Prevalence of repetitive and reward-seeking behaviors in Parkinson disease

Repetitive and reward-seeking behaviors (also known as compulsive or impulse-control behaviors) in patients with Parkinson disease (PD) such as pathologic gambling (PG),^1,2 pathologic hypersexuality (HS),³ binge eating,⁴ and compulsive shopping (CS)^5,6 are probably related to aberrant dopaminergic stimulation. Compulsive dopaminergic medication use^5,6 and punding (complex, repetitive, excessive, non-goal oriented behaviors)⁷ have also been described. We recently reported our tertiary clinic survey results documenting PG lifetime prevalence at 3.4% and current prevalence at 1.7%. PG was associated with dopamine agonists as a class effect without any association with dose or agonist type.²

HS in PD was recently reviewed and reported along with a retrospective database review for case ascertainment. The authors noted an association of HS with dopamine agonists.³

In this study we sought to determine prevalences of HS and CS, overall prevalence of these behaviors in PD, and medication associations in a case-control study.

Methods.

We screened consecutive follow-up patients with idiopathic PD (Queen's Square Brain Bank criteria) who attended the Toronto Western Hospital Movement Disorders Center in Toronto, Ontario, Canada, over 3 months. Exclusion criteria were atypical parkinsonism, dementia within a year of motor symptom onset, or inability to complete the survey. Patients were encouraged to complete the 15-minute patient-rated screen with spousal assistance. The survey included a modified South Oaks Gambling Screen (SOGS)²; Lejoyeux's CS questionnaire⁸ (score ≥ 6); and a specifically designed HS questionnaire (table 1A). Patients scoring above a set threshold for any of the three sections of the survey were contacted and referred for assessment with a psychiatrist (V.V.) experienced with PD behavioral symptoms as previously described.²

CS was diagnosed using McElroy's criteria.⁹ For methodologic consistency and to differentiate from medication-induced libido increases, we established a priori provisional HS diagnostic criteria (table 1B). The criteria were based on 1) sexual disorders criteria categorized under Impulse Control Disorders and Paraphilias in the Diagnostic and Statistical Manual of Mental Disorders, Version IV (DSM IV); 2) clinical experience (V.V., J.M., A.E.L.); and 3) DSM diagnostic formats including symptoms of excessive or atypical sexual behaviors, change from baseline, duration, and consequences. Subsyndromal hypersexuality was diagnosed if patients fulfilled all criteria except for consequences (Criteria C).

Other psychiatric diagnoses were made using the Structured Clinical Interview for Diagnosis of Axis I disorders (table 2). Medication types and doses concurrent with the behavior were recorded. Sex, age, age at PD onset, and medication types and doses of all screened patients were recorded.

The unpaired t test and Fisher exact test were used to compare patients with compulsive behaviors with PD controls. Significance was set at p < 0.05. The study was approved by the Research Ethics Board at the University Health Network and patients gave informed consent.

Results.

A total of 297/396 patients with PD (75%) adequately completed the survey (≥75% of questions answered). Surveys inadequately completed included patients with known cognitive deficits.

Fifty-five patients with elevated scores within each survey subsection were phone interviewed (PG 16; HS 32; CS 7); 4 HS and 1 CS patients could not be reached. Nineteen misinterpreted the subsection (PG 3; HS 14; CS 2). Two patients with likely HS refused assessment. Thirty patients were psychiatrically assessed (PG 12; HS 14; CS 4). Four could not be assessed due to distance or failure to show.

The behaviors all had onset after medication onset. HS lifetime prevalence was 7/297 (2.4%) and current prevalence was 6/297 (2.0%). A total of 4/297 patients had subsyndromal HS (table 1B) of which the only identified female patient became retrospectively aware of excessive sexual behaviors that caused mild distress only after her agonist had been stopped. The current prevalence of CS was 0.7% (2/297).

The percentage of all surveyed patients on different medication subtypes was previously presented.² The prevalences of PG,² pathologic and subsyndromal HS, CS, and the combined behaviors were more frequent on agonist than levodopa monotherapy (figure, A).

