Reproductive Ecology and the Endometrium: Physiology, Variation, and New Directions docx

Clancy* Department of Anthropology, University of Illinois, Urbana-Champaign, Urbana, IL 61801 KEY WORDS reproductive ecology; endometrium; ecology; energetics; inflammation ABSTRACT Endo

Reproductive Ecology and the Endometrium: Physiology, Variation, and New Directions

Kathryn B.H Clancy*

Department of Anthropology, University of Illinois, Urbana-Champaign, Urbana, IL 61801

KEY WORDS reproductive ecology; endometrium; ecology; energetics; inflammation

ABSTRACT Endometrial function is often overlooked

in the study of fertility in reproductive ecology, but it is

crucial to implantation and the support of a successful

pregnancy Human female reproductive physiology can

handle substantial energy demands that include the

pro-duction of fecund cycles, ovulation, fertilization,

placen-tation, a 9-month gesplacen-tation, and often several years of

lactation The particular morphology of the human

endo-metrium as well as our relative copiousness of

menstrua-tion and large neonatal size suggests that endometrial

function has more resources allocated to it than many

other primates The human endometrium has a

particu-larly invasive kind of hemochorial placentation and

trophoblast that maximizes surface area and maternal–

fetal contact, yet these processes are actually less

effi-cient than the placentation of some of our primate

rela-tives The human endometrium and its associated

proc-esses appear to prioritize maximizing the transmission

of oxygen and glucose to the fetus over efficiency and

protection of maternal resources Ovarian function

con-trols many aspects of endometrial function and thus

var-iation in the endometrium is often a reflection of

ecologi-cal factors that impact the ovaries However, preliminary

evidence and literature from populations of different

reproductive states, ages and pathologies also suggests

that ecological stress plays a role in endometrial

varia-tion, different from or even independent of ovarian

func-tion Immune stress and psychosocial stress appear to

play some role in the endometrium’s ability to carry a fetus through the mechanism of inflammation Thus, within reproductive ecology we should move towards a model of women’s fecundity and fertility that includes many components of ecological stress and their effects not only on the ovaries, but on processes related to endo-metrial function Greater attention on the endometrium may aid in unraveling several issues in hominoid and specifically human evolutionary biology: a low implanta-tion rate, high rates of early pregnancy loss, prenatal investment in singletons but postnatal support of several dependent offspring at once, and higher rate of reproduc-tive and pregnancy-related pathology compared to other primates, ranging from endometriosis to preeclampsia The study of the endometrium may also complicate some

of these issues, as it raises the question of why humans have a maximally invasive placentation method and yet slow fetal growth rates In this review, I will describe en-dometrial physiology, methods of measurement, varia-tion, and some of the ecological variables that likely pro-duce variation and pregnancy losses to demonstrate the necessity of further study I propose several basic ave-nues of study that leave room for testable hypotheses in the field of reproductive ecology And finally, I describe the potential of this work not just in reproductive ecol-ogy, but in the resolution of broader women’s health issues Yrbk Phys Anthropol 52:137–154, 2009 V V C 2009 Wiley-Liss, Inc.

The uterus is the site of many physiological processes

related to pregnancy, starting at implantation It is the

endometrium that is invaded by the trophoblast, and the

endometrium that in part determines the degree of

maternal–fetal contact Human female reproductive

physiology and behavior have evolved to handle

substan-tial energy demands and determine not only the viability

of conception, but also 9-month gestation and often

sev-eral years of lactation, with babies that are larger and

larger-brained than all other primates (Mace, 2000)

Only humans have such invasive fetal burrowing to

maximize the transfer of glucose and oxygen from

mother to fetus that, in some cases of pathology, the

placenta can breach the uterine wall (Bischof and

Campana, 1996)

The physiology and cyclic changes of the

endome-trium and placentation vary broadly across the

prima-tes (Martin, 2003) Where the strepsirhines have

epitheliochorial placentation and relatively low

mater-nal–fetal contact, haplorhines have hemochorial

placen-tation with a high degree of maternal–fetal contact

Human hemochorial placentation and endometrial

dif-ferentiation is characterized by the highest degree of

maternal–fetal contact known, where the interhemal

barrier (the cell layers separating maternal and fetal blood) narrow to a single-cell layer by the third trimester

This allocation of resources in humans to the fetus required a reorganization of endometrial tissue and a greater allocation of resources to endometrial function Although the ovaries control much of the proliferation and secretary processes of endometrial function through the menstrual cycle and can thus be constructive in understanding variation in fecundity, variation in con-ception rates cannot be explained by ovulation alone (Lipson and Ellison, 1996; Kosmas et al., 2004; Ulug

et al., 2006) This does not signal that ovarian function and endometrial function are not linked, but that

ovar-*Correspondence to: Kathryn B H Clancy, Department of An-thropology, University of Illinois, Urbana-Champaign, 607 S Math-ews Ave., 187 Davenport Hall, Urbana, IL 61801, USA.

E-mail: kclancy@illinois.edu

DOI 10.1002/ajpa.21188 Published online in Wiley InterScience (www.interscience.wiley.com).

YEARBOOK OF PHYSICAL ANTHROPOLOGY 52:137–154 (2009)

ian hormone concentrations explain a portion of

varia-tion in endometrial funcvaria-tion rather than all of it More

direct studies of endometrial functioning are necessary

to understand variation in reproductive function within

and between human populations, especially if we are to

understand reproductive processes that occur after

ovu-lation The lens of reproductive ecology is useful here, as

the factors that produce variation in ovarian function

are likely to exert some effect on the endometrium,

indi-rectly via the ovaries if not also diindi-rectly Population

var-iation in lifestyle, ecology, and developmental conditions

produce significant variation in ovarian hormones

(Ellison et al., 1993) As it is a target tissue of the

ova-ries, but a tissue also responsive to inflammation and

possibly insulin, age, energetic factors, and also

immu-nological and psychosocial stress, ecological factors will

be examined as possible sources of variation in

endome-trial functioning

I will review the physiology of the endometrium from

basic form to changes through the menstrual cycle,

including some comparative review across the primates

to provide the reader with basic information about

endometrial processes and functioning I will then

describe some of the methods available for measuring

endometrial function, and synthesize the current

litera-ture on endometrial variation and its proximate

determi-nants This information will help to inform a set of topics

in the context of reproductive ecology that I propose,

which will create the framework for hypotheses for

future testing

The central question of this review is to what extent

does the endometrium mediate reproductive success due

to its responsiveness to ovarian hormones that are

themselves mediated by ecology, and how much of

endo-metrial function is independent of the ovaries, that is,

impacted directly by ecological factors? Ecology produces

patterns of variation in ovarian function, which in turn

affects endometrial function; this undoubtedly produces

some effect Inflammatory processes also exert

significant effects on endometrial function (Sebire, 2001;

