Data-driven biomarker discovery: harnessing cognitive and physical health phenotyping from longitudinal health data to understand the mechanisms of healthy ageing

Population ageing poses major global public health challenges, with increased numbers of individuals living longer with multiple chronic morbidities and declining health. The burden for health systems and society is substantial. Genetic and environmental factors influence healthy human lifespan duration. Cognitive and physical function, along with resilience to various diseases being key determinants. Understanding the mechanisms of healthy ageing, identifying people resilient to health decline and establishing biomarkers that predict resilience, will facilitate development and testing of effective interventions.

Emerging areas relevant to cognitive and physical decline are glycaemic control (maintains health throughout the life-course) and the sirtuin system. We have demonstrated MRI detected brain lesions associated with above-normal glycated haemoglobin levels. Elevated glucose is associated with cognitive and physical dysfunction [Pappas 2016]. Aberdeen researchers recently reported interactions between diabetes treatment and Alzheimer’s disease in mice. Glucoregulation out with clinical diabetes is less well understood, but links to cognitive function supporting a role in healthy ageing.

The sirtuin system, SIRT1-7 and enzymes synthesising (from dietary derived sources) their cofactor NAD+, regulate diverse biochemical processes, including responses to glucose. SIRT1, has a key role in cognition and metabolic processes associated with health and ageing. Recently our lab identified extensive interactions between the sirtuin system and metabolic health [Drew 2014]. Our ongoing work exploring the transcriptome using next generation sequencing (NGS) further supports SIRT1 as a health status biomarker, linking dysregulated SIRT1 molecular networks to cell senescence, ageing, lipid metabolism, cellular homoeostasis and cellular response to nutrient levels.

APPROACH
To better understand the role of glycaemic control and sirtuin associated molecular networks in healthy physical and cognitive ageing, this project will exploit comprehensive genetic, phenotypic, and environmental data from people enrolled in the Aberdeen Children of the 1950s (ACONF) study.

ACONF includes 500 participants who joined the Generation Scotland project (ACONF:GS) and received detailed cognitive assessment and biological sampling. Their DNA, analysed by high-density genome-wide chip genotyping is research ready. The cohort annotation includes extensive early-life data for each participant (birth records; childhood cognitive test results) and longitudinal data on education, occupation, socioeconomic status, lifestyle, physical activity and diet. These co-variate data will be incorporated into genotype, healthcare and biochemistry phenotype association models. Links between biological ageing, lifestyle, physical activity and diet will be assessed using our existing data for a priori hypothesis driven analyses.

High resolution variant screening (targeted NGS) will be performed of gene networks we propose as candidate drivers of healthy lifespan using ACONF:GS DNA. Mapping genotype to phenotype, linking Generation Scotland data to detailed routine health care and biochemistry data, will allow characterisation of longitudinal physical health. Our existing work on trajectories in kidney function in ACONF will permit linking of laboratory data and clinical records. Phenotype maps for 40+ clinical conditions extractable from routine hospital data are being validated. Repeated cognitive assessment in childhood and adulthood, together with detailed brain imaging being conducted will permit characterisation of longitudinal cognitive health [Murray 2016].

These hypothesis driven analyses will permit investigation of the role of glycated haemoglobins and the sirtuin system in long-term healthspan. Additionally, GWAS will be performed on existing ACONF:GS genome-wide genotypes. This will allow exploration of recently enhanced phenotypic data, demonstrating how altered cognitive and physical health with age can be used to annotate phenotypes, enabling mechanistic insight from linkage to biological data. In parallel, funding will be sought to expand our existing pilot transcriptomic data for investigation of candidate eQTLs within significantly associated QTLs identified in the hypothesis driven genotyping analyses; RNA-seq analysis of the blood transcriptome will be performed in the GS:SHFS STRADL cohort for which physical health, brain MRI and questionnaire data will be available.

TRAINING
The successful student will join a multidisciplinary team drawing expertise from four areas of research excellence: Centre for Genome-Enabled Biology and Medicine, Farr Institute for Health Informatics Research, Rowett Institute for Nutrition and Health, and Aberdeen Biomedical Imaging Centre. They will focus on hypothesis-driven analysis of genetic, metabolic, and cognitive datasets, with large-scale health record linkage, with the opportunity to explore genome wide associations, developing skills in advanced statistical modelling to describe trajectories in longitudinal protocolised and non-protocolised health data. Genes, transcripts, biological pathways and molecular networks associated with age-related cellular processes will be explored. The student will benefit from University of Aberdeen high performance computer cluster, our established bioinformatics training programme, and comprehensive training and support from the team in bioinformatics, systems biology, health informatics and data linkage, together with support and guidance provided by our vibrant research community.