Maternal obesity, metabolic disease, and allostatic load

Physiology & Behavior 106 (2012) 22–28

Contents lists available at SciVerse ScienceDirect

Physiology & Behavior journal homepage: www.elsevier.com/locate/phb

Maternal obesity, metabolic disease, and allostatic load Michael L. Power a, b,⁎, Jay Schulkin a, c a b c

Department of Research, American College of Obstetricians and Gynecologists, 409 12th St SW, Washington DC 20024, United States Conservation Ecology Center, Smithsonian Conservation Biology Institute, National Zoological Park, Washington DC 20008, United States Department of Neuroscience, Georgetown University, Washington DC 20057, United States

a r t i c l e

i n f o

Article history: Received 11 May 2011 Received in revised form 6 September 2011 Accepted 7 September 2011 Keywords: Maternal obesity Placenta Adipose Metabolism

a b s t r a c t Maternal obesity is a risk factor for many metabolic diseases for the mother, both during gestation and post partum, and for the child in later life. Obesity and pregnancy both result in altered physiological states, significantly different from the state of the non-obese, non-reproductive adult female. The concept of allostasis may be more appropriate for understanding the physiology of both pregnancy and obesity. In pregnancy these altered physiological states are adaptive, in both the evolutionary and physiological senses of the word. Obesity, however, represents a state outside of the adaptive evolutionary experience of our species. In both cases the altered physiological state derives at least in part from signals from an active endocrine organ. In obesity this is adipose tissue, and in pregnancy it is the placenta. The signaling molecules from adipose tissue and placenta all have multiple functions and can affect multiple organ systems. Placenta acts as a central regulator of metabolism for both the maternal and fetal compartments, in essence acting as a “third brain” during pregnancy. Both adipose tissue and placenta express many proinflammatory cytokines; obesity and pregnancy are states of low-grade inflammation. Both obesity and pregnancy are also states of insulin resistance, and maternal obesity is associated with fetal insulin resistance. We argue that obesity during pregnancy leads to sustained and inappropriate activation of normally adaptive regulatory circuits due in part to competing and conflicting signaling from adipose tissue and placenta. This results in allostatic load, leading to the eventual break down of regulatory mechanisms. The result is impaired metabolic function of the mother, and altered development of metabolic systems and potentially altered neural appetite circuits for the offspring. © 2011 Elsevier Inc. All rights reserved.

1. Introduction Metabolic diseases (e.g. obesity, type 2 diabetes, hypertension and associated cardiovascular disease) have proportionately increased in the modern human population. Partly this is due to the decline in infectious disease and increases in life expectancy caused by modern medical advances; but the interaction of human biology with the modern human environment appears to have led to an absolute increase in metabolic disorders as well. Not only are metabolic diseases increasing in prevalence, they also are being diagnosed at earlier ages. Diseases that were associated with middle-to-older age are becoming increasing common in younger adults and even children. Our modern environment appears to result in a mismatch with many of our evolved, adaptive systems [1, 2] that increases the risk of metabolic disease. Metabolic diseases, such as diabetes and hypertension, represent conditions where the regulatory physiological processes have broken down. We argue that a true understanding of the pathophysiology of the metabolic syndrome requires an integration of physiology with an evolutionary perspective. Normal, adaptive metabolic responses are, in

⁎ Corresponding author. Tel.: + 1 202 314 2326. E-mail address: [email protected] (M.L. Power). 0031-9384/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2011.09.011

effect, causing disease because these metabolic responses are inappropriate, ineffective, or merely sustained for an excessive period of time due to environmental inputs that are outside of our historical norms. Metabolic syndrome is a disease of chronic insult, not acute insult. The concept of allostatic load [3] is a framework in which to understand the pathophysiology of metabolic diseases that generally do not arise due to any acute incident, but rather as an accumulating degradation of physiological function. In this essay we examine the concepts of homeostasis, allostasis, and allostatic load in the context of understanding metabolic disease. We focus on what is unfortunately becoming an increasingly significant source of human metabolic disease: maternal obesity. Obesity during pregnancy results in an increase in metabolic disorders for the mother, and also in altered metabolic development for her offspring, leading to later development of metabolic disease for both. 2. Homeostasis, allostasis and regulatory physiology Homeostasis is a simple but powerful concept. The internal milieu must remain constant, despite external challenges. In the words of Cannon [4]: “…the organs and tissues are set in a fluid matrix… So long as this personal, individual sack of salty water, in which each one of us lives and moves and has his being, is protected from change,

