The forecast aging of the United States—the doubling of the share of the population over 65 years of age in the next 75 years1—will have profound effects on the medical system. Pessimists worry that the increase in the elderly population will bankrupt the Medicare program; they urge the restructuring of Medicare coverage in order to accommodate the needs of this population. 2 Optimists note that medical spending is related to disability more than to longevity. If, as recent data indicate, the aging population not only is living longer but is also healthier, 3 then spending for medical care will decline over time. 4 Proposals to alleviate the economic impact of aging by restructuring the Medicare program to encourage enrollment in managed-care plans or by raising cost sharing are profoundly different from proposals to focus on health promotion. The article by Lubitz and colleagues in this issue of the Journal5 sheds light on this debate. Lubitz et al. simulated lifetime medical spending for elderly cohorts in different states of health at 70 years of age. They show that, with the exception of persons who are institutionalized at base line, who therefore incur greater costs over their lifetime than those who are not institutionalized at the age of 70, lifetime medical costs among the elderly are about the same, regardless of initial health status. Persons who are healthier at 70 years of age live longer but do not incur greater expenditures. Expenditures at the end of life have a central role in these findings. A large proportion of elderly persons become disabled at some point. Disability, and the institutionalization that often results from it, are very expensive. Only one third of the years lived after the age of 70 are lived with a disability, but the costs for care in those years account for two thirds of lifetime medical spending. Those who are healthier at base line will reach the disabled state at a later point but will still spend considerable time in that state. As a result, the lifetime spending for those persons will be the same as for persons in poorer health at base line.

The findings of Lubitz et al. support a key tenet of the optimists’ argument: the fact that people who live longer do not incur greater costs for medical services in total suggests that longer life in itself will not add to Medicare costs. Current Medicare projections, which implicitly assume that longer life costs more, 6 are thus substantially overstated. But the glass is half empty as well. The findings of Lubitz et al. suggest that improving the health of the cohort of persons who are becoming eligible for Medicare will not result in major reductions in medical spending; health promotion can go only so far in making Medicare affordable. Furthermore, although medical costs do not rise with improved health early in life, Social Security costs increase with longevity. In total, therefore, aging will result in increased public spending. A slow progression from good health to increasing levels of disability is not the only possibility. Many elderly persons die suddenly, with essentially no disability beforehand. Others, particularly persons with some forms of cancer, may be disabled for only a short period before death. Studies suggest that about half the deaths among the elderly are not preceded by a prolonged period of disability. 7 Among the other half, death results from chronic organ failure or frailty and is preceded by increasing disability and very costly care. Both medical and nonmedical interventions could well change this pattern by reducing the incidence of chronically disabling diseases and the disability that results from them. Medical advances have substantially reduced the disability associated with many diseases; cataract surgery is a