Sci. Aging Knowl. Environ., 28 June 2006
Vol. 2006, Issue 10, p. pe16
[DOI: 10.1126/sageke.2006.10.pe16]


Aging On the Job

James Levine, Justin Heet, and Barbara Burlingame

The authors are at the Mayo Clinic, Rochester, Minnesota 55905, USA (J.L.), the Hudson Institute, Indianapolis, Indiana, USA (J.H.), and the Food and Agricultural Organization of the United Nations, Rome, Italy (B.B.). E-mail: Levine.james{at} (J.L.)

Key Words: workforce • labor • non-exercise activity • physical activity • work


One of the legacies of the 21st century will be the aging of the world's population and the concomitant decline in the labor force. Not only are people on average living longer but also the number of older individuals represents a greater proportion of the population as a whole (see, for example, Hoffmann Perspective). This trend is occurring in almost every country (1). For example, in high-income countries, people over 60 years of age accounted for 12% of the population in 1950 and 20% today. In 50 years, the number is likely to reach 30 to 40% (1). This increase in the proportion of aging persons could dramatically impact the global workforce.

How likely is it that there will be a concomitant increase in the participation of older persons in the labor force? Public policy is tightly linked to this question. After World War II, social security support radically reduced labor-force participation by the older populations in wealthier countries. For example, in 1972, France and several other European countries reduced the age of eligibility for public pensions from 65 to 60. This measure resulted in a dramatic decrease in the labor participation of French men aged 60 to 64 from 69% to 16% in 5 years (Fig. 1A). According to current projections, workforce participation by individuals 65 and older in the year 2010 will be approximately one-third the level in 1950 (Fig. 1B).

Figure 1
View larger version (20K):
[in this window]
[in a new window]
Fig. 1. Participation in the workforce and age. (A) Workforce participation in France from 1962 to 2002 for men and women aged 30 to 34 and 60 to 64. (B) Workforce participation in developed regions by age in 1950, 1990, and 2010. (C) Work burdens for women and men from the Ivory Coast. (D) Distribution of the world labor force by age in 1962 to 2002. (E) Distribution of the French labor force by age in 1962 to 2002. (F) A theoretical composition of the projected workforce participation in developed regions in 2040.

In the developed world, the aggregate effect of declining participation in the workforce by older populations has been largely offset by unprecedented increases in labor-force participation by women. If projections hold, the ratio of economically active males to females will drop from 1.8 in 1950 to 1.2 in 2010. But beyond 2010, the increasing participation rates of women in the workforce, regardless of how dramatic the trend is, will not be able to compensate for the declining participation of older workers, although we recognize that these predictions do not take into account the role of technological advance on improved labor productivity. In general (despite exceptions in some subpopulations), although the aging population is increasing numerically, global workforce participation by the elderly does not appear to be increasing overall.

An intuitive solution to the problem might be to reverse recent history and adopt public policy measures that would augment workforce participation by older workers. In 2010, about 8% of individuals older than 65 years of age are projected to participate in the workforce. If this rate is unchanged, in 2040 there will be 26 million older workers. But if workforce participation by older persons returns to the rates that existed in 1950, there would be 75 million older workers in 2040. Thus, greater workforce participation by the growing older population could have substantial economic impact, with implications for social, health care, and fiscal planners.

As policy makers plan for the socioeconomic implications of the growing aging population, they have largely overlooked the impact of human biology. In a variety of species, including worms, flies, bees, rodents, dogs, cats, nonhuman primates, and humans, levels of physical activity decline with age (2-10). Data in humans are derived by measuring physical activity levels (PAL), defined as total daily energy expenditure divided by basal metabolic rate, thereby allowing a person's metabolic rate to be corrected for their size. PAL values decline with increasing age after the age of 60 (2, 11) and may be associated with muscle wasting, termed sarcopenia (see Hepple Perspective). Thus, spontaneous activity (that is, activity that does not include exercise) declines with age.

