Goh Jor Ming, M.Sc., Ph.D., Brian Kennedy, Ph.D., and Dean Ho, M.Sc., Ph.D. | National University of Singapore
Competition Sponsor: Ministry of Health and National Research Foundation of Singapore
Awardee Year: 2020
Exercise exerts physiological effects on multiple organs and tissues and is a powerful “drug” that improves healthspan, depending on the “dose” (frequency, intensity and duration) administered. A significant heterogeneity of inter-individual physiological responses to any standardized exercise regimen exists, however. Some individuals demonstrate adverse or no response to exercise training. Furthermore, adverse reactions may occur independent of improvements in cardiorespiratory fitness. Hence, to maximize public health efforts for primary prevention and healthspan, exercise prescription needs to be personalized.
CURATE.AI is a neural network-based optimisation platform that correlates inputs- in this case, exercise, and its respective doses or intensities, to outputs, which can include treatment efficacy or training-induced adaptations. This correlation is based on a quadratic algebraic series previously validated in over 40 disease indications that span the in vitro through clinical/human stages of development. These have included solid cancers to infectious diseases as well as cognitive training, among others.
We will compare 2 “doses” of exercise intensities and measuring physiological parameters e.g. heart rate (HR) and molecular biomarkers (e.g. IL-6, FGF-21, lactate) that are responsive to dynamic changes in metabolic demands. Our feasibility study cohort will comprise young (21-30 years) and older (50-64 years) family members (mothers and daughters) undergoing 3-weeks of high, or low-intensity exercise training. After a 2-week washout, subjects will repeat another 3-weeks of exercise of either low, or high-intensity training, whichever was not performed previously. Importantly, at the end of each week, CURATE.AI will optimize the subsequent week’s training session to either increase, maintain or decrease exercise intensity, based on the subject’s i) VO2, ii) HR, and iii) lactate concentrations during exercise. The exercise-induced changes in biomarkers and physiological responses will be assessed by CURATE.AI to generate unique parabolic curves for each subject. These individualized parabolic responses will yield further insight into personalized responses to exercise.
To learn more about this proposal, email healthylongevity@nas.edu.