If Reproduction Is What It’s All About, Why Does It Stop?
An evolutionary perspective on women’s biology across the life span ultimately runs up against menopause, a phenomenon that does not initially make sense if the currency of evolution is reproductive success. If increasing reproductive success is what evolution is about, then how could natural selection favour terminating reproductive functioning relatively early in a woman’s lifetime? In the historic and prehistoric past, women may not have lived very many years beyond menopause, but today this event occurs when women have as much as one third of their active and healthy lives ahead of them. Why have we not seen an increase in the age of menopause just as we have seen a decrease in the age of menarche as health conditions have improved in the last 200 years? In fact, there are two questions to address when considering menopause from an evolutionary perspective: why does ovulation cease at all, and why do women live so long after they stop reproducing? I will address the first in this article and save postmenopausal longevity for the next.
Is Menopause Unique to Humans?
In the medical literature, menopause is the end of menstruation, defined retrospectively after a full year has passed without a menstrual cycle. However, this definition is based on the assumption that the ‘‘normal’’ state for a woman is menstrual cycling. As we have seen, for most of human history women were pregnant or breastfeeding children and were rarely menstruating, so paying attention to how long it had been since the last menstrual period may not have been meaningful for explaining any biological processes. Because most other mammals do not menstruate it is meaningless to ask if they have a menopause. Thus, it is safe to say that menopause, when the definition depends on regular menstrual cycling, is unique to humans. As has been pointed out, however, the only meaningful definition of reproductive termination that works across species is the cessation of ovulation and this is the more common meaning found in the study of primatology. To understand how other species experience the end of reproductive cycling, it is perhaps more useful to ask if there are other examples of females who regularly cease ovulating before death. Virtually all female mammals that live long enough show signs that look like the approach to menopause for humans: increasing length of ovarian cycles and percentage that are an ovulatory; increased inter birth intervals; decreased sexual activity; and biochemical measures like hormonal increases and decreases, egg depletion, and decreased bone mass. It is argued that by these measures, and looking at menopause as a process rather than an event, chimpanzees and many other primates serve as good models for human reproductive decline. Another line of inquiry is to examine whether other species have long post reproductive lives. Although some degree of post reproductive life has been reported for many mammals (including some species of whales and the Asian elephant), no other primates live such a high percentage of their lives after reproductive termination. Female chimpanzees and monkeys experience decreased fertility in their later years, but most continue cycling until their deaths. Furthermore, in cases when they do cease ovulating before death, they are usually in very poor overall health, indicative of being much further along in the aging cycle than human females usually are when they cease ovulating. Primatologist Toshisada Nishida reports that at least five female chimpanzees at Mahale survived an average of 10 years after they last gave birth, but despite occasional reports of extended life beyond reproduction in apes and monkeys, it is far from the routine and expected phenomenon that it is in humans. One estimate is that fewer than 5% of chimpanzees and baboons in the wild live past the age of reproduction; even in human foraging populations, as many as one third of adult females live beyond age 50. Lee estimated that 10% of the !Kung that he worked with were over 60 and most of them continued to make significant economic contributions to their bands.
Why Cease Reproducing at about Age 50?
