Early Life Energy Balance: The Development of Infant Energy Expenditure and Intake in the Context of Obesity

Abstract

Purpose of Review

This review aims to provide a summary of the current knowledge on measurement tools and most recent evidence for prenatal and postnatal modulators of energy balance in young infants.

Recent Findings

The prevention of pediatric obesity depends upon curating the perfect imbalance of energy intake to energy expenditure, taking into consideration the energy needs for healthy growth. We summarize the recent evidence for the programming of fetal and infant metabolism influenced by maternal preconception health, prenatal metabolic milieu, and physical activity behaviors. In the early postnatal environment, caregiver feeding behaviors shape the extent of energy imbalance through dictating quantity and modality of infant energy intake.

Summary

There are biological and behavioral contributors to improper infant energy imbalance. Furthermore, caregiver and clinician education on overfeeding and clinical tools to prescribe and monitor infant overgrowth are absent. Ultimately, the lack of high-quality and modern research of infant energy expenditure underpins the lack of advancement in clinical guidelines and the needed prevention of pediatric obesity.

Introduction

Obesity is a multifactorial disease caused by an imbalance of energy intake (i.e., calories consumed) and expenditure (i.e., calories expended). Pediatric obesity is a public health concern, affecting 37 million children globally [1]. When present during childhood, obesity increases the risk for the early onset of noncommunicable diseases including type 2 diabetes and cardiovascular disease, and has adverse consequences on quality of life and mental health. Furthermore, the risk of maintaining obesity as an adult is 80%, exacerbating insult throughout the lifespan and leading to premature death [2].

During adulthood and states of weight maintenance, energy intake should match total energy expended – that is, the combination of resting metabolism and physical activities. The pediatric period is unique in that energy intake needs exceed total expenditures to account for the energy cost of growth – the energy deposited in newly accrued tissues. During infancy, the cost of growth is estimated to be 38%, declines to 1–4% of total energy requirements by two years of age and remains low through adolescence [3]. When energy intake fails to meet expenditures plus the cost of growth, under nutrition and stunted growth is likely. On the other hand, in the case of overnutrition, intake chronically exceeds energy expenditures and the energy required for growth (Fig. 1).

Fig. 1.

Energy Imbalance in Infants

Landmark studies show that components of energy expenditure are genetically inherited and a low energy expenditure during infancy results in overgrowth during the first few months of life [4, 5]. Furthermore, infants born to individuals with overweight and obesity have reduced energy expenditure, placing them at risk for the perpetual development of obesity [6]. Studying energy expenditure during infancy likely reflects innate human biology since individual behaviors such as physical (in)activity are not yet engrained. On the other hand, energy intake in infants, while may also be somewhat innate, is largely determined by external factors or parent/caregiver behaviors. Low energy expenditure and increased drive to eat have been identified as two independent obesity phenotypes [7]. Therefore, identifying infants at risk of developing obesity from birth may be a novel approach for the prevention of this widespread disease.

The first 1000 days of life – encompassing conception through two years of age – are crucial in the development and prevention of obesity [8–11]. Prenatal exposures in utero may promote changes to energy intake and/or expenditure through indirect mechanisms. Then following birth, caregiver behaviors shape infant energy intake, eating and physical activity behaviors. Without knowledge of energy expenditure to inform energy intake recommendations, over feeding during infancy is likely. Overfeeding during the first year of life would promote rapid weight gain, thus contributing to later obesity [9, 12, 13]. In support of this hypothesis, preclinical models and observational studies show the relationship between overfeeding in the neonatal period to later overweight and obesity [14, 15]. In this review we will discuss the assessment tools and the prenatal and postnatal modulators of energy expenditure and intake of infants in the first six months of life, as well as present shortcomings of current clinical guidelines to promote optimal pediatric energy imbalances in the context of preventing obesity.

Assessment of Energy Expenditure and Intake

Measurement of Infant Energy Expenditure

The combustion of energy for human metabolism is not a perfectly efficient system. Energy expenditure may be measured as direct heat loss through metabolic processes (i.e., direct calorimetry) or through the consumption of oxygen and production of carbon dioxide as a metabolic byproduct (i.e., indirect calorimetry). The latter is used in clinical research and healthcare settings. Daily energy expenditure can be broken down into four primary components: (1) sleeping energy expenditure, (2) resting energy expenditure – which encompasses the energy cost of arousal and basal thermogenesis, (3) the thermic effect of food – estimated at 10%, and (4) energy expenditure to support physical activities including those to support daily living or for exercise. Adult and pediatric methodologies may be applied to infant populations, with special considerations (Table 1).

