Dopaminergic nigrostriatal denervation and widespread intracellular α-synuclein accumulation are neuropathologic hallmarks of Parkinson's disease (PD). A constellation of peripheral processes, including metabolic and inflammatory changes, are thought to contribute to neurodegeneration. In the present study, we sought to obtain insight into the multifaceted pathophysiology of PD through the application of a multi-marker discovery approach. Fifty older adults aged 70+, 20 with PD and 30 age-matched controls were enrolled as part of the EXosomes in PArkiNson Disease (EXPAND) study. A panel of 68 circulating mediators of inflammation, neurogenesis and neural plasticity, and amino acid metabolism was assayed. Biomarker selection was accomplished through sequential and orthogonalized covariance selection (SO-CovSel), a multi-platform regression method developed to handle highly correlated variables organized in multi-block datasets. The SO-CovSel model with the best prediction ability using the smallest number of variables was built with seven biomolecules. The model allowed correct classification of 94.2 ± 3.1% participants with PD and 100% controls. The biomarker profile of older adults with PD was defined by higher circulating levels of interleukin (IL) 8, macrophage inflammatory protein (MIP)-1β, phosphoethanolamine, and proline, and by lower concentrations of citrulline, IL9, and MIP-1α. Our innovative approach allowed identifying and evaluating the classification performance of a set of potential biomarkers for PD in older adults. Future studies are warranted to establish whether these biomolecules could serve as biomarkers for PD as well as unveil new targets for interventions.