Perhaps we cannot defend menopause as a unique feature of humans, but that still leaves open the question of why it exists at all in any species, given the value placed on increased reproductive success. One theory about menopause is that it is just the result of the extension of the human life span, a by-product that has been ‘‘uncovered’’ in the last several hundred years as longevity has increased. This view claims that the maximum life span of the mammalian egg is about 50 years (remember, all the ova are already present before birth) and that no mammals can reproduce after the eggs have been depleted, no matter how long they live. By this reasoning, although the human life span has increased over the past several hundred years, the reproductive life span has not because the eggs simply cannot live much longer and the ovary just runs out of eggs. In the language of evolution, cessation of reproduction at about age 50 is an ancestral characteristic that we share with most other long-lived mammals, but a long post reproductive life (for women) is the derived or recently evolved characteristic that appeared after humans and apes diverged from a common ancestor. With regard to menopause, we are like apes. With our long lives, we are an unusual primate. Another proposal to explain menopause is that so much energy is needed ‘‘up front,’’ for reproduction in the early years, that there is nothing left over by the time a woman reaches 50. This is the concept of pleiotropy, which argues that genes that increase survival and reproductive success at early ages will be favourably selected even if they have negative effects on survival at later ages. Anthropologist Jocelyn Peccei proposes that each new infant is increasingly costly to a mother, especially for humans who produce large-brained and extremely dependent infants who need direct maternal care via lactation for three to four years. Even a few infants will deplete a woman’s energetic reserves to the point that her reproductive success may be greater if she invests her increasingly limited resources in her most recent offspring rather than giving birth again. Because it takes about 12 to 15 years for a child to become independent, it is argued that females who lived that long had greater reproductive success than those who died before their children entered their teens.
Early theories for menopause proposed that it was the object of selection itself because women who ceased reproducing early would have greater reproductive success through providing higher quality care for their current young than they would through having more, potentially lower quality young, who are less likely to survive. This has been difficult, however, to explain mathematically, based on the assumption that producing one or more offspring that share 50% of your genes would always be better than helping other young, even if they had as much as 25% of your genes. Furthermore, anthropological evidence is inconsistent with the idea that older women provision juveniles at levels that would increase fitness more than having their own offspring. Finally, if ceasing to reproduce early is advantageous if it leaves you with 15 to 20 years of healthy life left, it suggests that the age of menopause would have gradually increased in the past 100 years as living conditions and health have also increased. Despite incontrovertible evidence that these improved conditions have driven the age of onset of reproduction downward in almost all populations, there is no evidence that the age of termination of reproduction has been increased in the same way. In order for natural selection to act, there must be variability in a trait. Certainly there is variation in the age of menopause, but the average age is somewhat stable across human populations and does not show evidence of trending upward, even in health-rich populations. This suggests that cessation of reproduction is something that we, as mammals, are ‘‘stuck with,’’ and that it is part of our heritage that has limited ability to respond to natural selection afforded by improved ecological circumstances.
Childbirth is challenging for bipedal humans who give birth to large-brained infants, and the risks are not negligible. It is estimated that more than half a million women die of complications related to childbirth every year, and the risks increase at older ages with women over 40 having five times the risk of dying in childbirth as women 20 to 24. By another estimate, at ages greater than 40, the risk exceeds 30 times that at ages 20 to 29. Given these risks, it makes sense that pregnancy and birth would become increasingly rare beyond age 45, to the point that the chances of dying in childbirth exceed the chances of giving birth to a healthy infant, say at about age 50. Furthermore, a 50-year-old woman several hundred years ago would have probably been in poorer health than her equivalent today (at least in health-rich populations), suggesting that the risks may have been even greater. In particular, the risks of postpartum haemorrhage and blood clots increase with age, and placental and uterine problems increase with number of pregnancies. If a 50-year-old woman gives birth, she and her infant may die. If she dies, chances are that her other young children will also face compromised health and may even die or fail to reproduce. Thus, one could argue for cessation of ovulation early enough in life to ensure that you had several good years left for raising your young. Couple this with an increased ability to care for your grandchildren once freed from reproducing yourself, and the selective value of ceasing to reproduce may outweigh the negatives at the population or species level. There is far from agreement on this proposition. As I have repeatedly emphasized throughout this aximum fertility (quantity) is not necessarily the best way of achieving reproductive success. Humans, with such huge demands placed on them for raising offspring, seem to have adopted a strategy that puts quality before quantity, meaning that the optimum number of offspring is far below what can theoretically be achieved. This strategy appears to be particularly successful in situations where resources are constrained and there is a lot of competition for them, as probably characterized most of human history and most populations today. The same argument can be applied to reproducing beyond age 50. As already noted, energy limitations make it selectively advantageous for women to stop cycling when conditions for pregnancy are not good. These include athletic amenorrhea, illness, psychosocial and emotional stress, starvation, and, perhaps, being older than 50 years.
