How sex hormones boost – or cut – intellectual ability
Doreen Kimura
HOW Sex Hormones Boost — or Cut — Intellectual Ability
There is arguably no area of scientific inquiry as political and controversial as research on sex differences. So it should not have been surprising last November — when two biological psychologists reported landmark work on fluctuations in intellectual abilities linked to hormone changes — that the findings made national headlines that infuriated many who read them. Harvard professor of biology Ruth Hubbard, for example, who has written extensively on the subject of gender differences, dismissed the reports as “absurd.”
Here, for the first time, the chief investigator on that study explains in her own words the details of the extensive research behind those headlines — and her personal interpretation of the findings. In a companion piece, science writer and PT contributing editor Beryl Lieff Benderly focuses her attention on the press goofs that turned the original announcement into a media circus. She explores the process by which complicated scientific reports are communicated, and miscommunicated, to the public. — THE EDITORS
Most people accept the idea that hormones influence how we behave in explicity sexual situations, and even that they can affect our moods. But evidence is growing that sex hormones influence a wider range of behaviors, including problem-solving or intellectual activities. Some of the observed male-female differences in cognitive ability appear to be determined by sex hormones working on brain systems.
Much of what we know about hormonal control of behavior has been discovered in nonhuman animals, mostly rats and mice. The parallel between human and animal mechanisms has been good enough that we can often fill in the knowledge gaps about human beings by looking at studies in rodents. They tell us that fetal hormones not only determine whether male or female sexual genitalia will be formed, but also whether the adult will act sexually like a male or female. Merely having an XY (male) or XX (female) chromosome does not in itself ensure male or female development. If no male sex hormones (androgens) are present in the fetal life of an XY individual, or if there is a serious abnormality of androgen systems, a female will be formed.
Similarly, we may have XY individuals who do not merely look like females, but appear to have many “sex-typical” female behavioral characteristics, show female play preferences, and the like. In another kind of hormonal anomaly, XX individuals may be mildly androgenized by early exposure to natural or synthetic androgens. Even if they are raised as girls, report Melissa Hines at UCLA and June Reinisch at the Kinsey Institute in Indiana, such individuals are tomboyish, engage in rough and tumble play, and show masculine preferences in toys and other matters. This illustrates the general principle that early exposure to sex hormones has lifelong effects on behavior of both men and women. This type of influence is referred to as an “organizational” effect, in contrast to “activational” effects — the short-term influence of fluctuations in hormones.
One way to observe organizational effects on cognitive function is to look at the differences which exist, on average, between men and women. Men excel on certain spatial tasks, especially those requiring accurate orientation of a line or pattern, imaginal rotation, or discerning a figure embedded in a background pattern. Men are also better on tests of mathematical reasoning. Women, on the other hand, are better at certain verbal tasks, particularly verbal fluency and articulation as well as some fine manual skills. They also appear to be faster in scanning a perceptual array, especially to make an identity match.
How Hormones Affect Average Gender Differences
One problem with attributing these differences to early hormonal influences is that men and women are also treated differently all their lives. Many people would argue vociferously that the cognitive differences are not large, that they can be diminished by intensive training, and that there is great overlap between men and women in each ability. Without denying any of this, it is nevertheless true that sex hormones have been shown to influence cognitive function in ways not readily explained by differential experience.
For example, Daniel Hier and William Crowley in Chicago found that males in whom puberty did not develop at normal age — a condition related to abnormally low testosterone levels and reduced testicular size — were poor on spatial tasks. The researchers compared them to a group of males who had developed puberty normally, but who subsequently suffered lowered testosterone levels as a result of disease. The males with early testosterone deficiency performed worse on tasks of spatial construction and orientation, and on disembedding. The groups were equal on several verbal intelligence tests.
Rat studies offer direct evidence that manipulating the hormonal environment can affect spatial problem-solving ability. Christina Williams and her coworkers at Barnard College found that in running a radial-arm maze, male rats used only geometry cues while females employed both geometry and landmark cues. In rats, sex hormones can still alter brain organization in the immediate postnatal period, when the genitals have already been formed.
Williams found that castrating males during this period, so that gonadal hormones could no longer affect the brain, caused the males when adults to adopt a more “female” strategy on mazes. Since the rats were neither reared differently by their parents nor exposed differently to mazes, hormones clearly had a powerful, lasting effect on a nonreproductive behavior.
Hier and Crowley’s study might appear to suggest that the higher the androgen levels, the higher the spatial ability, but keep in mind that they were comparing normal and subnormal levels. Within the normal range of androgens, it appears that males with lower levels actually perform better on spatial tasks. Valerie Shute and coworkers at the University of California, Santa Barbara, found that, comparing men in the bottom and top 25% in terms of testosterone levels, those with lower levels scored higher on spatial tests. We have recently confirmed this in our laboratory.
Findings from these and other studies suggest a curvilinear rather than a direct relationship between androgen levels and spatial ability. This is consistent with the finding that in females — who of course have lower levels of testosterone than males — those with higher levels of testosterone score higher in spatial ability. The situation is complicated by the fact that testosterone is probably converted to estradiol in the brain before it affects the nervous system. Helmuth Nyborg of the University of Aarhus in Denmark has proposed that estrogen actually is the critical determinant of spatial ability.
