A Natural History of Menopause

Menopause is defined as an ending, a lack: you can't know you're in it till it's over. That's the dance, said a doctor I asked. To medicine, it's the one-year anniversary of a woman's last period. Just that day, that 24-hour spot of time.

To most women, the word carries much more weight. Hot flashes. Mood swings. A feeling of falling apart. Aches in the teeth, the joints, the eyes. Osteoporosis. The end of fertility—and of desire, too? The transformation from "babe" into "crone." The door to old age.

I fear beginning to look like my mother, with a very large stomach, said a woman in one survey. Others noted:

Wrinkling, pot-belly, old-looking skin.

It's a negative image of a body "drying up."

Weight gain is a constant struggle.

I dislike feeling old.

"Notable was the absence of any sense of achievement, or gained status, associated with becoming menopausal," noted the researchers who conducted the survey. "Rather, the anticipated benefits had more to do with the termination of currently annoying or troubling conditions."

Recently an anthropologist did find a biological reason for "the change": as a way to secure the health of the youngsters in a family while their mother is nursing a newborn. Post-menopausal females are prodigious food-gatherers, she found, and keep their grandchildren well fed. The news was uplifting: menopause may be a good thing for the species. But the anthropologist's work carried a downside as well. She named her idea "The Grandmother Hypothesis." (At menopause, I wonder, will I become white haired and apple-cheeked? Will I learn patience and pie-baking?)

Doctors talk about the perimenopause, those years (and we're talking five, more or less) of fluctuating fertility, when a woman's once punctual periods start wandering over the days and weeks. When it comes, the blood may gush—a sign frighteningly like that of some cancers. Just the irregularity of it all takes some getting used to. Especially when no one can tell you why.

"Why do women continue to feel they don't know anything about menopause when there are so many books on the bookshelves now?," asked Phyllis Mansfield, who had conducted the survey cited above. "Because every woman's menopause is unique," she answered herself, "and no study has validated that uniqueness."

Mathematics, oddly, may give women that validation: mathematics and almost a quarter of a million urine samples. These are the telling points of a $2 million study, funded by the National Institutes of Health last July, joining Mansfield and anthropologists Jim Wood, Darryl Holman, and Kathleen O'Connor at Penn State, with demographer Maxine Weinstein at Georgetown University and nurse-physiologist Ann Voda of The Tremin Trust at the University of Utah.

Mansfield is a psychologist, a health educator, and a scholar of women's studies. She has a reassuring manner and a soothing voice, yet she's effusive and energetic when talking about what drives her; for an activist she's unusually optimistic. She offered me chocolate when I came to her office: someone baked brownies for a birthday treat. When she spoke of her new project she sat very still. "All my research is done with the aim of giving women the information that will empower them. I want to help women. I want to help women manage a very scary transition—and it's scary because they don't know what to expect."

"It's a hot topic," said anthropologist Jim Wood of the menopause. "Because of the aging of the Baby Boom and also the development of new interventions for infertility that aren't just "treatments.' It's now possible for a woman to become pregnant after menopause, with donor eggs, a simulated hormonal milieu. There's no overwhelming technical reason why she shouldn't be able to do this." Wood has the kind of enthusiasm that is barely contained in his frame. His gestures are large, his vocabulary lively. He does professorial stuff: writes flow charts on the blackboard, knits his fingers behind his head, yet he can discuss without blushing such things as breastfeeding and menstrual blood. His research landed him in New Guinea, just out of graduate school, collecting urine and trying to unravel, by reading hormonal signs, why women are sometimes fertile, sometimes not, why their cycles suddenly end. "To discover the dearth of information on these questions was absolutely mindboggling," he exclaimed.

Darryl Holman was one of Wood's graduate students and is now a postdoctoral fellow in the lab. He rivals Wood's enthusiasm (and though a younger man, beats him hands down on length of beard), but his is a more gentle, inquiring manner. He's quick to blame his own bad assumptions for the trick his data from Bangladesh played on him. With a team of Bangladeshi field workers, he collected 20,000 urine samples one year. He was assessing rates of fetal loss—how often women became pregnant but lost the fetus before they knew it—and menopause factored into his tests. "If I'd only known," he mused. "If I'd not read the literature, I would have done a better study." He laughed and explained: "I went to Bangladesh believing women over 43 had a good chance of being post-menopausal. There'd been a study in the early "80s in this same area, and it found the mean age at menopause was 43.6. That's on the low end, but it's consistent with studies of menopause in developing countries," the theory being that menopause has something to do with poor nutrition or stress. "So I only collected women up to age 48," Holman said. "By chance I got a few older than that because the field worker messed up." A statistical analysis including these strays showed the mean age at menopause to be between 49 and 51. "My conclusion now is that this belief that menopause occurs earlier in developing countries is not true. It's hooey. I went back and reviewed the literature very carefully, and in all the studies that showed a really early age at menopause, the statistical analyses were flawed."