Factors Affecting the Onset and Experience of Menopause
Menopause is directly determined by the number of eggs a woman is born with, but several factors seem to influence its timing, perhaps by influencing the rate at which the eggs decrease through time. Geographic location, income, education, and marital status have been related to age of menopause. For example, single women experience earlier menopause than married women. This effect may be related to sexual behaviour, social environment, or other psychosocial aspects of the marital bond such as social support or stress reduction. One group of researchers suggests that the mechanism may be pheromonal—the presence of a man and male pheromones may influence ovarian function and oestrogen levels in ways that delay menopause. Factors that seem to be related to earlier age at menopause include childhood under nutrition, low socioeconomic status, low education levels, rural living, never having given birth, short menstrual cycles, and lifestyle factors such as smoking and alcohol consumption. All of these factors are interrelated, of course, so it is not possible to say which has the strongest effect or if any have independent effects. Age of menarche and use of hormonal forms of birth control do not seem to impact age of menopause. How parity influences age at menopause is unclear, with some studies showing an effect and others showing none. High dietary intake of fat, cholesterol, and coffee were associated with later menopause in a Japanese population. A number of lifestyle and environmental factors seem to have effects on symptom reporting at menopause and include body weight, body constitution, and exercise. A number of studies have found a positive association between BMI and risk for moderate to severe hot flashes compared with low BMI. Aerobic exercise can reduce both psychological and physiological responses to psychological stress. There is evidence that a woman’s experiences with and attitudes toward sexuality may affect her experience of menopause. There is also a reported association among increased irregularity of menstrual cycles, hot flashes, declining estrogens, and declining frequency of sexual intercourse. There is a report of associations between tobacco usage and both severity and frequency of hot flashes. Current smokers were more likely to experience hot flashes than past or never smokers in another study.
In some cultures it is believed that menopause is an emotionally and psychologically challenging time for women, and several studies have examined the relationship of mood and other psychological variables to the menopause transition. Depression, for example, has often been reported by women during this time, but most studies suggest that hormonal changes are not necessarily responsible. In one longitudinal study low mood and physical symptoms were shown to co-occur in midlife but have different causes. In two other studies, earlier life experiences were better predictors of distress at midlife than menopausal status and women approaching menopause were more susceptible to developing depression than were younger women. Socioeconomic status was associated with symptom reporting in a large multi-city, multi-ethnic study of 16,065 women in the United States. Those who reported difficulty paying for basic needs also reported more symptoms. Finally, there are many differences in the experiences of women during the menopause transition across cultures, even in the ‘‘physical symptoms.’’ Hot flashes, for example, are reported by many women in North America, Thailand, Norway, India, Nigeria, and Tanzania but are rarely reported for Japanese women, Navajo women, Mayan peasant women, and Sikh women living in Canada. This is further evidence that the relationship between biological aspects of the menopause and psychosocial and physical aspects is mediated by the lived experience of each woman, not the least of which is her cultural milieu.
Endocrinology of Menopause and the Late Reproductive Years
The reproductive life spans of women and men are quite different: in women, reproduction stops somewhat predictably at about age 45–50. Although the ability in men to reproduce declines with aging, it does not completely cease, at least at a predictable age. Evolutionary biologist George Williams suggested that this difference is due to the relative costs of reproduction to the sexes, with the costs to the female generally being much greater than the costs to the male. Moreover, the risks for an older woman from pregnancy and birth are greater than the risks to males producing sperm, and as we have seen, they may be greater than the benefits. Reproductive hormones like oestrogen and progesterone production begin to decline toward the end of the reproductive years until ovulation (and thus menstruation) ceases altogether. Recall that after ovulation, the corpus luteum secretes oestrogen and progesterone, which inhibit FSH and LH. At the end of the reproductive years, there is no egg, and so there is no corpus luteum, nor high levels of oestrogen and progesterone; and LH and FSH are not inhibited. This means that FSH and LH gradually rise and remain high in the postmenopausal years and it is these hormones that can be used to signal that the transition from reproductive to post reproductive state is underway. Highly frequent menstrual cycling is implicated in various reproductive cancers, so it seems that the cessation of cycling with menopause may be protective because it puts an end to high exposure of breast tissue to estrogens and high cell turnover rates over the course of 30 years of menstrual cycling with few interruptions. If 30 years of cycling put a woman at risk of breast cancer, imagine what 40 or 50 years would do. It is also important to consider the energy costs of continuing to cycle late in life. As noted earlier, each successive pregnancy uses a greater proportion of maternal energy reserves, meaning that the sixth or seventh pregnancy is relatively more expensive than the first or second.
