Fetal growth restriction results from a failure to achieve a higher growth potential and has been associated with many maternal conditions, such as chronic diseases (infections, hypertension, and some cases of diabetes and obesity), exposures (tobacco smoke, drugs), and malnutrition. This early adversity induces a series of adaptive physiological responses aimed at improving survival, but imposing increased risk for developing chronic nontransmittable diseases (obesity, type II diabetes, cardiovascular disease) in the long term. Recently, mounting evidence has shown that fetal growth impairment is related to altered feeding behavior and preferences through the life course. When living in countries undergoing nutritional transition, in which individuals experience the coexistence of underweight and overweight problems (the “double burden of malnutrition”), fetal growth–restricted children can be simultaneously growth restricted and overweight—a double burden of malnutrition at the individual level. Considering food preferences as an important aspect of nutrition security, we will summarize the putative neurobiological mechanisms at the core of the relationship between fetal growth and nutrition security over the life course and the evidence linking early life adversity to later food preferences.
The term food security has been used in many contexts, but its most recent definition describes the situation “when all people, at all times, have physical and economic access to sufficient safe and nutritious food to meet their dietary needs and food preferences for a healthy and active life.” As was wisely noticed by Pinstrup–Andersen, the addition of “food preferences” goes beyond the matter of accessibility to enough food, including access to food that is preferred. This implies the existence of an individual level of food security, in which people with equal access to food, but different food preferences, could show different levels of food security.
Food preferences and dietary behaviors are established during childhood, and are the result of a complex mixture of genetic, environmental, psychosocial, cultural, and cognitive influences. However, a recent growing body of evidence suggests that spontaneous food preferences may be affected by events happening even earlier, during fetal development. In the late 1980s, large retrospective studies were accumulated, providing evidence that perinatal events, such as being born with low birth weight, could persistently affect an individual's physiology and health/disease patterns throughout the lifetime. Impaired perinatal growth seems to be an important risk factor for ischemic heart disease in adult life,[4-8] for impaired glucose tolerance and type II diabetes in adulthood,[9-11] as well as for increased prevalence of metabolic syndrome.[12, 13] Therefore, fetal adversity seems to be a critical risk factor for the development of chronic noncommunicable diseases that are classically related to food insecurity. This article expands these ideas considering that fetal growth–restricted individuals have specific food preferences toward highly palatable/energy-dense/nutritionally poor foods. Therefore, the very concept of food security may depend on the individual's neonatal history.
Moreover, as the concept includes food preferences, it implies that food security is inevitably affected by neurobiological processes. The programming of specific neurobiological pathways and mechanisms that regulate feeding behavior, food preferences, metabolism, and energy expenditure by the nature of environmental cues in utero or during early postnatal life seems to be key to the link between fetal growth restriction and nutrition security across the life span.
An interesting observation from Nancy Howell, a demographer who lived 2 years with the indigenous !Kung from the Kalahari Desert, immediately prompted us to think about different vulnerable populations. She observed that “despite more than 25 years of ethnographic observation of the !Kung, I believe that their attitude toward food has not yet been thoroughly understood by any of the anthropologists…. What we observers have trouble understanding is what chronic hunger feels like, and how human appetite responds to the availability of a kilo or so of rather tart berries or somewhat fibrous roots.” Fetal growth restriction may be seen as a period of physiological chronic hunger, as the organism is challenged with a condition that impairs growth, and its biology and metabolism consequently suffer adaptations to guarantee survival in such an adverse environment. As this event happens in a period of intense developmental plasticity, it can lead to persistent “programming” effects, especially in the brain circuits related to energy intake and expenditure. The “thrifty phenotype”[4, 12] message will per pass the entire life of a fetal growth–restricted individual, and will be likely responsible for their permanent increased risk for many diseases, such as obesity and the metabolic syndrome. Therefore, by analogy to the observations of the !Kung, it is possible that we still do not understand the attitude of a growth restricted individual toward food, as their particular behaviors and preferences just recently started to be described. In other words, the sensation of food security may have a very distinctive meaning for a subject that was deprived from essential nutrients during fetal/early life.
