Evaluating the pathways linking complementary feeding practices to obesity in early life

Abstract

The complementary feeding period, when solids and liquids other than breast milk and formula are added to the infant diet, is an important development window shaping infant growth and weight gain, metabolic development, and lifelong eating practices. Yet, relatively little is known about how the timing and types of foods offered to infants at this stage may shape their growth and subsequent risk of developing obesity. This narrative review describes the existing literature on complementary feeding practices, discusses potential biological and behavioral pathways linking complementary feeding practices to the development of obesity, and offers potential avenues for intervention. While further research is needed to more fully understand optimal complementary feeding practices, existing evidence supports the importance of healthful early feeding practices in the physiological and behavioral regulation of growth and metabolism and the need for early intervention to prevent the development of obesity.

The prevalence of overweight and obesity in early childhood is increasing globally and the number of preschool-aged children who are overweight or obese is expected to reach 60 million worldwide by 2020.¹ Given the tendency of overweight to track from early childhood into adulthood,^2–4 early identification of modifiable factors, such as infant feeding practices, and how these influence weight gain, is critical, as is intervention on such factors. Much attention has focused on the potential role of breastfeeding in protecting against obesity,^5–8 but less research has examined the association between complementary feeding practices and later health outcomes. Yet, the complementary feeding period – ie, the stage when solids and liquids are added to the infant diet – is an important developmental window, when nutrient needs for growth are high, metabolic physiology is shaped, and food-related preferences and eating behaviors develop.^9,10

Despite recommendations from the World Health Organization that infants should not receive solid foods until 6 months of age and from the American Academy of Pediatrics (AAP) that introduction should occur “around 6 months of age,” research has shown almost all children in high-income countries are introduced to solid foods before 6 months,^11–13 with a varying proportion introduced to complementary foods considerably earlier than 6 months of age. In the United States, for example, around 40% of infants receive complementary foods by 4 months of age.¹⁴ Here, infant diets tend to contain energy-dense, nutrient-poor foods and beverages, with high intakes of refined sugars, protein, saturated fats, and sodium, and lower-than-recommended amounts of fruits and vegetables.^15–18 Such complementary foods and liquids may be problematic for both short-term and longer-term health. The early introduction of foods may increase morbidity from concurrent gastrointestinal and respiratory infections,^19–23 while poor-quality early foods may increase weight gain and contribute to risk of future obesity.^18,19,24

Local food availability, sociocultural norms, and a host of family and household sociodemographic characteristics shape complementary feeding practices. Consequently, diets followed during the complementary feeding period show wide variation in the timing and sequencing of foods offered; the types, amounts, and diversity of complementary foods; and the balance between commercially- and home-prepared foods.²⁵ Relatively little research has assessed the health impacts of these differing feeding strategies or the biological and behavioral pathways linking diet composition and quality to long-term obesity risk. Thus, this narrative review describes the association between complementary feeding practices and obesity, discusses possible mechanisms linking early diets to long-term health, and suggests possible avenues for intervention.

METHODS

This narrative review discusses results from observational studies, randomized control trials, systematic reviews, and narrative reviews indexed on PubMed or Google Scholar and/or listed in the references of other included papers. The review of the literature focused on the timing, patterning, and nutrient content of complementary solids and liquids in relation to pediatric obesity and the evidence, from human studies, of the behavioral and biological pathways linking complementary feeding to pediatric obesity. The following studies were excluded: qualitative studies, studies focused predominantly on non-growth outcomes (eg, cognitive development, micronutrient status), and studies where child diet was measured only after infancy. Since relatively little work on complementary feeding and the development of obesity originates from low- and middle-income countries, where undernutrition remains an important problem alongside the growing prevalence of overnutrition and the existing research has been reviewed elsewhere,^18,26 this review focuses on studies from high-income countries. Similarly, studies included in the review primarily evaluated healthy, term infants since there is a lack of research examining complementary feeding practices in preterm infants in relation to the development of obesity (though see a recent systematic review by Vissers et al²⁷).