Modugno et al., 2005; Agic et al., 2006; Puder et al.,

2006) As the endometrium is largely engaged with

processes of implantation and future gestation, energetic

or more broadly ecological conditions may affect the

ovaries and endometrium in different ways This

varia-tion could not be detected if only ovarian funcvaria-tion were

measured

This review will show how the study of the

endome-trium will not only help answer current questions in

reproductive ecology, but also lead to new questions

about how adult and childhood energetic condition

affect reproductive functioning, and discuss the

possi-bility of more than one ecological pathway to lead to

variation in endometrial function An understanding of

several important aspects of hominoid and human

reproduction may be impacted by information

regard-ing variation in endometrial function, includregard-ing a low

conception rate, high rates of early pregnancy loss,

prenatal investment in singletons but postnatal

sup-port of several dependent offspring at once, and higher

rate of reproductive and pregnancy-related pathology

compared to other primates, ranging from

endometrio-sis to preeclampsia Finally, the inclusion of the

endo-metrium into the study of human reproductive ecology

has implications not only for women’s health but also

opens avenues for future research into nonhuman

primates

ENDOMETRIAL PHYSIOLOGY Physiology and hormonal control of the

endometrium

The endometrium lines the corpus (body) of the uterus It is one of the fastest-growing tissues in humans, composed of two layers: the functionalis and basalis Although the basalis does not respond to the hormonal changes of the cycle, the basalis gives rise to the functionalis (Heller, 1994) The functionalis responds

to hormonal action and proliferates, maintains, differen-tiates, or sheds its cells based on these signals (Heller, 1994) Not all mammalian endometria behave in this way: rodent endometria decidualizes only in the pres-ence of a blastocyst, where human endometria do it as a matter of course (Finn, 1974)

Epithelial glands and cellular stroma compose the endometrium, both of which change morphologically across the menstrual cycle Where increasingly coiled glands and vessels, increased gland complexity, and mitotic division of the stroma characterize the prolifera-tive (follicular) endometrium, the secretory (luteal) endo-metrium is characterized by subnuclear vacuoles lined

up along the glands to maximize secretion, and stromal edema (swelling) at the time of the window of implanta-tion (Heller, 1994) The spiral arterioles are maximally coiled after this point (Heller, 1994), and towards the end of the cycle the entire stroma decidualizes (becomes

a dense cellular matrix to control trophoblast invasion)

If human chorionic gonadotropin (hCG) from an embryo had not signaled imminent implantation, the endome-trium would break down and hemorrhage from its differ-entiated state, which then leads to menstruation (Heller, 1994)

The main hormones that act on the endometrium are ovarian sex steroids (estradiol and progesterone), insu-lin, hCG and luteinizing hormone (LH), prolactin, and oxytocin Androgens and glucocorticoid receptors are also found in the endometrium (Jabbour et al., 2006) Cortisol may have a role in the endometrium, as it is often acti-vated as an anti-inflammatory response to the inflamma-tory mechanisms of menstruation and implantation (McDonald et al., 2006) Cortisol binds to the glucocorti-coid receptor and has a high affinity for the mineralocor-ticoid receptor in the endometrium; high cortisol concen-trations can interfere with mineralocorticoid signals and can cause disorders (McDonald et al., 2006) Further, cortisol is an important indicator of HPA activation and higher cortisol concentrations are correlated with preg-nancy loss (Nepomnaschy et al., 2006); chronic psychoso-cial stress is also associated with low birth weight babies

in a sample of low income women (Borders et al., 2007) HPA activation can increase levels of matrix metallopro-teinases, which are involved in degrading the extracellu-lar matrix (ECM) in tissue remodeling (Yang et al., 2002) This is important to the creation of spiral arteries, decidualization of the endometrium, implantation, and early gestation (Curry and Osteen, 2003)

Estradiol promotes the actions of the proliferative phase of the endometrium, and primes progesterone receptors for their role in the secretory phase; progester-one receptors cannot be expressed without first being primed by estradiol (de Ziegler et al., 1998) Progester-one, secreted by the corpus luteum, inhibits some of estradiol’s proliferative effects, and it maintains the endometrium through the implantation window in the mid-secretory phase (Brar et al., 1997) Whether the

endometrium responds to ovarian hormones in a

thresh-old (some minimum concentration is required for action)

or dose–response model (the amount of action varies by

hormonal concentration) is unclear In vitro fertilization

studies, where hormone concentrations are several times

the physiological norm, sometimes demonstrate a dose–

response model, where increased estradiol

concentra-tions are associated with a thicker endometrium

(Ran-dall et al., 1989; Milligan et al., 1995; Zhang et al.,

2005); this relationship holds in some normal cycles as

well (Randall et al., 1989; Bakos et al., 1994) Should the

endometrium prove to operate in a threshold model

simi-lar to testosterone and spermatogenesis, endometrial

thickness and function may not be as functionally

rele-vant as previously thought, or other factors could be

im-portant to the production of variation other than ovarian

hormones And if the endometrium operates in a dose–

response model, then greater inter and intrapopulational

variation may be expected, as has been found in ovarian

hormone concentrations

Hormone concentrations vary with energy

expendi-ture, nutritional status, and other ecological factors, and

thus ecology indirectly affects endometrial function (for a

review see Ellison, 2001) But the presence of insulin,

in-sulin-like growth factor-1 (IGF-1), and insulin-like

growth factor-1 binding protein (IGF-1 BP) receptors in

endometrial tissue (Strowitzki et al., 1993; Corleta et al.,

2000) suggests that some energetic factors could directly

affect the endometrium, because insulin is involved in

energy storage and release Insulin receptors are most

present during the secretory phase, where IGF-1

recep-tors are present throughout the reproductive cycle and

are modulated by IGF-1 BPs (Strowitzki et al., 1993)