Is Menopause a Medical Concern?
It appears that menopause is simply the end of reproduction for women, a phenomenon that potentially occurs for all mammals and is not necessarily something that needs to be ‘‘explained.’’ But why does menopause have to be a ‘‘bad’’ thing? This question derives from a fairly Western perspective that has pathologized the period of time when a woman is experiencing the physical and emotional transitions to a post reproductive state. In fact, the idea that menopause is an illness that needs to be treated has been part of Western understanding for more than 200 years. It is often likened to diabetes as a disorder characterized by hormonal deficiency. In fact, menopause has been called an ‘‘oestrogen deficiency disease.’’ Furthermore, it is often listed as a risk factor for a variety of diseases and disorders of aging, reinforcing the idea that it is a medical ‘‘problem.’’ Much of the medicalization-of-menopause phenomenon is based on the idea that women did not live longer than about age 50 in the past and that, therefore, prolonged life is ‘‘abnormal’’ and requires medical intervention to maintain health. This was part of the rationale for having women take hormone replacement therapy to enable their reproductive hormones to last as long as their lives. According to this view of menopause as an ‘‘oestrogen-deficiency disorder,’’ external sources of oestrogen needed to be added back into the system to keep mental and physical health at levels seen during the reproductive years. If you provide insulin to diabetics so they can maintain healthy lives, why wouldn’t you provide oestrogen to postmenopausal women? An obvious answer is that diabetes is not a universal phenomenon of the human life course, whereas cessation of ovulation is. This life course approach to menopause leads one away from seeing it as a medical problem resulting from a ‘‘deficiency.’’But recasting menopause as a developmental rather than a degenerative phase does not help women who experience stressful symptoms they perceive to be associated with menopause.
Horrible hot flashes that keep you awake at night and interfere with your daily activities are not going to be relieved simply by positive thinking or mantras that ‘‘this is normal, this is normal.’’ Like everything else that is evaluated from an evolutionary perspective, a normal, healthful physiological response (such as elevated body temperature and withdrawal of iron when exposed to disease) can become abnormal and even pathological when it is excessive. But a modification in the way of thinking about the problems many women experience at menopause may help to separate aging from normally occurring hormonal changes. Medical interventions may be necessary and desirable to treat osteoporosis, hot flashes, and cardiovascular disease, but they are treating these specific disorders, not menopause per se. In fact, useful and appropriate treatments are much more likely to be developed when the specific disorders are considered rather than focusing on the life phase transition that all women experience if they live long enough. In this way, the problems will be treated, not the life phase. This will avoid the problems that occurred when all postmenopausal women were advised to adopt HRT. Often terms used to describe menopause and the menopause transition reinforce the idea that it is a medical issue, including using the term symptoms to describe phenomena such as irregular menstrual periods and vasomotor changes. Anthropologist Emily Martin and others have pointed out that the metaphors we used to describe bodily functions are reflective of societal and medical views of those functions and, in turn, shape the way they are experienced on the individual level. Terms like symptoms, withdrawal, decline, and cessation usually have negative connotations; Martin suggests that it would be helpful if we could develop a new vocabulary to describe menopause.
