Do Women’s Faces Change Color Over the Menstrual Cycle?
It’s not difficult to tell when a female chimpanzee is in heat. As she nears ovulation — the point in her cycle when she’s most fertile — her bottom swells up like a balloon and turns bright pink.
Humans are obviously different. We don’t make a show of how fertile we are. But does this mean that human females have evolved to conceal ovulation?
Women are most fertile during the late follicular phase of their menstrual cycle, which starts about a week after their period begins and ends a week later with ovulation. At this time women experience subtle changes in their mating psychology, behaviour, and physiology that are similar to the changes we see in other primates.
You may have heard of Geoffrey Miller’s infamous lap-dancing study from 2007. Miller asked professional exotic dancers to keep a record of their nightly tip earnings for two months. The women also reported when their periods began and ended, so Miller could calculate when they were most fertile. He found that the dancers received about US$67 per hour when they were near ovulation, but only US$52 at less fertile times of the month (and US$37 during menses).
This suggests that women are sufficiently more attractive at peak fertility to persuade men to part with their hard earned cash. But what was it about the dancers that men found so enticing?
We don’t know for sure, but it was likely a mix of signals. Research has shown that as ovulation approaches, women’s voices rise in pitch, their body odor becomes more sexually attractive, and their clothing reveals more skin. Men find all of these changes appealing. So, not a far cry from male chimps, who prefer to get jiggy with big-bottomed females.
The face of fertility
There is also some evidence that women’s faces are more attractive near ovulation. In one study, women were photographed once at peak fertility and again later in the cycle when their fertility was lower. The pairs of photographs were shown to men, who were asked to choose the photograph they found more attractive. Men preferred the photographs taken during the fertile cycle phase.
What was it that changed in the photographs from one half of the month to the other? Well, the attractiveness effect was weaker when the women’s clothing and hair were obscured in the photograph. So clothing and hair are clearly important, but they’re not everything.
The same effect has been found by several other research teams (but not this one), so we can be fairly sure it’s genuine.
I and my research collaborators wondered whether women’s faces might be changing color across the month. This isn’t as far-fetched as it sounds. Women don’t vary in attractiveness over the cycle if they’re photographed wearing makeup, which implies that makeup conceals natural changes in skin appearance. And other primates, such as rhesus and Japanese macaques and mandrills, develop a redder face when they’re most fertile.
Perhaps our own species experiences a similar — if less noticeable — change in facial redness. This could certainly explain the attractiveness effect: women with redder faces are more attractive.
Are women like macaques?
Twenty-two young women visited our laboratory at the University of Cambridge every weekday for a month. It wasn’t easy wrangling an average of 13 repeat appointments for so many women, but we really wanted to take as many photographs as we could. Most researchers only photograph their volunteers twice, and until now one photograph per week for five weeks was the most anybody had managed.
Please help me!
I am running some new online experiments on attraction and it would be great if you could take part. They’re quick and fun! Click here to see a list of the current studies.
We photographed the women without makeup and asked them to wear a black hairdressers’ smock so that the color of their clothes wouldn’t be reflected onto their face (these three experiments show that women are more likely to wear red or pink clothes when they’re most fertile). Then we used a computer program to cut out patches of skin from the cheeks on each photograph.
Because cameras don’t represent color in the same way as the human visual system, we made adjustments to our camera so that it would more accurately record color. We then converted the results into the colors seen by the human eye and brain. This important step allows us to say whether any changes in color are biologically relevant: can they be detected by a person as well as a camera?
The women told us when their periods started and used Clearblue ovulation tests so we knew how fertile they were when each photo was taken. The average length of a woman’s menstrual cycle is 28 days, but most women vary somewhat in their cycle length. We used a formula to squash or stretch each woman’s cycle to fit the average of 28 days, so that we could spot patterns of color change regardless of whether a woman’s periods were 20 or 36 days apart.