M.L. Power, J. Schulkin / Physiology & Behavior 106 (2012) 22–28

we are freed from serious peril.” The concept of stability, of resistance to change is fundamental to homeostasis. Restraint, negative feedback, is the hallmark of homeostatic processes. However, homeostasis is not synonymous with regulatory physiology, and stability is perhaps a misleading word when considering evolved physiological adaptations. Much within the “internal milieu” is constantly changing and adapting. Some of the changes are programmed, such as circadian or seasonal rhythms, or the physiological changes associated with pregnancy and lactation. Others are acute responses to challenges. Many are anticipatory rather than reactive, occurring before the need has arrived. This is not an inherent contradiction of homeostasis; Cannon [4] himself quoted Richet “We are only stable because we constantly change.” But the concept of homeostasis must either be stretched, or other terms and concepts must be added to the lexicon of physiological regulation. Physiological systems serve the survival and reproductive capabilities of the organism (fitness). Stability is not the currency of evolutionary success. We argue that viability, defined as the capability of success or ongoing effectiveness, is a better concept. In an evolutionary context, this means the ability to pass on genetic material. Physiological regulation to maintain viability requires regulation of set points under some conditions, and abandonment of set points under others. There must be physiological processes that are not homeostatic, and that oppose, at least temporarily, stability. The concept of allostasis was proposed to account for regulatory systems that appeared to fall outside of the classic concept of homeostatic processes [5, 6]. For example, regulatory systems in which there are varying set points or no obvious set points at all (e.g. fear), and/or where the behavioral and physiological responses are anticipatory, and do not simply reflect feedback from a monitored parameter. Sterling [7] includes an adaptive perspective, a focus on anticipatory physiology, and central nervous system involvement as inherent in the theory of allostasis. We agree that those concepts are fundamental to understanding regulatory physiology; however, none are foreign to homeostasis. Homeostasis and allostasis can be considered as complementary components of physiological regulation. Homeostatic processes maintain/regulate physiology around a set point, and allostatic processes change the state of the animal, including changing or abandoning physiological set points. Homeostatic processes are associated with negative restraint and resistance to perturbations. Allostatic processes are associated with positive induction, perturbing the system, and changing the animal's state. We would modify Sterling and Eyer's [6] original definition of allostasis from “achieving stability through change” to “achieving viability through change”, and define homeostasis as “achieving viability through resistance to change”. Thus, neither should have primacy in regulatory physiology. 3. Allostatic load The concept of allostasis originally derived from the attempt by Sterling and Eyer [6] to explain patterns of modern morbidity and mortality with socioeconomic circumstances considered to be stressful. Boulos and Rosenwasser [8] state that “…implicit in Sterling and Eyer's model is the notion that allostatic regulation entails a price…” Although Sterling [6], with his emphasis on the adaptive nature of allostatic regulation, likely would disagree with that statement, many researchers studying the allostatic perspective appear to accept it. McEwen [3] specifically extended the concept of an allostatic state to regulatory systems that were vulnerable to physiological overload, with the resulting development of disease. Many regulatory systems evolved to respond to acute or at least shortlived challenges to viability. Their activation generally results in a change of state that alleviates the challenge and thus allows the regulatory response to cease, at least temporarily. Sustained activation of these regulatory circuits is outside of the evolutionary experience. If these regulatory systems are chronically activated, either due to competing

23

imperatives, conflicting signals, or the failure of the regulatory physiology to resolve the challenge, the associated costs can accumulate and lead to a lessening of health. The term allostatic state refers to chronic activation of regulatory systems either due to dysregulation/dysfunction of physiology or to conflicting, competing, or opposing demands [9]. The term allostatic load refers to the strain on physiology and regulatory capacity due to sustained activation of regulatory systems. However, allostasis and allostatic regulatory mechanisms do not imply pathology to any greater extent than does homeostasis. The activation of regulatory systems can have both short and long-term metabolic costs, and as such can lead to pathology under certain circumstances; but under many if not most circumstances they result in physiological adaptation that maintains health. One other mild criticism of the term allostatic load is that there is no particular reason why the regulatory circuits being chronically activated have to be allostatic in nature [10]. Homeostatic systems that become chronically activated due to sustained external pressures, or simply because the regulatory system is failing to achieve the desired homeostatic state, will potentially have the same general long-term consequences, slowly degrading physiology and lessening health. Allostatic load might be more properly termed regulatory load or metabolic load. The basic concept rests upon the fact that any regulatory system that remains continuously activated or activated to a level outside of its norm will eventually break down. The example we have chosen to examine in this essay is maternal obesity, and its resulting effects on maternal and child health. Both obesity and pregnancy can be considered allostatic states, and the combination increases the vulnerability to allostatic overload, as we will argue below. 4. Maternal obesity and metabolic disorders Obesity during pregnancy has dramatically increased in the United States [11]. Obesity during pregnancy is associated with greater mortality and morbidity of both the mother and her child [12, 13]. Obese women experience a higher risk for a number of pregnancy complications, including hypertension, preeclampsia, and gestational diabetes mellitus (GDM) [14]. Excessive weight gain during pregnancy and postpartum retention of pregnancy weight gain are significant risk factors for later obesity in the mother [15]. Additionally, maternal metabolic disease can have a significant impact on the in utero environment, and thus on fetal development and possibly the future risks of metabolic disease of the child later in life [16]. Women with a BMI b 30 have an odds ratio of 1.0 for acquiring GDM, whereas women with a BMI of 30–34.9 have an odds ratio of 2.6 and women with BMIs ranging from 35 to 39.9 kg/m 2 have an odds ratio of 4.0 [17]. One meta-analysis [18] of twenty studies on maternal obesity and risk of gestational diabetes mellitus (GDM) showed that when compared to normal-weight controls, the unadjusted odds ratios of developing GDM were 2.1, 3.6, and 8.6, among overweight, obese, and severely obese pregnant women. A recent study [19] found that weight gain, even at a rate of 1.1 to 2.2 kg per year in the five years prior to becoming pregnant, can increase the risk of acquiring GDM. This was especially true for women who were not initially overweight. Catalano and Ehrenberg [20] compared the changes in maternal insulin sensitivity throughout pregnancy in women with a pregravid BMI b 25, 25–30 and N30. In late pregnancy, all women have decreased insulin sensitivity, however at any time period obese women have lower insulin sensitivity than women of normal weight. There has been a significant increase in childhood and adolescent obesity in developed nations [21]. In the US both mean maternal weight and mean birth weight have been steadily increasing [22] (Fig. 1). Maternal obesity is a strong predictor of childhood obesity [23], higher fetal fat mass [24], fetal insulin resistance [25], and eventual metabolic syndrome in her offspring [26]. Overweight/obese

24

Maternal weight at delivery (lbs)

a

M.L. Power, J. Schulkin / Physiology & Behavior 106 (2012) 22–28

190

185

180

175

170

165 1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

Year

b

3340

Mean birth weight (grams)

3320 3300 3280 3260 3240 3220 3200 3180 1975

1980

1985

1990

1995

2000

2005

Year Fig. 1. Maternal weight at delivery (a) and infant birth weight (b) have steadily increased at Metro Health Medical Center, Cleveland, Ohio. Data from Catalano et al. [22].

