (15%) than Canadian patients (9.8%). We agree that prescribing practice, study conduct, and patient behaviors are possible explanations.

Second, levodopa was independently associated with ICDs (odds ratio, 1.51; P = .01), which the authors acknowledge as less robust. A total of 2682 patients were prescribed levodopa including 991 who took levodopa alone, while 59 cases did not take levodopa or a DA. Considering the median Parkinson disease duration of 6.5 years, patients who did not take levodopa or a DA would likely have shorter disease duration and reduced exposure to the cumulative risk of developing ICDs. The observed dose effect for levodopa but not DAs is consistent with patients with ICD who are taking a low-dose DA and patients with ICD as part of the dopamine dysregulation syndrome (DDS) who are taking high-dose levodopa. The study primarily tested within dopamine agonists and appropriately did not capture DDS data. Some of the levodopa association and dose effect is likely to relate to cases with DDS. The risk ratio of 1.5 for ICDs for levodopa should therefore be interpreted cautiously.

This is relevant to treatment choices in managing ICDs, the most studied being pathological gambling, which persists unless DAs are stopped (occasionally reduced) but resolves despite compensatory increases in levodopa (eg, from 615 to 881mg over 21 months).² The relative risk of ICDs for levodopa, and the levodopa dose effect identified, appear to overstate this clinical experience. Accordingly, levodopa remains a first- or second-line (after monoamine oxidase B inhibitors) choice in patients at high risk of ICDs and an essential treatment to maintain antiparkinson efficacy in patients who develop ICDs on DAs.

Correspondence: Dr Grosset, Institute of Neurological Sciences, Southern General Hospital, Glasgow G51 4TF, Scotland (d.grosset@clinmed.gla.ac.uk).

Financial Disclosure: Dr Grosset reports being a consultant for Vectura, Advisory Boards, Archimedes, Ipsen, Honoraria, Teva, GSK, Boehringer Ingelheim, and Roche and receiving grants from GSK; Dr Cardoso reports receiving honoraria from Boehringer-Ingelheim, Ipsen, Novartis, Roche, Sanofi-Aventis, and Teva; and Dr Lees reports being a consultant for Genus, serving on the advisory boards of Novartis, Teva, Meda, Boehringer Ingelheim, GSK, Ipsen, Lundbeck, Allergan, and Orion, and receiving honoraria from Novartis, Teva, Meda, Boehringer Ingelheim, GSK, Ipsen, Lundbeck, Allergan, Roche, and Orion.