Why We Get Sick

by Randolph M. Nesse

  THE MOST WORRISOME QUESTION

  Another puzzling aspect of allergy, at least respiratory allergy, is the apparent recency of its appearance as a major medical problem. John Bostock originally described his own symptoms of hay fever for the Royal Society in 1819 and later reported that he could find only twenty-eight cases after investigating five thousand patients in all of England. Records imply that hay fever was essentially unknown before 1830 in Britain and 1850 in North America. In Japan its incidence was negligible in 1950, but it now affects about a tenth of the population. If the increase is real and not just an artifact of inadequate records, what novel environmental factor of the last century or two can account for this alarming phenomenon?

  One clue comes from studies of the factors that seem to sensitize predisposed individuals, mainly exposure to antigens in the first two years of life. In one study of 120 infants with high susceptibility to allergy on the basis of their IgE levels at birth, 62 were raised as a control group without any intervention, while the mothers of 58 in the experimental group were taught how to keep their homes relatively clean of allergens, prevent mites, and avoid giving potentially allergenic foods to their infants. At age ten months, 40 percent of the control group had developed allergies compared to only 13 percent of the experimental group. Perhaps part of the increasing rate of allergy results from living indoors with drapes and wall-to-wall carpets, which provide breeding places for dust mites.

  When Eric Ottesen, head of the Clinical Parasitology Section at the National Institute of Allergy and Infectious Disease, studied the six hundred people who live on Mauke, an atoll in the South Pacific in 1973, only 3 percent of them had allergies. By 1992, the rate was up to 15 percent. He suggests that institution of treatment for worm infestations during the intervening years left the IgE system with no natural target, so that the usual mechanisms that downregulate the system are inactive and the IgE begins to attack harmless antigens.

  Breast-feeding decreases the incidence of allergies, so bottle-feeding may also contribute to the rise in allergies. Perhaps babies deprived of maternal antibodies make more immunological mistakes in coping with antigens on their own. Or perhaps crowded, mobile modern societies expose infants to a greater diversity of viral respiratory diseases and thereby greater exposure to miscellaneous allergens. The increased quantity and variety of atmospheric pollutants may foster increases in both helpful allergies (if such there be) and harmful ones, perhaps because chemical damage to the respiratory mucosa may admit antigens that would otherwise be kept out. Food allergies, although perhaps not as clearly on the increase, may have become more troublesome because we now have so little control over what we are really eating. Eggs, wheat, soybeans, and other possible allergens may be present in a great variety of commercially prepared foods and be extremely difficult to avoid, even by people who know they are allergic to them.

  What are we doing today that is different from what we did just a century ago and that makes us so much more vulnerable to so many diverse allergies? We desperately need real answers. Respiratory allergies affected less than 1 percent of people in industrial societies in 1840. Now, a hundred and fifty years later, it afflicts 10 percent. What might the future hold if we remain as ignorant as we are now?

  12

  CANCER

  On March 5, 1992, The New York Times carried an obituary for well-known actress Sandy Dennis, a cancer victim at fifty-four. That same day, the eighty-three-year-old actress Katharine Hepburn was enjoying her autobiography’s twenty-fifth week on the Times’s best-seller list. An obvious question is, Why did cancer strike Sandy Dennis? What caused her to miss out on the long life that her fellow actress enjoyed?

  This obvious question is morally and medically a good one, but there is a more profound biological question: How is it possible that any of us can live several decades without dying of cancer? Cancerous cells are merely cells doing their normal thing: growing and proliferating. How could so many cells do such an abnormal thing as inhibit their growth for many decades? Obviously they must; otherwise everyone would die of cancer at an early age. This, of course, is the ultimate explanation. Those least likely to die at an early age, from any cause, will be most likely to survive, reproduce, and have their cancer-delaying adaptations at work in future generations. This sort of evolutionary explanation can help us understand the workings and origins of our cancer-preventing adaptations and the prodigious accomplishment they represent.

  Confucius once said something like: A common man marvels at uncommon things; a wise man marvels at the commonplace. To marvel at the commonplace of not having cancer and at the mechanisms that make this possible may be the key to understanding how to make cancer even more uncommon.

  THE PROBLEM

  The magnitude of the problem of avoiding cancer may be appreciated from considering the long-term history of any cell in our bodies. A cell now contributing to normal functioning in the liver of some Hollywood star arose by the growth and division of some preexisting cell, probably one closely similar to itself. That parental cell arose from another before that, and so on. As we trace the ancestry of the liver cell, we find cells that look ever less like liver cells and ever more like undifferentiated embryonic cells. Some years back in the cell lineage we come to the fertilized egg from which the entire individual arose.