mothers have decreased insulin sensitivity, which results in increased lipid availability for fetal growth. Gene microarray analysis profiles of tissue from placentas from women who are overweight and have GDM show increased gene expression related to lipid metabolism as opposed to glucose metabolism [27]. Maternal adiposity predicts infant fat mass, but not infant lean mass [24], indicating that maternal obesity changes the quality and pattern of growth as opposed to merely increasing growth. Although increased risk of macrosomia is more strongly linked to maternal weight than diabetes status, GDM is related to significantly increased fat mass and body fat percentage in neonates, and significantly larger skin folds at all areas of measurement (tricep, subscapular, flank, thigh, abdomen) [28].

comes from adipose tissue [30]. Adipose tissue expresses aromatase, 3α-Hydroxysteroid hydrogenase type 3 (3α-HSD3) and 17β-HSD5, which are involved in androgen metabolism. These enzymes are increased in obesity [31]. Adipose tissue also expresses 11β-HSD1 which converts cortisone to cortisol [32] and 5α-reductase enzymes [31, 33] which convert cortisol to 5α-tetrahydrocortisol (5α-THF). Thus adipose tissue regulates the local concentrations of glucocorticoids [33, 34] and contributes to metabolic clearance of glucocorticoids [35]. Obesity is associated with both increased adrenal glucocorticoid production and higher glucocorticoid metabolic clearance, which appears to result in normal plasma concentrations. In obese individuals, 11β-HSD1 activity is reduced in liver and the inactivation of cortisol by 5α-reductase is enhanced [35, 36]. However, 11β-HSD1 activity is enhanced in adipose tissue of both obese men and women [35, 37]. Obese individuals have increased hepatic inactivation of cortisol, which is generally balanced by increased regeneration of cortisol in adipose tissue. Production of cortisol from cortisone via 11β-HSD1 can make a significant contribution to both local and circulating cortisol concentrations. The effect appears stronger in women compared to men [35], possibly due to the higher fat mass in women for a given BMI. The placenta is also an endocrine organ intimately involved in regulating metabolism, signaling to both maternal and fetal tissue, and regulating metabolism, growth and development of both. The placenta serves as a central regulator of maternal–placental–fetal metabolism and produces perhaps the broadest array of information molecules (hormones, cytokines, and all other classes of signaling molecules) of any other organ except for brain [38, 39]. The developing fetal brain is incapable of performing all of the necessary regulatory functions. Instead, signals from maternal organs passed through the placenta and signals generated by placenta assume many of these functions. Placenta also signals to maternal organs, including brain, in order to coordinate maternal physiology to benefit the growing fetus. There is a significant placental–brain axis for both fetal and maternal physiology. The late, eminent reproductive endocrinologist Samuel Yen referred to the placenta as the “third brain” in pregnancy [40] in recognition of the regulatory nature of placental function. In many ways pregnancy may be the canonical normally occurring allostatic state for human beings. A new organ (placenta) is developing and maturing, and taking over many of the regulatory functions from maternal tissue, including brain. For example, growth hormone secretion from the maternal pituitary essentially shuts down by the second trimester, with a placental form of growth hormone taking its place [41, 42]. A key brain function is to allocate resources within the body in order to defend its own viability [43]. Oxygen and glucose flow to brain is tightly regulated compared with the flow to other organs. Placenta also acts to allocate resources, in this case from the maternal to the fetal compartment. A well regulated flow of oxygen and glucose is vital to healthy placental function and fetal growth and development as well. Regulatory control is fundamentally altered during pregnancy, in part because metabolism and physiology must account for two organisms instead of one; but also because the regulatory signaling has been allocated to maternal brain and placenta. Many physiological set points are adjusted, often many times, during gestation. For example, circulating levels of glucose, insulin, cortisol and leptin all increase. Pregnant women are termed hyperglycemic,

5. Adipose tissue, placenta, and endocrine function Adipose tissue is not just a passive fat-storing organ. Besides the lipid-containing adipocyte, adipose tissue is comprised of connective tissue containing stroma-vascular cells and macrophages. Adipose tissue produces and secretes numerous cytokines and peptide hormones, such as leptin, adiponectin, and many of the interleukins (e.g. IL-6, IL-8, and IL-10). Adipose tissue also produces and secretes numerous steroid hormones and the enzymes involved in steroid hormone metabolism [29, 30]. For example, estrone is converted to estradiol in adipose tissue. Indeed, most if not all circulating estradiol in postmenopausal women

Table 1 Placental weights (grams) have increased in the US. Data from Naeye [78] and Swanson and Bewtra [44]. Year

10th percentile

Mean

90th percentile

Source

1959–1966 1995 2004

350 390 410

446 499 537

545 537 710

Naeye [78] Swanson and Bewtra [44] Swanson and Bewtra [44]

M.L. Power, J. Schulkin / Physiology & Behavior 106 (2012) 22–28

hypercortisolemic, hyperinsulemic, and hyperleptinemic. From the terminology one might be forgiven for thinking pregnancy was a metabolic disease. Maternal insulin resistance during pregnancy is not pathology, however, at least as long as it remains within the viable limits that characterize the pregnant metabolism phenotype. Indeed, maternal insulin resistance can be considered an example of allostatic regulation, in which physiology is continually adapting to changing circumstances in order to maintain viability. In evolutionary terms, viability requires passing genes on to the next generation. It is not surprising if understanding pregnant metabolism requires more than an understanding of the metabolic strategies for survival. It is also not surprising that obesity during pregnancy can lead to metabolic dysregulation. Both of these endocrine organs are signaling to maternal metabolism; but their signaling is not coordinated toward a common goal. Interestingly, there is a recent trend toward increasing placental weight in human pregnancy (Table 1), which is in parallel with the increase in maternal adipose tissue [44]. Both these potent endocrine organs are, on average, larger now than they have been in our past. It is reasonable to hypothesize that placenta size is increasing at least in part due to the interactive signaling between placenta and adipose tissue. For example, leptin is produced by adipose tissue, but also by the placenta in many mammalian species, including humans, baboons, bats, rodents, pigs, cows and sheep. Leptin is a molecule intimately linked with fat and feeding behavior. It also is linked to the insulin resistance of both pregnancy and obesity. However, in addition to its role as a regulator of energy intake and adiposity, leptin appears to have important functions in many reproductive processes, including important roles in fetal growth and developmental processes [45]. This illustrates another biological truism; biologically active molecules often have multiple functions, and are active in many physiological systems. Placental weight is correlated with placental leptin mRNA; cord serum leptin is correlated with placental leptin mRNA, maternal serum leptin, and with fetal mass [46]. Large for gestational age fetuses have higher than normal leptin, small for gestational age fetuses have lower leptin. In humans (and baboons) placental production contributes to both maternal and fetal leptin concentrations [45]. Because leptin is strongly associated with a measure of maternal nutritional status (fat mass), it is a plausible candidate for being an important metabolic signal for the maintenance and duration of pregnancy. Low leptin levels are associated with pregnancy loss in humans. Leptin levels may be abnormally high in pregnancies complicated by conditions such as diabetes mellitus and pre-eclampsia. Leptin is considered to be permissive of pregnancy, but not required. It may serve as a signal that maternal condition is satisfactory for reproduction. Of course,

