The term "oldest old" refers to persons at the upper segment of the age pyramid. By a conventional demographic definition those age 80 years and older are included in the oldest-old population, although, as noted below, alternative definitions are also in use. Since the end of World War II, in most countries the oldest old has been the fastest growing component of the population at large. This trend will almost certainly continue in the decades to come, with improving survival rates at very high ages.
The relationship between indicators of average lifespan, such as life expectancy at birth, and the age structure of populations is often misunderstood by statisticians, policymakers, journalists, and other users of demographic data. Increases in life expectancy are welcomed as a sign of progress and improvement in living conditions. The countries that are most advanced in the demographic transition are proud that their life expectancy at birth exceeds 80 years, at least among women (who, especially in low-mortality countries, enjoy an appreciable advantage over men). This was the case by the end of the twentieth century in Japan (the country which in the period from 1995 to 2000 had the highest female life expectancy at birth: 83.8 years) and in Hong Kong; in Finland, Norway, and Sweden; in most countries of southern and western Europe; and in Canada and Australia. (In the United States, female life expectancy during the period from 1995 to 2000 was, at 79.4 years, slightly short of this mark.) On the other hand, increases in the numbers of nonagenarians and even octogenarians are often viewed with alarm, as portending an increase in the burden of health care and old-age support. Most people, of course, do not die in the years narrowly bracketing the average life expectancy at birth.
The increasing number of old people is not a new phenomenon, although some still find it surprising. One reason for this is the way in which a final open-ended age group is used in tabulating age distributions and mortality rates: in the past, 65 years and over; now, 85 years and over. But 65 years and over or 85 and over tends to be popularly interpreted as approximately 65 years and approximately 85 years, not age brackets ranging from 65 or 85 to 100 or 110 years. The upper end of the distribution, the oldest-old population, is thus incorrectly visualized.
The Problem of Definition
The oldest old as an identifiable category within human populations always existed even if its definition in terms of biological, social, or chronological age was varying and fuzzy. During the seventeenth century, the absence of a concept of the oldest old was not very important because with life expectancy at birth at levels of 30 to 35 years, only a small minority–about 10 to 15 percent–of the population reached the age of 70.
The three great ages of humankind–youth, adulthood, and old age–correspond to the history of life from a biological point of view, with youth a phase of growth and development, adulthood a phase of reproduction and childrearing, and old age a phase of physical decline, beyond the selective pressures of evolution. The division also corresponds to the economic and social organization of the industrialized countries during the twentieth century around school, work, and retirement, reflected in the three broad age ranges of many statistical tables: 0 to 14 years; 15 to 64 years; 65 and over. But these socioeconomic ages have little reality for most developing countries, where very few old people enjoy retirement. And in the developed countries furthest advanced in the demographic transition, the period of old age, as defined by lessened or no economic activity, lengthens at both ends–as labor force participation tends to decline above the age 50 and as life expectancy at age 65 increases. This explains the logic of dividing old age into two parts: the "young old" and the "oldest old."
But on what criteria? Chronological age is a rather crude marker here. Everyone can agree that those aged 60 to 70 are "young old" and that those over 85 are oldest old, but what about those between 70 and 85? Criteria other than age might be considered.
Biodemographic approach. A simple biodemographic classification would consist of numbered generations or "ages": a child belongs to the first age, parents to the second age, grandparents to the third age, and great-grandparents (the oldest old in this scheme) to the fourth age. But this simple classification presumes reproductive success and a high degree of uniformity in the reproductive life cycle.
Functional approach. An approach largely popularized by the historical demographer Peter Laslett (1996) is that the fourth age–the age of the oldest old–starts when a person becomes physically dependent on another. Thus the third age starts when the person is released from the constraints of adulthood (paid work and education of children) and finishes when the loss of physical independence begins, without precise age limits. Indeed, the concept of loss of autonomy of old people is not easily pinned down as it involves three main criteria: physical dependence–the need for the assistance of another to perform the activities of daily life; cognitive autonomy–the capacity to make decisions for oneself; and social integration–the capacity to remain integrated in a community. Three indicators of life expectancy by health status (active life expectancy, dementia-free life expectancy, and institution-free life expectancy) are helpful in defining the boundary of the fourth age congruent with this approach.
Gerontological approach. In their model of successful aging, John Rowe and Robert Kahn (1997) distinguish four states, namely the absence of appreciable risk of degenerative disease or disability, the presence of risk, the presence of actual degenerative disease, and the presence of disability. Successful aging would imply prolongation of the period without accumulation of the risk of developing a degenerative disease or becoming disabled. The third age is generally associated with successful aging and the fourth age with unsuccessful aging. The fourth age would begin with the accumulation of risks of degenerative diseases and disability. The risk of "unsuccessful" aging is an important concept in gerontology, drawing on the concept of frailty.