Does Ability Fluctuate During Menstrual Cycles?
All these studies were probably sampling the organizational (chronic) effects of sex hormones among groups of people. About six years ago, I decided to examine the possibility that fluctuations in hormones within an individual might also be reflected in cognitive changes. The most readily available subjects for this approach were women — young women undergoing natural fluctuations in the course of the menstrual cycle, and older post-menopausal women who were taking hormone replacement therapy.
To maximize the likelihood of finding intellectual changes, we selected tests that measured either abilities favoring males or those favoring females. We assumed that higher levels of female sex hormones (estrogen and progesterone) might enhance “female” abilities, such as verbal fluency and manual skill, but have no effect, or even a negative effect, on “male” abilities such as spatial skill.
The menstrual-cycle study was done with Elizabeth Hampson, a doctoral student in my laboratory. Initially, we did not have the capacity to analyze blood from our subjects, so we picked periods in the menstrual cycle which could be identified without such assays — the midluteal, about 7 to 10 days before onset of menstruation, when both estrogen and progesterone were high; and days 3 to 5 after onset of menstruation, when levels of all sex hormones are low. (Days 1 and 2 were avoided because some women experience physical discomfort then.) Half the women (we studied 45 in all) were tested first in the high or midluteal phase, and half in the low, or menstrual phase. About 6 weeks elapsed between the two test periods.
We used tests that had previously shown sex differences, either in overall score or in brain organization: spatial tasks, perceptual speed tasks, verbal fluency (saying as many words as possible beginning with a particular letter), rapid articulation and speeded manual learning tasks. I had found that the latter two, although dependent on the left hemisphere in both men and women, showed a different intrahemispheric organization in the two sexes We also administered a mood inventory, since it is known that there are some mood changes throughout the month, though these are usually greatest in the immediate premenstrual period, which we deliberately avoided.
Female Hormones Make “Female” Skills Better, “Male” Skills Worse
The results were definitive. Comparing women with themselves on tests usually done better by women than men, they did significantly better in speeded articulation and manual skill tasks in the high or midluteal phase than in the low phase. Verbal fluency and perceptual speed were also generally better in the high phase.
In contrast, performance on the spatial tests, in which men are usually better, showed the reverse trend: The women did better in the low phase than in the high. This effect was significant for the Rod-and-Frame test, in which the subject simply makes a judgment about whether a rod is vertical or not. Since practice effects from the first to the second session sometimes obscured the finding, Hampson then looked only at the first test-session results of individuals who were in either the high or low phase. This produced a clearer difference between tests favoring males and tests favoring females: The paper-and-pencil spatial tests now also showed a significant enhancement in the low phase.
The fact that scores went in opposite directions on the two types of tasks suggests very strongly that we are seeing more than a generalized change in ability during one phase or the other. The same was true of the mood inventory. We saw no significant changes in mood between the two phases, and only one significant correlation between any of the mood components–which sample, among other things, depression, fatigue and vigor, and performance on the various tasks.
These were exciting findings, but we were concerned that the spontaneous fluctuations of hormones during the menstrual cycle might be secondary effects; perhaps the cognitive changes were due to concomitant fluctuations in other systems.
The Effects of Hormone Therapy
Fortunately, I had already begun to look into the possible effects of exogenous hormones in post-menopausal women. With them, sex hormones don’t fluctuate spontaneously but are administered on a regular basis to relieve discomfort during or after menopause and lessen the risk of osteoporosis. Most of the various regimes of hormones therapy now include progesterone along with estrogen. This treatment can raise plasma estradiol levels into the midluteal range of the natural menstrual cycle, and often improves the feeling of well-being–but we know almost nothing of its intellectual effects.
The 33 women we studied were all over 50 years of age and had not menstruated for at least a year. Twenty-two of them were on estrogen alone, the rest on combined estrogen and medroxyprogesterone. They all took estrogen alone (orally) for at least the first 15 days of the month. Some them took progesterone as well for 5 to 10 days. For the rest of the month they took nothing.
We tested them twice, once 10 days after they had begun estrogen (but before any progesterone was taken), and again when they had been off all medication for at least 4 days. To reduce the effect of practice, we tested half of the women first in the “on” phase, and half in the “off” phase, with approximately 6 weeks between testing. Any effects we saw should have been due primarily to estrogen, since the few who took progesterone took it immediately before the off period.
We again gave a variety of tests, including vocabulary to provide a measure of general intelligence; a manual skill task in which a sequence of hand movements had to be carried out quickly on a Manual Sequence Box (see below); an articulatory task–a tongue twister; perceptual speed tests; and two spatial tasks.
The Selective Effects of Estrogen
We found that estrogen affected the motor skills strongly: In the manual task and to a lesser degree the tongue twister, performance was significantly better under the “on estrogen” condition than in the off period. On the perceptual speed tasks, one of the comparisons showed improvement under estrogen; on the other there was no effect. On the spatial tasks, there were no significant changes related to hormonal status. The results of these studies confirmed that the effects of estrogen, even when administered orally, were selective, in that not all the speeded tasks were improved, but we did not replicate all the findings from the menstrual study.