• Download figure
• Open in new tab
• Download powerpoint

Figure. (A) Percentage of patients with Parkinson disease (PD) with pathologic gambling, hypersexuality, compulsive shopping, or any compulsive behavior on levodopa, agonist monotherapy, or adjunctive agonist. Black bar, any behavior; white bar, pathologic gambling; light gray bar, hypersexuality; dark gray bar, compulsive shopping. *p < 0.001, **p < 0.0001 vs PD patients with behaviors on levodopa monotherapy. (B) Percentage of patients with PD with hypersexuality according to use of adjunctive dopamine agonists. Black bar, pathologic hypersexuality (7); white bar, pathologic and subsyndromal hypersexuality (11).

All patients were on medications and none had excessive medication use. The medication association is reported for both pathologic and subsyndromal HS to avoid a bias due to limited sample sizes. There were no differences in the prevalences of HS and pathologic and subsyndromal HS within different agonist classes (figure, B (p > 0.05, Fisher exact test). Compared to PD patients without HS, total levodopa dose equivalent (LEDD) but not LEDD calculated for agonist only was higher for the combined pathologic and subsyndromal hypersexuality groups (p < 0.0001, unpaired t test) (table 2).

HS was associated with male sex (p < 0.05) and early PD onset (p < 0.001) (table 2). CS was associated with early PD onset (p < 0.001).

Discussion.

The overall lifetime prevalence of PG, HS, or CS was 6.1% and 13.7% in any patients on dopamine agonists. The lifetime prevalence of PD patients on dopamine agonists was 7.2% for PG, 7.2% for hypersexuality, and 1.4% for CS. The DSM-defined PG lifetime prevalence in Ontario, Canada, is well-documented at 1%.¹⁰ The prevalence of hypersexuality or CS in the general population is less well-established. The prevalence in untreated patients with PD is not known.

Study strengths included the systematic nature of the screening, case-control comparison of medication doses, and use of a psychiatric assessment. Although the use of a rigorous definition with our provisional operational diagnostic HS criteria was useful for methodologic purposes and as a clinically useful cutoff, this may have limited the yield. Furthermore, the questionnaire sensitivity or the diagnostic criteria of hypersexuality in women is not known. The study was limited by the lack of a clinician-interview assessment. Spousal assistance may have facilitated impairments in insight; however, hidden behaviors may not have been detected. Cognitively impaired patients who inaccurately completed the survey may have been missed. Pathologic eating was not screened as the behavior had not been recognized in the PD literature at the time of the study.

HS was associated with dopamine agonists although not with any specific agonist. However, HS has been reported in patients on levodopa monotherapy.³ Furthermore, in two patients in this study, HS occurred either on levodopa monotherapy or prior to adjunctive agonist therapy. Higher LEDD doses were also observed in the combined pathologic and subsyndromal hypersexuality groups as a result of higher levodopa rather than agonist dose suggesting a levodopa dose effect. However, interpretations are limited by the small sample size and high SDs in the identified patient group.

In the patients with HS, six of seven had comorbid depression. Whether depression is secondary to the behaviors, has similar pathophysiologic substrates to HS and hence co-occurs simultaneously, or mediates the behaviors is not known. Depression may also play a role in the type of interests or behaviors expressed in the context of dopaminergic medications. Treatment of the depression with antidepressant monotherapy improved HS in two patients. However, whether antidepressants improved HS by improving depression which may be mediating HS, had an effect through its known action on decreasing libido, decreased obsessional symptoms that may be comorbid with the behaviors, or had a direct effect on HS is not known.

The timing of the onset of the depression in relation to the onset of the HS was not clearly established in this study. In contrast, depression onset preceded CS onset in both cases.⁹ In one patient, treatment of the depression with antidepressants improved the CS behavior. However, the rationale for the improvement, as with that observed with HS, is not clear.

The rates may be influenced by medication practices in different centers, higher levodopa doses, treatment setting, and culture and environment.