Facial attractiveness in women was negatively correlated with age at menopause and positively correlated to current fecundity

Żelaźniewicz A, Nowak-Kornicka J, Zbyrowska K, Pawłowski B (2021) Predicted reproductive longevity and women’s facial attractiveness. PLoS ONE 16(3): e0248344. https://doi.org/10.1371/journal.pone.0248344

Abstract: Physical attractiveness has been shown to reflect women’s current fecundity level, allowing a man to choose a potentially more fertile partner in mate choice context. However, women vary not only in terms of fecundity level at reproductive age but also in reproductive longevity, both influencing a couple’s long-term reproductive success. Thus, men should choose their potential partner not only based on cues of current fecundity but also on cues of reproductive longevity, and both may be reflected in women’s appearance. In this study, we investigated if a woman’s facial attractiveness at reproductive age reflects anti-Müllerian hormone (AMH) level, a hormone predictor of age at menopause, similarly as it reflects current fecundity level, estimated with estradiol level (E2). Face photographs of 183 healthy women (Mage = 28.49, SDage = 2.38), recruited between 2nd - 4th day of the menstrual cycle, were assessed by men in terms of attractiveness. Women’s health status was evaluated based on C-reactive protein level and biochemical blood test. Serum AMH and E2 were measured. The results showed that facial attractiveness was negatively correlated with AMH level, a hormone indicator of expected age at menopause, and positively with E2, indicator of current fecundity level, also when controlled for potential covariates (testosterone, BMI, age). This might result from biological trade-off between high fecundity and the length of reproductive lifespan in women and greater adaptive importance of high fecundity at reproductive age compared to the length of reproductive lifespan.

Discussion

In contrast to the research hypothesis, the result of this study showed that facial attractiveness of women at reproductive age is negatively related with AMH level. Simultaneously, we found a positive correlation between face attractiveness and estradiol level, a hormone predictor of current fecundity [2], which was also shown in previous studies [6, 40; but see also for negative results 41]. Facial attractiveness was also negatively related with BMI what has been also shown in the previous studies [42,43].

Our results contradict the results obtained by Bovet et al. [16], showing a positive correlation between face attractiveness and predicted length of reproductive lifespan, estimated based on maternal age at menopause. Although, the most recent data on secular trends in age at menopause in Europe are scarce and difficult to compare, there seem to be no major difference between European countries, including Poland and France [44,45], that could explain the contradictory results of the studies. This difference in the study outcomes may be explained by different methods to estimate expected age at menopause employed in the two studies. Although, there is a positive association between mother’s and daughter’s age at menopause, existing estimates of the heritability of menopause age have a wide range [21,22,46]. Also, reported mother’s age at menopause may not be accurate due to the potential risk of recall bias [47]. Furthermore, previous research showed that AMH level is a better predictor of a woman’s TTM, compared to mother’s age at menopause [48,49], due to several reasons. AMH level is influenced by environmental factors that are also related with menopausal age, such as smoking or diet [50,51]. Also, a mother’s age at menopause is determined by genetic factors, that are shared by a mother and a daughter, and by environmental factors acting only on a mother, but not on a daughter [49]. While, a daughter’s age of menopause is influenced both by genetic and environmental factors, with genetic component reflecting not only maternal but also paternal genetic contribution [46,52]. Therefore, whilst information from mother’s age at menopause only reflects the maternal half of genetic influence, AMH level may reflect the sum total of genetic and environmental influences [50], and thus correlates more strongly with actual age at menopause [49]. Additionally, maternal age at menopause may only predict a daughter’s at menopause, whereas women’s fertility decline earlier, what reduce the chance of a successful pregnancy a few years before menopause. The age of the onset of a period of subfertility and infertility that precede menopause differs among women as well [46], and this should be indicated by AMH level (marker of diminishing ovarian reserve) but not by maternal age at menopause.