Although alterations in appetite and physical activity have been acknowledged and studied as important factors contributing to the developmental origins of adult diseases,[15-24] only recently has evidence emerged suggesting that fetal growth impairment is associated with specific feeding behavior and preferences during development. Considering food preferences as an important aspect of nutrition security as discussed above, we will summarize the putative neurobiological mechanisms at the core of the relationship between fetal growth and nutrition security over the life course and the evidence linking early life adversity to later food preferences. More specifically, it will address the impact of fetal growth restriction on feeding behavior regulation, as well as its modulation by reward systems, executive function, and the stress response, considering the implications for food preferences and health later in life. The article includes the reasonably well-established evidence from animal studies and more recent emerging evidence from human studies from developed and developing countries.
Understanding vulnerability—the size of the problem
Intrauterine growth restriction (IUGR) is a condition in which the fetus does not grow at a rate that will lead to its full potential size at birth. The diagnosis of this condition is only possible by repeated measurements of fetal growth by ultrasonograpy during pregnancy, and as most of the time in practice there is only information on birth weight, most small-for-gestational-age (SGA) newborns (defined as those with birth weight equal or below the 10th percentile for birth weight using populational sex-specific parameters) are considered to have IUGR. However, not all fetuses that are SGA are pathologically growth restricted and, in fact, 40% of them may be constitutionally small. Similarly, not all fetuses that have failed to meet their growth potential are in less than the 10th percentile for estimated fetal weight, or can be classified as low birth weight (i.e., less than 2500 g). For instance, a boy born at 41 weeks of gestation at 3000 g could be growth restricted, although he would be considered neither SGA nor low birth weight. Therefore, clinically, the functional definition of IUGR is very difficult to establish.
Regarding worldwide prevalence, a historical perspective has to be taken into account. The quarter century from 1956 to 1980 brought more progress and more promise for the low birth weight/premature infant than any similar previous amount of time. Better perinatal medical care accounted for the majority of the improvement, with a consequent reduction in mortality rates and a complete change in previously established epidemiological associations and known risk factors for morbidity in this population. Therefore, while mean birth weight varied little across the years in different continents,[29-34] perinatal survival drastically improved mainly due to reductions in birth weight–specific neonatal mortality, both in developed and developing countries in the last 50 years,[35-41] which intriguingly preceded the worldwide rise of adult chronic noncommunicable diseases. In low- and middle-income countries, infectious disease, undernutrition, and a rapid upsurge in noncommunicable disease risk factors, such as obesity and overweight, coexist in the same setting. Inadequate prenatal nutrition/care (especially in developing countries), increased survival of preterm/low birth weight individuals (especially in developed nations), fetal programming effects, and the subsequent exposure to high-fat, high-sugar, energy-dense, micronutrient-poor foods, are likely players in this scenario. Impressively, some studies suggest that 25–63% of adult diabetes, hypertension, and coronary heart disease could be attributed to the effects of low birth weight with subsequent weight gain during development.
Although yet to be systematically studied, it is likely that these diverse scenarios affect childhood growth and the programming of food intake and energy expenditure in a similar fashion.[44-46] In other words, different pathways to low birth weight are associated with rapid gain in fat mass during childhood, consequently leading to a persistent programming of the individual's feeding behavior and preferences toward palatable foods.
These behavioral features in fetal growth–restricted individuals interact with the food environment to determine adult health risks. At first, one may suppose that the availability of highly palatable foods is extremely different among high- and low/middle-income countries, but this has been shown not to be the case. Recent studies examining the trends in the purchase and sales of ultra-processed foods (attractive, hyper-palatable, cheap, ready-to-consume food products that are characteristically energy-dense, fatty, sugary, or salty, and generally obesogenic) in 79 high- and middle-income countries, show that these products dominate the food supplies in high-income countries. In addition, they are progressively more consumed in middle-income countries, in a fastest rate of increase and a level of penetration comparable to the one that has occurred historically in high-income countries.
Surprisingly, as opposed to earlier studies, urbanization no longer seems to be a strong risk factor for greater consumption of these food commodities at the population level. Moreover, the earlier general impression that in higher income countries greater obesity rates were found in rural areas and among the poor, with the reverse pattern seen in lower-income countries, is also changing.