Health psychologist Paula Derry notes that menopause and senescence are often conflated so that any disease or disorder that happens to women several decades after menopause is referred to as being ‘‘postmenopausal.’’ For this reason, osteoporosis is often associated with menopause, even if it occurs two to three decades later. It would be equivalent to referring to health events of the 30s and 40s as ‘‘postmenarcheal’’ or postpubescent. Clearly this is true in one sense of the term, but it is not a useful designation for the same reasons that referring to something that happens to a woman in her 80s as ‘‘postmenopausal’’ would not be very meaningful. Occasionally one will run across an article about ‘‘male menopause.’’ The very concept of ‘‘male menopause’’ used to describe the psychological challenges some men face at midlife suggests that female menopause is all about psychological issues. Clearly men do not experience the hormonal and reproductive changes that occur in women at this time, nor do they cease menstruating, so the concept makes sense only if menopause is seen as a psychological problem. For men, the decline in reproductive function that occurs late in life is highly variable and cannot be measured by a simple concept as one year from a given event (the last menstrual period in women).
Even the medical view of menopause is highly variable across cultures. For example, in Japan, ‘‘menopausal syndrome’’ is recognized clinically, but it is seen not as a result of hormone decline, but as a result of moral decline that occurs in women who ‘‘have too much time on their hands’’ and focus on themselves rather than on their families and communities. In a sense, physicians are blaming women for the ambivalence they feel when their children, in whom they are expected to invest so heavily, grow up and leave home, a time that usually coincides with menopause. The idea that a woman’s identity is closely tied to her role as mother is not unique to Japanese culture, of course, and is commonly reported as a reason that menopause is stressful in Western societies. Timing of menopause is subject to being defined as normal or abnormal. If menopause occurs before age 40, it is referred to as ‘‘premature ovarian failure’’ or POF. Obviously this somewhat arbitrary cut-off age would be meaningless in a non contracepting population where women may become pregnant in their late 30s and never resume cycling after several years of lactation. Whether perceived as a medical ‘‘problem’’ or not, the menopause transition is associated with a number of physiological changes that are experienced by women at this time and are usually explained by the changes taking place in ovarian function. Some of the most commonly reported sensations in the US include hot flashes, sweating, insomnia, vaginal and urinary discomfort, headaches, fatigue, irritability, depression, moodiness, and weight gain. Often these are divided into "physical symptoms" (hot flashes/sweats, and vaginal atrophy) and "psychological symptoms" (all others), and the frequency with which menopausal women report them is highly variable. Oestrogen replacement therapy (ERT) has been found to alleviate hot flashes and vaginal discomfort, whereas there is no evidence of an effect on other symptoms. One conclusion from this is that hot flashes and vaginal symptoms are related to the hormonal changes that accompany menopause, whereas the other symptoms co-occur with psychosocial changes many women experience during this time. It has been estimated that more than half of all menopausal women in the United States experience hot flashes at some time during the perimenopause, although they are reported more often in African American than in Euro-American women. On the other hand, hot flashes are frequently reported in women 10 or more years before menopause and before significant changes in menstrual cycling.
Hot Flashes and Night Sweats
Anthropologist Lynette Leidy Sievert assembled a table of ‘‘menopause complaints’’ reported by women from different populations around the world. Here are the ones listed most frequently: Philippines, Singapore, Taiwan—headache; Thailand— dizziness in one study, hot flashes in another; UK—depression; US (various populations)—aches and stiff joints or tenseness. Only two populations on this table listed hot flashes or night sweats in the top four complaints, calling into question the idea that hot flashes and night sweats are expected to occur at menopause, are stressful, and can be explained on the basis of hormonal changes alone. It is clear from extensive cross-cultural research that hot flashes are not universal phenomena of menopause and are experienced and reported under conditions where environmental factors (climate and altitude), culture (marital status, religion, attitudes toward menopause and aging, diet, smoking, and reproductive history), and biology (genes, hormone levels, sweating patterns) all intersect and influence their expression. As with other menopausal complaints, the evolutionary medicine take on hot flashes calls into question whether or not they were commonly experienced in the past, and if they did occur, whether they were stressful or interfered with daily function. Recalling the unusually high levels of reproductive hormones that Western women have throughout their lives, it may be that withdrawal of these hormones has a greater impact on the vasomotor system than it does for women with lower levels and fewer lifetime menstrual cycles. Perhaps a woman who starts menstruating at age 17, has 5-6 pregnancies, nurses for several years, and has only 60–100 menstrual cycles in her lifetime experiences fewer physiological disruptions at menopause than a woman who experiences 350–400 menstrual cycles with higher oestrogen and progesterone levels.