The results of the study also showed a negative correlation between AMH and E2 levels, what is in line with previous research [53,54]. Experiments in vitro showed that E2 down-regulates AMH expression in primary cultures of human granulosa cells (what in vivo may facilitate reduction of ovarian reserve), and when estradiol concentration reaches a certain threshold, it is capable of completely inhibiting AMH expression through ERβ receptors [55]. This, together with the results of our study, may suggest an existing trade-off between current fecundity, length of reproductive lifespan and a woman’s capability to invest in morphological cues of both. Life-history theory predicts that evolution of fitness-related traits and functions is constrained by the existence of trade-offs between them. Trade-offs are ubiquitous in nature, their existence is explained in the context of resource limitations [56], and may be observed not only between different traits and functions (e.g. immunity and fertility), but also within one function, e.g. different components of reproductive effort. Possibly, there is also a trade-off between high fecundity at reproductive age (the likelihood of fertilization within the cycles at reproductive age) and the length of reproductive lifespan (allowing for reproductive profits in a long-term perspective).

The existence of such trade-off may be confirmed by research showing that older age at menopause is related with using hormonal contraception for longer than a year [57,58; but see for contradictory results: 59,60] and occurrence of irregular cycles before age of 25 [58], which are often anovulatory [61]. Also, some research show that the number of children correlates negatively with AMH level in young women, what may suggest that more fertile women have shorter TTM [62,63]. On the other hand, some research show a positive correlation between AMH level and number of children [64] and that childlessness is linked with younger age at menopause [57,65,66]. However, this correlations may be caused by other variable (e.g. genetic factors or some disease), that causes both low fertility and earlier ovarian failure [66], and thus do not exclude the possible existence of the trade-off between high fecundity at reproductive age and length of reproductive lifespan.

Furthermore, sexual selection may act more strongly on male preferences toward cues of high fecundity at the reproductive age compared to cues of long reproductive lifespan. This presumption might explain the observed finding of a negative relationship between attractiveness and AMH and a simultaneous positive correlation between attractiveness and E2. Firstly, although humans often live in long-term pairbonds, remarriage is common after spousal death and/or divorce, resulting in serial monogamy [67]. Thus, as adult mortality was higher and the expected lifespan was shorter in our evolutionary past [68], men would profit more from mating with highly fecund women compared to mating with women with longer reproductive lifespan. Furthermore, many women (also in traditional societies) give last birth long before the time of menopause, not fully profiting from the length of their reproductive lifespan [69]. Pregnancy in older age is related to a higher risk of pregnancy complications, miscarriage [70], and maternal death [71], what might contribute to an earlier cessation of reproduction [69]. Also, many environmental and life-style factors may impact age at menopause [51,72], influencing the relationship between morphological cues of long reproductive lifespan at younger age and the actual age at menopause. Thus choosing a potential partner based on the cues of current fecundity may bring a greater fitness pay-off, compared to choosing a partner with a potentially long reproductive lifespan.

Finally, some limitations of our study need to be addressed. Both AMH and E2 levels were assessed only at the between-subjects level, based on a single measurement. Although AMH level has been shown to vary across menstrual cycle [73], the extent of variation is small and sampling on any day of the menstrual cycle is expected to adequately reflect ovarian reserve [74]. However, E2 level predicts most reliably a woman’s fecundity if based on repeated sampling across menstrual cycle [75]. Thus, it would be worth to verify the results of our study with repeated AMH and E2 measurements, using longitudinal, rather than cross-sectional design, to assess the relationship between these hormones and a woman’s facial attractiveness.

This is the first study investigating the relationship between AMH level and facial attractiveness in women. The results showed that women perceived as more attractive are characterized by lower AMH, hormonal predictor of age at menopause, and higher E2 levels, hormonal indicator of current fecundity. This might result from biological trade-off between high fecundity and the length of reproductive lifespan in women and greater adaptive importance of high fecundity during reproductive age compared to the length of reproductive lifespan.