In addition, obesity during pregnancy is associated with a range of gestational complications. Offspring of obese women are two to four times more likely to be large for gestational age or macrosomic.[51-53] On the other hand, they can also suffer from a failure to achieve a higher growth potential (i.e., an unbalance between excess energy supply and limited space availability in the maternal pelvis or placental functionality) and therefore a subset of obese women may have newborns that are growth restricted. As fetal growth impairment is associated with obesity in adulthood, when the female offspring reaches reproductive age she may be obese as well and again give birth to a growth-restricted newborn; this could be one mechanism contributing to the transgenerational transmission of obesity risk.
As commented above, aberrant fetal growth has been linked to the development of metabolic disease in later life. Interestingly, many authors describe a U-shaped curve linking both low and high birth weights with these conditions.[44, 45] Experimental and human evidence suggests that rapid weight gain in infancy is related to increased risk of overweight, elevated blood pressure, or impaired fasting glucose in young adulthood,[55-58] and this is also seen in low- and middle-income countries. Therefore, the promotion of rapid weight gain in growth-restricted children has to be seen with caution. In developing countries, the compensatory growth following a period of faltering growth is associated with reduced morbidity[59-61] and better cognitive development, but an important concern is whether the benefits of faster growth in these settings outweigh the possible long-term risks.
The biological mechanisms that link nutrient imbalances in utero and later behavior/disease risk may reflect the long-term consequences of fetal adaptive responses and have always been present (throughout history). At first, one could argue that these mechanisms would therefore neither easily explain nor contribute to the sudden increase in obesity rates and associated morbidity that we face in recent times. However, it has to be considered that over the past several decades dramatic global behavioral shifts in eating, drinking, and exercising have taken place, which seem to be specifically important to trigger these biological consequences.[64-70] Chronic stress, pollution, lack of physical activity, and poor quality food intake in large urban areas,[19-21, 46, 71-78] as well as micronutrient imbalances, lack of prenatal care, and tobacco smoke during gestation in the developing world[79-85] all seem to be environmental triggers that affect obesity risk in children, and may be particularly important for vulnerable individuals.
Messages to predict survival and the challenging environment
The developmental origins hypothesis proposes that an increased long-term risk for disease is induced through adaptive mechanisms in the fetus or infant in response to cues from the mother about the surrounding environment, her health, or physical state. The mother is able to transmit such information through the placenta and/or through breast milk during lactation. The developmental cue is not limited to current nutritional availability, being a summation of maternal (and perhaps even of previous generations’) experiences.
The developing organism has many ways to respond to a changing early environment; these responses operate during critical time windows and some of them become irreversible. For instance, IUGR reduces the nephron number, with a subsequent long-term increased risk of renal disease. Another example from a rodent model of IUGR shows that this group has a reduction in the number of cardiomyocytes, which may lead to compromised cardiac function later in life. Most of such responses are meant to confer immediate survival advantage, even if there may be subsequent postnatal costs. Some of these responses are meant primarily to have future adaptive value (predictive adaptive responses, PAR). They initiate a cascade of effects according to the nature, size, and time of the environmental cues.[90, 91] Therefore, PAR arise during development with the objective of optimizing the phenotype for the probable environment that the organism will face later in life. The adaptive capacity of PAR is dependent on the probability that the choices made during development are adequate for the environment that the individual actually experiences throughout maturation. Where there is a match between the predicted and actual mature environment, these PAR are appropriate and promote survival. Conversely, inappropriate predictions (mismatch) would increase the risk of disease.
