From disparities in mortality trends to future health challenges – DIMOCHA

A set of major industrialized countries where appropriate data are available has been selected: the United States, Japan, and Russia, as the biggest, plus 5 largely populated countries of the EU (Germany, France, England & Wales, Spain, and Poland).
Such a selection enables us to consider different stages of the Health transition, back to the 1950s: - the fall of infectious diseases (pioneers: England & Wales; laggards: Russia, Japan),
- the cardiovascular revolution (pioneers: the USA, France; laggards: Russia, Poland),
- the fight against old-age pathologies (pioneers: Japan, France, Spain; laggard: the USA)
The project will include first the reconstruction of consistent long-term cause-of-death series (freed from the disruptions caused by changes in the International Classification of Causes of Death). This work already started for some of the selected countries while it could be made easier for some others by the availability of bridge coding operations. Second, the development of new decomposition methods is required to link changes in age and cause-of-death patterns to longevity trends measured through life table indicators (like the mean, mode, dispersion and lifetime losses measures). It will result in both interpretation of observed changes and hypotheses for future mortality patterns and trends.

The project will give technical and scientific returns in at least 4 areas:
- Databases
All reconstructed historical series of deaths by cause will be made accessible to the public through an on-line user-friendly database.
Discussion is engaged with the University of Berkeley about the link to be made between this database and the cause-of-death component under construction in the framework of the more general Human mortality Database (HMD).
- Technical tools
Repeating the same type of patient work of time series reconstruction for various countries in different contexts leads to identify or define and implement new tools to make the work easier. Each new strategy/new tool will be clearly documented and shared by all participants of the project through a shared website. At the end of the project they will be made accessible to other colleagues (either for research or for teaching).
In the same time each main piece of work (time series reconstruction, new or refined methods for data analysis) will be recorded in technical reports, which could help for further investigators in the same field.
- Scientific publications (see below)
- Theoretical reappraisals and explanatory findings
One of the ultimate goals for undertaking so many empirical works is to question more precisely the current theory of divergence-convergence cycles: a crucial input to foresee future trends. The result could be a partial reassessment or impairment of the theory.
Also the special focus that will be made on health policy impact through the building of various scenarios of cause-of-death mortality prospects will give materials for demographic and socio-economic forecasts.

As final output, the project will provide various scenarios of possible future changes in population health according to selected public policy objectives. Those will include various options such as a longer mean duration of life, a greater number of individuals reaching the current mean duration of life, a greater concentration of deaths around the modal age. This tool will also help determine how changes in mortality from a specific cause of death could serve these different objectives. For instance, it will show how a fall in mortality from one specific cause (external causes or cancer or Alzheimer’s disease, for example) will affect the mean age at death (life expectancy), but also the age-distribution of deaths and the number of people reaching a certain age. It will also determine whether this result will be identical for all selected countries and will measure the impact on the general health of the population, including gender inequalities, with focus on dependency levels and population ageing.

The project will result in various type of scientific publications: monographic (to analyze long term trends in causes of death in each selected country), comparative (contrasts/similarities between selected countries), explanatory and theoretical (around health transition, health policy impact). About 30 papers will be presented in three years at various international meetings (PAA meetings, European Conference, IUSSP general conference, Asian population conference) a major part of them being submitted to scientific journals afterwards.

Why do some countries and population groups experience continuous improvement in health and survival while others are lagging behind? This question remains poorly understood and will be addressed in the project. Its main purpose is to better understand historical divergence in mortality within the developed world in order to identify factors of mortality change and to foresee future health challenges. Special attention will be paid to information on detailed medical causes of death as a principal source for unrevelling unravelling the causes of the observed changes in human survival and longevity.

The project will build a large international database of consistent series of cause-specific mortality functions (available on line at the end of the project) and will develop original techniques of decomposition to analyze them. Three main tasks structure the project: 1) finding out how to link cause of death structures and measures of survival and longevity using existing consistent mortality series to show how higher or lower mortality from particular diseases influence inter-country differences and changes in length of life; 2) reconstructing consistent mortality series by cause for nine large industrialized countries, including the United States, Japan, Russia, Germany (West and East), England and Wales, France, Spain, and Poland; and 3) analysing analyzing trends and differentials based on all countries cause-of-death series by age and sex to propose revisions to the health transition theory, to delineate possible mortality futures and to help design appropriate health policies and interventions.

New methodological developments will be brought to the following existing techniques: the method for fixing ruptures to correct for disruptions introduced by the revisions of the International Causes of Death classification, the multiple-decrement life table framework for decomposing life expectancy and lifetime loss, measures systematically estimating the effects of eliminating specific causes of death on the survival function and life-table aging rates, and universal decomposition methods to to split decompose a change in length of life or any other aggregate measure by age and cause.

Built on a very rich past experience of collaboration between INED and MPIDR, the project will bring together dedicated teams of senior and junior researchers in the two countries. Regular meetings will be organized to foster exchanges among the team members, and diverse communication tools, through on-line support in particular, will be implemented to promote such exchanges.