Statistical analyses are Holman's cup of tea. He talked of being "the statistical interpreter" of the menopause group, of being interested in "building etiologic models, mechanistic models, of how we think the biology is working," of being "quite proud of the follicular depletion model" that he and Wood have developed to explain menopause. "But I want to apply it as well," he added. "We're trying to understand what menopause is."

"There's another reason why women want to do this, to be a part of this study," said Kathleen O'Connor. O'Connor, also a postdoc, actually runs the lab. She's the one who has to go down to Walmart to collect the 200 clipboards, 200 Tupperware freezer containers, and 1200 styrofoam boxes. She's the one who'll be in charge of packing and sending the stuff out to the women volunteers and organizing hormonal assays on the quarter of a million urine samples they return. She's assayed her own urine as well. "It's fun experimenting on yourself," she kidded. There's a quickness, a cut-to-the-chase efficiency about O'Connor that's lacking in the other three Penn Staters on the team. She sorted through a stack of files looking for something, grabbed a napkin when she couldn't come up with a blank scrap to draw a diagram on. She showed me the chart: two hormones fluctuating like out-of-sync waves across a span of 30 days. "It's so cool to see that," she said, "when you get a profile of your hormones across your cycle, it's so cool. You learn what your body is doing. It's a reward in itself."

"I'm taking urine samples myself, too," noted Mansfield. "Kathy and I can sit around and say, You know what it's like when you can't open the freezer box because your hands are too stiff first thing in the morning?" She smiled. "It's very important to know what you're asking other people to do."

The "other people," in this case, are a group of 150 to 200 35- to 55-year-old women selected from the Tremin Trust Menstruation and Reproductive History Program. Beginning in 1934, when Alan Treloar of the University of Minnesota enrolled the 2350 women of cohort 1, and continuing with the 1600 women of cohort 2, who were recruited in the 1960s and "70s, the Tremin Trust women have kept menstrual calendars, recording the beginning and ending dates of each of their periods on a standardized card and, on its reverse, any "unusual events" that might have affected their cycles, such as pregnancies, births, abortions, surgeries, illnesses, use of medication, or lapses in their record-keeping. A year-end health report recorded their living arrangements and household composition, education, medical and smoking histories, exercise routines, and sources of stress or support. "Since 1967," Wood and his colleagues wrote in their grant proposal, "analyses drawing upon the Tremin Trust database have made significant contributions to our understanding of the human menstrual cycle. . . . Most of what we know about age patterns of menstruation, including during the menopausal transition, is based upon the Tremin Trust sample."

Mansfield met Ann Voda, current director of the Tremin Trust, at a conference the year after the Trust moved to the University of Utah in 1984. Both were asking questions about the menstrual cycle: "Not medicalizing," Mansfield said, "but asking different questions, saying, What do women experience? We were not making assumptions that menstrual events were signs of illness. They may be different from what women experience at other times, but not illness."

In 1990, the two started a study using a subset of the Tremin Trust women, those aged 35 to 55 who were still menstruating. They focused on heavy bleeding. "We asked the women to report not just when their periods started and stopped, but we devised a scale to measure how much bleeding was occuring. Women do notice if their period is longer or shorter, but what they most notice is this heavy bleeding." The women's yearly health report was expanded to include questions about changes in their bodies, their cycles, moods, sexual response, hot flashes, and other conditions related to menopause. "We asked them how they negotiated decision-making with their physicians on hormone treatments, and where they got their information on menopause." Voda and Mansfield also asked, Why do you stay in this project? "Mostly to help other women, the subjects said. They felt they had been so much better informed by being in the project," Mansfield noted. Some had listed Edith Bunker of TV's "All in the Family" as one of their chief sources of information before joining the study. I can't talk to my physician about this, one said. Another: Two doctors told me when I asked about when to expect menopause, "Ask your mother.' Thanks anyway. A third: My mother told me nothing about the menstrual cycle or menopause. It was hush-hush.

Independently of Mansfield, Wood and Maxine Weinstein had contacted Voda about using the Tremin Trust to look at the hormonal profile over menopause. It was Voda who linked Wood and Mansfield, both at Penn State. "It was natural for us to put together a grant proposal," said Mansfield. But it took several rewrites before a federal agency agreed to fund it. ("Our fear was that the government would take so long that the women would be past menopause," Mansfield said. "This was going to be the last try, they're getting old.") Funding came through in July "97, and Mansfield admitted that the protocol was improved by the review process. "We did a quick pilot study," she explained, "with women in this area collecting their urine in a device Darryl devised. We were able to show the government that we did have a method women would easily comply with."