Table 2 Expression of inflammation-related genes in human white adipose tissue and term placenta. Modified from Hauguel-de Mouzon and Guerre-Millo [57]. Molecule

Adipocytes

Stroma vascular fraction of adipose

Placenta

Adiponectin Leptin TNF-α IL-1b IL-1Ra IL-6 IL-8 IL-10 Resistin MCP-1 MIF PAI-1 VEGF Cathespin S

++ ++ +

0 0 +++ ++ ++ + ++ ++ ++ ++ + ++ ++ ++

0 + +

+ + + + 0 + ++ + + +

+ + + + + + + + + +

25

abnormal leptin levels may reflect abnormal placental function. For example, low leptin levels in pregnancy loss may reflect placental failure. Leptin is also a proinflammatory molecule, as are many of the hormones and cytokines secreted by both adipose tissue and placenta. Inflammation is an adaptive response, but one that carries costs. 6. Inflammation Pregnancy is an inflammatory state [13]. Human pregnancy has been termed a state of low grade albeit controlled inflammation [47]. Proinflammatory cytokines are expressed by uterine epithelium in the pregnant state and by the placenta (Table 2). The underlying cause of the inflammation of pregnancy is not well understood. There are probably multiple effectors giving rise to increased monocyte activation during pregnancy. Possible triggers include a variety of molecules released into maternal tissue by the placenta (e.g. cytokines and anti-angiogenic factors); but also necrotic debris shed by the syncytiotrophoblast, the multinucleated syncytial outer layer of the placenta that is contact with maternal blood. This debris includes what are termed synctiotrophoblast membrane microparticles (STBM). These microparticles are continuously shed from synctiotrophoblast into maternal circulation, and have been shown to result in monocyte activation in vitro [48]. The inflammatory response results in an influx of leukocytes into uterine tissue. In normal pregnancy the immune environment is skewed towards Th2 cytokine expression, especially in early pregnancy [49, 50]. A shift toward a Th1 cytokine expression is associated with spontaneous abortion, preterm delivery, and preeclampsia [50]. As normal pregnancy progresses there is an overall decrease in proinflammatory cytokines and an increase in counter-regulatory cytokines [51]. Obesity is also a state of low grade inflammation. Adipose tissue secretes many of the same proinflammatory cytokines as does placenta (Table 2). Obesity generally enhances adipose tissue secretion of proinflammatory cytokines, and is associated with lower secretion by adipose tissue of the anti-inflammatory hormone adiponectin [2, 13]. Adipose tissue in obese individuals is characterized by increased recruitment of adipose tissue macrophages (ATMs) with a shift towards a more proinflammatory ATM phenotype [13, 52-54]. Infiltration of adipose tissue by proinflammatory T-cells precedes the recruitment of and phenotypic changes in ATMs [55, 56]. This change in adipose tissue immune environment appears to be driven by Th1 cytokines, differing from the placental condition. The combination of proinflammatory signaling from excess adipose tissue and from placenta risks upsetting the regulatory balance [57]. Obese pregnant women exhibited increased systemic levels of C-reactive protein and IL-6 and higher gene expression in adipose tissue for IL-6, IL-8, and TNF-α [58]. Obesity appears to increase expression of proinflammatory cytokines from placenta, just as it does from adipose tissue. An increase in placental macrophages with proinflammatory cytokine expression has been demonstrated for obese pregnant women [59] and baboons [60]. Inflammation, both from obesity and pregnancy, is linked to insulin resistance. Among obese pregnant women circulating leptin, C-reactive protein and IL6 are increased [13, 25, 61]. Increased maternal circulating leptin is associated with maternal insulin resistance [62] and cord leptin is associated with fetal insulin resistance [25]. The proinflammatory signals have been shown to affect placental function, for example, altering function of the system A amino acid transporter [13, 63]. Placental expression of leptin receptor is decreased in obese women [63], perhaps in response to the high maternal circulating leptin. It is reasonable to hypothesize that the maternal expression of regulatory molecules will affect placenta by modifying its development and gene expression, and thus affect the development and gene expression of the fetus. However, as modern obesity is sufficiently removed from the norm of our evolutionary past, these placental adaptations may not

26

M.L. Power, J. Schulkin / Physiology & Behavior 106 (2012) 22–28

be effective, and possibly even counterproductive when viewed from the long-term health outcomes of the infant. The altered signaling from adipose tissue in maternal obesity may result in allostatic load in the placenta and fetus as they attempt to respond and adapt to those signals.

