Home Contact us Site Map

Cette page existe uniquement en anglais

Home > Activities > Committees >Longevity and Health > Seminar Report

Third IUSSP seminar on longevity and health

Rockefeller University, New York, 20-22 October 2003
Organised by the IUSSP Committee on Longevity and Health



The IUSSP Committee on longevity and health, in collaboration with the Laboratory of Populations at Rockefeller University, held a seminar, "Increasing Longevity: Causes, Consequences and Prospects", in New York, on 20-22 October, 2003. The seminar was supported by the IUSSP, the National Institute on Ageing (NIA) and Rockefeller University. It was divided in three parts: (1) recent trends of old-age mortality decline and theoretical and methodological issues in longevity increase ;(2) causes and differentials of old-age mortality decline and its future prospects; and (3) recent trends of old-age health improvement and well-being of the oldest old in the context of old-age mortality decline. Around 55 participants, worldwide specialists of human longevity, physicians, biologists, demographers, epidemiologists, economists, gerontologists and statisticians were gathered and 26 papers were presented in the seminar.

The seminar was opened by the Chair of IUSSP Committee on Longevity and Health (Jean-Marie Robine) and welcomed by Robert Butler (President, International Longevity Center), Joseph Chamie (Director, Population Division of the United Nations), and Joel Cohen (Head, Laboratory of Populations, Rockefeller University). The President of the IUSSP (Jacques Vallin) gave a positive remark on the performance of the Committee of Longevity and Health in the last few years and thanked the NIA for continuous financial support. Jean-Marie Robine expressed his gratitude to the hosting institution and the committee members and addressed the impact of books and papers written by Vanio Kannisto, a committee member who passed away in 2002, on the field of mortality and longevity. Thanks to the Committee, the unpublished paper of Kannisto was kept after his death.

The first day of seminar on October 20, 2003, covered four sessions. The first two sessions focused on the recent trends of old-age mortality decline including East Asia. Two afternoon sessions presented a part of well-being of the oldest old in the context of old-age mortality decline and discussed the theoretical and methodological issues in longevity increase.

Session 1: Recent trends of old-age mortality decline (A)
James W. Vaupel opened this session by discussing the trends in the latest data available on mortality after age 80. The central question in his paper "Continuing Improvements in Survival at Older Ages" addressed whether the remarkable improvements in survival that occurred between 1950 and 1995 have continued since 1995. His research showed that the population of aged 100+ remarkably exploded from 1950 to 2000. The rate of increase is 2.5 times from 1990 to 2000. He highlighted the average annual improvement in mortality between successive ten-year in 8 countries indicating there is no evidence of deceleration for the improvement in mortality for females.
Jean-Marie Robine presented the second paper in this session authored by Jean-Marie Robine, Graziella Caselli, Yasuhiko Saito, Carol Jagger, F. Paccaud, F van Poppel, CNJ Ruisdael, and James Vaupel: "An unprecedented increase in the number of centenarians", adressing the emergence of centenarians in 8 countries - Denmark, England and Wales, France, Italy, Japan, the Netherlands, Norway and Switzerland. The authors firstly described census data and annual population counts respectively and later presented the Centenarian Doubling Time (CDT) and commented on the mortality data. More importantly, they introduced the Centenarian Ratio (CR) i.e. the ratio of the number of centenarians to the number of the people aged 60 years forty years before - per 10,000 -, for assessing the importance of the number of centenarians in many countries with simple empirical data. In the final section they discussed the results and the health status of this new segment of the population.