TIMING OF SOLID FEEDING AND RISK OF OBESITY

Several recent systematic reviews^28–30 have examined whether early (variously defined as <2 months, <3 months, or <4 months) or later (>6 months) introduction of solid foods is associated with overweight in infancy or childhood. Overall, these reviews find that there is no consistent association between the timing of solid food introduction and overweight, body mass index (BMI), or other measures of body composition in children. One systematic review²⁹ revealed 13 out of 21 studies found no association between the timing of solid food introduction and overweight; 3 found no significant results after adjustment for maternal, infant, and/or household characteristics, such as race/ethnicity, socioeconomic status, and maternal obesity; and 4 studies found significant associations between solid food introduction before 3 or 4 months of age and later overweight in adjusted models. A limited number of studies further suggest that later introduction of solid foods (>6 months) may be associated with a greater risk of overweight in childhood.^31–33 These inconsistent results likely stem from the variety of outcomes measured (BMI vs weight-for-age), range of ages measured at follow-up (12 months to adulthood), and the type of milk feeding (breast-, formula-, or mixed-feeding). The significant associations seen in the systematic reviews and in several more recent papers are from studies conducted with 3000 or more participants,^34–39 suggesting that large sample sizes may be needed for sufficient power to detect an association. Significant associations have also been seen in studies assessing weight outcomes at younger ages,⁴⁰ suggesting that effects may not persist past infancy or that childhood factors are more important in the etiology of obesity.

Importantly, these findings may differ by type of milk feeding – a factor that was not controlled for in all studies. Breastfeeding appears to attenuate the association between the timing of solid feeding and later weight outcomes in some,^34,35 but not all, studies.^31,33,41 In their large sample of American infants, Huh et al³⁴ found that formula-fed infants who received solid foods before 4 months of age or after 6 months of age were more likely to be obese and have higher weight, weight-for-age z-scores, BMI, and BMI z-scores at age 3 years.³⁴ No significant association between the timing of solid introduction and body size was observed among breastfed infants. Contrastingly, Sun et al³³ found that the introduction of solids prior to 4 months was associated with later overweight in both fully and partially breast- and formula-fed infants, while the late introduction of solids (≥7 months) was associated with later overweight only in breastfed infants. Taken together, these studies suggest that while very early introduction of solids (prior to 4 months), or delayed introduction (after 6 months), may be associated with increased obesity risk, introduction between 4 and 6 months has little effect on later adiposity. These effects may differ for breast- vs formula-fed infants, though additional research is needed to better understand contrast effects in exclusively breastfed, exclusively formula-fed, and exclusively mixed-fed infants. Similarly, with the exception of the work by Huh et al,³⁴ the reviewed studies did not define “solid feeding” and whether this includes beverages, suggesting that more fine-grained analysis of the timing of specific food types may be warranted.

ENERGY AND MACRONUTRIENT INTAKE DURING INFANCY AND RISK OF OBESITY

Differences in energy intake between milk-fed infants and those receiving complementary foods, and the energy and nutrient content of different complementary foods, may underlie the association between solid feeding practices, both the timing of introduction and the diet quality, and infant weight gain. Limited evidence suggests that the early introduction of complementary foods may add extra energy to the infant diet. The introduction of complementary foods before 4 months was found to be associated with higher energy intake at 4 months and a higher BMI and percent body fat (%BF) in childhood in British children,⁴² though this effect was only significant in formula-fed infants. Findings from the EU Childhood Obesity Programme (CHOP) cohort of formula-fed infants revealed that energy intakes were higher among infants receiving solid foods in the first 8 months of life, even in months when energy intakes from solid foods were themselves negligible.⁴⁰

Energy intakes naturally increase across the complementary feeding period as infants’ diets transition to include a greater proportion of table foods. However, the slope of this increase varies by diet quality. A study of Australian infants documented that the energy-density of the infant diet increased by 36% from 9 to 18 months of age as infants consumed increasing amounts of sugar-sweetened beverages, meat products, and savory and sweet snack foods.¹⁶ While this study did not examine the association between these energy intakes and obesity risk, the findings that the consumption of these energy-dense, “non-essential” foods increased across time raise concerns that these “extra” calories may be displacing healthier foods in the diet. The Infant Care and Risk of Obesity Study, a prospective cohort study of low-income African-American infants, documented that age-inappropriate feeding of solids and liquids (defined as not adhering to age-specific AAP recommendations) was associated with a mean energy intake across infancy that was 100 kcal/d higher than the AAP-recommended daily intake,²⁴ with solids and juices contributing as much as 200 kcal above estimated needs, in relation to daily energy intakes, as early as 3 months of age.⁴³ A recent longitudinal study examining energy intake from shortly after birth to age 5 years in over 5500 children from a multi-country study documented that a greater energy intake in the first year of life was a strong predictor of overweight at age 5,⁴⁴ suggesting that early excess energy intakes may contribute to obesity risk in childhood.

