Abstract
Olfactory dysfunction is a well-established prodromal symptom of Parkinson's disease (PD) and may occur years prior to PD diagnosis. This study assessed validity of a shortened 5-item smell test in a large cohort of PREDICT-PD participants and investigated its associations with previously described prodromal and demographic factors. Participants without PD aged over 60 years-old completed a 5-item smell test and online platform including validated questionnaires - such as the REM Sleep Behaviour Disorder Single-Question Screen (RBD1Q) and Unified Parkinson's Disease Rating Scale (UPDRS) - as well as subjective olfaction assessment, and motor function assessment via the BRadykinesia Akinesia INcoordination (BRAIN) tap test. Log odds for PD were derived from smell test performance. Data was analysed by non-parametric tests in R-studio. A total of 1472 participants were included. Smell test performance declined with age and was poorer in males. Log odds for PD derived from the smell test correlated with RBD screening and bradykinesia as measured by BRAIN tap test but not motor impairment subjectively assessed by UPDRS part 2. Subjective olfactory assessments correlated weakly with objective smell test performance, with only fair agreement. Notably, subjective smell ratings did not correlate with motor function, whereas objective smell test results did. This study supports the feasibility of a shortened olfactory test for PD risk stratification in prodromal populations. Some previously described associations with prodromal and demographic factors were reaffirmed. Findings highlight the association between early motor dysfunction and olfactory impairment, emphasising the need for objective olfactory testing in research and clinical practice.