Session 2: Recent trends of old-age mortality decline (B): with focus on East Asia
The paper "Recent Mortality Trends and Prospects: Hong Kong and Taiwan" presented by Edward Tu and co-authored by Wang Jian Ping examining the level, rate and past changes of mortality in Hong Kong and Taiwan based on the historical series of the mortality evolution. In their paper, they used Hong Kong and Taiwan time series mortality data from the period of 1971-1999 and 1970-2001 to project future mortality. The results showed that the forecasted life expectancy at birth in 2050 is 85.5 years and 89.1 years for males and females respectively in Hong Kong with a 95% confidence interval. Similarly, the forecasted value of life expectancy at birth is 83.2 years and 88.8 years in Taiwan, respectively for males and females. They concluded that the age pattern of mortality varies over time and older ages gradually make greater contribution to mortality decline.
Zachary Zimmer, Linda Martin, and Hui-Sheng Lin's paper "Determinants of old-age mortality in Taiwan" used Gompertz regression to model mortality and constructed time-varying covariates in multiple survey waves. Their results showed that education effects were attenuated after the introduction of some health indicators. Functional and global assessments of health had stronger associations with mortality than do reports of health behaviors or particular chronic conditions such as diabetes. Mainlanders had higher survival than others. The survival of older adults with the greatest number and severity of functional limitations improved over the 1990s, suggesting a possible beneficial influence of the insurance program.

Session 3: Well-being of the oldest old in the context of old-age mortality decline (A)
Linda Fried and Luigi Ferrucci illustrated the new paradigm of measuring "Frailty in older persons" . They discussed how the concept of frailty can be used in geriatric care to identify persons who are ideal candidates for prevention interventions. They gave an overview of the current knowledge on the frailty syndrome from the etiological and physiological point of view and reviewed possible approaches to pharmacological and non pharmacological treatments. They highlighted that physical exercise has a positive effect on all modulator systems and is also the only intervention for which they have some evidence of disability prevention. They concluded that exercise increases the level of anabolic hormones such as testosterone and DHEAS, substantially reduces the circulating levels of inflammatory biomarkers, stimulates the production of free radical scavengers and, in some instances, improves the function of the autonomic system.
Zeng Yi presented a paper entitled "A new method for correcting the underestimation of disabled life expectancy inherent in conventional methods: application to the oldest old in China", co-authored with Gu Danan and Kenneth Land demonstrating that the widely used disabled life expectancies based on conventional methods (i.e. multi-state life table approaches or the Sullivan method) are significantly underestimated due to assuming no functional status changes between age x and death. They proposed a new method to correct the bias and apply it to the longitudinal survey data of about 9,000 oldest old Chinese aged 80-105 collected in 1998 and 2000. In their application the age trajectories of ADL status - specific death rates and ADL state transition probabilities of the oldest old were investigated for the first time in a developing country.

Session 4: Theoretical and methodological issues in longevity increase
John Bongaarts' presentation of the paper, "Estimating mean lifetime" by John Bongaarts and Griffith Feeney, demonstrated that the conventional estimates of period life expectancy (i.e. the life expectancy implied by current age-specific mortality rates and calculated with life table methods) are affected by an undesirable "tempo effect." The tempo effect is positive when the mean age at death is rising and negative when it is declining. The authors provided formulas applicable to populations with high life expectancy to adjust for this distortion. Estimates of the effect for females in three countries with high and rising life expectancy range from 1.6 years in the US and Sweden to 2.4 years in France for the period 1980-1995. The authors underline the importance of distinguishing between current observed death rates and current mortality conditions when interpreting the findings. They also argued that tempo effects distort both the observed death rates and the corresponding life expectancy, so that their values give a misleading indication of current mortality conditions.
Jay Olshansky presented "Biological evidence for limits to the duration of life" by examining temporal aspects of biological phenomena in three mammalian species to assess the biological plausibility. This examination revealed that: 1) physiological declines associated with reproduction consistently occur at ages that are less than one-third of the median age at death, 2) physiological parameters associated with aging in humans lose eighty percent of their functional capacity by age 80, and 3) young versus old individuals can be distinguished by the pathologies detected at death. The biological evidence suggests that organisms operate under warranty periods that limit the duration of life of individuals and the life expectancy of populations. These findings are used to discuss the issue of limits to the duration of life and the validity of mathematical models forecasting human longevity.
In their paper entitled "Lengthening of life: do current divergences contradict or support a general principle of convergence?" Jacques Vallin and France Meslé tried to integrate Abdel Omran's original epidemiologic transition unchanged into a broader scheme that encompasses both new advances in health and the extraordinary diversity of situations that prevail across the world. They argued that Omran's epidemiologic transition should be taken into account as the first stage of a global process of health transition which develops into several stages depending on different major changes in health strategies. Because not all societies are equally prepared to innovate or draw the benefits of innovation from outside, such major changes naturally result first in a more or less acute process of divergence followed by a process of convergence when late-entering countries become able to catch up the pioneers.