Estrogen receptors are necessary for IGF-1 to stimulate

a proliferative response in the follicular phase (Klotz

et al., 2002; Curtis Hewitt et al., 2005) and significant

cross-talk occurs in this process; IGF-1 BP also

modu-lates embryo implantation (Fluhr et al., 2006) Further,

insulin resistance is associated with thick endometria

in-dependent of reproductive pathology (Iatrakis et al.,

2006), and insulin inhibits differentiation in the

endome-trium in vitro (Giudice, 2006) Insulin and related

hor-mones are most active, therefore, around the window of

implantation, but insulin also plays some role in

endo-metrial proliferation and the downregulation of

decidual-izing mechanisms These receptors and hormones are

downstream mediators of ovarian function on the

endo-metrium (Klotz et al., 2002), and so are not fully

persua-sive evidence of direct ecological effects on the

endome-trium; however, the relationship between inflammatory

processes and insulin suggests, at the least, that

inflammation in the body can disrupt some of these

mechanisms (Pradhan et al., 2001)

hCG and luteinizing hormone (LH) act on the same

receptors; broadly, hCG signals the presence of an

embryo to the endometrium, and LH triggers ovulation

Endometrial tissue contains HCG/LH receptors and

mRNA (Licht et al., 2003) The expression of hCG/LH

receptors is affected by cycle phase, in that mid-secretory

phase endometria have full expression of their mRNA

but downregulation of full-length hCG/LH receptor

mRNA occurs in the late secretory phase and early

preg-nancy (Licht et al., 2003) Where maternal processes

may protect against late implantation through receptor

downregulation, which would be in a suboptimal

endo-metrial environment for successful pregnancy, fetal

proc-esses appear to promote maintenance of decidualized

endometrial tissue, as hCG both rescues the corpus luteum and affects prostaglandin synthesis HCG exhib-its a dose-dependent inhibition of IGF-1 BP and prolac-tin (Fluhr et al., 2006), and prolacprolac-tin affects endometrial function Prolactin is present in the window of implanta-tion and beyond in the secretory phase of the endome-trium, and it is necessary for embryo implantation (Fluhr et al., 2006) through the maintenance of secretory phase estradiol receptors (Basuray et al., 1983; Frasor and Gibori, 2003)

Finally, although oxytocin is best known for its dual roles as a major actor in parturition and as the ‘‘bonding hormone,’’ oxytocin is also present in the nonpregnant endometrium, most strongly at mid-cycle (Steinwall et al., 2004) Oxytocin, like its synthetic partner pitocin, stimulates muscle contractions Steinwall et al (2004) suggest that oxytocin production is upregulated by estra-diol and downregulated by progesterone, as this is the pattern seen for oxytocin production in the hypothala-mus Locally produced oxytocin in the nonpregnant endometrium could produce myometrial contractions that support sperm and egg transport, menstruation, and implantation (Steinwall et al., 2004)

It is neither the case that only estradiol and progester-one control the endometrium, nor that promotion of pro-liferation and decidualization are the only important actions on endometrial function Other hormones act to promote sperm transport and implantation, as well as allow the possibility of other direct effects on endome-trial function such as those by insulin; some of these are regulated by ovarian hormones, but some may be regu-lated by other factors This implies endometrial function

is impacted not just by ovarian function but by several factors acting in concert to maximize chances for conception

Menstrual cycle behavior of the endometrium

In addition to the broad proliferative and secretory shifts that occur in the endometrium across the men-strual cycle described earlier, the endometrium exhibits some specific behaviors at the periovulatory phase, the implantation phase, and the end of the cycle (menses) The periovulatory and menstrual phases are described next, and implantation receives its own separate discussion

Periovulatory phase The endometrium responds to ec-ological and ovarian signals, but it also plays its own role in fertility In natural and IVF cycles, the endome-trium produces periovulatory waves from cervix (the neck of the uterus that leads to the vagina) to fundus (the top of the uterus, at the other end of the corpus) These waves are quite literal; the muscles of the uterus contract in such a way that the endometrium moves in a wavelike, directional motion, that varies in frequency, direction, and intensity at different phases of the cycle (Bulletti and de Ziegler, 2005) Cervix to fundus waves predominate over other types of waves in conceptive cycles (IJland et al., 1997, 1999) IJland et al (1997 1999) suggest that these ‘‘inward’’ waves encourage semen to travel towards the egg and increases the chan-ces of conception In their examination of spontaneous, natural cycles, IJland et al (1997) showed a greater

‘‘outward’’ (fundus to cervix) waves in nulliparous women in a nonconceptive cycle than parous women in a nonconceptive cycle or women in a conceptive cycle

139

REPRODUCTIVE ECOLOGY AND THE ENDOMETRIUM

Tracking IVF patient cycles, significantly greater

endo-metrial wavelike activity (91% vs 71% of time observed)

and wave frequency (8.29 waves/min vs 3.99 waves/min)

were observed compared to spontaneous cycles (IJland

et al., 1999); other research teams also document this

(Lesny et al., 1998) Seventy-three percent of these

spon-taneous cycles had a wave direction switch from

‘‘outward’’ to ‘‘inward’’ at the time of ovum pickup

(perio-vulatory period) (IJland et al., 1999) The earlier this

wave direction switch occurred, the lower the chances of

conception; the findings suggest that the persistence of

‘‘outward’’ waves until hCG administration (34 h before

ovum pickup) is a frequent precursor to pregnancy

(IJland et al., 1999) The endometrium may guide

unwanted endometrial debris, pathogens or other

sub-stances out of the uterus until the periovulatory period,

when it switches to encourage sperm transport and

pre-vent embryo expulsion Therefore, endometrial behavior

factors significantly in conception

Menstrual phase Menstruation is a result of tissue

remodeling, and also an inflammatory process The

men-strual phase may be susceptible to HPA activation

because of the importance of matrix metalloproteinase

action in the breakdown of the decidualized

endome-trium (Yang et al., 2002) The end of the secretory phase

of a nonconceptive cycle is associated with ‘‘secretory

exhaustion,’’ that is, the endometrium has prepared for

conception, and without an embryonic signal to maintain

it, begins to break down (Heller, 1994) The withdrawal

of ovarian steroids stimulates prostaglandin production,

then prostaglandins aid in menstruation and stimulate

contractions to remove endometrial debris and blood

(Sugino et al., 2004); these prostaglandins are

differen-tially transported away from the endometrium at other

phases of the menstrual cycle (Kang et al., 2005) Luteal

phase defects (defined by reduced corpus luteum

func-tion and/or luteal phase shortening) are associated with

energetic constraint (De Souza et al., 1998; Rosetta

et al., 1998; Williams et al., 1999; Warren and Perlroth,

2001; De Souza, 2003) and affect progesterone levels,

which may affect prostaglandin production Thus, luteal

maintenance of endometrial tissue is costly, and it is

likely impacted by energetic and more broadly ecological

variation (Strassmann, 1996b) The degree of withdrawal

of ovarian steroids, where a higher degree of withdrawal

is derived from having a higher concentration to begin

with, may also prove to be important in future research

on this topic

Implantation and invasion of the endometrium

Should an embryo successfully implant and produce

sufficient hCG to rescue the corpus luteum, ovarian

ste-roid withdrawal, prostaglandin production, and other

processes associated with menstruation do not occur

(Baird et al., 2003), and the endometrium changes to

prepare for implantation Progesterone concentrations

increase and are essential to the processes described

next In addition to corpus luteum rescue (Csapo and

Pulkkinen, 1978; Baird et al., 2003), subepithelial

capil-lary permeability increases to provide greater access to

maternal blood flow (Tabibzadeh and Babaknia, 1995)