Table 1.

Techniques for Assessment of Energy Expenditure and Intake

Technique	Advantages	Disadvantages	Participant Burden	Cost
Energy Expenditure
DLW	• Easily administered • May be used in low resource settings • Allows for habitual assessment over several days	• High expertise required • Availability of isotopes • Requires multiple time points/days of measurements • Requires caregiver competency with biospecimen collections	Medium	High
Hood/Canopy	• Equipment readily available in most research laboratories • Non-invasive • Short procedure length • Easy to analyze	• Only appropriate for sleeping energy expenditure • Requires participant cooperation	Medium	Medium
Size-adapted Metabolic Chamber	• Ability to obtain all four components of energy expenditure • Non-invasive	• Specialized equipment required • Longer assessments	Medium	High
Prediction Equations	• Easily obtained in large populations • Can be used by clinicians	• Low precision	Low	Low
Energy Intake
Direct Weigh	• Tools are readily available	• Disruption of mother infant interaction • Night feedings are often neglected or challenging • Precision of scale is limiting factor	Medium	Low
Deuterium Oxide Dose to Mother	• Easily administered/sample collection • Can be used in low resource settings • Doesn’t interfere with habitual feeding	• High cost of analysis • Availability of isotopes • Specialized equipment • Multiple hours for the procedure	Medium	High
Recalls and Questionnaires	• Easily obtained in large populations • Give insight to behaviors • No specialized equipment needed	• Cannot quantify intake consumed directly from the chest • Time intensive • Subjective	High	Low

Doubly Labeled Water

The Doubly Labeled Water (DLW) method is the gold-standard approach for measuring “free-living” total daily energy expenditure (TDEE)[16]. DLW captures habitual TDEE over the course of several days (typically 7–14 days) and with minimal participant burden. With DLW procedures, special water is consumed in the form of two stable, non-toxic, non-radioactive tracers (deuterium [²H] and oxygen-18 [¹⁸O]) and their independent rates of elimination are tracked via saliva, urine, or plasma samples. In the case of infants, samples can be obtained non-invasively through saliva (i.e., oral swabs) or urine (i.e., cotton swabs or sterile urine collection bags placed in diapers) collections [17, 18]. Agreement between DLW method and indirect calorimetry has been shown to be as small as 1%, although the individual variability has the capacity to be large [19]. Therefore, the error of DLW is estimated to be up to 5% [20, 21].

The DLW method has limited constraints and can be applied throughout the human lifespan, including infants in the neonatal intensive care unit (NICU) with chronic medical conditions and healthy infants in both laboratory and field settings [22, 23]. DLW can be prepared and transported to locations of study and urine or saliva samples can be obtained by parents or remote researchers. For this reason, the International Atomic Energy Agency (IAEA) recommends the use of stable isotopes in low resource settings for the study of malnutrition [24]. Importantly, DLW can be paired with sleeping energy expenditure to estimate infant activity expenditure [25]. To do so, the difference is computed between the two 24h extrapolated expenditure values as: $ActivityEnergyExpenditure(kcals∕day)=TDEE(kcals∕day)-Sleepingenergyexpenditure(kcals∕day)$. Of course, the measurement tools are not the same (i.e., stable isotopes versus ventilatory gas analysis) and errors ranges are likely to differ.

Ventilated Hood

Sedentary energy expenditures may be measured using ventilated hood approaches. In adult populations, a ventilated canopy is placed over the head of the patient to collect respiratory gases (oxygen and carbon dioxide) analyzed by a nearby metabolic cart. In adults, standard practice is to obtain measurements in a supine position during rest, but not sleep. Measurements are typically 20–45 min in duration and with usable minutes averaged (kcal/min) and extrapolated to 24 h. Due to practical considerations, canopy measurements require an infant to be asleep [26, 27]. Due to the short nature of the measurement and the confinement to sleeping energy expenditure only, other energy expenditure components cannot be measured with a ventilated hood. Infant arousal, including crying, feeding patterns, and activity, has significant day-to-day and interindividual variability, leading to results with low precision if resting energy is obtained with this method [28]. Additionally, infant movement under the canopy may compromise the seal of the device leading to inaccurate measurements, although some groups are using a canopy positioned over an infant sleeping in a car seat [29].