Along with increasing energy intake, the macronutrient composition of the infant diet changes dramatically across the first years of life. Concerns have been raised that an imbalanced diet during this period may program metabolic dysregulation, contributing to greater weight gains and subsequent obesity risk.^45–47 Results from observational studies examining the association between specific macronutrients and weight and adiposity outcomes have been mixed, however. A systematic review of complementary feeding and childhood obesity risk found little evidence that macronutrient intakes, particularly those of carbohydrates and fat, during the first year of life were associated with infant weight gain or later risk of obesity.^44,48 In one of the few studies to measure carbohydrate intake, for example, no association was found between intake in the first 12 months and child BMI or adiposity at age 6 in an Icelandic cohort.⁴⁹ Similarly, fat intake from 6 to 12 months has been shown to have no effect on growth velocity or body fatness in infants or children,^44,47,50 though some evidence suggests that higher fat intake in the second year of life may be associated with a larger BMI-SDS (BMI standard deviation score) and %BF at age 7.⁵¹

In contrast to carbohydrate and fat intakes, protein intakes have received considerably more attention. Several studies, including a systematic review, document that higher protein intake before age 2, and particularly around 12 months, is associated with a higher rate of growth and higher subsequent BMI in childhood.^52–55 Other studies looking at intake earlier in infancy (ie, at 6 or 9 months) or after the first year⁴⁴ have found no significant associations,⁴⁴ however, suggesting that the timing of intake may be important and/or the impacts of early protein intake may not persist beyond early childhood.^44,48 Supporting the importance of the timing of high protein intakes, a longitudinal cohort study of over 200 infants from Dortmund, Germany, found that higher protein intake at 12 months, but not at 6 months, was associated with higher child BMI and %BF at age 7.⁵¹ Similarly, in the Generation R cohort of nearly 3000 children from the Netherlands, protein intake at 12 months was associated with greater BMI and fat mass at age 6⁵⁴ and age 10,⁵⁵ even when controlling for current diet and protein intake.

Results from these and other studies indicate that the type of protein consumed may also be important. Several studies have found that protein from animal sources, but not plants, is associated with higher weight gain^48,54; findings from studies comparing protein from dairy vs meat sources are less consistent. In one observational study, higher protein intakes from dairy (cow’s milk and cow’s milk formula) were associated with significantly higher weight, length, and weight-for-length compared to breast milk, beginning in the third month of life,⁵⁶ and were independently associated with a higher %BF at age 7.⁵¹ Two small randomized control trials further support the importance of the quantity and source of protein in the complementary diet.^57,58 In the first trial, conducted in breastfed infants from the United States,⁵⁸ infants weaned onto a higher protein, meat-based complementary diet had greater gains in weight-for-age and length-for-age than infants fed a low-protein, cereal-based diet. In the trial involving formula-fed infants,⁵⁷ infants receiving dairy-based protein had greater weight-for-length (WLZ) scores and a greater risk of overweight than infants receiving meat-based protein – a difference that persisted until age 2.⁵⁹ Conversely, other large observational studies have found no association between differences in child BMI or overweight risk and the type of animal protein source.⁵⁴ Higher protein intakes from dairy sources at 18 months, for example, were associated with higher BMI z-scores (BMIz scores) at age 7, while higher protein intakes from meat and fish were associated with greater BMI z-scores at ages 7 and 11, and overweight at age 7, in a large Danish cohort of over 22 000 children.³⁹ Taken together, these results suggest that protein intake may indeed shape weight gain and adiposity risk but that the timing of excess intakes and protein source (plant vs animal) likely moderate these effects.