Embryo implantation is then a tissue remodeling process

of adhesion and implantation similar to that found in

other processes of the body such as inflammation and

tu-mor invasion (Bischof and Campana, 1996; Bulletti and

de Ziegler, 2005) Implantation—paracrine cell-signaling

and adhesion—is one of the oldest processes in multi-celled organisms and a critical step in their development

is the ability for cells to communicate and adhere non-randomly For embryonic implantation, the embryo moves to the uterus, orients itself so that the inner cell mass is facing the endometrial lining, and dissolves its zona pellucida The embryo then apposes, adheres, and invades the endometrial epithelium At this point, troph-oblast syncytia (cell-like structure containing many nuclei) proliferate to invade the ECM of the endome-trium; the embryo digests its way through the ECM to implant, which best occurs when the cells are quiescent (rather than experiencing frequent or intense wavelike activity) (Beier and Beier-Hellwig, 1998) Finally, cytotrophoblastic cells migrate within the forged syncy-tia pathway, leading placental villi formation (Fig 1) (Bischof and Campana, 1996)

In preparation for the receptive period or ‘‘implanta-tion window,’’ the endometrium changes its adhesion molecule, cytokine, and key endometrial protein expres-sion (Tabibzadeh and Babaknia, 1995) The cytokines present during endometrial receptivity are leukemia in-hibitory factor (LIF) and the interleukins, especially interleukin-1 (IL-1) (Lindhard et al., 2002) These cyto-kines coordinate implantation with the embryo under the influence of sex steroid hormones (Lindhard et al., 2002) LIF and IL-1 also are present during inflamma-tory processes generically in the body, suggesting that implantation and inflammation are evolutionarily linked The apical plasma membrane of the surface epithelium

is non-adhesive until it is specifically altered during receptivity; then, the plasma membrane acquires the ability to form reflexive gap junctions, or targets where cells can attach (Tabibzadeh and Babaknia, 1995) On its surface, the endometrial epithelium forms pinopodes, which are secretory membrane elements (Tabibzadeh and Babaknia, 1995; Beier and Beier-Hellwig, 1998), and are important to adhesion of the embryo during implan-tation (Norwitz et al., 2001)

While described earlier as important during the perio-vulatory period, endometrial wave activity also plays a functional role in the implantation window In spontane-ous cycles, a quiescent endometrium in the midluteal phase and conception are associated While ‘‘outward’’ waves characterize the early to mid follicular and late luteal phases, and ‘‘inward’’ waves characterize the peri-ovulatory period, the implantation window tends to have the lowest wave activity (IJland et al., 1997) The picture

is a bit more complicated when wavelike activity in dif-ferent regions of the uterus during an IVF cycle is meas-ured; there, the uterus tends to have ‘‘inward’’ waves in the isthmocervical region (neck of the uterus) and ran-dom or opposing (both ‘‘inward’’ and ‘‘outward’’) waves in the corpus (IJland et al., 1999) In artificial cycles of women with mostly female-origin subfertility (78%), where wave activity has a greater amplitude and higher frequency than in spontaneous cycles, there is some indi-cation that the endometrium guides the embryo to the main body of the uterus and uses ‘‘inward’’ waves close

to the cervix to prevent embryo loss (IJland et al., 1999) The decidualization of the endometrium, its thickness, its wave activity, and its synthesis of a suite of cytokines and hormones together establish a specific, optimal envi-ronment for conception and implantation Following ini-tial invasion, the trophoblast sends additional cells responsible for further remodeling of the endometrial environment during the first trimester These cells

promote arterial reorganization to increase access to the

maternal blood supply, suppress immune function, and

signal to the endometrial glands to create the required

combinations of cytokines, nutrients and growth factors

for fetal nourishment through at least 10 weeks (Burton

et al., 2002; Hempstock et al., 2004)

Beneath the implantation site, the once-thick

endome-trium drastically thins to decrease the trophoblast’s

sep-aration from maternal energy (Hempstock et al., 2004)

Initially, the endometrium creates a hypoxic

environ-ment most suitable to early fetal growth (Jauniaux

et al., 2000, 2003a; James et al., 2006) As the first

tri-mester ends, the placenta takes over the nourishment of

the fetus and much of the endometrium’s activity ceases;

however, the endometrial glands continue to

communi-cate with the spaces between placental villi containing

maternal blood, which suggests they could continue to

provide additional nourishment or some other role

(Jau-niaux et al., 2003b)

The endometrium’s ability to provide a suitable

environment for conception, implantation, and early

ges-tation and placenges-tation relates critically to pregnancy

and fertility Insufficient endovascular invasion can lead

to hypertension, preeclampsia, and inadequate fetal

growth, whereas unrestricted trophoblast invasion can

lead to placenta accreta (when the placenta attaches

itself too deeply to the uterus), hydatidiform moles (mass

on the trophoblast that usually does not contain

tropho-blast cells), and choriocarcinoma (cancer germ cell

con-taining trophoblast cells) (Bischof and Campana, 1996)

Pathological trophoblast invasion is increasingly thought

to be a problem of the immune system and the

regula-tion of inflammatory processes (Norwitz et al., 2001;

Challis et al., 2009) These pathologies are not commonly

found in other animals; literature searches on typical

laboratory animals or nonhuman primates yielded

no results Next, I review the broad anatomical and

physiological differences in the endometria of primates

to highlight some of the adaptations particular to humans

Nonhuman primate endometria

Long follicular phases characterize primate reproduc-tive cycles and differentiate them from nonprimate ani-mals; these follicular phases include estradiol priming of the endometrium and dominant follicle (or follicles in some cases) development (Barnett and Abbott, 2003) Most primates give birth to singletons, with exceptions

in strepsirhines and, notably, the callitrichids within pla-tyrrhines (Harvey et al., 1987) After that, aspects of uterine and endometrial physiology diverge within the primates in at least four ways First, uterine type diverges: strepsirhines and tarsiers have bicornuate uteri where the uterus has two ‘‘horns’’ but is fused in its lower two-thirds leading to one cervix and vagina, and the rest of the haplorhines have unicornuate uteri with one body, cervix, and vagina (Gelder, 1969) Second, the type of arteries formed to support a fetus varies: strepsirhines and platyrrhines have straight arteries, where the arteries of the catarrhines have spiral arteries (Hernandez-Lopez et al., 1998) Third, strepsirhines do not menstruate visibly but most haplorhines do (all catarrhines and most platyrrhines), and menstruation generally increases in copiousness as one moves through these categories (Hrdy and Whitten, 1987; Strassmann, 1996a,b)

Forms of placentation are the fourth main way endo-metrial physiology of primates vary: while hemochorial placentation has been suggested as the ancestral form for eutherian mammals and for primates (Wildman et al., 2006), endotheliochorial placentation has also been suggested to be ancestral in primates (Martin, 2008) Strepsirhines and haplorhines diverged in their placen-tation types, where strepsirhines use epitheliochorial placentation and haplorhines use hemochorial (Martin,

Fig 1 The process of embryo implantation 1, Transport; 2, orientation; 3, hatching of the zona pellucida; 4, apposition; 5, adhe-sion; 6, invaadhe-sion; 7, syncytialization; 8, villous formation Please see the section entitled Implantation and Invasion of the Endome-trium for a more detailed description Reproduced from Bischoff P, Campana A 1996 A model for implantation of the human blas-tocyst and early placentation Human Reproduction Update 2(3):262–270, by permission of Oxford University Press.