The second day of the seminar on October 21, 2003 comprised four sessions focusing on causes, differentials and future prospects of old-age mortality decline.

Session 5: Causes and differentials of old-age mortality decline (A)
Shiro Horiuchi opened this session presenting James Riley's paper "Mortality decline at ages 65-84". In his paper, the author considered that the most substantial gains in early old age survivorship occurred in many countries before the introduction of the medical remedies and lifestyle modifications that have drawn most research. For the most part gains since 1960 that have occurred in one country have generally not been shared even by neighbouring countries. Jamaica and Costa Rica began improvements in early old age survivorship later than the US, England and Wales, Sweden, Norway, and France, but caught up. France gained survivorship in these ages at a particularly rapid pace, while Denmark lost relative position. The author highlighted that it may be more revealing to compare less educated and lower income groups across countries. Demographers may also be able to contribute in the particularly thorny problems involved in the timing of disease and death. The author suggested that new research is needed on changes in behaviour adopted for other reasons that may nevertheless have paid off in health.
Joanna Lahey presented a paper on "Becoming oldest old: evidence from historical US data" by Dora Costa and Joanna Lahey, illustrating that indicators of environmental insults in early childhood and young adulthood as well as the quarter of birth, residence, occupation, wealth, and the incidence of specific infectious diseases affected older age mortality. They found that the effect of quarter of birth on older age mortality has diminished over the twentieth century and that the declining impact of quarter of birth explains 16 to 17 percent of the difference in ten year mortality rates of American age 60-79 in 1900 and in 1960-1980. They estimated that at least one-fifth of the increase between 1900 and 1999 in the probability of a 65 year old surviving to age 85 may be attributed to early life conditions.
Russell Luepker's paper entitled "Changing patterns of mortality from major chronic diseases" presented trends in mortality, morbidity, risk factors and cost of care associated with cardiovascular disease. The author briefly discussed international trends and commented on their implications. He concluded that cardiovascular diseases were the result of an improved and more affluent lifestyle in an aging population. Medical research and healthcare communities met this new epidemic with a variety of strategies. The discovery of risk factor associations led to healthy lifestyle promotion characterized by improved eating patterns, physical activity and tobacco cessation. High-risk individuals, by virtue of elevated blood pressure and blood cholesterol, were identified by screening programs and treated. These approaches reduced the level of risk and subsequently prevented premature events, resulting in the reversal of the age-adjusted rates of cardiovascular disease by the mid-1960s. That reversal was mainly manifest by the delay of these chronic diseases, pushing them into later years of life. The positive result has been a continued increase in life expectancy in the last half of the 20th century.

Session 6: Causes and differentials of old-age mortality decline (B)
Eileen Crimmins presented her work on "Social Differentials in Mortality and Health at the Older Ages" and investigated socioeconomic differentials in a variety of health outcomes among people over the age of 70. The aim of her paper was to examine how such differentials change with age and how they differ by cause. In addition her paper related current health among this aged group to characteristics of the cohort's past, including survival and exposure to childhood illnesses.
Kenneth G. Manton, Igor Akushevich, Alexander Kulminski, and Kenneth C. Land's joint paper on "Modeling Regenerative Processes in Aging Human Populations" was presented by Kenneth C. Land. The authors proposed a new model to detect regeneration in longitudinal human population data. They suggested to analyse health changes at later ages by extending the well-known Fokker-Planck-Kolmogorov equation for random walk processes. The non-linear form of the mode describes state variable field effects on ageing dynamics and mortality. This model provides a biologically more informed basis for projecting future population growth, and the future distribution of health trends, in the elderly U.S. population for both federal policy concerns and to project health care needs for private markets.