DIET PATTERNS AND RISK OF OBESITY

Energy and macronutrient intake are shaped by the “whole diets” infants consume, and thus, understanding the feeding patterns associated with healthy development is critical for designing interventions to prevent obesity. Several studies have created diet quality indices,⁶⁰ performed factor analysis,⁶¹ or used latent variable methods^62–64 to categorize the range of foods infants are fed, and describe “healthy” infant diets. For example, the Avon Longitudinal Study of Parents and Children (ALSPAC) in the UK validated a 14-component “Complementary Feeding Utility Index” to assess breastfeeding, responsivity of feeding practices, timing of solid feeding, and quality of complementary foods.⁶⁰ Factor analysis of 633 one-year-old infants from 5 European countries identified 2 predominant diet patterns: one labeled as “core foods,” with strong positive factor loadings for vegetables, potatoes, fish, olive oil, and meats, and a second labeled “poor-quality fats and added sugars,” with strong positive loadings for saturated spreads, added sugars, and dessert foods.⁶¹ Using a latent variable approach, the Southampton Women’s Study (SWS) in the UK conducted a principal components analysis (PCA) to identify 2 main dietary patterns at 6 and 12 months of age: an “adherence to complementary feeding guidelines” pattern, characterized by high consumption of fruit, vegetables, cooked meat and fish, and other home-prepared foods and low consumption of commercial baby foods, and an “adult foods pattern,” characterized by a high consumption of bread, snacks, sweets, breakfast cereals, and squash, and a low consumption of breastmilk, baby rice, and cooked or canned fruit.⁶⁵ Similarly, an Australian study used PCA to identify “core” and “non-core” food patterns at 14 and 24 months of age from a large number of foods and beverages consumed.⁶² As in the SWS, the “core” patterns were distinguished by a “healthier” pattern of high fruit, vegetable, grain, and meat consumption and a less desirable pattern characterized by a high consumption of white bread, spreads, sweetened beverages, and snacks. In contrast to these “variable-centered” approaches, several recent studies have used latent class analysis (LCA) to identify distinct patterns within individuals and then group infants with similar patterns.^63,64 Using this approach, Hohman et al,⁶³ for example, identified the following 5 classes of infant diets based on the combination of milk feeding and complementary foods at 9 months: (1) Breastfed, Fruits and Vegetables (BFV), (2) Breastfed, Low Variety (BLV), (3) Formula, Fruits and Vegetables (FFV), (4) Formula, Low Variety (FLV), and (5) Formula, High-Energy Density (FHED).They were also able to show that the probability of being in these classes was associated with maternal characteristics, such as age, income, marital status, education, and BMI, and that participation in a responsive parenting intervention was associated with belonging to the FFV class, which represented one of the more “healthy” patterns.⁶³

Despite these multiple approaches, examination of infant diets tends to yield similar patterns. Nearly all studies identified a “health-conscious pattern,” characterized by higher fruit, vegetable, whole grain and/or healthy fat intakes, and a less optimal pattern, characterized by higher intakes of energy-dense foods and added sugars. An explicit comparison of 3 methods – an a priori diet quality index, a PCA, and a reduced rank regression with body composition as an outcome – all identified “health-conscious” patterns associated with higher fruit, vegetable, whole grain and vegetable oil intake in 1-year-old infants.⁶⁶ The similarities of patterns across studies suggest that the dietary pattern approach provides an important advantage over studying individual foods or nutrients, especially when studying a multifactorial process such as the development of obesity.⁶⁷

Relatively few of the studies assessing infant diet quality, however, have examined the association between these measures and concurrent or later adiposity or overweight. Overall, those that have examined the anthropometric outcomes associated with complementary feeding patterns have found only modest effects,⁶⁸ which vary based on age of follow-up and the anthropometric outcome assessed (see Table S1 in the Supporting Information online). These results, however, remain unchanged when accounting for maternal effects on infant diet or size. Not surprisingly, studies with anthropometric outcomes in infancy and early childhood tend to show more consistent effects. For example, higher scores for adherence to complementary feeding guidelines in the SWS described above were associated with greater weight and skinfold thickness gain from 6 to 12 months, but not weight or length at 12 months⁶⁵; with greater lean body mass but not BMI or fat mass at age 4⁶⁹; and inversely with greater fat mass but not BMI at age 6 compared to infants with lower adherence scores.⁷⁰ Similarly, in the Infant Care and Risk of Obesity Study, age-inappropriate feeding of solids and liquids was associated with an increased risk of infants having a weight-for-length (WFL) greater than the 85th percentile at 3–18 months of age.²⁴

Other studies show more persistent effects of early diets. A study of 1800 American mothers and infants participating in the Infant Feeding Practices Study (IFPS) II found that infants fed a “mixed” diet, characterized by breastmilk and/or formula along with higher-energy density foods such as French fries and sweets, at 9 months were more likely to be overweight at 12 months,⁶⁴ and that this group continued to have a higher prevalence of overweight at age 6.⁷¹