141

REPRODUCTIVE ECOLOGY AND THE ENDOMETRIUM

2008); strepsirhines have less maternal–fetal contact,

with six cell layers dividing them, and haplorhines have

more contact, with only two cell layers; humans lose one

more cell layer by the third trimester of pregnancy

(Abitbol, 1990)

Several points need to be summarized here First, the

primate ancestral condition was likely invasive

placenta-tion, but strepsirhines evolved less invasive means to

grow fetuses Second, primates that give birth to

multi-ples have epitheliochorial placentation with the

excep-tion of the callitrichids, and have straight arteries

including the callitrichids Thus, epitheliochorial

placen-tation is in strepsirhines because it is more efficient for

the carrying of multiple fetuses Third, menstrual

copi-ousness increases with placentation invasiveness In

haplorhines, however, placental invasion only increased,

which suggests greater investment in their singletons

Primates have greater fetal brain growth than other

mammals (Martin, 1996), which may explain the variety

of attempts made to conserve maternal resources,

increase maternal–fetal contact, or otherwise find an

efficient means of negotiating the trade-offs between

current and future reproduction, and reproduction and

survival

More resources are then allocated to endometrial

growth and function, and to fetal growth, than in other

animals, with particular emphasis on the catarrhines

and hominoids The endometrium and its associated

reproductive processes have transformed significantly

across the primates, where more derived adaptations

indicate increased secretory mechanisms and increased

maternal–fetal contact In particular, human

endome-trial processes imply an increased embryonic and fetal

role in directing maternal energy, and an increased need

for maternal–fetal contact alongside the need for

protec-tion from maternal immunological defenses Martin

(1996) has suggested that mammals have the largest

brains they can in the context of maternal metabolic

resources during gestation and lactation; further, Martin

et al (2005) demonstrate relationships between basal

metabolic rate (BMR), gestation period, body mass, and

brain mass that suggest a trade-off between BMR and

gestation period in the development of relatively large

brains Thus, the close maternal–fetal contact found in

the human placenta and endometrium may be a way for

the fetus to take maximal advantage of maternal energy

for brain growth

Measuring endometrial function

Several methods exist to measure endometrial

func-tion and morphology; these different methods generate

different kinds of data that indicate different aspects of

endometrial functioning Asking women to record

dura-tion of menstrual bleeding is the least invasive; asking

them to rate their perceived menstrual copiousness is

also possible, and a pencil-and-paper scale has been

recently described (Mansfield et al., 2004) The difficulty

in creating universal agreement across subjects for what

constitutes menstrual copiousness impairs this method,

as well as the difficulty in determining how many days

one menstruates when the beginning and ending of

cycles does not occur at the same time of day, and the

‘‘end’’ of menses can be difficult for a subject to interpret

(for a review see Belsey and Farley, 1987) Further,

sub-jects often have some difficulty recalling menstrual cycle

dates, perhaps in part due to cultural discomfort with

this biological process (Roberts et al., 2002; Andrist

et al., 2004) The MVJ pencil-and-paper scale (Mansfield

et al., 2004) correlated with menstrual blood loss meas-ured from used sanitary napkins at r 5 0.683 which, while a strong relationship, may not be sufficient for testing mechanistic, physiological hypotheses

Menstrual fluid can be more precisely measured through a few different methods of collection, from weighing used sanitary pads collected in sealed plastic bags (Mansfield et al., 2004), to performing an alkalin-hematin method to assess blood in used sanitary pads involving soaking the used pads in solution and using a photometer to determine heme (Hallberg and Nilsson, 1964; Newton et al., 1977), and to collecting menses in a menstrual cup (Morrison and Brown, 2008) All these methods miss the menstrual blood that is lost through sanitary practices and urination Both the method of subject appraisal of menstrual blood loss and menstrual fluid collection assess the same thing: the amount of en-dometrial tissue and blood that was left at the end of a reproductive cycle; variation in this measurement could indicate the degree of endometrial proliferation, endome-trial maintenance after proliferation, or both

The most invasive method of assessing endometrial function is through an endometrial biopsy, which requires entry into the endometrium through the cervix and the collection of a small amount of endometrial tis-sue; this requires the most significant clinical support and can be uncomfortable for the participant This tissue can be tested for various molecular and biomarkers of endometrial activation and receptivity including gene (Riesewijk et al., 2003), pinopode (Nardo et al., 2002), and integrin expression (Thomas et al., 2003)

Transvaginal ultrasonography, which is ultrasound using an endovaginal probe, balances useful, quantita-tive information with comfort and invasiveness for sub-jects Abdominal sonography does not yield consistent enough results in assessment of nonpregnant reproduc-tive organs in humans (though it is sufficient for smaller primates), so while transvaginal ultrasound may at first seem more daunting, it is the more reliable and compa-rable method, as it is used clinically for diagnostic and research purposes Abdominal sonography also tends to require a full bladder to adequately view reproductive organs; this can be more time-consuming and uncomfort-able than transvaginal ultrasound Transvaginal ultrasound can be measured multiple times, even daily, during a menstrual cycle, which allows observation of changes in endometrial thickness In one survey of women, though they anticipated significant discomfort before experiencing transvaginal ultrasound, they found

it significantly less uncomfortable than mammography and Pap smears (Kew et al., 2004)

Transvaginal ultrasound makes possible the measure-ment of endometrial thickness, endometrial pattern, the functionalis/basalis ratio, and endometrial volume Endo-metrial thickness is the double thickness measurement of the endometrium on the sagittal plane at its widest point Endometrial pattern is an assessment of the degree of echogenicity of the endometrium, which is thought to reflect the degree of decidualization and receptivity of the tissue Endometrial thickness and pattern in particular are useful assessments of morphology, both because of their frequency in the literature and the relationships that have been found between these measurements and pregnancy success; results are highly reproducible between sonographers (Epstein and Valentin, 2002)