Session 7: Future prospects of old-age mortality decline (A)
Shripad Tuljapurkar's paper on "Changes in the variance of age at death: analysis and implications for forecasts and models" began with the observation by Wilmoth and Horiuchi that there is a secular trend in the variability of the age at death that accompanies the trend to longer life. He examined several aspects of these trends: first, the extent to which the trend in variance is affected by the actual distribution of deaths at very high ages (over 90); second, the relationship between this trend and alternative models of the age-pattern of death at high ages; third, the implications of this trend for the relative importance of intracohort selection in the time-evolution of mortality patterns; and fourth, the implications of this trend for forecasts of mortality using alternative methods of projection.
John Wilmoth presented his paper on "The persistent near-linear increase of life expectancy at birth in industrialized countries: explanation and prospects for the future" and built four hypothetical scenarios of future mortality trends. Two scenarios hold the age pattern of mortality decline constant over time, while the other two assume a gradual shift in this pattern toward older ages. These calculations illustrate that the aging of the mortality decline could be a key mechanism driving the near-linear increase of life expectancy at birth. In light of these results, he suggested that a revision of standard mortality forecasting methodology should be considered.

Session 8: Future prospects of old-age mortality decline (B)
Hania Zlotnik and Thomas Buettner's co-authored paper on "Prospects for increasing longevity as assessed by the United Nations: What new methodology can tell us?" giving an overview of future prospects for the increase of longevity at the world and regional levels according to the most recent assessment of mortality prospects prepared by the United Nations Population Division. She discussed the development of a new methodology for the estimation and modeling of mortality change by age and over time which can bring insights to the projection of mortality over the long run, particularly in countries with deficient information on mortality levels and trends. The new methodology was illustrated by application to selected countries and comparisons were made between the results obtained and those embodied in the official estimates and projections of the United Nations. Comparisons were made not only with respect to measures of mortality levels but also with respect to changes in the expected age distributions of ageing populations brought about by the use of the new methodology.
Lloyd Demetrius presented his work on "The mortality plateau: predicting the age-distribution of the oldest-old" and addressed an increase in human mortality from the age of 35 years to 90 years, then a deceleration until age 110 and finally a decrease . He highlighted the questions: can demographic theory predict the age at which the deceleration in mortality rate occur? Can demographic theory predict the age distribution of the oldest old and predict the change in human centenarian rate? He discussed that the variability of the age specific fecundity and mortality is the result of evolutionary forces and social and environmental constraints. Mortality trajectory of post-reproductive is not due to random drift but is critically determined by the effort of natural selection. He proposed the Entropic paradigm to describe demographic stability (i.e. the sensitivity of the population numbers to perturbation in the birth and death rate). Large entropy indicates late sexual maturity, broad reproductive span, small litter size whereas small entropy suggests early sexual maturity, narrow reproductive span, large litter size.

The last day of seminar on October 22, 2003 mainly discussed recent trends of old-age health improvement and well-being of the oldest old in the context of old-age mortality decline.

Session 9: Recent trends of old-age health improvement
Vicki Freedman presented her paper on "an overview of late-life health and disability trends in the U.S." assessing the quality, quantity, and consistency of recent evidence on U.S. trends in the prevalence of self-rated old age disability and physical, cognitive, and sensory limitations during the late 1980s and 1990s. The author also evaluated the evidence on trends in disparities by major demographic groups. The prevalence of any disability declined significantly for older U.S. adults during the 1990s from -1.55% to -0.92% per year on average. However, these improvements did not evenly distribute among all specific measures of disability. To the contrary, late-life disability declines concentrated among IADL limitations. Estimates of the average annual rate of decline of the former ranged from -2.74 to -0.40. The prevalence of severe cognitive impairment may also have declined. Limitations in hearing appeared to have been constant over the last decade.
Jacques Légaré, Robert Bourbeau, Bertrand Desjardins and Chad Deblois's joint paper on "Variation in cohort size and lower mortality in the elderly: Implications for pay-as-you-go healthcare systems" presented by Jacques Légaré provided strong arguments in favour of adding a funded healthcare plan that would be fairer for future generations. The pay-as-you-go funding of healthcare is inherited from a period in time where each cohort was larger than its predecessor; it is now outmoded in a world where demographic dynamics are quite different. The phenomenon of imbalance between succeeding cohorts in terms of aging and increased longevity, together with the use of costly technologies at the end of life, could seriously impede our ability to fund public pension plans and the healthcare and social services sector. Implementing a health fund that conserves social gains and ensures intergenerational fairness is becoming one of the most pressing issues for today's policy-makers in industrialized societies.