  That cell had a history too, a lineage through various oocytes and oogonia back to the embryonic cells that developed into the Hollywood star’s mother. Likewise, the sperm that did the fertilizing came from a lineage of spermatocytes and spermatogonia back into the embryonic cells of our star’s father. Thus back through the mother’s and father’s original zygotes into the grandparental generation, and so on in endless repetition of ever-dividing embryonic and reproductive cells. Never in these sequences of cell divisions, for the billion years or so since the origin of the first real cells, was there ever one that did not divide, and nowhere in these lineages was there anything that looked like a liver cell.

  We offer Figure 12-1 as an aid in understanding this essential fact of life. All our ancestors had livers, but none of the cells of these ancestral livers gave rise to any of our liver cells, or to anything else in our bodies. We arose entirely from a line of endlessly proliferating germ-line cells. This picture, of an eternal germ plasm giving rise to elaborate somata of individuals, which are always genealogical dead ends, was first presented by August Weismann, a nineteenth-century Darwinian.

  Now, for the first time in these eternal lines of descent and after dozens of the cell divisions needed to create an adult soma from a single cell, we find a cell, say a liver cell, that must play a specialized role in the life of a multicellular individual. This liver cell must do something none of its ancestors ever attempted: it must stop dividing. If there is an injury to the liver, the cell may be called upon to divide again. This sort of growth and division must be in precisely the amount and pattern required for normal liver function and must cease as soon as this machinery is fully restored. If ever, in any one of the billions of cells of the liver, the growth and division process is turned on inadvertently and proceeds unchecked, a tumor develops and eventually causes a lethal disruption of some physiological function.

  FIGURE 12-1. Germ plasm concept of Weismann. The eternal line of germ cells gives rise to individual bodies with a limited life span.

  The individuals diagrammed can be of either sex.

  From this perspective, life seems rather precarious. It suggests that we must have some really superb anticancer mechanisms acting in our favor. As American marine biologist George Liles observed, “the cells and organs that make life possible had better be well designed, because the job of living is formidable. Living beings—plants and animals, bacteria and slime molds and fungi—every animate entity faces a set of challenges that would give pause to the most inventive designer.” He was led to this remark in considering what might seem a rather simple sort of problem, the proper routing of water through the feeding machinery of a mus
sel. How much more formidable is the challenge of avoiding cancer for several decades in the collection of ten trillion cells that make up a human being!

  Biologists today more or less universally believe that multicellular organisms, such as ourselves, arose from some group of the protozoa, in which each cell was a functionally independent individual. Most of their reproduction was asexual, with one cell dividing to form two new ones. In some modern protozoan species these two new individuals do not break completely apart but stick together in pairs. In others, the offspring of pairs stick together in filaments or sheets called colonies. In a few, the colonies may differentiate into germ cells and somatic cells, as shown in Figure 12-1. This means that some previously independent cells, apparently voluntarily, give up reproduction and become genealogical dead ends. They devote themselves entirely to supplying nutrients and protection to the few germ cells that ultimately participate in sexual reproduction. Some such sequence of developmental events, as observed in the much-studied colonial protozoan Volvox carteri, must have characterized some remote ancestor of all multicellular animals.

  Can this acceptance of a sterile, servile role be explained by natural selection? The answer is obviously no, if this process means selection among cells for those best able to survive and reproduce. The answer is yes if the selection is among the genes best able to get themselves into future generations. If the reproductive and somatic cells of a Volvox colony have the same genes, it does not matter which cells actually do the reproducing and which become sterile. All that matters is that the sterile cells, in their strictly somatic roles, make the colony’s reproduction of genes identical to their own more successful than if they too formed eggs or sperm. If colonies with ten reproductive cells and a hundred sterile ones reproduce more successfully than those with eleven and ninety-nine, the tendency for most of the colony cells to assume a somatic service role will be perpetuated.

  A colony of a hundred cells, all derived in a short time from a single original cell, may well be all of about equal health and vigor and will almost certainly be of the same genotype. The resources needed to produce a hundred cells from one may all be shared equally, and all cells have elaborate mechanisms for protecting the genetic material from damage or alteration. But what about a thousand or ten thousand cells? Would colonies that big be asking for trouble? Might there not be occasional mutations that would make cells behave in ways other than those that maximally benefit the colony as a whole? For instance, might not such a mutant cell start appropriating more than its maintenance requirements for nutrients and start growing and reproducing, even though this might be harmful for the colony? Such large colonies surely need special adaptations for maintaining discipline among the many component cells.

  THE SOLUTION

  How about a colony the size of an adult human body? What sort of special adaptation would be adequate to maintain discipline among ten trillion cells? From an engineering perspective, it is difficult to imagine how any quality control system would be equal to the job. An auto manufacturer faced with turning out a mere ten thousand vehicles, not one of them with a serious flaw, would be well advised to quit the business. A single living cell is incomparably more complicated than any automobile.