ENDOMETRIAL FUNCTION VARIATION

In a fecund cycle, the endometrium proliferates in the

follicular phase, and then is generally assumed to be

maintained at about the same thickness while it

differ-entiates through the luteal phase (Johnson and Everitt,

1988; Baerwald and Pierson, 2004) This means there

are two main ways endometrial thickness can vary: in

the degree of follicular proliferation, and in the degree of

luteal maintenance Most endometrial thickness studies

are in assisted reproduction, when it is measured at the

time of hCG injection or ovum pickup, which

approxi-mates midcycle in a natural cycle; this means current

literature only has information on potential variation in

follicular endometrial proliferation Any variation in the

endometrium through the window of implantation is

thus not assessed, when its functioning is most relevant

to achieving pregnancy What follows is a review of the

recent literature Most work was carried out in medical

settings, and thus the populations are categorized

clinically: normo-ovulatory women, women undergoing

assisted reproductive treatment, postmenopausal

women, and women with endometrial pathology

Normo-ovulatory women

Menstrual bleeding duration was used as a biomarker

for endometrial function in a study that examined

ener-getic correlates to variation in reproductive functioning

Duration of menstrual bleeding was shorter in the

pre-harvest hunger season than the pre-harvest season in Lese

women (Bentley et al., 1990) Cycle length, but not

dura-tion of menstrual bleeding has been shown to vary with

work-related physical activity in US workers (Sternfeld

et al., 2002) And in a subcohort of women from that

study who had participated in the Michigan Bone Health

Study, recreational physical activity was negatively

asso-ciated with duration of menstrual bleeding (Sternfeld

et al., 2002)

Other studies examined endometrial function using

endometrial thickness as its proxy Clancy (2007a,b)

used a single luteal endometrial thickness measurement,

and found that mean endometrial thickness did not

dif-fer between urban US and rural agricultural Polish

women sampled, and that endometrial thickness was

dependent on luteal phase day in Polish women but not

US women Endometrial thickness was positively

corre-lated with C-peptide concentrations (a biomarker for

in-sulin) and negatively correlated with age in the Polish

sample (P 5 0.05 and P 5 0.04, respectively); a negative

trend was found with energy expenditure calculated in

METs (kcal/min) (P 5 0.09) (Clancy et al., in press)

Only one journal article addressed breastfeeding women,

and it found that recently postpartum breastfeeding

women have less endometrial activity as assessed by

en-dometrial pattern than those women who bottle-feed

their infants (Freedman et al., 1976)

The richest data on endometrial function study

nor-mal, spontaneous cycles, using transvaginal ultrasound

to measure endometrial thickness repeatedly throughout

the cycle These data provide longitudinal information to

determine population variation in endometrial

prolifera-tion and maintenance across the cycle Cycle-long

stud-ies of endometrial thickness in natural menstrual cycles

exist on populations in economically developed countries

(Canada, Sweden, the UK), and they align their subjects’

data by ovulation day (Randall et al., 1989; Bakos et al.,

1994; Baerwald and Pierson, 2004; Raine-Fenning et al., 2004) These data are briefly described below and illus-trated (Table 1 and Fig 2)

Ovarian development occurs in waves several times through the menstrual cycle, with waves defined as a group of follicles growing synchronously; most women have two or three waves per cycle (Baerwald et al., 2003) Baerwald and Pierson (2004) measured endome-trial thickness, area, volume, and pattern in Canadian women to test their hypothesis that women with differ-ent follicular wave patterns would exhibit differdiffer-ent endo-metrial dynamics through the menstrual cycle They found endometrial thickness increased earlier during the follicular phase in women with two over three waves, and within women with two waves increased earlier in women with major (a dominant follicle was selected) ver-sus minor (no dominant follicle selection detected) waves, and no differences were found in these groups during the luteal phase (Baerwald and Pierson, 2004) Follicular waves did not appear to impact luteal phase endometrial thickness, and these groups were pooled for the following analysis They described a plateau in endo-metrial thickness during the luteal phase that lasts until just before menses, but their data demonstrate noticea-ble variation: while the first few days after ovulation do remain constant, there is a visual drop in endometrial thickness 4 days after ovulation and then a second pla-teau that lasts until day 12, with statistical analysis of this variation forthcoming (Clancy et al., in preparation) Bakos et al (1994) described changes in the endome-trium through the menstrual cycle in 16 Swedish women

to demonstrate the usefulness of sonography in sponta-neous and artificial cycles They found a positive rela-tionship between estradiol and endometrial thickness when the entire follicular phase was analyzed, but not when analyzing the late follicular phase (Bakos et al., 1994) Endometrial thickness varied significantly between women and displayed a similar two-plateau effect found in the luteal phase of the Canadian sample,

at a slightly higher overall thickness, though the quali-tative, rather than statistical, quality of this analysis must be stressed

In a sample of English women, endometrial thickness did not appear to change appreciably through the luteal phase (Raine-Fenning et al., 2004) These subjects were measured every 4 days, and the lower measurement fre-quency may account for the lack of variation found Ran-dall et al (1989) measured estradiol and endometrial thickness in three groups of Scottish women trying to conceive: women with unexplained infertility, normal women with male factor infertility, and women with tubal occlusion The results from the normal women are described here Estradiol and endometrial thickness pos-itively correlated, and endometrial thickness increased through the luteal phase; however, luteal measurement frequency was only every 5 days (Randall et al., 1989) Aligning cycles at ovulation rather than at the end of the cycle provides information about endometrial thick-ness in the follicular phase and early in the luteal phase, which supplies important evidence about the influence of estradiol on endometrial thickness Through the luteal phase, endometrial thickness appears to vary more sig-nificantly, but the presentation of the data make quanti-tative assessment challenging: aligning at ovulation allows for comparisons around ovulation, but as women even in the same population experience wide variation

in luteal phase length the decline in endometrial

thick-143

REPRODUCTIVE ECOLOGY AND THE ENDOMETRIUM

ness cannot be adequately assessed this way Future

work will assess the original data in a way that allows

for better assessment of the late luteal phase, through

the alignment via days before menses, an alignment

typ-ical to the study of population variation in progesterone

The variation in mean endometrial thickness in these

populations—from 6 to over 13 mm—suggests the

capacity for significant variation in normal,

premeno-pausal ovulatory cycles

Assisted reproductive treatments

In vitro fertilization (IVF) studies demonstrate

rela-tionships between the endometrium, hormone

concentra-tions, and pregnancy, though the hormonal

manipula-tions of these cycles can make it difficult to apply the

findings to normal, spontaneous cycles Endometrial

thickness (ET) has been shown to positively correlate

with implantation rate after IVF (Noyes et al., 1995;