Size-Adapted Metabolic Chamber

Whole room calorimeters, or metabolic chambers, are constructed air-tight rooms that capture all gases expired by human participants [30]. Unlike metabolic carts, in which participants are tethered to an examination table/car seat and a specific position, a metabolic chamber allows for free movement within the 18,000 to 27,000 L room. Because a participant can perform all typical daily activities (e.g., sleep, rest, activity, eating), all four components of energy expenditure may be quantified with procedures lasting 12 to 24 h. Therefore, whole room calorimeters are the gold-standard method for measuring 24h energy expenditure (24hrEE). These methods have been extended to the development of specialized, size-adapted metabolic chambers (~ 800L) to measure energy expenditure in infants with good accuracy and precision [31, 32]. Infant metabolic chambers employ similar principles but are built on a smaller scale and allow for caregiver interaction through the use of airtight portholes for uninterrupted routine infant care – such as feeding, diaper changes, and infant soothing. Because single measurement periods can be significantly longer than that of ventilated hoods, resting metabolism can be obtained as the sum of all awake energy expenditures and sleeping energy expenditure [31]. Due to practicality of separating an infant from their caregiver, infant metabolic chamber measurements are still considerably shorter than adult 24h metabolic chamber measurements. Only one study to date has measured infant energy expenditure over the course of 24 hours [33]. This eloquent study in ten infant participants found 24hrEE of 4–6 month olds to be approximately 75 kcal/kg/day. Importantly, this seminal study also showed that energy expenditure measurements for 4 and 6 h could be extrapolated to 24hrEE and were within 3.4% and < 1%, respectively of actual 24h measurements. We recently conducted a validation study of a novel infant metabolic chamber and found that reliability was excellent for both infant extrapolated 24hrEE and sleeping energy expenditure compared to TDEE from doubly labeled water and prediction equations [32].

Prediction Equations

Prediction equations for infant energy expenditure have been published by few organizations and research groups. The most commonly known are those by the National Academy of Medicine Dietary Reference Intakes for Energy (DRI) [34, 35]. Equations for infant energy expenditure were first published in 2005 (Box: Equation 1) and included infants birth through the end of two years of age [35]. Due to lack of observed sex differences at the time of publication, there was a single equation for both males and females. Equations were recently revised with the recent 2023 DRI and now include a sex-specific equation for infants of the same age range (Box: Equations 2 and 3) [34]. In a small modern cohort we recently found the accuracy of prediction equations to have a mean difference of only 6% in infants aged 1–3 months compared to measured TDEE through DLW [32]. However, as with many population based prediction equations, precision is modest, at best. Prediction error ranged from approximately −100 to + 125 kcals, which amounts to nearly 25% of total daily infant energy expenditure. Despite improvements to the prediction equations over the past 20 years, we hypothesize three primary reasons for low precision with the recent updates: (1) age of reference data (most data are greater than 20 years), (2) the range of infant age included in a sole equation is large, despite varying energy costs of growth from birth to three years of age, and (3) the lack of inclusion of physical activity or mobility. Unlike all other populations included in the DRI, activity quantification is glaringly absent in infant equations. While infants indeed have limited activity, we found that the addition of daily physical activity improves precision. However, we acknowledge that differences in activity could likely be accounted for by age, since mobility has a tremendous increase from birth to three years of age.

Equations 1–3: Estimation of Energy Expenditure for Infants Aged 0–2 yrs

Equation 1: 2005 DRI Equation for Energy [35]

$$TEE=89\times \left(weight\right)-100$$

Equation 2: 2023 DRI Equation for Energy for Males [34]

$$TEE=-716.45-(1.00\times age)+(17.82\times height)+(15.06\times weight)$$

Equation 3: 2023 DRI Equation for Energy for Females [34]

$$TEE=-69.15-(80.0\times age)+(2.65\times height)+(54.15\times weight)$$

Weight is entered in kilograms; height is entered in centimeters; age is entered in years

Measurement of Infant Energy Intake

Assessment of infant energy intake is challenging, particularly when consumed directly from the lactating individual (Table 1). While bottle feedings can be more easily quantified, there are several assumptions including the estimation of the energy density of human milk and proper formula preparation (i.e., water to formula ratio).

