Circulating Multi-Omics Predictors Of Cognitive And Motor Function In Sedentary And Habitually Active Adults

Older age is associated with declines in cognitive and motor functions. A goal of geroscience research is to preserve these functional domains and extend “healthspan,” the number of years spent healthy and disease-free. In this context, there is a growing interest in using “omics” to determine predictors of healthspan. PURPOSE: To identify circulating predictors of cognitive and motor performance in sedentary and habitually active adults using a multi-omics approach. METHODS: Whole-blood samples and NIH-Toolbox Cognitive and Motor function test performance measures were obtained from young sedentary adults (27 ± 6 years, n = 12), older sedentary adults (70 ± 7 years, n = 12), and older habitually active adults (68 ± 5 years, n = 16). DNA and RNA were isolated from blood samples for whole-genome bisulfite sequencing and RNA-seq, respectively. Plasma was isolated for proteomics analyses. Weighted gene co-expression network analysis (WGCNA) was performed to identify genes/transcripts that correlated with cognitive and motor function scores. These genes were then filtered to include only those that were differentially methylated (DNA) and expressed (RNA) with age and exercise, and detectable via proteomics. Penalized stepwise regression models were used to assess the predictive value of the identified proteins. RESULTS: WGCNA identified 2,053 transcripts that were negatively correlated with cognitive and motor function test scores. Of these transcripts, 239 were differentially methylated (DNA) and differentially expressed (RNA) with age and exercise, and 18 of these were detectable via plasma proteomics. Regression models using expression of the 18 proteins were able to accurately predict age (R2 = 0.70, p < 0.005), executive function (R2 = 0.63, p < 0.05), episodic memory (R2 = 0.53, p < 0.05), processing speed (R2 = 0.62, p < 0.05), grip strength (R2 = 0.53, p < 0.05), and 4-meter walk time (R2 = 0.43, p < 0.05). CONCLUSIONS: We identified 18 functionally relevant proteins and found that their expression predicted multiple important domains of physiological function. To our knowledge, this is the first study to collect health, cognitive and motor function, transcriptomic, epigenomic, and proteomic data in the same individuals using age and habitual exercise as a framework for identifying mediators of healthspan.