Affiliations: Department of Neurology, King George Hospital, Barking Havering & Redbridge University Hospitals Trust, Romford, Essex, UK | Department of Neurosciences, Royal London Hospital, Whitechapel, UK
Note:  Correspondence to: J.H. McAuley, Department of Neurology, King George Hospital, Barley Lane, Ilford, Essex IG3 8YB, UK. Tel.: +44208970827; E-mail: [email protected]
Abstract: Background: Olfactory hallucinations are known to occur in idiopathic Parkinson's disease (IPD) but are much less well-described than visual hallucinations. Objective: To report the prevalence, clinical features, response to treatment and prognosis of olfactory hallucinations in IPD. Methods: 205 consecutively reviewed IPD patients and 205 non-IPD control patients attending the local hospital were surveyed for the presence of olfactory hallucinations by specific questioning; the IPD patients were followed up for at least three years. Results: Of 188 patients who had a clinical course remaining consistent with IPD, four were initially found to have olfactory hallucinations, yielding a prevalence of 2.1% (95% confidence interval 0.4–5.4%). Two further patients developed such hallucinations later during the study. Olfactory hallucinations were not always accompanied by other hallucination modalities. The patients had a long duration of disease treated with dopaminergic medication, loss of sense of smell typical for IPD, no dementia or features suggestive of non-idiopathic PD after three years follow-up, a lack of insight into their hallucinations with consequent failure to report them spontaneously, and a good and lasting response to modest doses of atypical antipsychotics. Conclusions: This study confirms the significant prevalence of olfactory hallucinations, describes their clinical features and indicates that they may occur in isolation and not predict other psychotic or dementing features. The nuisance that such hallucinations cause, their ease of treatment and their failure to be volunteered as a symptom means that specific questioning for their presence should be included in routine assessment of patients with IPD.