Direct Weighing

Direct weighing represents the simplest way to assess milk intake objectively. The method is straightforward, inexpensive, does not utilize specialized equipment and involves weighing an infant using a calibrated scale prior to and immediately following a feeding. One major benefit of direct weighing is the technique can be applied to all ingestive means (e.g., human milk consumed directly from the body – herein referred to as “chestfeeding” – or bottle, formula, gastric tube feedings). The accuracy of direct weighing is good at an average of 1.3 mL per feeding, but the precision is poor. One study found −12 to + 30 mL difference across approximately 100 infants [36]. Rationale for these discrepancies include the lack of sensitivity in scales to detect small changes in infant weights and the underestimation due to production of urine and spit up during feeding. This has led investigators to recommend that direct weighing not be used in clinical practice [36]. Furthermore, accurate representation of habitual intake may require several feedings, including overnight feedings, since variability in intake is common [37].

Deuterium Dose to Mother

The deuterium oxide dose-to-mother (D₂O DTM) technique can be used to measure the quantity of human milk that an infant consumes [38–40]. Importantly, the D₂O DTM technique can be used during exclusive chestfeeding or chestfeeding concomitant with infant formula or solid foods supplementation. The D₂O DTM technique assumes the mother and child as a system to estimate the flow of a non-radioactive labelled water tracer (deuterium) from the mother to the infant through human milk. Assessment of deuterium in the infant can be obtained similar to DLW procedures, requiring both isotope dosing and collection of both saliva and urine samples from not only the infant but also the mother. Due to similarities in procedures, the IAEA also recommends the D₂O DTM technique to understand nutritional needs in low resource settings [24].

Food Recalls and Questionnaires

Food recalls may be used to document the quantity of human milk or formula consumed through a bottle. While it’s not possible to utilize these tools to quantify milk intake directly from the chest, questionnaires and recalls may be used to determine frequency and duration of feedings to estimate intake. Energy estimations are more easily calculated for formula intake as opposed to human milk, however the assumption that the formula was prepared as intended is necessary. Thus, using a food recall method has the potential to under- or overestimate energy intake if bottles are prepared using a slightly different powder to water ratio than manufacturer instructions.

Preestablished questionnaires are also used to estimate intake and ingestive behaviors. For example, the Questionnaire on Infant Feeding was developed to categorize infants on the basis of nutrition source – that is, own biological parent human milk, non-biological/donor human milk, infant formula, or other, feeding duration, as well as feeding mode – at the chest/body or via bottle [41]. Other questionnaires, such as the Baby Eating Behavior Questionnaire (BEBQ), gathers information about infant feeding behaviors [42]. The BEBQ is adapted from the Child Eating Behavior Questionnaire, which is a validated questionnaire for children aged 3–13 years and assesses eating behaviors across eight subscales [43]. The BEBQ assesses four appetitive traits during the first six months of life: (1) enjoyment of food, (2) food responsiveness, (3) slowness in eating, and (4) satiety responsiveness. Questionnaires such as the BEBQ that measure ingestive behaviors are a valuable tool to utilize alongside objective measures of energy intake.

Prenatal and Postnatal Modulators of Energy Expenditure and Intake

Fetal developments in utero together with nutrition during the postnatal period modulate the physiology of energy expenditure. On the surface, feeding variety to the newborn appears limited, given the offerings of only human milk or infant formula milk.. However, recent data support nutritive (referring to macronutrients and micronutrients) and non-nutritive (referring to molecules that act upon metabolism in an indirect pathway such as signaling hormones) components found in human milk may act upon energy balance systems. Furthermore, feeding behaviors of caregivers begin to mold behaviors of energy intake as early as day one of life.

Fat-Free Mass

Fat free mass (FFM) and energy expenditure are strongly related. In adults, FFM is the primary regulator of energy expenditure, accounting for 60–90% of the variance in resting energy expenditure (REE) [44, 45]. In infants, we and others have found a correlation between FFM and energy expenditure, albeit to a lesser degree than adult populations [32, 46]. A recent analysis of data from over 6,000 individuals from the International Atomic Energy Agency (IAEA) DLW database, identified infancy as a period with exceptionally high FFM-adjusted energy expenditure. During the first year of life, FFM adjusted energy expenditure is nearly 50% higher than values observed in adults [47]. This is likely attributed to the immensely high energy cost of growth during this phase of development.