For growth-restricted individuals, independent of the mechanism that lead to impaired fetal growth (e.g., gestational malnutrition, chronic maternal diseases leading to placental insufficiency, smoking), the general environmental message transmitted during pregnancy is most likely that of future paucity of nutritional resources. However, the actual environment challenges these individuals, as global overconsumption of cheap calorie-dense, nutrient-poor foods has taken place.[92, 93] Excess consumption of processed foods, which are generally inexpensive, highly palatable, and energy-dense, also displaces the consumption of healthier whole foods, contributing to the development of overweight. Interestingly, this is more related to variations in socioeconomic status within countries than to the levels of economic development of the different countries.[94-96]
Moreover, the constant advertising of unhealthy foods, including many breakfast cereals and pastries, directed selectively at children also contributes to the obesity epidemic in general; the average American child watches 10,000 television advertisements for food each year, most of them for sugary cereals, candy, fast foods, and soft drinks, increasing food consumption. Industrial manipulation of food flavor, texture, color/appearance, and package attractiveness and size,[102, 103] influences food intake and choices. Besides, associations between altered preferences and/or intake and the rewarding aspects of the food environment (lighting, music, habit),[104, 105] social engagement, peer acceptance, knowledge of other people's choices, and toy giveaways, are observed. Finally, availability[110, 111] and the perception of availability may also influence food consumption.
Fetal growth restriction seems to be an important contributor to stunting and wasting in children, with estimates that about one fifth of childhood stunting could have its origins in the fetal period, as shown by being born with IUGR. Interestingly, epidemiological studies show that children and adolescents living in countries undergoing nutrition transition can be simultaneously growth retarded and overweight,[114-120] which is called a “double-burden of malnutrition” at the individual level. Some factors determining the coexistence of overweight and stunting in preschool children are low socioeconomic status, lower maternal age, level of education and perceived social status, and postnatal deficiencies. Therefore, even when living in a nutritionally poor environment, fetal growth restriction is associated with malnutrition, stunting, and subsequent overweight and its consequences, although these can be seen as adaptations to guarantee survival/reproduction in extremely harsh environments. Figure 1 summarizes the multilevel setting of environmental cues that influence development, especially in the context of nutrition and risk for adulthood overweight/obesity.
Neurobehavioral aspects of nutritional security
Energy intake is essential to homeostatic maintenance and survival, being finely tuned through intricate and complex mechanisms. Didactically, the regulation of food intake can be divided into homeostatic, hedonic, and executive components. In addition, an intense communication between peripheral signals and different brain regions allows a more detailed control through modulation of specific neurotransmitter systems by circulating hormones.
Homeostatic control of appetite
Central homeostatic mechanisms involve mainly the hypothalamus, a key region that consists of more than 40 distinct areas. The main appetite regulatory site, the hypothalamic arcuate nucleus (ARC), receives inputs from other hypothalamic regions,[122, 123] as well as from extrahypothalamic regions.[124-126] Its neurons are anatomically positioned near the fenestrate capillaries, placing them in close contact with important nutrients such as glucose and hormones such as leptin, insulin,  and ghrelin, for which they have receptors.[130, 131] Two subtypes of neurons were identified in the ARC, both containing the inhibitory neurotransmitter, gamma aminobutyric acid (GABA).[132, 133] One of these neuronal populations expresses proopiomelanocortin (POMC) and the cocaine and amphetamine–related peptide (CART), and when activated leads to decreased appetite and increased energy expenditure.[134-136] By contrast, the other cell population contains neuropeptide Y (NPY) and the agouti-related protein (AGRP) and is associated with an orexigenic response and decreased energy expenditure. Leptin and insulin increase the activity of POMC neurons and inhibit NPY neurons, while ghrelin acts in the opposite way.[139, 140]
In experimental studies, fetal growth–restricted rat pups nursed by control dams demonstrate significantly increased food intake with rapid catch-up growth and adult obesity.[141, 142] The obese phenotype is associated with reduced satiety responses to leptin and impaired ARC signaling responses to leptin,[143-146] as well as with increased responses to appetite stimulatory factors, such as ghrelin, and an increased ratio of appetite to satiety gene expression.[24, 148-150] Altered POMC mRNA expression seen in different studies suggests that this polypeptide is a key target for developmental programming in this IUGR model.[145, 151] These alterations were similarly seen in other animal models, such as baboons, lambs, and piglets. Maternal obesity or exposure to a high-fat diet during gestation also results in pups that demonstrate increased food intake, adult obesity, and leptin resistance.[155, 156] Thus, either maternal under- or overnutrition may program offspring hyperphagia and metabolic syndrome.