7. Epigenetics and in utero programming of disease Epigenetics is an old concept that precedes our modern understanding of genetic mechanisms. The term and original concept was introduced by C.H. Waddington [64], combining epigenesis from embryology with genetics, to describe how genotype interacting with environment gives rise to phenotype. A more recent conception of epigenetics is heritable changes in gene expression without underlying change in DNA sequence [65], and refers primarily to mechanisms such as DNA methylation and histone modifications that alter the functional qualities of the DNA sequence (e.g. ability of transcription factors to bind) but leave the DNA sequence unchanged [66]. This may be a fundamental mechanism for changing set points, as it changes gene expression in tissue. Regulation is the key to survival. Animals are constantly adjusting their physiology and metabolism in order to remain within the bounds of viability. The new and exciting understanding of genetics is that, at the level of DNA, regulation is also the key to viability. Our new understanding of genomics brings it closer to that of regulatory physiology. There is metabolism at all levels in an organism: the organism level, the organ level, the cellular level, and the genome level. Many of the changes to fetal metabolism in response to the in utero environment of maternal obesity are likely due to epigenetic changes; i.e. changes in DNA expression via DNA methylation or demethylation, histone modifications, or other changes to the structure of the chromosomal DNA without any actual change of DNA sequence. These epigenetic changes can activate or silence genes by such mechanisms as recruiting methyl-CpG binding proteins which then block transcription factor from binding to the promoter sites or change chromatin structure enhancing heterochromatin formation [67]. The placenta is a hot spot for epigenetic regulation, being generally the tissue with the lowest overall levels of DNA methylation. Placenta (and brain) also expresses a large number of imprinted genes, another form of epigenetic regulation. Fetal programming of physiology is a rapidly expanding area of biological knowledge that has critical relevance for human metabolic diseases. Many modern human diseases can be classified as metabolic disorders. Obesity, diabetes, hypertension, cardiovascular disease and so forth are fast becoming the leading sources of mortality and morbidity across the globe. Epidemiological evidence, first from Anders Forsdahl on differences in mortality rates due to heart disease among Norwegian counties [68], and then by Barker [69] in England, led to the hypothesis that early life experience, including in utero experience, can lead to later predisposition to heart disease. The placenta occupies a critical place in this transmission of developmental information from the mother. Prebirth, clues about the nutritional environment come from the mother's physiological responses to her environment, through the placenta. Fetuses may have to adapt to the supply of nutrients crossing the placenta, either a deficit or an overabundance, and these adaptations may permanently change their physiology and metabolism [16]. These programmed changes can have a significant impact on offspring later in life, and may be the origins of a diverse array of diseases, including heart disease, hypertension, and noninsulin-dependent diabetes. In fact, because of fetal programming, obesity may become a self-perpetuating problem. Daughters of obese mothers may themselves be vulnerable to becoming obese and more likely to have offspring that share this vulnerability, a form of “inheritance of acquired characteristics” from mother to child.

8. Obesity and brain These epigenetic changes likely affect brain development. The brain is a hot spot for imprinted genes and epigenetic effects [70]. Many of the information molecules produced by placenta and/or passed through placenta from maternal circulation are potent modulators of behavior in later life. For example, leptin has significant developmental effects on neural circuits related to feeding behavior; the same circuits that leptin activates later in life. Interestingly, leptin has no affect on appetite in neonatal rats or mice, at least for the first two weeks of life [71]. However, circulating leptin increases dramatically in the first and second weeks of life in rodents. This time period corresponds to a key brain developmental period; during this time the arcuate nucleus makes connections with other hypothalamic nuclei. These connections appear to be stimulated by leptin. Leptin deficient mice have poorly developed connections between the arcuate and other hypothalamic nuclei, and develop hyperphagia as they age [71]. A post natal peak in circulating leptin is also observed in lambs born to normal weight ewes, but this peak is missing in lambs born to ewes fed to obesity [72]. Compared to lambs from normal weight ewes, lambs from obese ewes had higher cortisol and leptin at birth, indicating disruption of the normal developmental pattern. Lambs of obese ewes are vulnerable to hyperphagia and obesity as adults [72]. Rats and mice are born in an altricial state; the first two weeks post partum are the equivalent of the final weeks of gestation in humans. Thus, in humans it is likely that these leptin-associated neural developmental events occur in utero, while the placenta can still affect development due to its endocrine signaling. In humans leptin is secreted by placenta into both maternal and fetal compartments. Maternal circulating leptin increases in pregnancy, but is still reflective of maternal adiposity [73]. Thus, obese pregnant women will have higher circulating leptin than will normal weight pregnant women. Leptin expression by placenta does not appear to be affected by the increased maternal leptin due to obesity; however, expression of leptin receptor by placenta was decreased in obese women [63]. Maternal leptin levels are correlated with cord blood levels; thus fetuses of obese mothers on average will be exposed to higher leptin levels. In addition to peripheral metabolic changes associated with increased cord blood leptin, such as fetal insulin resistance [25], human infants of obese mothers may have altered appetite circuits due to the effects of altered placental signaling on fetal brain development. Maternal adipose tissue signaling, in conjunction with placental signaling, may work to alter her child's future feeding behavior. Maternal diet and maternal obesity have been shown to affect future feeding behavior of offspring in rodent models [74]. Obesity in humans is associated with increases of amygdalar and hippocampal volume [75], though it is not clear when these changes might have occurred, and which is cause and which is effect. Interestingly, breast milk provides a post natal source of exogenous leptin to infants [76]. Leptin concentration of breast milk is negatively correlated with infant growth; however, there does not appear to any association with maternal adiposity [77].

9. Conclusion In our conception of physiological regulation the key underlying principles are: 1) Regulatory physiology is best understood from an evolutionary perspective. Regulation serves produce a viable organism, which includes reproductive success as well as survival. 2) Anticipatory responses are fundamental adaptations for evolutionary success. Animals anticipate needs, and modify their physiology in an attempt to achieve the most appropriate metabolic state for the anticipated challenge.

