A dramatic rise in life expectancy can hardly be described as the worst challenge that a society can ever have faced. Ask many parts of Africa, where dramatic falls in life expectancy are now common. Yet the rises reported in the journal Science this week pose jaw-dropping questions. Even existing projections were already placing unprecedented demands on health, social service and social security plans. Yesterday's reports that these projections were far too modest will redefine everything again.

Consider current projections first. These already suggest that fewer births and longer life mean the number of people aged 65 and over will increase at 10 times the overall rate of population growth in the next 40 years. Where there were only 270 centenarians in 1951, there are now 6,000. Projections suggest there could be 45,000 by 2030. Even current projections have big implications. Health and social support costs rise asymmetrically with age. People in their first decade of retirement cost the NHS twice as much as people of working age but, by 85, they are costing four times as much. Social service costs rise even more alarmingly. Pension managers, meeting this week, were forecasting a gloomy future even before the new projections.

Now along come two scientists writing in Science, who suggest that current projections are far too low. The current estimate is that male life expectancy will rise in Britain from 75 to a mere 79 by 2025, and from 80 to 83 for females. The scientists make no estimates for Britain, but their methodology suggests that female life expectancy here could rise to 100 by 2085. One of the authors suggests that female babies in the two countries with the longest life expectancy - Japan and France - already have a 50:50 chance of reaching 100.

These are not lone voices. In last year's Reith lectures, Professor Tom Kirkwood, a gerontologist, argued that ageing was neither inevitable nor necessary. If the last century's achievement - extending life expectancy by 25 years - was extended into this one, then life expectancy would reach 100 within 100 years. Hopes remain high that new medical interventions will slash heart disease and cancer death rates, the two biggest killers.

Yet there will be many sceptics, and for good reasons. The last century's achievement owed as much to social interventions - clean water, good sanitation, better housing - as medical interventions. The most important medical interventions were immunisation programmes. Together, they ended the scourges of TB, typhoid, and cholera as well as the diseases which attacked children: diptheria, measles, dysentery, whooping cough and scarlet fever. The main reason why life expectancy rose so dramatically was the elimination of high death rates among young people. Life expectancy for 65-year-olds rose only a modest amount: from another 11 years in 1880 to 16 years in the 1990s. If this century matches the last, life expectancy for 65-year-olds would only rise to 88.

Population projections are an art not a science. Current projections provide more than enough evidence that policy planners are not paying enough attention to our ageing population as it is. This country needs therefore to grasp three thorny pension challenges: raising retirement age; compulsory company contributions; and mandatory individual contributions. Health and social services are better prepared than pensions. They are already familiar with the challenge. And at least the health service starts with the government's record recent rises in investment.