It's not a method easy on the lab. "We're getting daily samples from a large number of women over a long range of time," said O'Connor, "it's unprecedented. It'll open a window into the biology." It'll be a lot of work. Each woman will get a 15-pound "urine collection kit" each year: 185 "urine collection devices" (one a day for six months, then she can take a six-month break), a plastic freezer storage case, sheets of labels, sealing tape, seven 8-ounce refrigerant gel packs (one extra for travel), six postpaid polyfoam mailing boxes, and an instruction sheet on a clipboard. The easy-to-use urine collection device, which Holman hopes to patent, obviates the need for the subject to process urine in any way, e.g. to pour it from one container to another, according to the grant proposal. Each morning, the woman collects her own urine, seals the vial, attaches that day's sample identification number to the vial, and then places it into the plastic container provided for storing the specimens in her freezer. On the label she does "a little bit of record-keeping," noting if she's menstruating, taking birth control pills, eating large amounts of soy protein ("Some studies find an effect on the hormones we're looking at," said O'Connor. "It's controversial."), any medicines she might be using.

"Altogether," said O'Connor, "it should take her three to five minutes, max."

At the end of each month, she'll ship her 30-odd vials toUniversity Park via overnight express.

"We'll probably need 12 freezers here in the lab," said O'Connor. "We're hoping to get 40,000 urine samples a year." From each daily sample O'Connor and her assistants will take a one-milliliter aliquot, freezing the rest of the 10-to-15 milliliter sample. The urine in each aliquot will be used in four hormone assays, testing for forms of estradiol, progesterone, follicle-stimulating hormone (FSH), and luteinizing hormone (LH). "Each of the assays," said O'Connor, "takes about a week to run. At 36 samples per plate, 30 plates at a time, we'll be doing about 1,000 samples a week. A robotic pipetter will do a lot of the work for us," she added. "That'll help prevent carpal tunnel syndrome from repeat pipetting. But it will still be an enormous amount of work. That's why other people haven't done this before."

Over the last year, processing Holman's 20,000 Bangladeshi urine samples, O'Connor has streamlined the assaying. "Every day of the week we take out the plates, line them up, and do something to them. It's like working in your kitchen. The samples have a wide range of colors. They're quite pretty. There is a distinctive odor—" She laughed. "I get a lot of comments from the genetics people down the hall, especially when we're boiling it. But it's better than working with blood or feces."

Mansfield, meanwhile, is in charge of keeping the samples coming in. "My job," she explained, "is to make calls to every woman on the project and find out how she's doing. To make sure she doesn't have questions, that her morale is good." Mansfield will also be the one integrating the women's label jottings and yearly surveys with the hormonal assays—"that is, placing what we're finding in the lab in the context of women's lives: the stressful events, new jobs, divorces, deaths; what the women are taking, in terms of medical products or herbs; and how they're feeling: What is going on when their LH and FSH are sky high?

"We've learned that all changes are alarming to women," Mansfield said, citing her earlier work with Voda. "A woman who has a 32-day cycle will become totally alarmed if she has a 35-day period, when for another woman that's totally normal. No one has documented what all the changes are over the menopause. Treloar did some of it, the man who started the Tremin Trust."

Nearly three million office visits are made by women between ages 25 and 54 each year for complaints related to menstrual disorders; the annual visit rate for abnormal bleeding is 7.2 per 100 women, Mansfield and her colleagues note in their grant proposal. Changes in menstrual patterns have been linked to osteoporosis, cardiovascular risk, cancers of the breast and endometrium, thyroid and pituitary disorders, and other health problems. Yet no one knows the full range of "normal." Despite recent work on the biology of menopause, little is known at the population level about variation among women in the perimenopausal transition or how that variation relates to their earlier experiences.

"Here's the Reader's Digest Condensed Version," Wood began. He was giving me the lab's working model, their theory of what menopause is. When Holman had explained it to me, he used equations; O'Connor scrapped the math and pulled out illustrations and graphs. Wood began with the basics.