Kovacs et al., 2003; Zhang et al., 2005) In controlled

ovarian hyperstimulation cycles where estradiol is often

four times the physiologically normal concentration,

estradiol and endometrial thickness have been positively

correlated, though only 6% of the variation in

endome-trial thickness could be explained by estradiol

concentra-tions (Zhang et al., 2005) Through the exogenous

hor-mones administered that prepare the endometrium for

implantation, in vitro fertilization also affects

endome-trial receptivity and maturation in the luteal phase

(Kolb and Paulson, 1997; Tavaniotou et al., 2001), which

confirms a link between hormone concentrations and

en-dometrial function However, these data suggest that the

relationship between ovarian hormones and endometrial

function, while certainly crucial, is not a strict dose–

response model, meaning that incremental increases in

hormones do not necessarily correlate to an equal

increase in endometrial function

Assisted reproductive technologies research is a

con-tradictory array of information: some studies say

endo-metrial thickness bears no relationship to achieving a

successful pregnancy (Bassil, 2001; Dietterich et al.,

2002; Kolibianakis et al., 2004), another says increasing

endometrial thickness decreases chances of pregnancy

success (Weissman et al., 1999), and still other studies

suggest that endometrial thickness is positively

associ-ated with pregnancy success (Oliveira et al., 1993; Noyes

et al., 1995; Kovacs et al., 2003; Zhang et al., 2005) A

table is provided to demonstrate some of the main

differ-ences in methods and results in a representative sample

of the ART literature on endometrial thickness (Table 2)

Three main issues explain these different results: 1)

sta-tistical and grouping factors, 2) assisted reproductive

method used, and 3) calculation of success of the ART

method (i.e., successful implantation, chemical

preg-nancy, gestational sac, fetal heartbeat, ongoing

pregnancy, live birth)

In terms of statistical methods, some articles

com-pared endometrial thickness means between groups of

pregnant and not pregnant women, while others com-pared pregnancy rates between groups of high and low endometrial thickness This was often determined by the question the authors were asking; for instance, for those concerned that an artificially-induced thick endometrium (from hyperstimulation via exogenous hormone adminis-tration) could reduce pregnancy rates, the method was

to group according to endometrial thickness, usually above or below 14 mm Most of the articles that found that endometrial thickness had a positive relationship with pregnancy rates grouped subjects by their preg-nant/nonpregnant state rather than their endometrial thickness; thus, a threshold endometrial thickness prob-ably does not exist over or under which pregnancy is unlikely Another factor that complicates interpretations

of this literature are the different methods used to achieve pregnancy; for instance, clomiphene citrate stim-ulates ovulation but has been found to reduce endome-trial thickness (Randall and Templeton, 1991), whereas GnRh agonists are likely to impact endometrial thick-ness Comparisons of these results are challenging because the degree of exogenous stimulation is so differ-ent Finally, authors defined a successful outcome as chemical pregnancy (positive hCG test), clinical preg-nancy by ultrasound (gestational sac or heartbeat), or even ongoing pregnancy (pregnancy for at least 20 weeks) Sometimes subjects were in the pregnancy cate-gory even if they eventually miscarried, so long as they hit the milestone that study defined as successful (Richter et al., 2007), and sometimes the authors did not know the ultimate outcome of all the pregnancies of the included subjects (Kovacs et al., 2003)

Despite these methodological differences, the ART lit-erature has a lot to offer reproductive ecologists Because

TABLE 1 Characteristics of natural cycle endometrial thickness studies from Clancy et al (in press)

measurement day

Fig 2 Results of four studies of natural luteal phase endo-metrium (Sweden: Bakos et al., 1994; Canada: Baerwald and Pierson, 2004; England: Raine-Fenning et al., 2004; Scotland: Randall et al., 1989) Studies were aligned by ovulation day, with the knowledge that ovulation (as confirmed by ultrasound)

is 24 h after the LH surge Error bars were omitted Reproduced from Clancy KBH, Ellison PT, Jasienska G, Bribiescas RG.

2009 Endometrial thickness is not independent of luteal phase day in a rural Polish population Anthropological Science DOI: 10.1537/ase.090130.

endometrial measurements are standard procedure for

most ART, and the subjects have a stake in the outcome

and tend not to miss appointments, many authors have

successfully measured a large volume of cycles

retrospec-tively and prospecretrospec-tively So even though ART cycles are

exogenously stimulated, relationships between

endome-trial thickness and different calculations of implantation

or pregnancy can still inform our understanding of

vari-ation in endometrial function Thus, what these articles

together suggest is that a thicker endometrium largely

improves the outcome for ART, but that those that are

very weak or very strong responders to ART (with a

very thin or thick endometrium) may have less success

Other factors documented in ART research important

to pregnancy are differentiation of the endometrium,

en-dometrial pattern (Coulam et al., 1994; Sharara et al.,

1999), uterine contractility or endometrial waves (IJland

et al., 1996, 1997, 1999), and molecular indicators of

re-ceptivity (Paulson et al., 1990; Lessey et al., 1996; Beier

and Beier-Hellwig, 1998; Lessey, 2000; Lindhard et al.,

2002; Cavagna and Mantese, 2003) This body of

research implies that endometrial thickness, waves,

pat-tern, and receptivity are all relevant to achieving

preg-nancy, at least in stimulated cycles Endometrial

thick-ness, waves, and pattern are measured with noninvasive

transvaginal ultrasound; molecular indicators of

recep-tivity require a more invasive endometrial biopsy And while it is obvious that hormonal concentrations influ-ence endometrial proliferation and decidualization, these data do not resolve whether this relationship is one of a threshold model (where a threshold hormone concentra-tion produces an effect), a dose–response model (where increasing hormone concentrations produce increasing effects), or whether other factors additionally influence endometrial variation

Postmenopausal women

While literature on natural and artificial cycles rarely includes lifestyle or energetic information, other data on postmenopausal women and other study populations indicate that endometrial thickness varies with energy availability in a dose–response model (Shu et al., 1992; Douchi et al., 1998; Iatrakis et al., 2006) The postmeno-pausal endometrium is no longer influenced by active ovaries, and yet it varies with energy status Research-ers have found a positive relationship between BMI and endometrial thickness (Andolf and Aspenberg, 1996; Douchi et al., 1998), body weight and endometrial thick-ness (Andolf and Aspenberg, 1996), and obesity and endometrial thickness (Serin et al., 2003) A positive relationship between endometrial thickness and energy

TABLE 2 Representative publications on endometrial thickness from the ART literature

Number of subjects/cycles

Age of subjects (years)

Number of embryos

Long GnRH,

short GnRH,

GnRH antagonist

ET higher in conception cycles

in women under

35 yrs; age negatively correlated with ET

ET in conception

vs non-conception cycles

Long GnRH,

IVF, and ICSI

High and low ET

in nonconception cycles; not ss

that achieved pregnancy independent of age

rate positively associated

range 23-44

Follicular LA

with GnRH,

luteal LA

Women with ET over/under 14 mm had similar clinical pregnancy rates

[9 mm had higher implantation, clinical pregnancy, and ongoing pregnancy rates

477 subjects,

516 cycles

that achieved pregnancy independent of age

pregnant, 33.1

in not pregnant subjects

Mean 2.9 in pregnant, 2.6

in not pregnant subjects

Kovacs et al., 2003

women with ongoing pregnancies or

no pregnancies

GnRH, gonadotropin releasing hormone; IVF, in vitro fertilization; ICSI, intracytoplasmic sperm injection; ss, statistically signifi-cant; ET, endometrial thickness; LA, leuprolide acetate; CC, clomiphene citrate; IUI, intrauterine insemination.

