Who’d call dat livin’?

As a role model, Methuselah is not ideal. Apart from his 969-year lifespan, almost all we know about him is that his first child, a son, was born when he was 187, and that he subsequently ‘begat sons and daughters’. We don’t know whether those first 187 years included a protracted adolescence, or how he fared towards the end of his life. Ira Gershwin’s splendidly execrable rhyme: ‘Who’d call dat livin’/When no gall’l give in/To no man what’s nine hundred years?’, suggests only one of many unenviable scenarios. Longevity is desirable only if the prolonged life is a happy one.

The same proviso applies even more to immortality – for, like diamonds, immortality is for ever. If you find one of the mythical fountains of youth, and reckon you can face the embarrassment of meeting your family and colleagues afterwards, you’d do well to strip off and bathe. (You may wonder whether you need to strip, but the charming detail from Giacomo Jaquerio’s painting that adorns the jacket of Jay Olshansky and Bruce Carnes’s book suggests that this is the convention.) Immortality is not synonymous with eternal youth, however, and Olshansky and Carnes remind us of one of the less well-known stories in Gulliver’s Travels. It involves a tiny minority in the country of the Luggnaggians, who are born with red dots on their foreheads and are known as Struldbrugs. The Struldbrugs are immortal, but to his great surprise Gulliver discovers that this is not a blessing but a curse. After the age of thirty they grow melancholy and dejected; when they are ninety their memories begin to fail and their teeth and hair to disappear; and eventually they become unable to keep up with changes in the language and cease to communicate. The modern moral is obvious.

Olshansky and Carnes are demographers, and, in spite of their book’s title, they are more concerned with mortality than immortality. Since it is mortality that we’re more likely to have to face, this is fair enough. Demography, though, is not the most charismatic of subjects and ‘tables of mortality’ are hardly come-hitherish – even if you tabulate the chances of surviving rather than of dying, and call them ‘life tables’. All the more credit to the authors, then, for managing in this slim volume directed at a lay audience – it is innocent of both references and graphs – to make the subject fascinating, and life tables exciting.

If this sounds over the top turn to the appendix. A four-page table shows the expected ‘years of life remaining’, and the ‘probability of living to your next birthday’, for American men and women at each age from birth to 110, in the year 2000. Without looking at the table, ask yourself how old the average American man and woman would have to be for their chances of surviving the following year to have fallen to 99, 90 or 50 per cent. If you then look at the table you will almost certainly find that your estimates are wildly out and much too gloomy. To have no more than a 99 per cent chance of surviving another year, the average American man has to be 57 and the average woman 62. For 90 per cent, the figures are 83 and 88. Most remarkably, for the chance of surviving another year to drop below 50 per cent, the average man must reach 108 and the average woman 109. (If this seems unbelievable, recollect that even if your chance of surviving each year remained at 50 per cent, your chance of surviving four years would be only 1 in 16, that is 1 in 24.) The expectation of life at birth is 73.5 years for male babies, 79.6 for female. These figures are for the US, but those for Western Europe, Australia or Japan would not be very different.

How are such life tables produced? If you wanted to provide them for mice you would follow the fate of a cohort of mice from birth until they had all died. If you tried that method with people, however, you would be dead yourself before the task was completed, and when it was finished the results would be mainly of historical interest. Instead, you use census figures for a particular year, together with the records of deaths occurring during that year, to calculate the fraction of the population of each age who have died in the course of the year. From these figures you can calculate life expectancy at each age; though whether the predictions turn out to be true will depend on whether the risks at each age remain the same as in the year studied. Improvements in the standard of living or medical advances will tend to make the outcome better than predicted; epidemics, wars or natural disasters will make it worse. Even demographers can’t predict the unpredictable.