Hedonic component of food intake
Eating beyond homeostatic needs when facing caloric-rich palatable foods demonstrates that the perceived pleasantness of foods can modulate food intake indirectly by influencing the preference for certain foods. In addition, highly flavored foods have the ability to trigger learned associations between the stimulus and the reward (conditioning). The brain extracts information about quality, intensity, and hedonic value from gustatory neuronal responses; therefore, appetite for specific foods and nutrients is under complex neuroregulatory control. The forebrain plays a prominent role in the hedonic value that the brain attaches to gustatory activity originating from the oral cavity. The nucleus accumbens (NAcc) has been related to directive behaviors and appetitive instrumental learning[156, 157] and may provide an interface between motivation and behavioral action. Furthermore, neuroimaging studies in humans reveal that food-related cues activate areas of the brain associated with the processing of information related to the pleasurable features of stimuli; these areas include the ventral tegmental area (VTA), substantia nigra, amygdala, and orbitoprefrontal cortex,[158, 159] and are either involved in the synthesis and release of dopamine or are targets for dopamine projections.
Lately, researchers have increased awareness about the similarities between overeating and other addictive behaviors, such as tobacco smoking, alcohol consumption, and stimulant use,[160, 161] suggesting that brain circuits can be deranged with natural rewards like eating, just as they can with drugs.[162, 163] Foods, especially those that are highly palatable (sweet, salty, and fat), can enhance mood[164-167] and can be used for comforting purposes that exceed basic energy needs. Therefore, natural reinforcers such as food may prompt the incentive salience phenomena, involving two distinct neurobiological domains: “wanting,” the incentive motivation resulting in increased appetite, food cravings, and other behaviors related to motivation to obtain food; and “liking,” or the hedonic component that reflects the immediate experience or anticipation of pleasure.[168-170] Simply put, the main neurotransmitters involved in incentive salience are dopamine (wanting) and opioids (liking). The neural networks related to the liking hedonic component of reward include pathways involved in taste processing in the brainstem, pons, nucleus accumbens, ventral pallidum, amygdala, and prefrontal cortex (PFC).[172, 173] Mu-opioid receptor stimulation within the NAcc has been shown to increase the intake and preference for sweet and high-fat foods,[174, 175] and is associated with hedonic orofacial responses.
A possible mechanism by which the early environment could persistently affect the individual's food preferences is the programming of the sensitivity to the reward or pleasure associated with the ingestion of a palatable food. Interestingly, in adult rodents, prenatal protein malnutrition alters the response to reward. In addition, these offspring have increased expression of dopamine-related genes (tyrosine hydroxylase [TH] and dopamine transporter) in brain regions related to reward processing (VTA, NAcc, and PFC) and homeostatic control (hypothalamus), increased content of TH in the VTA and increased dopamine in the PFC, accompanied by behavioral alterations, such as altered reward processing, hyperactivity, and an exaggerated locomotor response to cocaine. Following chronic food restriction, there is a significant decrease in mu-opioid receptor binding in specific regions of the parabrachial nucleus. More specifically, perinatal undernutrition induces a decreased mu-opioid receptor density (Bmax) in the midbrain in adult life. Therefore, an imbalance in the relative importance of hedonic versus homeostatic signals may be associated with the overconsumption of palatable/rewarding foods, resulting in poor regulation of body weight in the long term.
Executive functions: perception and decision making
Executive function is a set of cognitive processes that are essential for higher order mental functioning. There are several domains of executive function, including (1) inhibitory control, (2) attention and mental flexibility, (3) reward sensitivity, and (4) working memory. These complex behaviors are mediated by prefrontal cortical function, being modulated by dopaminergic, noradrenergic, serotonergic, and cholinergic neurotransmission in response to changes in the environment.
The PFC may be involved in the conscious perception of some types of flavors and particularly in the integration of valuation and comparison processes (coding of rewards relative to other available rewards, general and specific satiety, temporal discounting and negative valuations, such as negative health consequences) that affect food selection. Obese children react to food stimuli with increased prefrontal activation, suggesting that reduced inhibitory control may play a role in excessive feeding behavior. Infants who were growth restricted have poorer executive functioning,[185, 186] increased vulnerability to addictive disorders and attention-deficit hyperactivity disorder; therefore, alterations in frontal brain regions could also play a role in their food choices.