Other research suggests that focusing only on infant and child size may limit the understanding of the impacts of infant diet patterns on later growth and health. Longitudinal assessment of growth in a cohort of French infants and young children demonstrated that infant feeding patterns – characterized through PCA of milk feeding, timing of food introduction, and types of solid foods – were associated with linear growth and weight gain from birth to age 1, ages 1–3, and ages 3–5, despite a lack of significant associations with infant or child anthropometry at those ages.⁷² Research looking at body composition at age 6^54,67 found that while “health-conscious” patterns were associated with fat-free mass in childhood, no diet patterns predicted fat mass after adjustment for sociodemographic factors and maternal BMI. These results highlight the importance of measuring body composition to better understand the associations between early diets and body size. Further, the attenuation of their observed effects after controlling for maternal obesity suggests that other genetic or behavioral factors may mediate the association between diet patterns and infant size.

Importantly, considerable research shows that these empirically derived diet patterns track across infancy and early childhood into later childhood,^61,71 potentially linking early feeding practices to later obesity independently of their direct effects on infant body size. Greater adherence to complementary feeding guidelines in infancy was found, among an Australian cohort, to be associated with higher scores on the “conscious diet” pattern and lower scores on “processed foods” in mid-childhood.⁶⁰ Latent classes of milk and complementary feeding derived at 9 months predicted differences in the intake of fruits, vegetables, and other foods at age 6 in the IFPS II as well, with children in the Breastfed, Fruits and Vegetables (BFFV) class consuming more fruits and vegetables than the other groups.⁷¹ Conversely, Saavedra et al⁷³ found that less healthy eating patterns, particularly the consumption of sweets and sugar-sweetened beverages and low intake of fruits and vegetables, were established by 18–20 months of age in young American children participating in FITS II (Feeding Infants and Toddlers Study II). Similarly, a longitudinal study of European children from 5 countries found that the dietary pattern characterized by added sugar, unhealthy fats, and low consumption of fish and olive oil tracked most strongly from its establishment at 1–2 years of life to age 8.⁶¹ IFPS II infants with poorer diet quality at 9 months (ie, a diet higher in energy density and lower in nutrient density) also continued to have higher-energy-dense diets including fried potato, soda, and sweet desserts at age 6.⁷¹ Such findings highlight the need to intervene early in the complementary feeding period to identify and encourage “healthy” diets.

MECHANISMS LINKING EARLY DIET TO OBESITY RISK

Along with the macronutrients consumed and dietary patterns established during complementary feeding, infants also receive important cues that may entrain both biological and behavioral pathways around eating and contribute to the development of obesity (Figure 1). Further, complementary feeding practices are embedded in social and household contexts that may increase the risk of, or conversely protect against, the development of obesity. Understanding these interacting pathways is critical for successful intervention.

Figure 1.

Pathways linking complementary feeding practices to pediatric obesity.

Biological pathways: Biologically, what and how infants are fed can influence later vulnerability to obesity by programming growth rates and body composition, altering metabolism, and promoting different colonization by the gut microbiota. We are only just beginning to understand how differences in feeding practices may shape these pathways. Differences in macronutrient and energy intake associated with different complementary diets may shape the neural circuitry regulating appetite and body composition.⁷⁴ The introduction of complementary foods may program appetite regulation through hormonally mediated hunger and satiety cues. In one animal model, the introduction of complementary food increased the secretion of ghrelin, a stomach-derived hormone that stimulates appetite, increases food consumption, and promotes adiposity,⁷⁵ potentially linking the timing of weaning to appetite signals and changes in body composition. Diet composition may also directly affect body composition. According to the “early protein hypothesis,” high protein intakes during milk and/or complementary feeding appear to enhance insulin and insulin-like growth factor secretion and to increase fat mass by up-regulating adipogenesis and adipocyte differentiation.^76,77 Other studies, however, revealed no differences in insulin-like growth factor (IGF)-I levels between infants receiving varying amounts or sources of protein,^57,59 indicating that other mechanisms, such as altered branched-chain amino acids and acetylcarnitine metabolism, may be important.⁷⁸

Preliminary research suggests that feeding practices may alter epigenetic programming in human infants. CpG loci methylation patterns observed at 9 months of life differed from those present at birth, indicating that epigenetic phenomena are malleable during the complementary feeding period.⁷⁹ Further, methylation of the leptin gene, a hormone associated with fat mass and satiety, was found to be associated with prenatal and postnatal exposures, including maternal diet during pregnancy, breastfeeding, and complementary feeding practices. These findings highlight the potential for the nutritive and bioactive constituents of breast milk and complementary foods to influence gene transcription and/or protein synthesis, with longer-term consequences for tissue function and responsiveness to environmental cues.⁸⁰

