15. Conclusions

Death rates at ages 80 and over have been calculated for the post-1950 period for 28 low-mortality countries that process sufficiently reliable data on deaths by single years of age to allow the use of the extinct cohorts method. The material available covers most of the countries representative of the low-mortality countries of the world.

        The study shows that mortality in old age has undergone in the developed countries during the post-war period a deep-going and fairly general transformation reaching much lower levels than have ever been recorded before. Unprecedented in known demographic history, this decline has made in the prevailing trend a break which had not been foreseen in population projections.

        The new wave of mortality decline has varied from country to country as to its onset as well as to rapidity and extent and has in some been interrupted by periods of stagnation or temporary increase. The age pattern of decline has varied from an essentially uniform one to one favouring either the relatively younger or the very oldest. The decline has reached the highest ages in which mortality can today be empirically measured, i.e. ages of 107-109 years.

        Among the great variety of national situations there has been on the whole a perceptible tendency of the decline to accelerate in more recent years. Yet, the development has not been leading towards convergence between countries, on the contrary. The only feature common to all 28 countries observed has been a development more favourable to the survival of women than of men, thus widening the already existing female advantage.

        Looking for the causes of this development we should note first of all that during the post-war period increasing numbers of persons have been reaching the age 80, the starting line of observation in the present study. The reasons for this increase lie partly in the past as advances in medicine and in living conditions have permitted people to reach higher ages. At the same time and perhaps because of it, old persons are receiving increased medical attention.

        While all this has created a long-term trend for improved survival, we have seen in the present study that important factors have come into play after 1950 and particularly in the last decades, reducing mortality even at the oldest ages in an unprecedented manner. These reductions have in most cases been comparable to and simultaneous with reductions in mortality among the younger elderly.

        These new factors which are distinctively period rather than cohort factors may include further advances in medical practice such as more generalized control of blood pressure but there is on the other hand solid evidence from many countries to the effect that an important and probably the main factor has been a change for a healthier lifestyle - a change which has been shown to have almost immediate effects on the probability of survival. Such behavioural changes prevent or delay the onset of morbid processes which would in time lead to serious health impairment or death. Maybe the aging process itself is thereby delayed.

        Studies from various countries have shown that the new lifestyle has been adopted first by the better educated, wealthy, urban sectors of the population and has been gradually filtering through to other social strata. The time lag in this process has almost certainly contributed to creating or aggravating social differentials in mortality which have been observed in many large-scale studies.

        The novelty and magnitude of the observed mortality decline seem to justify it being called a new stage in mortality transition. In contrast to the earlier stages in which the main benefits accrued to childhood and working age, the new stage is characterized by mortality decline in later life where the main beheficiaries are the elderly from their sixties to the eighties and nineties and, to a hardly lesser extent, even beyond.

        We do not, as yet, know the ultimate extent of the new mortality transition. But considering that the new lifestyle has not yet been anything like universally adopted, a vast potential exists for further improvement in health. And since this social process seems to be going on and as there is at the present time a strong downward momentum in death rates, the new transition may, barring unforeseen events, still continue for an extended period.

Updated by V. Castanova, 1 March 1999