Inequalities in later life frailty and wellbeing: an interdisciplinary approach to causality

The challenges posed by the ageing of populations have been repeatedly documented, with recognition of the need to minimise dependency and to maximise social engagement and wellbeing. The proposed research is concerned with examining causal processes relating to frailty and wellbeing (both broadly defined) at older ages, including factors operating across and at particular points of the life-course, socioeconomic inequalities, genetic, metabolic, psychological and social processes, and resilience and vulnerability. It will use an interdisciplinary approach, bringing together innovative methods and leading academics from across the biological, clinical and social sciences. The approach will also be comparative, using similarities and differences across nations in demographic, social, economic and policy circumstances to get a clearer understanding of policy influences. The work plan will ensure that the full range of users (older people, lay carers, practitioners, professionals, academics, Government and NGOs) are engaged in influencing the conduct of the research and the communication and application of findings.

The study will make use of data provided by the English Longitudinal Study of Ageing (ELSA), sister studies in the US and mainland Europe, and a similar study in Canada. ELSA is a multidisciplinary study involving repeat surveys with a representative sample of people aged 50 and older, giving six sets of interview data (each two years apart), three sets of biological data (collected four years apart) and a detailed life-history interview collected using robust methods. A wide range of data are available, including: stored DNA and serum suitable for additional genetic, metabolic and biomarker analysis; already analysed biomarkers; direct assessments of cognitive and physical function; and a range of self-reports covering demographics, economics, health, participation in social, civic and cultural activities, and social networks. Such data are unique in containing a breadth and depth of topics across disciplines, repeated measurement over time, and detailed life-histories. This allows the integration of different disciplinary approaches to provide unique analyses that will greatly enhance our understanding of the causes of positive and negative outcomes at older ages, and how these are distributed across the population. Our power to do this will be maximised by examining the influence of national context.

These data will be used to develop measures of frailty and wellbeing, model life-course trajectories, including later life events, examine relationships with socioeconomic position, examine genetic influences and their relationship with markers of metabolic processes, gene-environment interactions, and identify resilience and vulnerability to adverse events.

The challenges posed by the ageing of populations have been repeatedly documented, with recognition of the need to minimise dependency and to maximise social engagement and wellbeing. The proposed research is concerned with examining causal processes relating to frailty and wellbeing (both broadly defined) at older ages. It will use an interdisciplinary approach, bringing together key findings and innovative methods from across disciplines, will be comparative, using similarities and differences across national contexts to get a clearer understanding of policy influences, and will engage a full range of users in the conduct and dissemination of the research.

Core data sources will be the English Longitudinal Study of Ageing (ELSA), sister studies in the US and mainland Europe, and a similar study in Canada. ELSA is a multidisciplinary study involving repeat surveys with a representative sample of people aged 50 and older, giving six sets of interview data (each two years apart), three sets of biological data (collected four years apart) and a detailed life-history interview. Data coverage includes: stored DNA and serum suitable for additional genetic, metabolic and biomarker analysis; already analysed biomarkers; direct assessments of cognitive and physical function; direct assessments of physical health; and a range of self-reports covering demographics, economics, health, participation in social, civic and cultural activities, and social networks. These data will be used to: develop measures of frailty and wellbeing; model life-course trajectories; examine the influences of genetics, socioeconomic position, social factors, and gene-environment interactions; explore the relationship of these factors with markers of metabolic processes; and identify factors relating to resilience and vulnerability to adverse events.

Key hypotheses to be tested cover: the relationships between life-course trajectories, later life socioeconomic position and frailty and wellbeing; that socioeconomic effects will operate through the hypothalamic pituitary adrenal axis connected to processes relating to anabolic function and inflammation; that resilience and vulnerability can be identified and will result from environmental factors, genes, and gene-environment interactions; and that a detailed study of gene-metabolite-phenotypic interfaces will lead to the discovery of novel pathways to frailty and wellbeing. Traditional statistical approaches will be supplemented with innovative methods to handle missing data (selection models; pattern mixture models; shared parameter models; sensitivity analysis), longitudinal data (multilevel multistate competing risks models; growth mixture and structural equation modelling of trajectories; sequence analysis and optimal matching) and data from different national contexts (Integrative Data Analysis; latent variable approaches; and anchoring vignettes).