Modulation of central mechanisms that regulate food intake
Leptin, insulin, and ghrelin
The neuropeptides that regulate energy intake and expenditure (homeostatic processes) through the hypothalamus also modulate the activity of dopaminergic neurons and their projections to regions involved in the rewarding processes underlying food intake. It is suggested that this could be a mechanism by which overeating and the resultant resistance to homoeostatic signals impairs the function of circuits involved in reward sensitivity, conditioning, and cognitive control. As such, evidence suggests that the activity of dopaminergic VTA neurons that project to the NAcc can be modulated by peripheral energy status signals, including leptin, insulin, and ghrelin,[189-208] and both leptin and insulin are associated with a decrease in the response of the NAcc to food cues.[209, 210]
Several studies have shown that cord leptin levels are diminished in SGA infants,[211, 212] increase during catch-up growth, and decrease again in adulthood in the context of an excess of adipose tissue when corrected for body fat mass, gender, and fasting insulin, suggesting altered adipocyte function and leptin resistance in these individuals. Low birth weight also is related to impaired insulin secretion,[215, 216] decreased glucose tolerance in later life,[216, 217] and diabetes.[43, 218] Potentially, leptin/insulin modulation of central dopamine is altered in SGA individuals, leading to an altered reward response to food, with consequent increased palatable food ingestion and the development of obesity.
Brain-derived neurotrophic factor
Brain-derived neurotrophic factor (BDNF) is a member of the neurotrophin family and is expressed during development in excitatory synapses from immature neurons. Initially, it was identified as a growth factor that influenced survival of sensorial neurons, but many other roles such as neuronal survival, differentiation, and neuronal plasticity, as well as modulation of satiety signals, are also now acknowledged.
BDNF plays a critical role in the regulation of homeostatic food intake. Mice with a BDNF deletion have a dramatic increase in body weight, hyperleptinemia, hyperinsulinemia, and hyperglycemia.[221-223] Acute intracerebroventricular injection of BDNF reduces body weight and food intake,[224, 225] and infusion of BDNF in the dorsal vagal complex (DCV) diminishes caloric intake and results in decreased body weight. Melanocortin-4 receptor stimulation, systemic leptin, or cholecystokinin injections increase BDNF in the DCV, and therefore this area seems to be an important region for the interplay between BDNF and these hormones. BDNF facilitates the appetite suppressor effects of melanocortin, interacting with leptin in this area. BDNF also influences hedonic food intake, as its receptor TrkB is expressed in dopaminergic VTA neurons, PFC, and projection GABA neurons in the NAcc, and suggests that, in these regions, BDNF plays a role in reward mechanisms. BDNF depletion in the ventral tegmental area leads to defects in dopamine neurotransmission, increased high-fat diet intake and consequent obesity.
Interestingly, in humans, a functional polymorphism of BDNF (Val66Met) has been associated with vulnerability to obesity[230-232] and to different psychiatric diseases that are also related to food intake alterations, such as mood disorders, anxiety, psychotic states, and eating disorders.[236, 237] This polymorphism is characterized by a replacement of a valine by a methionine in the codon 66, leading to diminished intracellular transportation and secretion of BDNF.
The placental BDNF/TrkB system is differentially modulated in rats and humans during pregnancies with fetal growth perturbations, being affected by maternal energetic status. Besides, in rodents, growth restriction during perinatal life modifies both BDNF and its functional receptor TrkB gene expression in specific hippocampal and hypothalamic areas during sensitive developmental windows; TrkB is decreased in cultured neurons from the cerebral cortex of low birth weight rat pups, with a repression of BDNF-stimulated MAPK/ERK1/2 and PI3K/Akt pathways, and BDNF fails to prevent neuronal death.
The effects of acute stress–induced activation of the hypothalamic–pituitary–adrenal (HPA) axis on food consumption are well known, increasing the intake of highly palatable foods in humans and animals,[241-245] even in the absence of hunger.[243, 246] Emotional or stress-induced food consumption seems to be characterized by impulsivity and altered reward sensitivity, associated with dopamine dysregulation underlying incentive salience (see Ref.  for a review).