A major theme in the book is the way an individual’s risk of dying changes with age, and the way that pattern has changed in the past and may change in the future. The subject is topical because we have only recently completed a century that saw more dramatic changes in life expectancy than any since modern Homo sapiens first appeared, more than a thousand centuries ago. Even at the very end of the 19th century, 10-15 per cent of all babies born in the US died before their first birthday, deaths in childbirth were not uncommon, children – and not only children – were at risk from a host of infections, sanitation was often primitive, malnutrition was common and the expectation of life at birth was 45 years. Late 19th-century ‘survival curves’ (that is, graphs showing the fraction of the population surviving to each age) show a steep fall for the first few years, becoming much less steep between the ages of ten and 20, and then gradually but progressively steepening again during the rest of life. Modern survival curves in developed countries – with neonatal mortality very low, childhood infections largely controlled and childbirth safer – show scarcely any initial fall, though even in the US the first year of life remains slightly riskier than any subsequent year up to the late fifties. I remember being taught as a London medical student: ‘the first day of your life is the most dangerous – except, of course, for the last.’

What about the future? Olshansky and Carnes point out that, while preventing deaths that occur early in life obviously has a substantial effect on the expectation of life at birth, the prevention of deaths late in life can only have a much smaller effect, unless the lives of a substantial portion of the population are prolonged well beyond what is now thought of as the maximum span. This would require a successful attack on the biological processes that cause ageing – overcoming what Olshansky and Carnes confusingly refer to as ‘the inescapable reality of entropy in the life table’. (This is confusing because the inevitable increase in entropy – roughly translatable as disorder – predicted by the second law of thermodynamics applies only to closed systems; and living entities are not closed systems.) Such prolongation is not inconceivable, but there is no evidence that it is happening; it is therefore unlikely that changes in the expectation of life comparable to those seen in the 20th century will occur in the near future. In any event, any substantial progress must involve the later, increasingly steep, part of the survival curve; so why does it steepen, and why do we age?

This brings us to Benjamin Gompertz and his compound-interest-like ‘law of mortality’, published in 1825. Olshansky and Carnes give a good account of Gompertz’s ‘law’, and show that it fits the mortality curves of humans, and (with suitably adjusted time-scales) dogs and mice, even better if deaths from diseases not related to ageing are excluded. But they don’t say anything about the background to his discovery; so before clarifying the law let me fill that gap.

Gompertz was born in London, where his father and grandfather had been diamond merchants. Excluded from university because he was a Jew, he studied mathematics privately, becoming a frequent contributor to the Philosophical Transactions of the Royal Society and, in 1819, a fellow of the Society. In 1821 he applied for the job of actuary of the newly formed Guardian Insurance Office, but was rejected because of his religion. His brother-in-law, Sir Moses Montefiore, together with Nathan Rothschild, thereupon set up the Alliance Assurance Company with Gompertz as actuary. Finding an actuary for an insurance company must be common enough; but founding an insurance company for an actuary is probably unique. Anyway, the company flourished – and still flourishes as the Royal & Sun Alliance – and Gompertz did, too. He had, some time previously, prepared new tables of mortality for the Royal Society, and looking at them again he realised that, after puberty until about the age of 80, the risk of dying doubled every eight years or so. In other words, once the risks of birth and childhood were over, for a long period the risk followed a compound-interest law; in maths-speak, it increased exponentially. (The same law describes the chain reaction in the atomic bomb, or the growth of a rabbit population when there are no predators and food is unlimited.) Gompertz found similar behaviour when he looked at English, French and Swedish mortality data for different time periods, and later work has shown that similar patterns occur in some, but not all, other species.

Why the risk of dying should increase in this way is still a problem – a problem that, curiously, Olshansky and Carnes ignore. They do, however, consider the more fundamental question: why do we age? Until the middle of the 20th century, it was rather taken for granted that, with the exception of the reproductive or germ cells – the eggs and sperms that carry life forward to the next generation – all the cells in living organisms must ultimately die. But many plants can be propagated indefinitely without the use of germ cells, and some simple animals – the little Hydra that we learned about at school, and the more elaborate but related sea anemones – are thought to be able to survive indefinitely in favourable conditions. And though it is now clear that most normal body cells cannot go on living and dividing indefinitely, it is also clear that cancer cells can. There are, in many laboratories around the world, cultures of so-called HeLa cells, which are all the remote offspring of cells taken from a carcinoma in an unfortunate and now long-dead woman, from whose name HeLa is derived.* So why do we age?