M.L. Power, J. Schulkin / Physiology & Behavior 106 (2012) 22–28

3) The central nervous system is fundamentally involved in regulatory physiology, coordinating responses among organ systems, integrating information across organ systems, and of course using past and current information to anticipate challenges and initiate the responses. The concept of allostasis was proposed as an alternative to homeostasis at least in part to incorporate fully the above principles into the science of physiological regulation. However, they are not foreign to homeostasis, and, in our opinion allostasis does not replace homeostasis. The addition of an allostatic perspective broadens the concept of regulatory physiology to match the complexity of the adaptive biology to be studied. Both obesity and pregnancy result in significantly altered metabolic states compared to the more homeostatic states of the nonobese, non-pregnant human female. These altered states occur due to signaling from endocrine organs: adipose tissue in the case of obesity and the placenta during pregnancy. The altered metabolism of pregnancy represents an adaptive state, having arisen over many millions of years of evolution. In contrast, obesity was extremely rare in our species' past, and likely represents a novel state effectively unseen by natural selection. Pregnancy can be considered the canonical example of allostatic regulation, with set points and regulatory pathways being continually altered to meet the changing demands of the maternal–placental–fetal unit. The placenta is a potent endocrine organ that secretes a wide array of signaling molecules into maternal and fetal circulation. These signals alter maternal circulating insulin, leptin, cortisol and many other potent information molecules, leading to a state of adaptive hyperglycemia, and insulin and leptin resistance. The placenta is in effect acting as a central regulatory organ, analogous to brain regulatory function among organs, but instead coordinating maternal and fetal physiology and development. Placental signaling influences the function of all maternal and fetal organs, perhaps most importantly brain. The maternal brain and the placenta (and fetus) both require a stable source of oxygen and glucose. This implies that these fundamental organs of evolutionary success will often have competing and even conflicting demands that must be coordinated and accommodated through mutual signaling mechanisms. The placenta–brain axes may be the most important understudied aspects of pregnancy for understanding both adaptive and pathological phenotypes. Unfortunately, obesity also tends to produce many similar metabolic responses, though these may not be adaptive. Obesity specifically complicates glucose metabolism, potentially exacerbating the conflicts and coordination challenges for brain and placenta. We argue that the increased risks of metabolic disease in both mother and child due to maternal obesity can be framed under the paradigm of allostatic load. Competing and conflicting signals from adipose tissue and placenta result in inappropriate or over extended activation of regulatory systems. Over time these regulatory mechanisms begin to break down, leading to tissue damage in the mother (e.g. pancreas) and altered development in the fetus, possibly including altered neural circuits important in appetite regulation. These altered developmental changes also may include epigenetic changes that could be passed to the next generation, in effect a mechanism for the inheritance of an acquired characteristic (obesity).

References [1] Gluckman P, Hanson M. Mismatch: why our world no longer fits our bodies. Oxford University Press; 2006. [2] Power ML, Schulkin J. The evolution of obesity. Johns Hopkins University Press; 2009. [3] McEwen BS. Stress, adaptation, and disease: allostasis and allostatic load. Ann NY Acad Sci 1998;840:33–44. [4] Cannon WB. Stresses and strains of homeostasis. Am J Med Sci 1935;189:1–14. [5] Schulkin J. Rethinking homeostasis: allostatic regulation in physiology and pathophysiology. Cambridge: MIT Press; 2003.