  Consider the problem faced by an embryo of a hundred cells that gives rise to one of a thousand that produces one of ten thousand and so on to the ten-trillion-cell adult. Most of these cells will die and be replaced by others. All these cells are equipped with genes that turn out products essential to their division, and some genes are adjusted so as to stop making this product when local conditions indicate that the tissue is mature and no additional cells are currently needed. If one of these genes gets accidentally altered in a way that makes it heedless of these conditions and the gene goes on making its product, mechanisms of DNA editing and repair step in and correct the flaw—or at least they are supposed to. One out of about two hundred people has a gene that greatly increases the likelihood of colon cancer. Originally thought to be a gene that actually did something to cause cancer, it is now recognized as a defective form of a normal gene that acts in the detection and rectification of abnormal DNA structure. When this system is not working, DNA abnormalities accumulate and the chance of cancer increases drastically.

  Very few such flaws actually get a chance to express themselves. How few? Let’s assume that only one such gene in ten thousand cells makes its product when it is not supposed to. Starting with ten trillion cells, we can assume there are a billion altered cells, scattered through the body, that are capable of initiating a cancerous growth. This is not all that reassuring. But there is another kind of genetic safeguard in each cell: tumor-suppressor genes that actively inhibit cell growth, perhaps by destroying the product of a gene that makes a substance essential to division, when it is inappropriate. Let’s assume that this safeguard is also fantastically effective and that the daily rate of failure is only one in ten thousand cells. We can now assume we have only a hundred thousand cancers beginning in the body each day. If there were three equally reliable safeguards and abnormal cell division could not begin unless all three failed, there would still be ten new cancer cells formed each day. This is still not very reassuring.

  The situation is analogous to the problem of command and control of nuclear missiles. The risk of catastrophe from accidental firing is so great that the system is designed first and foremost to prevent accidental firing, even at the risk of sometimes not being able to fire when needed. This is the exact opposite of the smoke-detector principle we described for defensive responses. Control of cell division could be said to be based on the principle of “multiple safety catches.” The crew in the missile silo cannot fire the missile without a secret code. Even with the code, multiple procedures must be followed in sequence, including two people turning keys simultaneously in two different parts of the room. The system is designed so that any irregularity makes it impossible to fire the missile at all. Similarly, the body’s cells have multiple safety-catch mechanisms. If failure of these mechanisms is detected, other mechanisms stop cell growth. When, despite all previous safeguards, cells grow at an inappropriate rate, still other mechanisms cause the aberrant cells to self-destruct.

  A recently discovered gene called p53 is the best example. It makes a protein that protects against cancer by regulating the expression of other genes. In certain circumstances it can shut down cell growth or even make the cell self-destruct. If a person inherits one abnormal copy of the gene that makes this protein, anything that happens to the other copy can lead to catastrophe. The p53 gene is abnormal in fifty-one types of human tumors, including 70 percent of colon cancers, 50 percent of lung cancers, and 40 percent of breast cancers. As John Tooby and Leda Cosmides have pointed out, however, such genetic abnormalities are not necessarily present in the tumor. Cells are often studied after they have lived for years in tissue culture, an environment that may select for genetic abnormalities that increase the rate of cell division.

  In addition to these various anticancer mechanisms operating in cells, there are those that operate between them. They detect misbehavior in their neighbors and secrete substances that inhibit the misbehavior. Finally, there is the immune system, which may bring a host of weapons into play against an incipient maladaptive growth as soon as it finds a difference between it and normal tissue. A detectable cancer must somehow have achieved the highly improbable feat of getting past these many layers of defense. Unlike a parasitic worm or infectious bacterium, it cannot draw on a long history of accumulating its own defenses against the host’s defenses. It is entirely the product of chance alterations in the cellular regulatory machinery. What cancer has on its side is mainly the astronomical number of chances it gets to achieve success against the immense odds.

  CANCER PREVENTION AND TREATMENT

  To avoid contracting cancer, the first thing you want to do is to pick your parents wisely. Susceptibility to cancer, like so many other diseases, is hereditary. This is true both in
general and for particular forms of cancer, most notably for some rare childhood cancers and those of the breast and colon. Members of families in which such cancers have occurred frequently may have twenty to thirty times the likelihood of contracting them as those in cancer-free families. Even when controlled for family members’ tendency to experience similar environmental conditions, the evidence for predisposition for certain kinds of cancer is strong. Mice can be bred to form cancer-prone stocks in which one cancer-control mechanism is already missing in every mouse. This enormously increases the likelihood of one or more kinds of cancer. Some human cancers are inherited in the same way.

  Another good way to reduce the likelihood of cancer is to live dangerously: die young, and you are unlikely to get cancer. The fact of senescence means that the environment of any cell and its regulatory capabilities are deteriorating. Hormonal and local regulation of cell growth and proliferation, like all other aspects of adaptive performance, becomes less effective as we go through that terminal life-history stage known as adulthood. The cell itself ages, and as the cardiovascular and digestive and excretory systems deteriorate, it will be ever less well supplied with nutrients and other essentials and ever less effectively unburdened of its waste products. An inevitable consequence is that its potential for growth and cell division is ever less well regulated. Maladaptive growths become steadily more likely to occur and spread unchecked.