In keeping with energy homeostasis, it is logical that energy intake is also mediated by FFM. Indeed, food intake largely depends on FFM independent of fat mass in adults [48]. Using novel ad-libitum food intake tracking, FFM was associated with caloric intake, even after adjusting for fat mass [49]. The “Drive to Eat” hypothesis is based on the premise of the body’s desire to maintain energy homeostasis. Given that FFM is the largest contributor to REE, which is the largest contributor to energy expenditure, energy intake must be proportional to levels of FFM in order to prevent the wasting of muscle and preserve lean tissue [50]. The “Drive to Eat” hypothesis was recently proposed in infants [46]. Similar to adults, it was found that both quantity of milk intake (determined by test-weighing) and estimation of energy intake were correlated with FFM but not fat mass in 12-week-old infants [46]. Given the relevance of FFM to both energy expenditure and intake, understanding the prenatal and postnatal modulators of FFM will be highly relevant to understanding the etiology of early life energy balance.

Brown Adipose Tissue

Unlike white adipose tissue which serves as an energy reservoir in the postnatal environment, brown adipose tissue (BAT) is a multilocular, mitochondria rich, and highly thermogenic tissue mediated by the expression of the tissue-specific uncoupling protein 1 (UCP1). Fetal development of BAT begins in late utero to protect against hypothermia in the postnatal environment [51]. Infants have four primary BAT depots to increase heat production surrounding vulnerable structures including the perirenal, supraclavicular, subscapular, and paraspinous regions. Not all adults retain functional BAT after childhood, yet higher rates of energy expenditure, improved health, and a lower BMI are observed in individuals with detectable and functional BAT [52, 53]. While associated with reduced obesity and improved health in older children and adults, the link between BAT and infant health is currently unknown. However, it is hypothesized that BAT expends large amounts of energy at rest during early infancy. BAT is likely a meaningful contributor to energy expenditure in infants given histological characteristics suggesting a capacity for significant thermogenesis [54]. Despite the known properties and physiological relevance of BAT in neonates, the factors that contribute to BAT mass and activity at birth are only beginning to be unraveled. Moreover, while the measurement of BAT in infants has recently been validated, the extent to which neonatal BAT is activated and stimulates thermogenesis at birth has not been studied [55, 56].

Nutritive and Non-Nutritive Components of Milk

Unlike infant formula, which has a standard composition – human milk is dynamic. Human milk changes drastically over the first year of life starting with colostrum, (abundant in protein) to mature milk (abundant in fat) which typically develops around postnatal day six. At birth exclusively chestfed infants consume < 20 mL per day, progress rapidly to 500 mL/day by the time mature milk is developed, and further to approximately 750 mL/day by three to six months of age [37]. While quantity increases, landmark studies using bomb calorimetry show a decline in energy content over the first six months of life [57].

Human milk not only is responsive to infant age, but also to maternal physiological and behavioral factors. Lactating individuals with overweight and obesity have higher levels of glucose, insulin and lipid in their milk [58, 59]. These metabolites transfer to the infant and shape early metabolism. In preclinical models, when mice born to healthy lean dams are cross-fostered at birth to suckle milk from dams with obesity after birth have significantly different metabolic phenotypes than those pups remaining with lean healthy dams, including insulin sensitivity, body weight, and adiposity [60]. Independent of body mass index (BMI), maternal diet can also influence human milk directly – as in the case with many micronutrients – or through intertwined pathways [61]. For example, a maternal diet high in fat, and even more so, high in sugar, results in increased triglycerides in human milk, and at the expense of a concomitant protein reduction [62]. A high fat maternal diet may also increase caloric density and lipid content in human milk [63]. Correlations have been found between protein, carbohydrate, and total energy content in the maternal diet and corresponding levels of macronutrients in human milk [64].