Session 10: Well-being of the oldest old in the context of old-age mortality decline (C)
Gary Andrews's presentation on "The voices of the oldest old - how can they be heard? - Evidence from a longitudinal study" by Gary Andrews and Sandra Davis, explored an oldest old group of survivors in a longitudinal study of the oldest old in an Australian urban population. The findings show that there is a high level of general satisfaction with life, family and community along with a positive sense self worth and purpose. A number of predictors of achievement of what they defined as "healthy ageing" were identifiable from data obtained nine years earlier in the baseline data collected on the Australian Longitudinal Study of Ageing participants. These were: health and lifestyle factors including: no arthritis, regular alcohol consumption, fewer IADL disabilities, and greater activity in domestic chores and household maintenance; social factors and an economic factor of self-report that the level of assets were sufficient to meet the current financial needs.
Madeleine Rochon's paper "The adaptation of society and Baby Boomers to aging: two interacting processes" addressed the questions: what type of care is needed? How should care be provided and financed? From a historical perspective, four processes: aging, advancement of knowledge, replacement of generations, and continuous changes to social organization and health care systems were taken into consideration. She presented an overview of the knowledge on an increase in the length and quality of life and reviewed intergenerational changes, which are likely to affect the characteristics of the oldest population in the future. She also examined needs of the oldest old and their caregivers both in terms of quality and quantity of services. The effect of age in relation to other variables such as vital status and health status were presented. She presented a number of ongoing changes in the methods of assessing the population's needs, planning and organizing services for frail older persons. Finally, various problems of financing long term care based on existing systems and demographic, economic and health trends were discussed.

Session 11: Well-being of the oldest old in the context of old-age mortality decline (D)
Jenny De Jong Gierveld and Pearl Dykstra's paper "What impacts does longer life have on family, and informal care giving by children in particular?" focused on the patterns of interaction between older parents and their children as parents reach advanced ages. They looked at the impact of parents' growing older, and at the implications of partner status transitions in the lives of both the older and the younger generation for the quality of intergenerational relationships and support arrangements. Data came from the NESTOR 'Living Arrangements and Social Networks' Survey. In 1992, face-to-face interviews were conducted with 4494 respondents. They constituted a random sample of men and women born in the Netherlands between 1903 and 1937. The analyses focus on a subgroup of men and women aged 65 to 89 years living independently, who have at least one living child. In addition to variables mentioned so far, gender, health, educational level, number of children and travel distance to the nearest child are among the predictors of familial embedment and support.
Guo Zhigang's paper on "What impacts does demographic transition have on family support for the elderly in China?" presented the analyses on the family size, living arrangement of the elderly and its determinants, inter-generational wealth flows. Based on the 1982, 1990, and 2000 national population census data (according to both the officially published and the findings computed based on the original sample data), he outlined the population change and presented some results about China's current fertility policy and its implications to future population and family structure.

