Background: Elevated peripheral levels of different cytokines and chemokines in subjects with Alzheimer's disease (AD), as compared with healthy controls (HC), have emphasized the role of inflammation in such a disease. Considering the cross-talking between the central nervous system and the periphery, the inflammatory analytes may provide utility as biomarkers to identify AD at earlier stages.
Objective: Using an advanced statistical approach, we can discriminate the interactive network of cytokines/chemokines and propose a useful tool to follow the progression and evolution of AD, also in light of sex differences.
Methods: A cohort of 289 old-age subjects was screened for cytokine and chemokine profiling, measured in plasma, after a thorough clinical and neuropsychological evaluation. A custom algorithm based on Fisher linear discriminant analysis was applied to ascertain a classification signature able to discriminate HC from mild cognitive impairment (MCI) and AD.
Results: We observed that a joint expression of three proteins (a "signature" composed by IFN-α2, IL-1α, TNFα) can discriminate HC from AD with an accuracy of 65.24%. Using this signature on MCI samples, 84.93% of them were classified as "non-HC". Stratifying MCI samples by sex, we observed that 87.23% of women were classified as "non-HC", and only 57.69% of males. Indeed, in a scatter plot of IFN-α2 and IL-1α, the HC group was better separated from MCI and AD in women as compared with men.
Conclusion: These findings suggest that AD is accompanied by a peripheral inflammatory response that can already be present in MCI subjects, thus providing a mean for detecting this at-risk status and allow an anticipated intervention.
Keywords: Cytokines; dementia; inflammation; markers; sex.