Evidence suggests that properties of human milk may also influence infant energy expenditure through signaling hormones. While BAT indeed primarily acts as a thermogenic organ, recent research identifies BAT as having an endocrine function through signaling metabolite related to thermogenesis referred to as “BATokines” (i.e., BAT related lipokines) [65]. In recent years, BATokines have been identified in both infants and in human milk. One thermogenic activator, bone morphogenetic protein 8B (BMP8B) acts directly on BAT, increasing sensitization to sympathetic nervous system signaling [66]. BMP8B has been found in the neonatal blood stream and appears to be related to activation of infant BAT in humans, but mechanisms to increase this in utero or postnatally are yet to be discovered [67]. Another BATokine, 12,13-dihome acts to increase fatty acid uptake into BAT, thus increasing activation and reducing circulating triglycerides. Novel data show the presence of 12,13-dihome in human milk and its association with lower infant fat mass at 1 month of age [68]. 12,13-dihome may be modulated by behaviors of the lactating person. While it is not associated with maternal BMI, concentrations increase in human milk following an acute exercise bout, indicating a mechanism to increase BAT activity in infants through lactation. Lastly, preclinical models show that dams on a high fat diet during lactation downregulated BAT thermogenesis in offspring. However, these observed effects may be primarily due to aforementioned reduction to insulin sensitivity (resulting in decreased glucose transport to BAT) associated with a maternal high fat diet [69].

Influence of Maternal Exercise on Infant Energy Expenditure

Maternal exercise may influence infant energy expenditure through changes in fetal body composition (FFM development) or indirectly via whole body energy metabolism [29]. A prenatal lifestyle intervention incorporating dietary modifications and structured exercise has been shown to upregulate fetal FFM development. In a study of more than 300 mother-infant dyads, a prenatal lifestyle intervention prescribing a Mediterranean diet and physical activity produced infants with greater total body and abdominal lean mass than infants born to those without lifestyle intervention [70]. It is important to note that the effects of diet alone on infant FFM have not been reported. In individuals with prepregnancy overweight or obesity, a prenatal intervention to reduce gestational weight gain through dietary modifications and physical activity resulted in 98 g and 105 g higher FFM and lean mass, respectively [71].

Through studying mesenchymal stem cells (MSCs) harvested from infant umbilical cord tissue, researchers can explore cellular pathways and processes that underpin clinical outcomes. Using this novel approach, it is understood that infants born to mothers with obesity have umbilical MSCs with greater potential for adipogenesis, as shown by a 30% greater Oil Red O stain (ORO), 50% greater peroxisome proliferator-activated gamma (PPAR-γ) protein content. In return, infants born to those with obesity showed less potential for myogenesis, as shown by 10% lower total B-catenin protein content, a major driver of myogenic differentiation [72, 73]. However, exercise prescription during pregnancy has the potential to alter the fate of progenitor cells towards myotubes, thus increasing fetal skeletal muscle and FFM development. Indeed, MSCs of infants born to individuals prescribed structured exercise throughout pregnancy (prescription of 60–80% of maximal oxygen uptake for a minimum of 20 weeks) show improved insulin signaling, enhanced mitochondrial respiration, and higher rates of fat oxidation [74]. Taken together, these data indicate potential for intrinsically higher energy demand in infants exposed to exercise throughout pregnancy.

The benefits of exercise on human milk are only starting to be unraveled. A novel human milk oligosaccharide and lipokine, 3′-siallylactose (3’SL), is upregulated with maternal exercise training, crosses to human milk, and is transferred to the infant. Preclinical models reveal exciting data – cross fostered offspring born to non-exercise trained dams during pregnancy and transferred to exercise trained lactating dams, show reduced fat mass, improved glucose tolerance, and decreased fasting insulin [75]. These effects were glaringly absent in exercised-trained 3’SL knockout dams. In humans, 3’SL content in milk is positively correlated to average activity and steps per day, and negatively correlated to BMI. Whether 3’SL directly influences energy expenditure is unknown but a worthy avenue of exploration given it’s unremarkable influences on infant metabolism [75, 76].

Determinants of Infant Energy Intake and the Obesity Risk Debate

It is widely agreed that human milk is the optimal food source for all infants, regardless of maternal health status. Concomitantly numerous health, societal, cultural, and socioeconomical reasons exist as to why many parents are unable to feed their infant human milk. The risk of obesity development between infants fed formula milk versus human milk has long been debated. Data from a large systemic review including nearly 70,000 infants concluded that exclusive feeding of human milk results in a 22% reduction in childhood obesity risk compared to infants fed formula milk [77]. Infants fed formula demonstrate more rapid growth during the first year of life – a primary predictor of later childhood obesity – compared to those fed human milk [13, 77, 78]. It is hypothesized that obesity risk between the two sources of food ultimately rest upon three primary factors (1) hormonal (Lipokines, BATokines, Oligosaccaharides), (2) macronutrient differences, and (3) behavioral (Table 2).

