Modelling and forecasting health expectancy
So far the main focus of research has been on the trends of health expectancy in a given country or on cross-country comparisons between countries. Less emphasis has been put on determinants of health expectancy, such as obesity, smoking or educational level. Furthermore, studies aiming at predicting future health expectancy are rare. This is partly due to the lack of long time series data on health status. In addition, forecasting health expectancy is based on multiple dimensions and is thus more complicated than forecasting life expectancy which is based on a single dimension, mortality.
The main research questions addressed in this thesis aim to fill these gaps, and are as follows:
- What are the mechanisms and factors that influence health expectancy?
- What are the likely future trends in health expectancy?
The first part of this thesis, Chapters 2, 3 and 4, aims to answer the first research question by establishing the extent to which health expectancy is related to major risk factors of population health. In these chapters health is measured in various ways based on data from repeated cross-sectional survey collected in the Netherlands (Chapter 2) and from longitudinal multinational surveys (Chapter 3 and 4). Chapter 2 investigates whether the mortality risk related to disability can be explained by other risk factors, and differences are quantified in terms of life expectancy between the nondisabled and the disabled population. Chapter 3 estimates the extent to which overweight status and smoking status is related to life expectancy with disability. In Chapter 4 differences in disability-free life expectancy are estimated for different educational levels. In both Chapters 3 and 4 estimates are given for Western-European countries.
The second part of the thesis, i.e. Chapters 5 to 7, aims to answer the second research question on future trends of health expectancy by projecting past trends between either 1981 and 2007 or 1989 and 2007 until 2020/2030. The implications of future trends in population health, covered in this thesis, involve changes in the prevalence of selected health conditions and economic consequences on long term care expenditures. Chapter 5 presents a multistate life table framework that can be applied to forecast the health expectancy of a population. Such an application is presented for the Netherlands where the health states are ‘nondisabled’, ‘disabled’ and ‘dead’. Chapter 6 explores the effects of the future prevalence of disability on future long term care expenditures. In Chapter 7 future prevalence of overweight and obesity are estimated by taking into account the changing distribution of body mass index in the population.
Chapter 8 focuses on the main discussion points and elaborates on the policy implications of the likely trends.