Along with these hormonal and epigenetic effects, complementary feeding also shapes the colonization of the gut microbiota as infants are exposed to new foods.^81,82 While few human studies have examined the dynamics of microbiota composition in relation to the types and sequence of new foods added into infant diets,^83,84 animal models⁸⁵ and human studies^86–88 document a shift in the predominant phyla as the high-fat breast-milk diet is replaced by, or supplemented with, a carbohydrate-rich weaning diet, thereby introducing new substrates for the survival and growth of bacteria not previously supported by breast milk or formula.^81,82 The inclusion of new foods into the diet and the cessation of breastfeeding appear to be more important than the timing of solid food introduction.^87,89 The introduction of solids between 3 and 6 months, for example, was found not to be associated with patterns of colonization at 9 months in 2 large-cohort studies of Danish infants, entitled SKOT I & II.⁸⁹ Solid food introduction, however, has been associated with increased bacterial diversity in a number of studies.^86,89–91 The transition to “family foods,” characterized by a transition from breast milk, formula, and porridge to a higher content of meat, milk, cheese, animal fat, and rye bread, was associated with several measures of alpha diversity, including microbial richness and evenness, in the SKOT infants.⁸³ Among infants in the Australian BLISS (Baby-Led Introduction of SolidS) study, higher intakes of fruit and vegetables, breads and cereals, and dietary fiber at 7 months were associated with greater alpha diversity at 12 months.⁹¹

Although the evidence examining the effects of the specific components of the complementary diet on gut microbiota colonization and maturation remains limited,^83,84 the macronutrient content of the complementary diet is also likely important. In the SKOT I and II cohorts, protein and fiber intakes were positively associated, and fat intake negatively associated, with microbiota diversity at 9 months of age.⁸⁹ These patterns were thought to result from the increasing diversity and protein content of a diet that more closely resembled the family diet, rather than from any specific food or food group. Similarly, increased fruit & vegetable and bread & cereal intakes between 7 and 12 months were found to be associated with increased dietary fiber intake and changes in bacterial colonization, with decreases in the relative abundance of bacteria associated with milk consumption (eg, Bifidobacteriaceae and Veillonellaceae) and an increase in bacteria that play a role in polysaccharide degradation (eg, Lachnospiraceae and Ruminococcaceae).⁹¹ Other studies document differences in colonization between children consuming different food groups. The findings from a study of 5-month-old American infants randomized to receive meats vs micronutrient-fortified cereals as the primary complementary food until 9 months revealed differences in the prevalence of key bacteria, including Lactobacillus and Bacteroides, between the groups.⁹² Data from slightly older children, aged 2–3 years, has shown that specific intakes of dairy and plant-based foods (fruit, vegetables, soy, nuts, and pulses) were associated with distinct microbiota profiles.⁹³ In this cohort of Australian children, higher dairy intake, particularly of yogurts, was associated with lower species diversity and richness and greater prevalence of Firmicutes, while vegetable intakes were associated with a greater abundance of Lachnospira.

Since different bacteria differ in their energy-extracting capabilities and diet is a key determinant of bacterial colonization, differences in early diet have the potential to influence infant metabolism, weight gain, and subsequent obesity risk. A study of exclusively breastfed (EBF) and mixed-fed (ie, breast- and formula-fed) American infants followed longitudinally across the weaning period found that bacterial gene pathways for energy metabolism, specifically nitrogen and methane metabolism genes, were overrepresented in infants receiving solids compared to EBF infants (p ≤ 0.05), probably due to the higher protein content of solid foods compared to breast milk.^86,94 Longitudinal studies examining infant diet, microbiota, and weight gain across infancy and childhood are rare; however, preliminary evidence suggests that the composition of the microbiome may be also associated with weight gain and adipose deposition in infancy. Body mass increase between 9 and 36 months was found to be positively associated with changes in the relative abundance of Firmicutes and Clostridium and Eubacteria species, and negatively associated with bacterial strains of the Enterobacteriaceae family, in a study involving Danish infants.⁹⁵ Moreover, among Belgian infants, the proportion of Bacteroides strains at 3 and 52 weeks of age was found to be positively associated with, and the proportion of Staphylococcus strains at 3 and 26 weeks of age inversely associated with, BMIz scores in 1 and 3 year olds, controlling for infant feeding and a number of other potential confounders.⁹⁶ Similarly, overweight at age 7 was related to microbiota colonization at 6 and 12 months of age in Finnish children.⁹⁷