Session 12: Well-being of the oldest old in the context of old-age mortality decline (E)
Victor G. Rodwin and Michael Gusmano's joint paper on "Growing older in world cities: Models for the future from New York, London, Paris and Tokyo" presented by Victor G. Rodwin outlined and summarized the major findings from ILC-USA's World Cities Project on the living arrangements, population characteristics, and health of older persons living in the urban core of these world cities, and the issues they raise for urban aging and long-term care in the future. In his presentation, he highlighted several issues such as how do we use the distribution of elderly population to locate the number of nursing homes? Does neighbourhood really help when elderly need help for instance, a kind of personal assistance? How does neighbourhood affect the late life? He found that little research has been carried out on the current and previous neighbourhood and that the geographical area affects the activity of elderly.
Robert Butler presented his paper on Social goals and visions "What can we expect for the oldest old in the future in terms of well being and quality of life?" addressing factors (e.g. individual and population educational level) affecting well-being and quality of life. He highlighted health in the 21st century covering a variety of topical issues such as genomic/ proteomics, how to extend life, disease mission research, lifestyle research and prevention strategies as well as environmental and pathogen controls. He also discussed the trend in Total Fertility Rate (TFR) decline and longer life. What are the implications of depopulation when TFR continues at low level and longevity increase? What are the determinants of disease in older persons? What kind of health and medical services do we need? Lastly, he concluded by an old saying "if you have your health you have everything" for ending this session.

The seminar was closed by Jean-Marie Robine, Chair of the IUSSP Committee on Longevity and Health. He announced the progress of publication in the previous seminar that was held in Montpellier, 2000 and thanked the enormous effort of the scientific committee members and the participation of each speaker. This successful seminar inspired much discussion and debate on the increase in human longevity, in particular, the lengthening of survival at old ages. Most importantly, the seminar reviewed and discussed latest empirical findings and theories about causes, consequences, and future prospects of the longevity expansion, thereby stimulating and encouraging further investigation.
Siu Lan Karen Cheung

List of participants

Steven Albert
Columbia University

Gary Andrews
University of South Australia

John Bongaarts
Population Council

Thomas Buettner
United Nations

Robert Butler
International Longevity Center

Graziella Caselli
University of Rome

Joseph Chamie
United Nations

Siu Lan Karen Cheung (Junior Demographer)
University of Rome

Joel Cohen
Rockefeller University

Dora L. Costa

Eileen Crimmins
Southern California University

Danan Gu
Duke University

Lloyd Demetrius
Harvard University

Griffith Feeney

Luigi Ferrucci
National Institute on Aging

Vicki Freedman
Polisher Research Institute

Linda Fried
Johns Hopkins University

Jenny Gierveld
Netherlands Institute for Advanced Study

Guo Zhigang
Peking University

Michael Gusmano
International Longevity Center

Sabine Henning
United Nations

Shiro Horiuchi
Rockefeller University

Rene Jahiel
University of Connecticut

Fanny Janssen
University Medical Center Rotterdam

Iliana Kohler
University of Pennsylvania

Alexander Kulminski
Duke University

Joanna Lahey
Massachusetts Institute of Technology

Kenneth Land
Duke University

Jacques Légaré
University of Montreal

Russell Luepker
University of Minnesota

Laurent Martel
Statistics Canada

Linda Martin
Population Council

Dobrosav Matiasevic
DeWitt Rehabilitation and Nursing Center
Fax 212 772 8750

France Meslé
National Institute of Demographic Studies, France

M. Nizamuddin
Columbia University
United Nations

Jay Olshansky
University of Illinois at Chicago

Francois Pelletier
United Nations

James C. Riley
Indiana University

Jean-Marie Robine
National Institute of Health and Medical Research, France

Madeleine Rochon
Ministry of Health and Social Services, Canada

Victor Rodwin
New York University

Kathe Rogerson
Rockefeller University

Cheryl Sawyer
United Nations

Laura Shrestha
National Institute on Aging

Jacob Siegel

Richard Suzman
National Institute on Aging

Edward Jow Ching Tu
Hong Kong University of Science and Technology

Shripad Tuljapurkar
Stanford University

Jacques Vallin
National Institute of Demographic Studies, France

James Vaupel
Max Planck Institute for Demographic Research

John Wilmoth
University of California at Berkeley
Zeng Yi
Duke University

Zhongwei Zhao
ANU and Cambridge University

Hania Zlotnik
United Nations

Anatoly Zoubanov
United Nations

IUSSP, 3-5 rue Nicolas, F-75980 Paris cedex 20, France
Tel  +33 1 56 06 21 73 - Fax +33 1 56 06 22 04 - contact us