27

[6] Sterling P, Eyer J. Allostasis: a new paradigm to explain arousal pathology. In: Fisher S, Reason J, editors. Handbook of life stress, cognition, and health. New York: John Wiley and Sons; 1988. [7] Sterling P. Principles of allostasis: optimal design, predictive regulation, pathophysiology, and rational therapeutics. In: Schulkin J, editor. Allostasis, homeostasis and the costs of adaptation. Cambridge, UK: Cambridge University Press; 2004. p. 17–64. [8] Boulos Z, Rosenwasser AM. A chronobiological perspective on allostasis and its application to shift work. In: Schulkin J, editor. Allostasis, homeostasis and the costs of adaptation. Cambridge, UK.: Cambridge University Press; 2004. p. 228–301. [9] Koob GF, Le Moal M. Drug addiction and allostasis. In: Schulkin J, editor. Allostasis, homeostasis and the costs of adaptation. Cambridge, UK: Cambridge University Press; 2004. p. 150–63. [10] Power ML. Viability as opposed to stability: an evolutionary perspective on physiological regulation. In: Schulkin J, editor. Allostasis, homeostasis and the costs of adaptation. Cambridge, UK.: Cambridge University Press; 2004. p. 343–64. [11] Kim SY, Dietz PM, England L, et al. Trends in pre-pregnancy obesity in nine states, 1993–2003. Obesity 2007;15:986–93. [12] Leddy MA, Power ML, Schulkin J. Maternal obesity and the impact on maternal and fetal health. Rev Obstet Gynecol 2008;1:170–8. [13] Denison FC, Roberts KA, Barr SM, Norman JE. Obesity, pregnancy, inflammation, and vascular function. Reprod 2010;140:373–85. [14] Lynch CM, Sexton DJ, Hession M, Morrison JJ. Obesity and mode of delivery in primigravid and multigravid women. Am J Perinatol 2008;25:163–7. [15] Rooney B, Schauberger C. Excess pregnancy weight gain and long-term obesity: one decade later. Obstet Gynecol 2002;100:245–52. [16] de Boo HA, Harding JE. The developmental origins of adult disease (Barker) hypothesis. Aust N Z J Obstet Gynaecol 2006;46:4–14. [17] Weiss JL, Malone FD, Emig D, Ball RH, Nyberg DA, et al. Obesity, obstetric complications and cesarean delivery rate—a population-based screening study. Am J Obstet Gynecol 2004;190:1091–7. [18] Chu SY, Callaghan WM, Kim SY, Schmid CH, Lau J, England LJ, et al. Maternal obesity and risk of gestational diabetes mellitus. Diabetes Care 2007;30:2070–6. [19] Hedderson MM, Williams MA, Holt VL, Weiss NS, Ferrara A. Body mass index and weight gain prior to pregnancy and risk of gestational diabetes mellitus. AJOB 2008;198:409.e1–7. [20] Catalano PM, Eherenberg HM. The short- and long-term implications of maternal obesity on the mother and her offspring. BJOG 2006;113:1126–33. [21] Ogden CL, Carrol MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 1999–2004. JAMA 2006;295: 1549–55. [22] Catalano PM, Thomas A, Huston-Presley L, Amini SB. Phenotypes of the infants of mothers with gestational diabetes. Diabetes Care 2007;30(Suppl 2):S156–60. [23] Reilly JJ, Armstrong J, Dorosty AR, Emmett PM, Ness A, Rogers I, et al. Early life risk factors for obesity in childhood: cohort study. BMJ 2005;330:1357–9. [24] Sewell MF, Huston-Presley L, Super DM, Catalano P. Increased neonatal fat mass, not lean body mass, is associated with maternal obesity. Am J Obstet Gynecol 2006;195:1100–3. [25] Catalano PM, Presley L, Minium J, Hauguel-de Mouzon S. Fetuses of obese mothers develop insulin resistance in utero. Diabetes Care 2009;32:1076–80. [26] Boney CM, Verma A, Tucker R, Vohr BR. Metabolic syndrome in childhood: association with birth weight, maternal obesity, and gestational diabetes mellitus. Pediatr 2005;115:e290–6. [27] Radaelli T, Varastehpour A, Catalano P, Hauguel-de Mouzon S. Gestational diabetes induces placental genes for chronic stress and inflammatory pathways. Diabetes 2003;52:2951–8. [28] Catalano PM, Thomas A, Huston-Presley L, Amini SB. Increased fetal adiposity: a very sensitive marker of abnormal in utero development. Am J Obstet Gynecol 2003;189:1698–704. [29] Fain JN. Release of interleukins and other inflammatory cytokines by human adipose tissue is enhanced in obesity and primarily due to the nonfat cells. Vitam Horm 2006;74:443–77. [30] Kershaw EE, Flier JS. Adipose tissue as an endocrine organ. J Clin Endocrinol Metab 2004;89:2548–56. [31] Wake DJ, Strand M, Rask E, Westerbacka J, Livingstone DE, Soderberg S, et al. Intra-adipose sex steroid metabolism and body fat distribution in idiopathic human obesity. Clin Endocinol 2007;66:440–6. [32] Seckl JR, Walker BR. 11β-hydroxysteroid dehydrogenase type 1: a tissue-specific amplifier of glucocorticoid action. Endocrinology 2001;142:1371–6. [33] Tomlinson JW, Finney J, Hughes BA, Stewart PM. Reduced glucocorticoid production rate, decreased 5α-reductase activity, and adipose tissue insulin sensitization after weight loss. Diabetes 2008;57:1536–43. [34] Stimson RH, Andersson J, Andrew R, Redhead DN, Karpe F, Hayes PC, et al. Cortisol release from adipose tissue by 11β-hydroxysteroid dehydrogenase type 1 in humans. Diabetes 2009;58:46–53. [35] Rask E, Walker BR, Söderber S, Livingstone DEW, Eliasson M, Johnson O, et al. Tissuespecific changes in peripheral cortisol metabolism in obese women: increased adipose 11β-hydroxysteroid dehydrogenase type 1 activity. J Clin Endocrinol Metab 2002;87:3330–6. [36] Stewart PM, Boulton A, Kumar S, Clark PMS, Shakleton CHL. Cortisol metabolism in human obesity: impaired cortisone to cortisol conversion in subjects with central obesity. J Clin Endocrinol Metab 1999;84:1022–7. [37] Rask E, Olsson T, Söderber S, Andrew R, Livingstone DEW, Johnson O, et al. Tissuespecific dysregulation of cortisol metabolism in human obesity. J Clin Endocrinol Metab 2001;86:1418–21.

28

M.L. Power, J. Schulkin / Physiology & Behavior 106 (2012) 22–28

[38] Petraglia F, Florio P, Vale WW. Placental expression of neurohormones and other neuroactive molecules in human pregnancy. In: Power ML, Schulkin J, editors. Birth, distress and disease: placenta–brain interactions. Cambridge University Press; 2005. p. 16–73. [39] Power ML, Schulkin J. Introduction: brain and placenta, birth and behavior, health and disease. In: Power ML, Schulkin J, editors. Birth, distress and disease: placenta–brain interactions. Cambridge University Press; 2005. p. 1–15. [40] Yen SSC. The placenta as the third brain. J Reprod Med 1994;39:277–80. [41] Frankenne F, Closset J, Gomez F, Scippo ML, Smal J, Hennen G. The physiology of growth hormones in pregnant women and partial characterization of the placental GH variant. J Clin Endocrinol Metab 1988;66:1171–80. [42] Alsat E, Guibourdenche J, Luton D, Frankenne F, Evain-Brion D. Human placental growth hormone. Am J Obstet Gynecol 1997;177:1526–34. [43] Peters A, Schweiger U, Pellerin L, Hubold C, Oltmanns KM, Conrad M, et al. The selfish brain: competition for energy resources. Neurosci Biobehav Rev 2004;28: 144–78. [44] Swanson LD, Bewtra C. Increase in normal placental weights related to increase in maternal body mass index. J Mater Fetal Neonatal Med 2008;21:111–3. [45] Henson MC, Castracane VD. Leptin: roles and regulation in primate pregnancy. Semin Reprod Med 2002;20:113–22. [46] Jakimiuk AJ, Skalba P, Huterski R, Haczynski J, Magoffin DA. Gynecol Endocrinol 2003;17:311–6. [47] Sacks G, Sargent I, Redman C. An innate view of human pregnancy. Immunol Today 1999;20:114–8. [48] Messerli M, May K, Hansson SR, Schneider H, Holzgreve W, Hahn S, et al. Feto-maternal interactions in pregnancies: placental microparticles activate peripheral blood monocytes. Placenta 2010;31(2):106–12. [49] Chaouat G. Regulation of T-cell activities at the feto-placental interface — by placenta? Am J Reprod Immunol 1999;42:199–204. [50] Challis JR, Lockwood CJ, Myatt L, Norman JE, Strauss JF, Petraglia F. Inflammation and pregnancy. Reprod Sci 2009;16:206–15. [51] Denney JM, Nelson EL, Wadhwa PD, Waters TP, Mathew L, Chung EK, et al. Longitudinal modulation of immune system cytokine profile during pregnancy. Cytokine 2011;53:170–7. [52] Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante Jr AW. Obesity is associated with macrophage accumulation in adipose tissue. J Clin Invest 2003;112:1796–808. [53] Lumeng CN, DelProposto JB, Westcott DJ, Saltiel AR. Phenotypic switching of adipose tissue macrophages with obesity is generated by spatiotemporal differences in macrophage subtype. Diabetes 2008;57:3239–46. [54] Morris DL, Singer K, Lumeng CN. Adipose tissue macrophages: phenotypic plasticity and diversity in lean and obese states. Curr Opin Clin Nutr Metab Care 2011;14:341–6. [55] Kintscher U, Hartge M, Hess K, Foryst-Ludwig A, Clemenz M, Wabitsch M, et al. T-lymphocyte infiltration in visceral adipose tissue. A primary event in adipose tissue inflammation and the development of obesity-mediated insulin resistance. Arterioscler Thromb Vasc Biol 2008;28:1304–10. [56] Nishimura S, Manabe I, Nagasaki M, Eto K, Yamashita H, Ohsugi M, et al. CD8+ effector T cells contribute to macrophage recruitment and adipose tissue inflammation in obesity. Nat Med 2009;15:914–20. [57] Hauguel-de Mouzon S, Guerre-Millo M. The placenta cytokine network and inflammatory signals. Placenta 2006;27:794–8.

