| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||
COMMENTARY |
Institute for Research in Extramural Medicine, and Department of Psychiatry, Vrije Universiteit Medical Center, Amsterdam, The Netherlands.
Address correspondence to Prof Dr. D. J. H. Deeg, Vrije Universiteit Medical Centre, Institute for Research in Extramural Medicine/LASA, Van der Boechorststraat 7, 1081 BT Amsterdam, the Netherlands. E-mail: djh.deeg{at}vumc.nl
Robine and Michel (1) make a commendable effort to formulate a general theory on population aging. Theories help to understand what is going on—and are especially useful for understanding invisible processes such as those on the population level. It is, moreover, good to see that the Robine-Michel theory builds on the seminal ideas proposed by Myers a decade ago (2,3). My comment boils down to the statement: "Global theories are nice, but the local complements the global!" I give three arguments to support this statement.
First, the Robine-Michel theory starts from the observation that life expectancy is increasing globally. However, one cannot ignore the recent, tremendous decrease in life expectancy in Russia and other East-European countries. The local forces that are at work here should be included in any general theory.
Second, when interpreting time trends, we need to consider all three of the possible time-related factors: age, period, and cohort. The majority of explanations forwarded to explain time trends are based on cohort factors. Robine and Michel, for example, mention higher education, and adult lives in better conditions. However, time trends may be explained by period effects as well. Analysis of age, period, and cohort factors influencing disability trends at older ages in the 1990s in the Netherlands, for example, showed that period effects were the main explanatory factors (4). Notably, these trends were increasing and related to cut-backs in health services—a consequence of national policy making. Thus, local factors counteracted global trends. Note, meanwhile, that the Dutch disability level is among the lowest in developed countries (5,6).
This leads to my third argument. Robine and Michel attempt to classify countries in terms of phases of the disability transition. They start from the initial level of life expectancy. What time is best called "initial" is an open question. Regardless of when this "initial" level is chosen, I propose to start from the other side: from the initial level of disability. When a disability level is high, such as in the United States, there is room for improvement—in other words, compression. When a disability level is low, such as in Australia or the Netherlands, it may be sensitive to local period influences, and thus prone to increase—in other words, expansion. Again: the local complements the global.
A final comment places doubts on the increase in life expectancy from the opposite end of the age spectrum. We know that, in the past century, life expectancy owed its greatest increase to the enormous decline in mortality at the youngest ages. In our age, medical science continues to make advances in neonatal care of preterm births. Moreover, a sizable proportion of babies is born after in vitro fertilization. The question now arises whether the disability and mortality levels in these children will be the same as those in children born at term and after normal conception when they reach older age (7). What influence will these medical advances have on the development of life expectancy in the late 21st century? And what influence on the level of disability at older ages? A general theory of population aging should not ignore these developments.
References
| ||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