In 1952, Peter Medawar pointed out that until the middle of the 18th century most of our ancestors would have died from infections or injuries long before they were old enough to experience the effects of ageing. Genes that produced harmful effects late in life would therefore not have been eliminated by natural selection, and with time, random mutations would cause them to accumulate. A few years later, George Williams pointed out that, because genes can have more than one effect, a gene that had beneficial effects early in life and harmful effects later on could actually be spread by natural selection. Genes predisposing to cancers or Alzheimer’s disease could have useful roles earlier in life.

Both Medawar’s and Williams’s theories may be right, but it is not essential to postulate the accumulation of harmful genes to explain ageing. In 1977, Tom Kirkwood proposed what he called the ‘disposable soma theory’. Recalling the distinction that arises very early in the development of each individual between the germ cells, which will convey their genes to the next generation, and the cells of the rest of the body, the soma, he pointed out that, though it is vital that the machinery of the germ cells be near perfect, the dependability of the machinery in the body cells is less critical. Making very dependable machinery is expensive in energy and materials, and since there is evidence that most of our early ancestors would have been dead before they were 40 – from infections or accidents or attacks by predators or starvation – natural selection is unlikely to have favoured mutations that used limited resources to make machinery which remained dependable at more than twice that age. Hence the failures of old age now.

If this theory is correct, we are at risk of multiple failures when we are old, and a major extension of life expectancy is not going to be easy to achieve. The changes in mortality that have already taken place will, though, continue to alter the age structure of the population of the developed world, as children alive now or yet to be born grow into long-living adults. Rather than having the shape of a pyramid, with many young children at the base and a few very old people at the apex, as it did until the Second World War, the population structure will, by the middle of this century, become much more rectangular. This has frightening implications for the provision of care for the elderly. Olshansky and Carnes point out that when the United States Social Security Program was created in 1935, its designers estimated that the number of people qualified to draw benefits would never exceed 25 million. By the year 2000, the Program was supporting more than 38 million. (Perhaps it’s significant that it was in 1935 that Trotsky described old age as ‘the most unexpected of all the things that happen to a man’.)

Many readers of this book will be less interested in the demographic aspects of ageing than in how to delay its effects. So Olshansky and Carnes tell us about the relevant effects of free radicals – highly reactive small molecules, peculiar in having unpaired electrons – in our tissues (bad) and free radical scavengers (good); sugar and fat-rich diets (bad), fruit and vegetables and lean meat (good); smoking (bad), regular exercise (good); megadoses of antioxidant vitamins, blue-green algae, growth hormones, melatonin and all products marketed as stopping or reversing the ageing process (ineffective), herbal medicines (let the buyer beware). They are scathing about the ‘hucksters’ in the ‘prolongevity industry’ who peddle ineffective remedies to an anxious population which is encouraged to believe that everyone could enjoy a lifespan of 120 years, and that growing old and infirm is a personal failure which could have been avoided. Sometime in the future, the elimination of harmful genes or the introduction of helpful ones may have a role. They tell us of an experiment to introduce an extra gene for superoxide dismutase – a doughty scavenger of free radicals – into fruit flies. The fruit flies lived longer. But they also tell us that there are data linking the overproduction of superoxide dismutase to the dreaded motor neurone disease in humans and a related disease in mice.

So what are we oldies to do? Olshansky and Carnes give their recipe at the end of the book. It includes

daily vigorous exercise . . . plenty of fruits, vegetables, fibre, and moderate amounts of low-fat protein; a restful sleep every night; an intellectually rewarding, non-stressful job, or no job at all; daily body massage; sex at least once a day; and a regular indulgence in your favourite vice: chocolate, barbecue ribs, you name it. The frequency of the indulgence can rise with advancing age at a rate of one or two per week for every decade lived – that is, one or two indulgences per day by age 70.

With an old age like that to look forward to, who’s going to worry about immortality?