Study Design

The principal component of spontaneous activity in individuals who are employed is the energy expended during their work hours, and in fact PAL values are predominantly predicted by the type of work performed (2, 11). Because spontaneous activity declines with age, one would expect that work burden would also decline. Work burden is difficult to measure in high-income countries where pensions and employment laws might provide artificial constraints on work. Thus, to address the hypothesis that work participation declines with age, we examined data on work burdens in 1787 women and 1565 men living in three rural areas of the Ivory Coast geographically separated by more than 400 km: the Northern Savane, West Forest, and East Forest regions. This is a community that is minimally fettered by legal constraints (12)

These cross-sectional data were recorded over 7 days. The study period encompassed the full spectrum of agricultural activities. In Northern Savane, these activities included hunting, crop tending, and harvesting; in West Forest, food gathering, tree felling, sowing, and planting; and in East Forest, crop tending, harvesting, and land clearing. A trained enumerator observed each subject for all waking hours; every 15 minutes, the enumerator assigned one of 200 numeric codes to represent the subject's activity during that period. Activities for each individual were recorded separately and independently of cohabitants' activities. Data were compiled for the 22,397 person-days and 23 million data points.

We found a decline in workforce participation with increasing age for both women and men (Fig. 1C). In addition, the nature of the work performed by the elderly was different from that done by younger, gender-matched individuals; in both agricultural and domestic tasks, the elderly carried out more specialized, less thermogenic tasks. Because the data were gathered by an objective observer rather than by asking the subjects to rate their activity, we were intrigued that work practices indeed followed the decline in spontaneous activity associated with aging. This decline in work participation appears to mirror the decrease in spontaneous activity associated with biological aging. Until prospective studies are complete, it is not possible to define whether the physical aspect of aging is responsible for the decline in work participation or whether the more skilled older workers are assigned less industrious tasks that take advantage of their greater workplace experience. Regardless, the decline in workforce participation by the elderly subjects appears to mirror the decrease in spontaneous activity associated with biological aging.

To determine whether this trend could be generalized beyond the population we studied, we compiled labor-force data from 33 countries from 1962 to 2002 (13). Throughout the world, participation in the workforce declines with increasing age (Fig. 1D). This is true even of France, where the impact of the relatively short-term social policies described in Fig. 1A did not affect the overall work-distribution curve (Fig. 1E). These data are cross-sectional in nature and so biases may exist, such as selection for work because of beneficial health status. Thus, longitudinal data will be need to be confirm these observations. Nonetheless, these data suggest that biology may supervene social policy.


Reversing the trend of declining workforce participation by older workers would require not only public policy measures that encourage participation by this group of workers but also measures to somehow counter the biological phenomenon of decreased total physical activity with aging. If such a change were to occur, it would produce a colossal shift in the age distribution of the workforce, as illustrated in Fig. 1F.

Beyond biology, many elements of such a change have yet to be studied. For example, given that in many nations of the developed world health care costs are paid by private health insurance provided through employers, changing the work participation curve to include increasing numbers of older individuals could transfer large amounts of health care costs from the public to the private sector. The economic impact of such a change is worthy of investigation.

We speculate that refining job specifications for older workers may be key for increasing their participation in the workforce. Both in the Ivory Coast population and in animal models, older workers perform tasks that take advantage of the longevity of their experience while conceding to decreased physical strength (14). For example, honey bees transition from hive work to foraging as they get older (14), a transition remarkably similar to our observations of dwellers from the Ivory Coast.

In conclusion, as the world's population continues to age, more elderly people will be available to join the work force. For this aging population to seriously impact the workforce, there would need to be a dramatic shift in current work practices, for example, by offering elderly workers jobs that cater to the age-related decline in spontaneous activity and exploit their experience-enhanced skill set. Such shift would necessitate enormous social change and would defy the biological decline of physical activity with aging. As we try to project the economic impact of the aging population, we ought to bear in mind that biological factors may overlay and ultimately supervene economic planning. As we try to project the economic impact of the aging population, we ought to bear in mind that biological factors that affect the ability to work may overlay and ultimately supervene economic planning that aims to influence work practices. Thus, it may be helpful to determine how to reverse or acquiesce to the biological impact of aging on physical activity.