Some researchers propose that chronic stress and repeated food restraint seem to have independent and possibly synergistic effects on increasing the reward value of highly palatable foods. When having the option[241, 243] individuals will favor foods with high fat and/or sugar content—so-called “comfort foods”—during times of exposure to stress. Elevated HPA axis activation, palatable food intake, and the consequent accretion of abdominal fat may serve as feedback signals that reduce perceived stress, thus reinforcing the stress-induced food intake.
Animal research demonstrates the existence of stress-induced functional changes within the PFC, such as behavioral flexibility,[251, 252] working memory, and the recall of conditioned fear extinction.[253-255] Therefore, persistence of inappropriate behaviors, such as increased food consumption, could happen during exposure to chronic stress, as a result of stress-induced structural changes within the PFC.
Persistent programming of the HPA axis has been proposed as a mechanism to explain the association between impaired fetal growth and later development of the metabolic syndrome and hypertension in adulthood. Indeed, studies suggest that birth weight is related to the HPA axis response to acute stress, being increased or decreased in relation to normal birth weight individuals depending on the age and conditions of testing.[257-261] Nevertheless, one could surely propose that growth-restricted individuals have a differential HPA response to acute stress when compared to their normal birth weight counterparts. This alteration in HPA axis activity in low birth weight individuals may also influence their feeding behavior and food choices as discussed above. Figure 2 summarizes the neurobiological aspects involved in food intake and preferences.
Fetal programming of food preferences—emerging evidence in humans
In the past 5 years, evidence has emerged showing that besides homeostatic adaptations and their influence on energy intake/expenditure regulation and consequent increased weight gain, hedonic feeding may also be affected by variations in the early environment. That is, beyond appetite, spontaneous food preferences could also be programmed. For instance, adults exposed to the Dutch famine in early gestation are more likely to consume a high-fat diet (defined as more than 39% of energy from fat). In young adulthood, Brazilian women born with severe IUGR prefer to eat more carbohydrates than women born nongrowth restricted, and this finding persists after adjustment for several confounders, such as maternal income, smoking, and education level at the time of delivery, and participants’ smoking, education level, current body mass index (BMI), and physical activity. Last year, an investigation in individuals from the Helsinki Birth Cohort Study aged 56–70 years of age showed that as birth weight and/or ponderal index at birth increases, the intake of fruits and berries is also increased, while the percentage of intake from fat decreases. Recently, another study comparing young adults aged 19–27 years of age who were born at very low birth weight (VLBW) with term-born controls demonstrated that VLBW subjects have lower mean daily intake of vegetables, fruits, berries, and milk products, resulting in a lower daily intake of calcium, vitamin D, and cholesterol.
While the finding of an association between fetal adversity and food preferences in adulthood is of great interest, the direction of causality in this relationship could be questioned: it may be that events happening during fetal development trigger adaptive metabolic changes that secondarily influence eating behavior over time. Therefore, we have studied feeding behavior in infants and young children in relation to birth weight, before metabolic changes are manifested. We were able to see that among 3-year-old children with normal birth weights, girls show a greater ability to delay food rewards than boys; in contrast, among children with IUGR, there is no such differential ability between girls and boys. In addition, in these girls, impulsive responding toward a sweet reward predicts both increased consumption of palatable fat and higher BMI at 48 months of age. These findings suggest that in girls, the quality of fetal growth may contribute to impulsive eating, which may promote an increased intake of fats and consequently higher BMIs.
Perhaps more compelling evidence for the fetal programming of food preferences is a study performed in 1-day-old preterm infants that received 24% sucrose solution or water for evaluation of taste reactivity. The affective pattern of taste reactivity components reflects palatability or sensitivity to the hedonic signaling (i.e., pleasure) associated with the ingestion of a palatable food. A highly positive correlation was shown between fetal growth and the hedonic response to sweet solution, but not to water. The more intense the IUGR, the lower the frequency of the hedonic response observed, suggesting that IUGR is linked to a decreased sensitivity to the enjoyment elicited by the sweet taste, possibly leading to overconsumption of this and other types of palatable foods when trying to reach a higher degree of pleasure. Recently, two other studies have also described altered food preferences and feeding behavior related to IUGR.