Behavioral pathways: Along with these biological mechanisms linking feeding to obesity, behavioral aspects of the feeding environment are also important in shaping long-term food preferences and eating behaviors. Children’s diet preferences and intake largely reflect the foods that become familiar to them,¹⁰ so the types and amounts of foods offered during the complementary feeding period may shape food acceptance into childhood. Food preferences develop early, and acceptance of new foods in the first 2 years of life is associated with the variety of foods offered, the frequency with which foods are presented, and the sensory properties of foods.⁹⁸ Early exposure to vegetables, for example, may enhance infant liking, increase consumption in childhood, and contribute to lower adiposity and risk of obesity.^99,100 Along with taste, texture appears to play an important role in food acceptance.^101,102 In a study conducted by Coulthard et al, infant consumption of home-prepared fruits and vegetables at 6 months of age was found to be associated with greater intake of fruits and vegetables at age 7, but consumption of baby foods had little effect on later food preferences.¹⁰³ Consequently, concerns have been raised that the widely prevalent use of commercial baby foods – nearly half (45%) of UK 8- to 10-month-old infants¹⁰⁴ and over 50% of 6- to 9-month-old American infants receive preprepared infants foods daily¹⁰⁵ – may shape child dietary preferences and subsequent obesity risk.¹⁰⁶ In the DONALD (Dortmund Nutritional and Anthropometric Longitudinally Designed) cohort study, for example, a higher percentage of commercial food consumption in infancy was associated with lower fruit and vegetable intake in preschool- and school-aged boys, though not girls.¹⁰⁷ Work from this cohort furthers documents that higher exposure to commercial foods in infancy is associated with greater sugar intakes,¹⁰⁸ likely due to the higher added sugar content of commercial infant foods¹⁰⁹ and the inclusion of sweeter vegetables, such as carrots, in many infant foods.¹⁰⁷ This higher early sugar intake may enhance children’s sugar preferences and contribute to greater sugar intake in childhood.¹⁰⁸

While commercial baby foods may shape dietary quality in infancy, their association with later child obesity risk has received relatively limited attention. The consumption of home-prepared foods, compared to commercially prepared foods, was found to be associated with greater dietary diversity and lower adiposity at 12 months in breastfed Canadian infants.¹¹⁰ Moreover, the consumption of a commercial milk-cereal drink has been associated with a greater risk of overweight in Swedish infants, with an effect that persists into childhood.¹¹¹ However, since the use of commercial baby foods is also linked to household factors such as maternal employment and income,¹⁰⁶ the relative quality of commercial foods compared to that of the family diet must also be considered.¹¹² Among American infants participating in FITS survey,¹¹² for example, those fed commercial foods had higher fruit and vegetable intake, lower consumption of sweets, and higher micronutrient intakes than infants not receiving commercial food. These contrasting results highlight the importance of considering infant diets in relation to home food environments and family characteristics more broadly.

Along with what infants are fed, the way infants are fed during this sensitive period may also influence eating behaviors, such as responsivity to hunger and satiety, and psychological aspects of eating, such as eating in the absence of hunger. Parental feeding practices representing insensitivity to infant cues or inappropriate responses to infant behavior may contribute to obesogenic feeding practices. Infants perceived as fussy, for example, may be more likely to receive solid foods or juice as parents interpret fussiness as hunger.⁴³ Similarly, bigger babies, or those growing at a faster-than-average rate, may get complementary foods earlier²⁵ or be fed against recommendations.²⁴

Parental feeding styles – the attitudes that characterize parental approaches to maintaining or modifying children’s eating behavior and the quality of interaction during feeding^113,114 – that are more controlling or restrictive may lead to poorer appetite regulation in the infant. Several studies have shown that maternal restriction is associated with greater infant weight gain.^114–116 In the Infant Care Study, restrictive and pressuring feeding styles were found to be associated with greater infant energy intake, age-inappropriate feeding, and lower weight-for-age z-scores.¹¹⁴ Although restriction and pressuring had opposite effects – restriction was associated with better feeding practices and higher infant weight, and pressuring with poorer feeding practices and lower infant weight – both feeding styles were characterized by higher levels of maternal control. Such feeding styles may entrain the development of eating behaviors uncoupled from hunger and satiety.