"Up in your head you have the hypothalmus." He drew a box on the blackboard. "A little lower down in the head you have the pituitary gland. The hypothalmus releases gonadotropic releasing hormone, GnRH. That stimulates a couple of hormones known as FSH and LH, the gonadotropins, which in turn stimulate the ovary and are critical in maintaining regular cycles. These hormones select which follicle in the ovary is going to develop into an egg. What we think is happening with this selection process is that a lot of follicles begin developing, then randomly die by a process called atresia. Atresia is the default value. But the idea is that if a follicle has started to develop at such a time that it presents the right hormone receptor during a previous month's cycle, it will be selected—it will be rescued from atresia and go on to become an egg.

"The developing follicles in the ovary feed back to the pituitary gland by releasing steroids—estradiol, which is a form of estrogen, and progesterone are the most important." At the blackboard, Wood finished drawing boxes for the pituitary and ovary and linked them with a pattern of arrows. "Except for a brief period mid-cycle, these feedbacks are mostly negative, that is, they suppress the production of hormones. The exception is estradiol. Just before ovulation, it switches over in a way that's still mysterious to a positive feedback. But it's the negative feedback from the follicles at the beginning of the cycle that's critical for getting a normal cycle underway.

"We now know with fair confidence," Wood said, speaking for science in general, "that menopause is driven by this process of follicular depletion." Or, as O'Connor had put it, "Aging comes from the ovaries." It works like this: A girl is born with upwards of a million follicles in her ovaries. Like the cells in the brain and the eye, these follicles are never replaced when they die. Unlike brain cells, dying off is what follicles do naturally: by the time a girl reaches puberty and her hormones kick in, she will have lost all but some 100,000 of them. Every month after that, a dozen or more follicles will begin to grow, only one of which is usually selected to mature and be released at ovulation. By the time she is 45 (at least according to the handful of studies that have been done using autopsy cases or ovaries that had to be surgically removed), a woman's follicles will number only about a thousand. "What happens in menopause," said Wood, "is that once you've run out of follicles, all the hormonal feedback is removed. FSH and LH go up to very high levels and randomly fluctuate, which is what we believe is responsible for hot flashes.

"The big mysteries in this research have to do with the mechanisms of follicular depletion," Wood added. "But from the statistical information, it looks like a process that begins prenatally and randomly continues. There's a random exhaustion of follicles ending with menopause." That statistical pattern lends itself to "some fairly simple mathematical models," said Wood. (As Holman had noted. "It looks like an exponential process, like a radioactive half-life.") Wood continued, "As modelers, we thought, If the process is as straightforward as that, then we should be able to develop a mathematical model of follicular depletion that could tell us what the age distribution of menopause should look like, and it should tell us something about the variation in ages of when women reach menopause.

"So we developed a model. But as we started playing around with the model more, and the implications of the model, we thought, There really shouldn't be such a thing as menopause. It's an arbitrary definition: If you haven't experienced vaginal bleeding in 12 months, you're post-menopausal. But the model told us there should be a very long upper tail"—that is, that a graph of bleeding episodes over a woman's lifetime wouldn't show just monthly clusters of dots that suddenly stop; the dots would instead at some point start spacing out and looking erratic, tailing off little by little until they eventually gave out.

"But then, we asked, Why do we see these long lags between bleeding at higher ages? It turns out that it has to do with interarrival times in superimposed pure-death processes."

Huh? Wood saw the blank look on my face and laughed, a great guffaw. "This loss of follicles can be modelled in a way that statistical modellers call a "pure-death process.' That's a subclass of what statisticians call "birth-death processes,' but here you have a stock of something and the only thing that can happen to them is that you can lose them.

"It's a very strange biological reality.

"What our model predicted was that at any age you had a possibility of having a period of time where there was no follicular development. In a young woman, that period of time might be milli-seconds long, but it's there. In an older woman, it can be very long indeed. Menopause, in a way, is just a very long phase before follicular development. Which means that a woman well past "menopause' could still get some vaginal bleeding. This is what our model predicts."

He smiled. "It's a lot nicer than thinking that something very wrong is going on if you get bleeding after menopause," he added.

Back to the science: "So then we got to wondering if these interarrival times left any hormonal signature? We searched the literature and we found it.

"We used to think that if you saw a combination of low steroids and high gonadotropins that it was the signature of menopause. But if our model was right, there should be a period in any woman's cycle when no steroids were released and we would see high gonadotropins. And we found a handful of cases. There was no explanation given. These were just weird times when the women were not producing steroids and their gonadotropins went real high. One endocrinologist went so far as to label this the "hypoestrogenic-hypergonadotrophic period.' He only saw it in one woman.

"At the same time, we were doing a pilot study for our NIH grant proposal. Kathy was collecting urine in women of a wide age range, and as data came in, hey presto! There was our own hypoestrogenic-hypergonadotrophic phase. It looked like what you'd expect during one of our interarrival times.