Despite the different outcomes described in the various studies above, all these results point to the idea that the quality of the fetal environment affects the development of food preferences, influencing food choices persistently through to adulthood. In general, an adverse fetal exposure will be related to less healthy choices later in life: foods rich in fat or sugar, and fewer fruits and vegetables. This becomes of great importance, considering the developmental origins of adult overweight, type II diabetes, cardiovascular diseases, and metabolic syndrome discussed above. The chronic, persistent alteration in food preferences in individuals exposed to fetal adversity possibly contributes to the development of these diseases.
It is noteworthy that these studies were performed in different parts of the world, including developed and developing countries, in times of war or not, from different climate types, and very different ethnic backgrounds. This suggests that the biological strength of the association between fetal adversity and the development of food preferences may overcome the role of other important environmental variables, such as culture, food exposure, and habituation in influencing food choices (and the risk for obesity thereof) over the lifetime.
Large cohort studies and meta-analyses suggest that childhood obesity per se is not an independent risk factor for adulthood mortality or morbidity and for chronic diseases, such as type 2 diabetes, metabolic syndrome, hypertension, dyslipidemia, and carotid artery atherosclerosis.[268-270] On the other hand, the tracking of BMI across the life span does enhance the risk for these outcomes.[269, 271, 272] In other words, overweight or obese children who became nonobese by adulthood have a prevalence of these conditions that is similar to that among persons who were never obese. This emphasizes the importance of both primary and secondary prevention strategies aimed at reducing weight in childhood and persistently thereafter.
Very frequently, families and obese children are aware of the need for healthy eating and exercise but are unable to translate this knowledge into weight loss. There may be children who need special help to overcome obesity. Population-based measures such as decreased food advertisement, school meal reform and physical activity programs, exercise-friendly environments, and community-based exercise incentives have been shown to be successful to varying degrees, but it is not known whether the early life story influences the responses to these interventions.
Behavioral modification to overcome obesity has been shown to be poorly effective. Understanding of the neurobiological mechanisms that are clinically relevant for the modulation of food intake and preferences in individuals that suffer early deprivation may ensure the development of more efficient behavioral/nutritional interventions. In addition, very early counseling (e.g., childhood nutrition) may have potential for long-term obesity prevention.
The ideal perinatal nutrition after fetal growth restriction is another area to be researched in more detail. While survival and cognitive development are improved with optimized growth during this period,[60-62, 275] this should not be achieved through overfeeding or parental permissive behavior toward unhealthy foods, which have consequences on their own.[55-58, 276, 277] It has to be considered that about 90% of the children born after fetal growth impairment will experience compensatory catch-up growth in infancy. This necessarily involves increased caloric intake, which can be a result of the programming of appetite[22-24] and energy expenditure,[15-21] but also possibly increased food preferences toward highly palatable foods, as discussed above. In the pediatric follow-up of these children, careful nutritional advice should be given to families, encouraging breastfeeding (including durations of exclusive and total breastfeeding)[279, 280] and timely introduction of high-quality complementary foods (containing micronutrients and essential fats), as families of low birth weight children tend to feed them differently.
Improving maternal nutritional status during pregnancy to reduce fetal growth restriction seems to be the most important focus for policy making. Moreover, researchers show that the sensory environment in which the fetus lives changes as a function of the food choices of the mother, as dietary flavors are transmitted and flavor the amniotic fluid. Similarly, breast milk contains molecules derived from the mother's diet. Therefore, the maternal diet during pregnancy and/or nursing may also play a role in the programming of offspring feeding choices, and very early counseling—even prenatal—should reinforce the importance of healthy food intake and physical activity using a family-based approach.
However, it is still unclear if this type of intervention can effectively alter the long-term health outcomes in these children. The scientific evidence reviewed here justifies investments in prevention research and programs to support families and communities in nurturing healthy children, potentially offering healthier environments for the next generations.
We thank Professor Laurette Dubé for kindly reviewing a draft of the manuscript.