Investigation of Volatile Metabolites in Sebum as Prodromal Indicators of Parkinson's Disease

Background: Parkinson's Disease (PD) has been associated with a distinct odour, strongest in sebum-rich areas. Thermal Desorption Gas Chromatography Mass Spectrometry (TDGCMS) has revealed volatile signatures that distinguish individuals with Parkinson's Disease (PD) from healthy controls. Here, we applied the same method, including subjects with isolated REM sleep behaviour disorder (iRBD) to examine the volatiles in sebum and compare this with that found in PD subjects and control participants. Participants with iRBD have a high likelihood for conversion to overt clinical synucleinopathies like PD, Dementia with Lewy Bodies (DLB) or (less commonly) Multiple System Atrophy (MSA). Methods: Subjects with clinically established PD (n=16) or iRBD (n=9) as well as healthy controls (n=9) were included. Following methods established in our laboratory, sebum was sampled from each participant using cotton gauze and the headspace from these swabs, analysed directly with TDGCMS. Univariate and multivariate analysis was employed to probe the differences between volatile metabolites found for each phenotype. Putative identifications were assigned using spectral matching against the Golm metabolome and NIST spectral databases. Findings: We can completely classify each phenotype using the sampled volatilome from which we built models with logistic regression analysis. The classification between PD and control improved on previously published work, from 85% to 100%. Putatively annotated molecules include alkanes, aldehydes, fatty acid methyl esters (FAMEs), and three metabolites namely purine, tropinone and oleamide. Investigation of highly ranked features revealed 18 features that showed intermediate expression in samples from iRBD participants. Interpretation: TDGCMS can differentiate volatile metabolite signatures from sebum between PD, RDB and control samples. More than 70% of the identifiable metabolites that 2 permit this discrimination were putatively annotated as hydrocarbons and fatty acid methyl esters (FAMEs). Our prior work indicates that these components arise from larger lipid molecules that decompose during the experiment. Features putatively annotated as tropinone, oleamide and purine, have previously been linked with neuroprotection, sleep induction and antioxidation, respectively, are significantly different between the three groups of participants, along with FAMEs and hydrocarbons. ### Competing Interest Statement The authors have declared no competing interest.