Bone mineral density (BMD) declines at the time of menopause, a phenomenon that is supposed to be associated with declining oestrogen. Because of this association, the medical view is that providing external sources of oestrogen can reverse that trend and increase BMD. Not surprisingly, things are not that simple. In fact, BMD decline slows after about 5 years and there is evidence of what has been called ‘‘compensatory processes’’ whereby bones become wider to compensate for the decreased mineralization. This is another situation in which context is important: bone mineral loss in the context of smoking, low levels of exercise, high levels of fat intake, high reproductive hormones throughout life, and some disease processes may be related to increased risk of fracture. In the absence of these other factors, however, oestrogen decline and associated decrease in BMD may not have much of an impact on fracture risks. Evidence in support of this is seen in Mayan women who show decreased BMD following menopause but no increase in fractures. This may be another example of an inappropriate designation of ‘‘normal.’’ If ‘‘normal’’ mineralization is what pre-menopausal women have, then the decrease following menopause must be ‘‘abnormal’’ and warrant some sort of intervention. Certainly there are cases in which decreased BMD is associated with fractures, but in many instances, the fractures are more closely related to lifestyle factors (like smoking, obesity, low rates of exercise that improve balance) than bone mineral levels.
An old wives’ tale is that each pregnancy costs a tooth, referring to the amount of calcium that is sometimes diverted from a woman’s skeleton to her developing fetus. The ability to recruit calcium from the skeleton is advantageous for fatal development, but there is a cost later in life when a woman is at risk for developing osteoporosis. The ovarian hormones work with other hormones to regulate calcium balance in bones, blood, tissues, and organs. Progesterone in particular seems to increase bone mass, probably by enhancing the ability to absorb and use calcium, which is especially important during the last trimester of pregnancy. More important than pregnancy for bone development is lactation, a period when infant access to calcium depends on the mother’s calcium levels, and it may draw down her skeletal reserves even further. If the interval between lactation and the next pregnancy is too short, her skeletal reserves may not have time to recover. When the body allocates calcium, the needs of the fetus and nursing infant take priority over the mother’s needs. In fact, long periods of lactation take a toll on a woman’s bone mineral levels, especially if dietary sources of calcium are limited. If she breastfeeds right up until the time when reproductive hormones drop at the menopause, she may begin the postmenopausal period with a deficit. When the reproductive hormones decrease at menopause, this triggers considerable changes in metabolism of calcium and bone. The systems that enabled a woman to absorb more calcium from the foods she consumes and maintain high levels no longer work well now that their job (reproduction) is over. This means that calcium levels depend to a greater extent on what can be taken from bone. Osteologist Alison Galloway suggests that the drop in oestrogen and progesterone at menopause serves as a signal to release restrictions on mobilization of calcium from bone. In her words, the body ‘‘borrows’’ calcium from the bone, but ‘‘this is a debt that cannot be repaid.’’ Again, this is the price we pay for a body ‘‘designed’’ for reproduction, not for long and healthy lives. Thinking about osteoporosis this way leads us to understand that it is the result of a very successful adaptation and not necessarily pathological. For women in health-rich populations, however, the drop in ovarian steroids is more precipitous because the premenopausal levels of these hormones are high; this may mean that the effects on bone demineralization are even greater and perhaps move us into the realm of pathology requiring medical intervention.