One reason for the differences might be that, since exogenous estrogen in post-menopausal women is detectable in plasma even several weeks after treatment has ended, the difference in estrogen level between our so-called on and off phases may be quite small. It may be too small to produce changes in spatial ability, though clearly large enough to produce changes in motor and articulatory skill. That suggests that motor programming skills are more sensitive to fluctuating estrogen levels than is spatial skill. We have some indications from the menstrual study as well that this might be the case.
Another reason for the difference might be that older and younger women have different baseline levels of spatial ability, in which case changes in sex hormones would not have the same effects. Unfortunately, we didn’t use identical spatial tests, so we couldn’t compare our two samples.
How Hormones Affect the Brain
We can only guess what mechanism enhances motor skill during high estrogen states, since we have no direct information on what is happening in the brain then. Here again, however, the animal literature is suggestive. Jill Becker and coworkers at the University of Michigan in Ann Arbor found that female rats were better at walking a narrow plank in the estrus phase of their cycle, when estrogen and progesterone levels are high. She also found that implanting estradiol directly into the basal ganglia enhanced performance. The precise mechanism for this change has still to be worked out, but Becker suggested that estrogen might induce changes in the firing of dopamine neurons. And estrogen may affect other brain areas as well.
In our human studies, estrogen seems to have a strongly positive effect on functions which we know depend critically on the left hemisphere. Performance on the Manual Sequence Box and on speech articulation tasks requires an intact left hemisphere. What’s more, these functions are carried out differently in the left hemisphere of males and females: The left anterior part of the brain is more critical in females, suggesting that estrogen somehow facilitates activity of the frontal regions, of the left hemisphere, or of both.
Further evidence that the left hemisphere may be favored by high estrogen levels comes from Hampson’s finding that the right-ear superiority on dichotic listening is increased during the preovulatory estrogen surge that occurs in normally cycling women. We are currently also investigating the possibility that frontal and posterior regions of the brain are differently affected by estrogen levels.
People often ask why hormonal changes affect adult abilities, since such effects would seem to serve no useful function. We don’t really know the answer, but it seems reasonable that, whatever hormonal mechanism originally establishes these individual variations, it may simply continue to be sensitive to changes in the critical hormones throughout life. Although the fluctuations may merely be ripples on a preestablished baseline, they are sufficient to suggest that the adult human brain continues to be sensitive to such influences.
Hormones Affect Men as Well as Women
The question of how androgens and estrogen interact, both in originally organizing male and female behaviors, and in their adult effects, is being studied extensively. Roger Gorski of the University of California, Los Angeles, has shown that estradiol produces some aspects of brain sexual differentiation related to perinatal testosterone. Bruce McEwen at Rockefeller University, however, points out that estrogen can have different effects in male and female brains. Perhaps different programs are activated by the same hormone.
We need a great deal more information about the basic action of sex hormones before we can work out their specific contributions to cognitive patterns. For example, knowing the current blood levels of testosterone in either men or women is just one step in a chain of information which ultimately must include how much testosterone reaches the brain, what proportion is converted to estradiol, where the estradiol has its effects, and so on.
How Important Are These Differences?
What do our studies on the apparent fluidity of certain abilities mean for the day-to-day functioning of women? While the fluctuations we find are interesting and significant–they tell us something about how cognitive ability patterns are formed–they are not large. Also, up to now they seem most consistent for the kinds of things women already do well. So for most women, they aren’t an important factor. Of course, women vary widely in their sensitivity to these influences. For some, the changes may make them feel clumsier at some periods of the month than at others.
It’s important to remember that we studied women because it was easy to do. It turns out that men also undergo hormonal fluctuations, both daily and seasonal. We are looking into both types, encouraged by a study of wild voles by Steven Gaulin at the University of Pittsburgh and Randy FitzGerald at Montclair State College. They have found that sex differences on a maze task appear only in polygynous species of these rodents where males must roam a larger territory to find females. This sexually dimorphic type of vole has seasonal fluctuations in territorial range, with the range being larger in the mating season. If there is a parallel effect in humans, we might be able to detect seasonal variations in men for certain spatial abilities. This would give us another important link in understanding how individual differences in abilities evolved.
How Hormones Affect Manual Dexterity Because women tend to have better manual dexterity than men, we decided to investigate what aspects of dexterity might be affected by fluctuations in estrogen levels. The aspect most sensitive to estrogen turned out to be organizing several movements into a pattern, which we measure using the Manual Sequence Box (left).
The task is to press the top button with your index finger (top); pull the vertical handle toward you with all four fingers (middle); and press down on the bottom bar with the thumb (bottom). Each woman did a learning sequence and five timed trials during both her high-estrogen and low-estrogen periods. On average, each woman did markedly better with high estrogen levels. Women undergoing hormone therapy also did the task faster when they were on estrogen.
Doreen Kimura is a professor of psychology at the University of Western Ontario in London, Ontario.
COPYRIGHT 1989 Sussex Publishers, Inc.
COPYRIGHT 2004 Gale Group