June 28, 2006
  1. Department of Economic and Social Affairs, Policy Division, "World Population Ageing 1950-2050," (United Nations, New York, 2002).
  2. A. E. Black, W. A. Coward, T. J. Cole, A. M. Prentice, Human energy expenditure in affluent societies: An analysis of 574 doubly-labelled water measurements. Eur. J. Clin. Nutr. 50, 72-92 (1996).[Medline]
  3. T. B. Kirkwood, C. E. Finch, Ageing: The old worm turns more slowly. Nature 419, 794-795 (2002).[CrossRef][Medline]
  4. S. Shibata, Y. Minamoto, M. Ono, S. Watanabe, Aging impairs methamphetamine-induced free-running and anticipatory locomotor activity rhythms in rats. Neurosci. Lett. 172, 107-100 (1994).[CrossRef][Medline]
  5. J. H. Marden, B. Rogina, K. L. Montooth, S. L. Helfand, Conditional tradeoffs between aging and organismal performance of Indy long-lived mutant flies. Proc. Natl. Acad. Sci. U.S.A. 100, 3369-3373 (2003).[Abstract/Free Full Text]
  6. N. T. Papadopoulos, J. R. Carey, B. I. Katsoyannos, N. A. Kouloussis, H. G. Muller, X. Liu, Supine behaviour predicts the time to death in male Mediterranean fruitflies (Ceratitis capitata). Proc. R. Soc. Lond. B. Biol. Sci. 269, 1633-1637 (2002).[Medline]
  7. J. W. Vaupel, J. R. Carey, K. Christensen, Aging: It's never too late. Science 301,1679-1681 (2003).[Abstract/Free Full Text]
  8. F. Nemoz-Bertholet, F. Aujard, Physical activity and balance performance as a function of age in a prosimian primate (Microcebus murinus). Exp. Gerontol. 38, 407-414 (2003).[CrossRef][Medline]
  9. C. T. Siwak, P. D. Tapp, S. C. Zicker, H. L. Murphey, B. A. Muggenburg, E. Head, C. W. Cotman, N. W. Milgram, Locomotor activity rhythms in dogs vary with age and cognitive status. Behav. Neurosci. 117, 813-824 (2003).[CrossRef][Medline]
  10. S. E. Peachey, E. J. Harper, J. M. Dawson, Effect of ageing on resting energy expenditure in cats. Vet. Rec. 144, 420 (1999).[Free Full Text]
  11. K. R. Westerterp, Impacts of vigorous and non-vigorous activity on daily energy expenditure. Proc. Nutr. Soc. 62, 645-50 (2003).[CrossRef][Medline]
  12. J. A. Levine, R. Weisell, S. Chevassus, C. D. Martinez, B. Burlingame, W. A. Coward, The work burden of women. Science 294, 812 (2001).[Free Full Text]
  13. Organisation for Economic Co-operation and Development, "Labour Force Statistics" (Organisation for Economic Co-operation and Development, 2003).
  14. Y. Ben-Shahar, A. Robichon, M. B. Sokolowski, G. E. Robinson, Influence of gene action across different time scales on behavior. Science 296, 741-744 (2002).[Abstract/Free Full Text]
  15. The authors acknowledge helpful discussions with Richard Jackson, Program Director and Senior Fellow, Center for Strategic and International Studies Global Aging Initiative. This work was supported by NIH grants (DK56650, 63226, and 66270 and AG26117) and by the Robert and Arlene Kogod Foundation.
Citation: J. Levine, J. Heet, B. Burlingame, Aging On the Job. Sci. Aging Knowl. Environ. 2006 (10), pe16 (2006).

Science of Aging Knowledge Environment. ISSN 1539-6150