Sleep problems and insomnia are frequently reported by women during the menopause transition and often accompany aging in general. Women who report hot flashes and night sweats also report that these are disruptive of sleep, so the hormonal changes that seem to influence these symptoms may also underlie sleep problems. Unresolved, however, is which comes first, sleeplessness or hot flashes. In other words, does a hot flash wake you up or do you wake up and then have a hot flash? Most research supports a link between hot flashes and disturbed sleep, although the direction of the association is not fully understood. Furthermore, sleeplessness may also underlie mood disorders reported by women experiencing menopause, which may explain why mood disorders and hot flashes are associated in many studies. In 1977, two researchers suggested that oestrogen therapy might decrease both ‘‘psychological symptoms’’ and insomnia by relieving hot flashes, and referred to this possibility as a domino effect. In the present interpretation of the domino effect, hot flashes (perhaps caused by changes in oestrogen level) disturb sleep, and sleep disruption causes mood problems. In our research using daily reports of sensations, events, and mood, my colleagues and I were able to test for the sequence of sensations as a preliminary investigation of what comes first in the domino hypothesis. We found that hot flashes on one day were followed by reports of sleep problems, which in turn, were followed by mood disturbances. But that is apparently not the whole story. When we controlled for sleep problems, the predictive value of hot flashes on mood was not eliminated, supporting an independent effect of symptoms. Further, having sleep problems was a better predictor of mood than were hot flashes. Thus, hot flashes and sleep problems likely work both together and independently to influence mood at the time of menopause. Sleep itself can be viewed from an evolutionary perspective. There are clearly internal mechanisms that regulate sleep, but they can be easily disrupted by both internal and external events. In fact, ‘‘normal’’ sleep is, like many other aspects of our lives, culturally constructed and varies widely across cultures. For many of us, a good night’s sleep happens for about eight hours in a comfortable bed in a quiet, dark, slightly cool room.
Ideally, we are virtually dead to the world and alone or with our mate or infant. If we awaken frequently during the night because of noises, temperature changes, movements of bed partners, or internal states, we complain that we did not have a good night’s sleep. We are advised to go to bed at a set hour each night and wake at a set hour each morning. But consider the circumstances under which sleep took place in the past. Clearly some level of sleep is necessary for healthy physical and mental functioning (for instance, memories are formed during sleep), but there are risks from sleeping too soundly. Predators, enemies, and fires going out probably all favoured the ability to awaken quickly in order to respond to challenges. Sleeping close to others provided both warmth and a level of protection for adults as well as for infants. Although most sleep took place at night (we are diurnal animals, after all), it is highly unlikely that it took place between set hours. In Carol Worthman’s words, sleep was ‘‘fluid’’ and occurred within ‘‘fuzzy boundaries in time.’’ Furthermore, Worthman suggests that one of the reasons that some of us have more difficulty sleeping than others is that we minimize potential distractions, making it too quiet and giving us too much time to think and worry. Human sleep evolved in a context that included crackling fires, quietly breathing family members, and ‘‘munching of animals’’ that gave cues of security and enabled relaxation of vigilance. She refers to this as a ‘‘cue-dependency model of sleep’’ and suggests that without these cues we turn to other indicators of security, finding with our worried thoughts very little relief from cognitive stresses that disrupt sleep. Maybe our teens who like to sleep with the radio on are onto something. As wage, we find that we have more difficulty falling asleep and staying asleep. For the elderly, in particular, sleep problems are quite common. In fact, sleep aids and even powerful sleeping pills are often resorted to by people who have difficulty falling asleep or find themselves waking frequently in the night. Breathing interruptions known as sleep apneas become increasingly common with age, so some of the arousals may be adaptive in that by awakening, a person triggers a return to normal breathing. In cases like this, a person who takes a sleeping medication may find that he or she fails to awaken when respiration falters, leading to death. The evolutionary medicine view of frequent awakenings in old age is that they often serve the purpose of re-establishing respiration and are thus advantageous, defences rather than defects.