"Then we said—and this idea took a long time coming—we said, occasionally you'd expect to see one of these in a younger woman. We looked at our data on younger women and, by God, there it was. No steroids, and the gonadotropins would start to go up. They didn't get very far. As soon as you got hormonal evidence that another follicle was developing, the gonadotropins went right back down. When we looked at the full set of younger women, some showed no sign of this phase, some showed one day, some three or four days. But never very long phases.

"And we started to think the following brave thoughts. Let's call them hypotheses. We think these inactive phases will turn out to be a normal part of the perimenopausal transition. They will get longer—the age progression turned out to be quite regular. It makes sense to think of menopause as a very long inactive phase that could be followed by another round of follicular development if there are follicles left. And the fact that the inactive phase is getting longer, but that there's a lot of variability, explains why women's cycles are so different. It also explains a lot of what have been hitherto mysterious connections between hormone levels and menopause."

Just like women throughout the ages, I've long compared my cycles to the moon's. I wax and I wane. I come to a glowing fullness at ovulation, I get sharper and more prickly as I approach the dark respite and the cycle begins anew. The metaphor, woman as moon, doesn't bear up under too much analysis. Yet imagine this: If the moon were mortal, if she aged as I do, those dark nights that we call the new moon would increase. The month would linger. One night of darkness now, maybe two. Five nights when I'm 45. By the time I'm 55, years and years might go by without moonlight, and then, suddenly, a sliver of moon.

"We can predict, mathematically, how long it will take since the last cycle before an ovary begins putting out estradiol," Holman said, "and it depends on a woman's age. At 30 years of age, it's really short. At 45, it's really noticeable.

"We call this period the inactive phase of the cycle. We're proposing that there are three phases of the menstrual cycle, not two.

"The other two are the follicular phase and the luteal phase. During the follicular phase, the follicles are growing. The luteal phase is after ovulation, which means that the selected follicle has turned into a corpus luteum. That's when progesterone starts going up, to finish preparing the lining of the uterus for implantation." Sometimes in post-menopausal women, Holman noted, when their ovaries are removed for one reason or another, a corpus luteum is found inside. In other cases, a woman believed to be post-menopausal will get pregnant. A hysterectomy is often prescribed if a woman two or three years past menopause suddenly begins again to have vaginal bleeding, yet such a lag, in her case, could be normal: merely an extra-long phase before a last follicle develops.

To counteract some side effects of menopause, such as mood swings and hot flashes, as well as to protect against osteoporosis, physicians often suggest estrogen replacement therapy: According to the New York Times, Premarin, an estrogen replacement, is the most widely used prescription drug in the United States. "One of the things physicians typically do before a woman goes on estrogen replacement therapy," Holman explained, "is to take a blood test for hormones. If our model is right, a woman who has a single blood sample taken and looks menopausal may still have another bleeding episode. You need to know what part of her cycle the woman is in. It's not recognized that a woman can go six months in an inactive stage and yet suddenly have one of her follicles grow."

Plus, added O'Connor, "When is she going to go to the doctor? When she's unhappy. When she's in this inactive phase. Why does she feel lousy? Because her gonadotropins are high and her steroids are low. It's been a routine part of her cycle, but now it's become longer and more frequent because she's older. Those gonadotropin peaks used to be very narrow, only a day or less, before the steroids would also get high. Now it's a longer period of time."

"We're hopeful," said Wood, "that by gaining a clearer picture of the processes involved we can come up with much more reasonable guidelines for estrogen replacement therapy. For instance, how long might it be before a woman is likely to be making these hormones herself again? That's something we could predict."

Plus, added Mansfield, "There might be patterns in the way a woman's menstrual cycle changes during menopause. That's not interesting in itself, but what if we could predict patterns? For instance, what if every woman who had one child had a certain pattern, and it differed from that of a woman who had no children? Wouldn't that help women to know?"

"There might be other things that affect the rate of follicular depletion," said Holman. "We haven't identified specific ones, because the data are still being entered, but besides pregnancies, these might be the age at menarche, the use of birth control, the length of the cycle, even things like smoking. There's an epidemiological finding that women who smoke have earlier menopause. That might be purely a coincidence, nobody knows the mechanism, but smoking might kill follicles or it might affect a woman endocrinologically somehow. Right now there's not a lot of evidence that anything affects follicular depletion very much, but this study will let us look for factors that might somehow affect the underlying rate."

Said Wood, "Adding this inactive phase to our model of follicular depletion is really a powerful concept in understanding the aging of the reproductive system. It's an elegant way to explain some phenomena that haven't been tied together before. And it can make s