Trajectories of functional disability for the elderly in Britain


  • Robert John French Centre for Multilevel Modelling, University of Bristol
  • Fiona Steele Department of Statistics, London School of Economics and Political Science



Ageing, Activities of daily living, Health trajectories, Britain, British Household Panel Survey (BHPS), Structural equation model (SEM), Growth model, Measurement model, Temporal measurement invariance


This study uses an innovative approach to characterise trajectories of functional disability over the final stages of the life course. We use data from the British Household Panel Survey (BHPS), an annual household survey of all adults in a representative sample of British households from 1991-2008. The analysis focuses on the sub-sample of elderly household members who were aged from 65 to 74 in any of the 18 waves of data, with a final sample of 3,671 individuals contributing a total of 13,982 person years. As in previous research, we estimate latent growth curves, but extend the standard model to incorporate a measurement model for the latent outcome variable ‘functional disability’. We identify accelerating trajectories of functional disability for a representative sample of elderly individuals separately by gender. We show that socio-occupational classification is associated with the level of initial functional disability and to a less extent the change in functional disability with age. The contribution of this paper is to explore the use of a measurement model to exploit the variation between items in discriminatory power for identifying an individual’s functional disability. Further we are able to explicitly test for temporal measurement invariance in functional disability i.e. to what extent the items consistently measure the latent variable as people age.