145

REPRODUCTIVE ECOLOGY AND THE ENDOMETRIUM

status in postmenopausal women could indicate

aromati-zation of androgens to estrone by adipose tissue, insulin

and IGF-1 action, or both, on the endometrium Thus,

both an indirect pathway via steroids (estrone action

implicates this indirectly) or a direct pathway (via

insu-lin) are possible

Endometrial pathology

Obesity and postmenopausal endometrial cancer risk

positively correlate (Shu et al., 1992; Gull et al., 2001;

Kaaks et al., 2002; Lukanova et al., 2006; Setiawan

et al., 2006; Xu et al., 2006) Further, endometrial

thick-ness and BMI are positively correlated in recovering

anorectics (Andolf et al., 1997) These data further

sug-gest relationship between energy status and functioning

Women who experience multiple spontaneous

miscar-riages have not had their endometrial function explicitly

measured, but several aspects of this population imply

an endometrial origin to pathology Choriodecidual

inflammatory syndrome is a main cause of early preterm

delivery and second trimester miscarriage (Sebire, 2001),

and this and other inflammatory syndromes are

associ-ated with undiagnosed and untreassoci-ated gluten intolerance

(Rostami et al., 2001), which over time promotes

sys-temic inflammation As many endometrial processes are

inflammatory, it may be important in future research to

examine inflammation and immune function in the

con-text of the endometrium, as another important aspect of

ecology

Ovarian hormones, insulin and inflammation pull out

as the most relevant factors that produce variation in

endometrial function in the literature, with independent

and interrelated actions documented Ovarian hormones

are often the bearers of ecological information, insulin

can also inform on energy availability, and inflammation

can be produced by immunological or psychosocial stress

This article focuses on these factors for its remainder

ENDOMETRIAL FUNCTION AND REPRODUCTIVE

ECOLOGY

Several hypotheses have been suggested in the last

few decades to explain menstruation and endometrial

cy-clicity These hypotheses fall into three major categories:

menstruation as a cleansing process, energetic

explana-tions of menstruation, and physiological explanaexplana-tions for

menstruation Early researchers attempted to isolate an

elusive compound they called the ‘‘menotoxin,’’ a toxic

substance secreted in a menstruating woman’s sweat

that could cause harm to male babies and cut flowers

(Macht, 1924; Freeman et al., 1934; Macht and Davis,

1934; Davis, 1974; Reid, 1974; Bryant et al., 1977;

Pickles, 1979) As problematic as that initial work was,

the idea that menstruation cleanses the body persisted,

perhaps because of the strength of widespread cultural

beliefs in this purpose (Montgomery, 1974; Whelan,

1975): later work focused on the elimination of unwanted

embryos (Clarke, 1994) and sperm-borne pathogens

(Profet, 1993) Strassmann (1996a,b) and Finn (1996,

1998) offered alternatives to these ideas, with their

hypotheses of energy economy and terminal

differentia-tion, respectively Strassmann (1996b) suggested that it

was more costly to maintain the endometrium from cycle

to cycle, and that menstruation evolved to reduce the

energetic costs of fecundity Finn (1998) argued that

menstruation is a necessary consequence of the terminal

differentiation of endometrial tissue that occurs after es-tradiol priming and progesterone action (de Ziegler et al., 1998); the endometrium must start over once the tis-sue has differentiated beyond a point at which it can proliferate for the next cycle

Finn’s hypothesis gains support in the light of the physiology and comparative primatology of the endome-trium: the variability in endometrial and placental archi-tecture, and the particular architecture of human placentation demonstrate the specialized tissue the endo-metrium becomes in preparation for implantation Decidualization is a process that cannot be reversed, and

so endometrial tissue must be removed if it is to prolifer-ate again Further, the endometrium is maximally recep-tive through an implantation window in the luteal phase, after which implantation is unlikely regardless of embryo quality or stage Strassmann’s hypothesis loses support because of this, but also from evidence sur-rounding ecological variation in endometrial function described in the previous section of this article Terminal differentiation and menstruation’s other important pur-pose allows endometrium to respond to the ovaries and ecology from cycle to cycle; without this, the endome-trium could not respond to changing ecological conditions Strassmann’s important contribution to reproductive ecology is attention on energetics and the endometrium, without which the ecology of the endome-trium and the primary topics of this review might never have been explored Therefore, the following section will focus on ecology and endometrial function, synthesizing the existing literature, and describing new directions for research in reproductive ecology

While the relationship between ovarian function and reproductive success is obvious, the mechanisms that link them are not Inter and intrapopulation variation in ovarian steroids has been consistently documented (i.e., Ellison and Lager, 1986; Bledsoe et al., 1990; Lager and Ellison, 1990; Bentley et al., 1998; Jasienska and Elli-son, 1998; Rosetta et al., 1998; Warren and Perlroth, 2001; Vitzthum et al., 2002; Nu´n˜ez-de la Mora et al., 2007) These data demonstrate relationships of energy expenditure (Ellison and Lager, 1986; Bledsoe et al., 1990; Rosetta et al., 1998; Warren and Perlroth, 2001), energy balance (Lager and Ellison, 1990), nutritional status (Bentley et al., 1998), and developmental condi-tions (Vitzthum et al., 2002; Nu´n˜ez-de la Mora et al., 2007) with ovarian hormones

Combine the data demonstrating a relationship between energy and immunity and the endometrium, energy and ovarian hormones, estradiol concentrations and rates of pregnancy in ovulatory cycles, and estradiol and endometrial function, and it becomes clear that ovarian and endometrial function are important compo-nents of fertility that must be studied together in repro-ductive ecology Because the endometrium is a target tissue of ovarian steroids, it is the next place to look to better explain aspects of fecundity and fertility that remain unclear with ovarian function alone In particu-lar, endometrial function plays a role in variation in fe-cundity and fertility via variation in endometrial thick-ness and pattern, as well as variation in implantation rates and early fetal loss Ecology, ovarian function, and age are likely the prime determinants of endometrial and more general reproductive variation, though genetic variation is as yet largely unstudied and may also prove important Because the endometrium has a strong role

in implantation and early gestation, fetal loss is also of