Table 2.

Three primary hypothesized contributors to obesity risk between human milk and formula fed infants

Hormonal	• BATokines (discussed in 3.3) • Lipokines/Oligosaccharides (discussed in 3.4)
Macronutrients	• Protein content (discussed in 4.1)
Behavioral	• Overfeeding (discussed in 4.2) • Responsive versus scheduled feedings (discussed in 4.2) • Adherence to AAP guidelines (discussed in 5.0)

Macronutrients

The macronutrients of infant formula are distinctly different from human milk. Of relevance to the obesity debate is protein content. Mature human milk contains between 0.7–1.0 g protein/100 mL, whereas infant formula contains between 2.0–3.5 g of protein/100 mL [79]. Since infant protein requirements are based on human milk, protein intake in formula fed infants far exceeds recommendations [80]. This presents a major challenge. Recent evidence supports that the excess protein contributes to obesity risk in infants [81, 82]. In a randomized clinical trial, formula fed infants received either a formula containing high-protein or low-protein, but within recommendations, for the first year of life. Children who consumed the high-protein formula as infants had a significantly higher BMI and a 2.5 × greater likelihood of having obesity at age six, compared to those on the low-protein formula. Interestingly, anthropometric measurements were similar between children fed the low-protein formula during infancy compared to those who received human milk as infants [83]. One hypothesis to the protein overfeeding conundrum is the relationship between dietary protein and enhanced secretion of insulin and insulin like growth factor-1 (IGF1), both of which are upregulated in formula fed infants compared to infants fed human milk [84].

Behavioral

The protection of obesity from human milk feeding is partially mitigated when expressed milk is fed from a bottle, suggesting that feeding behaviors also play a critical role. A study of over 2,000 infant-mother dyads found that compared with infants fed human milk exclusively from the chest, BMI z-scores at 3 months were 0.12 higher in infants fed some expressed milk from a bottle, 0.28 higher for infants fed partially human milk, and 0.45 higher for exclusively formula fed infants [85]. Compared to infants fed human milk from the chest, infants fed human milk exclusively from the bottle gained an average of 89 g per month more over the first year, and weight gain was associated with the number of bottle feedings [86]. This is highly relevant given that the majority of “exclusively chestfed” infants receive expressed milk from a bottle [85]. If bottle feeding alone is a culprit in obesity development, infants fed formula are at a great disadvantage.

Self-regulation – the ability of an infant to respond to internal satiety and hunger cues – when fed from a bottle is limited, regardless of milk source. Infants bottle fed earlier in life are more likely to overconsume milk to empty a bottle later in infancy compared to infants fed directly from the chest [87]. Such infant behaviors increase the likelihood of being overfed. Mother-infant dyads participated in a randomized trial at 1, 2, 5, 7, 10, and 12 months of age to demonstrate capacity for self-regulation of energy intake. Dyads were randomized to (1) feed every hour for six hours or (2) feed based on infant demand (e.g., crying, hunger cues). Infants in the hourly feeding group were unable to regulate their intake, resulting in approximately 5 kcals/kg more milk being consumed per 6 h window [88]. Thus, responsive feeding – allowing the infant set their own feeding schedule based off hunger cues – is a strategy to prevent excess energy intake during the first year of life. Indeed, responsive feeding by caregivers is associated with positive eating behaviors in children, including increased self-feeding and enjoyment of a variety of foods, whereas scheduled and/or restrictive feeding – feeding based off caregiver cues – is associated with the development of obesity [89, 90]. Since infant formula can only be fed using a bottle, creative strategies such as utilizing opaque bottles to remove caregiver influence on feeding, is a novel strategy to reduce over feeding [91].

