Throughout the developed world, middle-aged and older people are living longer on average than ever before. Survival in later life has been increasing for 100 years but only recently has it been making a material contribution to the ageing of developed societies. Its earlier effects were swamped by the rising tide of people surviving from high birth rates and falling childhood mortality in the early 20th century - the so-called demographic transition.

There has been much debate over the implications of this increase in longevity. Pessimists have suggested that it is largely due to increased survival of older people with disability or chronic disease. Optimists, on the other hand, consider that increasing longevity represents more people reaching old age in a fitter and healthier state than did their predecessors. There is no reason why the two processes should not coexist, but obviously they differ in their implications for health and social services and national economies.

Epitomizing the optimistic view, Fries (1980)¹expounded a model of compression of morbidity. He suggested that longevity in the developed world was evolving through prevention of disabling illnesses, so lengthening healthy life but shortening the average period of disability before death. This process would outweigh any effect from prolongation of survival of people with chronic irremediable disease and disability. Fries further suggested that average longevity would come to be limited, at around the age of 85, by the maximum biological lifespan of the human species. In fact, there is no convincing evidence for a fixed limit to human longevity. Over a century of social change in the United Kingdom, the percentage increase in life expectancy of 100 year old women has been comparable with that of the young and middle-aged (Grimley Evans 1997)². Intrinsic, biological, ageing processes interact with extrinsic factors in environment and lifestyle; the rate at which the biological clock ticks depends on the environment.

Although at the time Fries was writing there was little consistent evidence that compression of morbidity was taking place, it now seems that, in the United States at least, it may indeed be under way. Manton and colleagues (2001)³ have reviewed the findings of successive samples of the National Long-Term Care Survey of the American population from 1982 to 1999. There are now two million fewer people aged over 65 with significant disability than would have been expected on the basis of the prevalence rates of 1982. Over the seventeen years there was an increase in life expectancy at all ages in the American population and a 30% increase in the numbers of people aged over 65, but the age-standardized prevalence of those with disability fell from 26.2% to 19.7%. Of particular significance, the proportion resident in nursing homes fell from 6.8% to 4.2% and this 38% fall represents an absolute decline in numbers of nursing home beds occupied by older people. The fall in disability prevalence accelerated from an annual rate of 0.26% 1982-9, through 0.38% in 1989-94, to 0.56% in 1994-9.

Is this too good to be true? Other American datasets give less clear-cut results (Crimmins et al. 1997)⁴ but the differences lie in the form rather than the direction of the trend over time. Is it happening in other countries? Probably in some, although Manton and Vaupel (1995)⁵ have shown that life expectancy at the age of 80 is higher in the USA than in England, France, Sweden, or Japan. Perhaps it would be better not to know; we do not want our politicians to relax into thinking that any possible problems of population ageing will simply go away. The moral to be drawn from the work of Manton and colleagues is that because what is happening in the USA is clearly due to extrinsic factors, it could be made to happen elsewhere. So the real question is how have the Americans done it?

Essentially we want to live longer but to die faster, and the table below lists ways in which this can be achieved:

Postponement as prevention	Lifestyle Knowledge, opportunities, incentives Public health and medical care
Disability-reducing interventions	Surgical Medical Rehabilitation
Less disabling environments	Wealth distribution Architecture and planning

A key approach is through postponement as prevention. Because of age-associated loss of adaptability the older we are when struck by a potentially disabling disease, such as stroke or coronary disease, the more likely we are to die rapidly rather than linger in a disabled state. Postponing disease is partly a matter of lifestyle - taking exercise, not smoking, eating a sensible diet and so forth. But healthy lifestyles are a societal as well as an individual issue. We have concentrated too much on health education as a means of persuading people to live healthier lives; health education increases knowledge but rarely changes behavior. We should think more about opportunities and incentives, such as making cities safer and pleasanter for pedestrians and cyclists rather than car drivers, or persuading supermarkets to take their profits from junk foods in order to subsidize fruits and vegetables.

A second element in postponement as prevention lies with the medical aspects of preventive medicine. Both primary and secondary prevention have less priority in our health systems than they merit, and older people are often excluded. This is perverse since we have good reason to believe that preventive benefits from drugs such as statins and ACE-inhibitors extend to a much wider range of contexts than emerged from their earlier use in high-risk conditions. Perhaps we cannot afford to hand them out unrestrictedly through health services, but have we really thought through the implications of a democratic commitment to the freedom of choice of an informed public? People are free to spend money on cigarettes if they so choose. Could they be given greater personal choice over spending their money on medically sanctioned over-the-counter purchase of preventive drugs such as statins?

A third element in the prevention of disability is the rational deployment of disability reducing interventions. Rehabilitation after acute illness in old age is one crucial element provided largely by departments of geriatrics. But there are many others. For older people the provision of disability-reducing interventions such as coronary artery surgery and angioplasty and joint replacements is much higher in the USA than in other countries. Many pharmacological treatments - including statins, ACE inhibitors and blood pressure treatment - are more effective in terms of deaths and disability prevented per thousand patients treated if given to older than to younger people. This is because treatment reduces the chance of death or morbidity by a constant percentage of background risk, and this rises with age. Because of ignorance or ageist prejudice, however, older people are less likely than younger to be offered such treatments.

But compression of morbidity is not merely a medical issue. Disability may be the result of an unsuitable environment as much as a person's physical or mental impairment. Quality of housing and household equipment and appliances are matters to a large extent related to personal wealth. Inequitable distribution of national wealth damages health as well as social well being. Architects and planners are not sufficiently aware of the design implications of the needs of increasing numbers of people with minor age-associated impairments. We should not be producing houses, public buildings or urban spaces that older people find challenging. There are a few research teams working on non-disabling environments but the design requirements they specify are not written routinely into briefs for architects and planners as they should be.

Compression of morbidity, bringing personal and economic benefits, is attainable. But the public need to demand it of their politicians as well as their doctors.