Demographic approach. A definition of oldest old based on loss of autonomy faces the difficulty that a large minority of older people will never lose their autonomy. Demarcating the category of oldest old by setting a plausible if necessarily arbitrary age when persons are thought to enter it, such as age 80 or 85, is an approach commonly used by demographers. It should be remembered, of course, that an arbitrarily selected advanced age does not have the same connotations in the early twenty-first century as it did in the past when attainment of that age was exceptional. Nor does it have the same meaning for men as for women. Other demographic definitions of the oldest old, or the fourth age, would be in terms of the proportion of survivors–the highest age reached by at least 25 percent of the population; or in terms of life expectancy–the age at which there remains ten years of expected life. The nominal ages corresponding to these criteria, however, would vary depending on the level and pattern of mortality; hence, they differ from one country to another and from one time to another.
Numbers and Trends
In virtually all countries the populations at ages beyond 80 years–the oldest old–have strongly increased in size since 1950. This is a consequence of declining mortality and also of the fact that as time passes the oldest old consist of survivors of increasingly larger birth cohorts (groups of individuals born at the same time). The proportions of the oldest old within the overall population have also been increasing, in part as a result of declining fertility,
which narrows the base of the age pyramid. Thus, for example, in the 15 countries of the European Union the proportion at ages 80 and older was 1.2 percent in 1950 and 3.8 percent in 2000 (more than tripling). In the United States the corresponding percentages were 1.1 and 3.2.
Within the oldest old, the number and proportion in the higher ages increased especially rapidly. In 1995 James W. Vaupel and Bernard Jeune showed that in Western countries the number of centenarians doubled approximately every ten years starting in 1950; the doubling time for the number of people 105 years old was just a little shorter. Of course, in 1950, although the numbers of those over 80 were already numerous, few were over 90, very few were centenarians, and almost none had reached 105. In the lowest-mortality countries, such as France and Japan, in the early twenty-first century, the fall in mortality at old ages appears to be accelerating and the "centenarian doubling time" is becoming shorter. As a consequence of these changes, the maximum reported age at death has been strongly increasing. The longest reliably recorded human lifespan (for the French woman Jeanne Calment, who died in 1997) is 122 years.
These trends are likely to continue in the coming decades. Table 1 shows the size of the oldest-old population in 2000 in the three most populous countries of the world and in the 15-nation European Union and the proportions of the oldest old within the total population. The table also includes anticipated figures of the oldest old in the year 2050, as projected by the United Nations (UN) on rather conservative assumptions as to the future evolution of mortality. Note that the absolute numbers of the oldest old in 2050 are affected only by mortality (and to a degree also by future international migration), because those aged 80 years and older will be survivors of persons already alive (survivors of those aged 30 years or older in 2000). The future proportions of the oldest old within the total population are of course affected by future fertility as well as mortality. The projected proportions shown in the table are based on the UN's "medium fertility" assumptions.
As the table indicates, the rapid expansion of the oldest-old population is not limited to the Western world. Developing countries are also faced with greatly increased numbers over age 80. In China, for example, during the first half of the twenty-first century the number of oldest old is expected to grow nearly ninefold, in contrast to a 15 percent increase of the total population. By 2050, the oldest-old population of China is expected to be some 99 million. As a proportion of the total population that number would represent 6.7 percent–well above the corresponding proportions in 2000 in the United States or in the European Union.
Women invariably represent a high percentage of the oldest old. For example, in the United States in 2000, some 67 percent of those above age 80 were women. And among centenarians, women accounted for 87 percent of the total. In these conditions, further calculations of health expectancies (for example, active life expectancy and disability-free life expectancy) are more and more relevant. They suggest an increase in healthy life expectancy in some–but not all–countries such as the United States and France. In Austria, for instance, life expectancy in good perceived health at age 80 increased by 1.3 years between 1978 and 1998, going up from 2.8 to4.3 years.
Doblhammer, Gabriele, and Josef Kytir. 2001. "Compression or Expansion of Morbidity? Trends in Healthy-life Expectancy in the Elderly Austrian Population between 1978 and 1998." Social Science and Medicine 52: 385–391.
Laslett, Peter. 1996. A Fresh Map of Life, 2nd edition. Houndmills, Basingstoke, Eng.: Macmillan.
Robine, Jean-Marie, Isabelle Romieu, and Emmanuelle Cambois. 1999. "Health Expectancy Indicators." Bulletin of the World Health Organization 77: 181–185.
Rowe, John W., and Robert L. Kahn. 1997. "Successful Aging." Gerontologist 37: 433–440.
Thatcher, A. Roger. 2001. "The Demography of Centenarians in England and Wales." Population 13(1): 139–156.
Vaupel, James W., and Bernard Jeune. 1995. "The Emergence and Proliferation of Centenarians." In Exceptional Longevity: From Prehistory to the Present, ed. Bernard Jeune and James W. Vaupel. Odense, Denmark: Odense University Press.