The analysis
When it came time to analyze our data, we didn’t want to simply select a window of days in the second week of the cycle and label it “high fertility”, and do the same for “low fertility” days in the third and fourth week. There is no strict agreement among scientists as to how large each window should be. There are considerable “researcher degrees of freedom” to choose windows that give us a more desirable result. If we decided to select days 7–14 as our “high fertility” window rather than, say, days 6–13, perhaps we would find an effect of cycle on face color that we otherwise wouldn’t. Of course, this would be unethical and extremely bad scientific practice.
There are other ways around this dilemma, but we decided to analyze continuous change in color over the cycle using a method called Fourier regression. To be honest, how this technique works is beyond my feeble brain. Thankfully, my collaborators are mathematical geniuses and they inform me that Fourier regression is perfect for analyzing data that vary repeatedly over a known period of time. It’s a technique more commonly used in the fields of epidemiology and climatology, but it works just as well with menstrual cycle data. Perfect!
The results
So what did we find? Well, like their primate cousins, women do get redder faces at the middle of the cycle as they approach ovulation. However, the level of redness stays relatively high through the remainder of the cycle when fertility is low. Then it dips in the first few days of the cycle (during menses).
This isn’t exactly what we had expected. Although women’s faces did vary in redness, it didn’t seem likely that men could use this change as a cue to a woman’s fertility. Women are much more fertile just before ovulation than just after, but the redness of their faces at those two times was almost identical. Macaques, we ain’t.
Next we checked whether the change in redness was perceivable by the human eye. We constructed two computer models of the human visual system and found that, even under perfect lighting conditions, it wouldn’t be possible for a human to detect these changes in redness.
It is therefore pretty doubtful that facial skin color is responsible for the effect of the menstrual cycle on women’s attractiveness. If our species ever advertised our fertility with noticeable changes in facial color, we don’t any more.
What’s going on?
So if it’s not face color, what is it that causes cycle-related shifts in attractiveness?
Some scientists have found that women’s faces change shape over the cycle, although I think it’s more likely that these changes are explained by subtle variations in posture or expression than genuine transformations in the contours of the face. Even in passport-style photographs, women are rated more attractive when they wear attractive clothes (invisible to the viewer), possibly due to barely perceptible micro-expressions. We already know that women wear more sexy, revealing clothes when they ovulate, and this may drive changes in expression and, therefore, face ‘shape’ and attractiveness.
It’s also plausible that there are more obvious fluctuations in facial skin color than those we detected. After all, we did only look at a small area of the cheek. Perhaps women’s lips become especially red at peak fertility, even without the help of lipstick (women wear more makeup near ovulation).
Some indicators of women’s fertility are stronger when women are more stimulated. Heterosexual women are more flirtatious when fertile, but only in the presence of attractive men. Men find dilated pupils attractive in a woman. Straight women’s pupils increase in diameter during the fertile phase, but only in response to photographs of the women’s boyfriends. It’s possible that facial redness works in a similar way. We detected a real though imperceptibly small change in redness, but maybe the blushing that accompanies arousal is more pronounced when women are fertile. We’ll need to do more research to find out.
Whatever it is that’s going on, women shouldn’t worry that they’re advertising their fertility status with a flushed red face. As the eminent evolutionary biologist Bernard Campbell once pointed out, “voluntary signaling by the female [may replace] the involuntary physiological signals of estrus”.
In other words, if anything’s going to give you away, it’s the revealing clothes, extra makeup, and kickass pole-dancing skills.
Burriss, R. P., Troscianko, J., Lovell, P. G., Fulford, A. J. C., Stevens, M., Quigley, R., Payne, J., Saxton, T. K., & Rowland, H. M. (2015). Changes in women’s facial skin color over the ovulatory cycle are not detectable by the human visual system. PLOS ONE. doi: 10.1371/journal.pone.0130093 (read full paper)
For an audio version of this story, see the 30 June 2015 episode of The Psychology of Attractiveness Podcast.
Do you have any questions about the project? Tweet me: @RobertBurriss