[58] Basu S, Haghiac M, Surace P, Challier J-C, Guerre-Millo M, Singh K, Waters T, Minium J, Presley L, Catalano PM, Hauguel-de Mouzon. Pregravid obesity associates with increased endotoxemia and metabolic inflammation. Obesity 2011;19: 476–82. [59] Challier JC, Basu S, Bintein T, Minium J, Hotmire K, Catalano P, et al. Obesity in pregnancy stimulates macrophage accumulation and inflammation in the placenta. Placenta 2008;29:274–81. [60] Farley D, Tejero ME, Comuzzie AG, Higgins PB, Cox L, Werner SL, et al. Feto-placental adaptations to maternal obesity in the baboon. Placenta 2009;30:752–60. [61] Radaelli T, Uvena-Celebrezze J, Minium J, Huston-Presly L, Catalano P, Haugel-de Mouzon S. Maternal interleukin-6: Marker of fetal growth and adiposity. J Soc Gynecol Investig 2006;13:53–7. [62] McIntyre HD, Chang AM, Callaway LK, Cowley DM, Dyer AD, Radaelli T, et al. Hormonal and metabolic factors associated with variations in insulin sensitivity in human pregnancy. Diabetes Care 2010;33:356–63. [63] Farley DM, Choi J, Dudley DJ, Li C, Jenkins SL, Myatt L, et al. Placental amino acid transport and placental leptin resistance in pregnancies complicated by maternal obesity. Placenta 2010;31:718–24. [64] Waddington CH. Canalization of development and inheritance of acquired characters. Nature 1942;150:563–5. [65] Jones PA, Takai D. The role of DNA methylation in mammalian epigentics. Science 2001;293:1068–70. [66] Crews D, McLachlan JA. Epigenetics, evolution, endocrine disruption, health, and disease. Endocrinology 2006;147:S4–S10 (suppl). [67] Heerwagen MJ, Miller MR, Barbour LA, Friedman JE. Maternal obesity and fetal metabolic programming: a fertile epigenetic soil. AJP 2010;299:R711–22. [68] Forsdahl A. Are poor living conditions in childhood and adolescence important risk factors for arteriosclerotic heart disease? Br J Prev Soc Med 1977;31:91–5. [69] Barker DJP. Fetal and infant origins of adult disease. BMJ 1990;301:1111. [70] Gregg C, Zhang J, Butler JE, Haig D, Dulac C. Sex-specific parent-of-origin allelic expression in the mouse brain. Science 2010;329:682–5. [71] Bouret SG, Simerly RB. Minireview: leptin and development of hypothalamic feeding circuits. Endocrinol 2004;145:2621–6. [72] Long NM, Ford SP, Nathanielsz PW. Maternal obesity eliminates the neonatal lamb plasma leptin peak. J Physiol 2011;589:1455–62. [73] Butte NF, Hopkinson JM, Nicolson MA. Leptin in human reproduction: serum leptin levels in pregnant and lactating women. J Clin Metab Endocrinol 1997;82: 585–9. [74] Shankar K, Harrell A, Liu X, Gilchrist JM, Ronis MJJ, Badger TM. Maternal obesity at conception programs obesity in the offspring. Am J Physiol Regul Integr Comp Physiol 2008;294:R528–38. [75] Widya RL, de Roos A, Trompet S, de Craen AJM, Westendorp RGJ, Smit JWA, et al. Increased amygdalar and hippocampal volumes in elderly obese individuals with or at risk of cardiovascular disease. Am J Clin Nutr 2011;93:1190–5. [76] Smith-Kirwin SM, O'Connor DM, Johnston J, De Lancey E, Hassink SG, Funanage VI. Leptin expression in human mammary epithelial cells and breast milk. J Clin Endocrinol Metab 1998;83:1810–3. [77] Dundar NO, Anal O, Dundar B, Ozkan H, Caliskan S, Büyükgebiz A. Longitudinal investigation of the relationship between breast milk leptin levels and growth in breast-fed infants. J Pediatr Endocrinol Metab 2005;18:181–8. [78] Naeye R. Do placental weights have clinical significance? Hum Pathol 1987;18: 387–91.