Over feeding has been shown to be established as early as day one of life and this is driven by formula recommendations [92]. At birth, the stomach capacity of a neonate is an average of 20 mL [93]. Chart reviews of over 1,000 formula fed infants during the first 24 h of life show that 93% were overfed (using a conservative approach of fed in excess 30 mL per serving) at least once. The amount of over feedings during the first 24 h of life was proportional to obesity risk. Infants overfed five and seven times out of the first seven feedings were five and seven times, respectively, more likely to have overweight and obesity at 4 years of age [14]. Using this information compared to colostrum fed directly from the chest (approximately 20 mL total on day 1 of life), formula fed infants may be setting the stage for the potential of a lifetime of overfeeding. Furthermore, we have recently shown that caregiver bottle preparation of infant formula contains up to 11% more calories per bottle due to the over scooping of powder and in simulation studies the extent of this over-dispensing could drive growth from the 50th to the 75th percentile for male and female infants [94]. More clear formula preparation instructions and/or pre-dispensed formula containers may help to reduce the likelihood of infant formula overfeeding [95].

Future Directions to Advance Infant Energy Balance and Move Towards Precision Nutrition

There are critical shortcomings in the guidelines for infant energy intake. The American Academy of Pediatrics (AAP) recommends exclusive human milk feeding for the first six months, followed by continued human milk feeding until at least 12 months, and the introduction of solid complementary foods at six months. While ideal, public health policies including the lack of lactation support coverage by insurance providers and prolonged paid leave limit the feasibility of achieving these guidelines by many in the United States (US). In a large diverse US cohort, 77% did not adhere to the AAP guidelines within the first year of life and 39% introduced complementary foods prior to 6 months of age [96]. An international study showed that 62% of infants were fed complementary foods prior to 6 months of age, of those, and 6% were introduced prior to 3 months [97].

Feasibility aside, energy prescription is also ambiguous, not tailored to individual needs, and therefore can create confusion among caregivers and pediatric care providers. The AAP provides recommendations on suggested ounce intake across age groups. According to the AAP, bottle fed infants should consume 3–4 oz per feeding across 6–7 feedings per day by one month of age and progress to 6–8 oz per feeding across 5–6 feedings per day [98]. Using these two recommendations alone, parents and caregivers must choose an appropriate amount of milk to feed their 1 month old infant between 18 or 28 oz per day (where the difference is nearly double the quantity) and 6 month old infant 30–48 oz per day. The alternative to these recommendations come from the 2023 DRI, which build upon energy expenditure models (discussed in 2.1.4) [34]. The DRI are more complex and provide individualized caloric intake prescriptions derived from weight, length, age, biological sex, and the estimated energy cost of growth. While more personalized, recommendations are far from ideal due to variation in energy expenditures, which are the premise of the equations. Indeed, the 2023 DRI committee recognizes the lack of available data and recommends more infant research to inform future updates to promote adequate infant growth. While beyond the scope of this review, it is important to note ambiguity in caloric prescription of human or formula milk beyond 6 months of age and upon the introduction of complementary foods.

Growth charts from the World Health Organization (WHO) and Centers for Disease Control (CDC) provide a tracking tool for infant growth. However, current usage of growth charts is primarily to ensure that infants are consuming enough nutrition (i.e., they are not underfed) rather than used as a tool to prevent overnutrition and to assist caregivers with developing healthy eating behaviors for children. The AAP suggests that infants maintain their trajectory along growth charts as an indication of adequate nutrition but provide little support for the use to identify overfeeding and rapid weight gain. Pediatric care providers lack the tools to prescribe energy intake to reduce infant obesity risk. Thus, it is not surprising that pediatric care providers are not comfortable with handling overfeeding and the management of rapid or excessive weight gain [99]. Precision infant nutrition could be tailored by linking energy intake prescription with infant growth charts. Given that these topics are not adequately discussed in clinical care, it is also not alarming that parents are more concerned about underweight than overweight in infants [100].

Conclusion

Obesity is one of the major challenges to human health. Human energy expenditures and intakes are established early in life and shaped by the intrauterine and postnatal environments. The maternal–fetal crosstalk in utero becomes the maternal-infant unit upon delivery and initiation of human milk feedings. Caregiver feeding behaviors have the potential to shape infant intake behaviors for the remainder of early pediatric years. To prevent the development of obesity and obesity-related morbidities in childhood and beyond, the first step is to conduct robust clinical studies in infants to advance the understanding of biology (parental influence, genetics, and uterine environment) and behaviors on the development of energy expenditure and intake and their independent and synergistic contributions to human health and metabolic diseases such as obesity. Once these are understood, early life interventions to optimize the energy balance for the prevention of obesity and promotion of healthy growth and development can occur. Infants are our most vulnerable population, and infant nutrition research demands more attention.

Data Availability

No datasets were generated or analysed during the current study.