Research comparing traditional spoon-feeding with baby-led weaning (BLW) approaches to complementary feeding has suggested that the BLW approach, which is characterized by reduced maternal control, no spoon-feeding of pureed foods, and greater infant involvement in determining what foods and how much food to consume, may be associated with greater sensitivity to satiety and better appetite control.¹¹⁵ While several studies have documented healthier feeding practices in families using the BLW approach, including better adherence to WHO feeding recommendations,^117,118 less picky eating,¹¹⁸ more diverse fruit and vegetable intake,^119,120 and greater food enjoyment,¹²¹ the effects on infant weight are less consistent. Two observational studies have reported BLW to be associated with lower infant BMI.^121,122 However, both of these studies relied on parental report of child weight and involved families with higher education and longer breastfeeding duration, potentially introducing bias and limiting generalizability. A more recent randomized control trial – the BLISS study, conducted among 206 Australian infants – found no difference in BMIz score or risk of overweight at 12 or 24 months between infants randomized to either a BLW or traditional spoon-feeding group.¹²³ In contrast to previous work,¹²² mothers in the BLW group reported that their infants were less responsive to appetite, though no differences in energy intake were found. While further research is needed, these contrasting results highlight the importance of understanding not just what, but how, infants are fed, in shaping appetite, energy regulation, and potential obesity risk.

Household and social context: The importance of both early diet quality and parental feeding styles highlights how embedded infant feeding practices are in larger household and social contexts. Household factors, such as family size, maternal education, and employment/marital status, are associated with both diet patterns and the timing of solid food introduction. Lower maternal education and income are associated with earlier introduction of solid foods¹¹⁶ and infant feeding practices that do not follow guidelines,^24,64 contributing to the emergence of a social gradient in diet quality by early childhood.¹²⁴ Since parental diets, themselves shaped by economic necessity and preference, in large part determine infant diets, obesogenic dietary patterns characterized by low fruit and vegetable intake and high snack and sugar-sweetened beverage consumption tend to be transmitted intergenerationally.¹²⁵ Parental diets and feeding practices also differ according to parental weight status. Overweight mothers, for example, introduce sweets, pastries, and sugar-sweetened beverages to infants earlier than do normal-weight mothers, and infants of overweight mothers consume these sweets more frequently if their mothers do.¹²⁶ Obese mothers may also model eating behaviors, such as eating in the absence of hunger, using food instrumentally to reward or control child behavior, and prompting their children to eat more during meals – factors that contribute to higher intake and greater obesity risk.¹²⁷

The complex and multiple interactions between biological, behavioral, and social factors in shaping complementary diets also render the pathways linking different complementary feeding practices to infant and child obesity difficult to delineate methodologically. Complementary feeding is linked to earlier milk feeding practices in a number of ways, making it difficult to identify the independent effects of solid feeding. For example, infants who are breastfed may be introduced to solid foods later than those who are formula fed.²⁵ In turn, longer breastfeeding duration has been associated with better adherence to dietary guidelines¹²⁸ and greater infant fruit and vegetable intake.¹⁵ The relationship between milk feeding and complementary feeding is also likely to be bidirectional, since the earlier introduction of solid foods increases the likelihood of breastfeeding cessation as solid foods displace breast milk.¹²⁹ Further, many of the factors that shape milk feeding practices, such as maternal education, income, and obesity, also influence the timing of introduction to, and diet quality of, complementary foods.^24,126 Such findings suggest that intervening early in milk feeding practices may have downstream effects on solid feeding as well.

CONCLUSION AND RECOMMENDATIONS

Overall, the evidence for the importance of early complementary diets in the development of obesity in infancy and beyond is both limited and mixed. The timing of solid food introduction appears to have minimal effect on later obesity, unless solid introduction occurs early in infancy (<4 months) and infants are also formula-fed. Higher energy and protein intakes during complementary feeding are consistently associated with greater weight gain during infancy and larger BMI and %BF in childhood, though their effects may depend on the timing of intakes and the specific foods infants receive. While relatively few studies have examined infant diets more holistically, those that have suggest that “healthier” diets associated with greater adherence to feeding recommendations, higher consumption of fruits and vegetables, and the consumption of home-prepared foods are linked to a reduced likelihood of overweight in infancy and early childhood and better diets in childhood.

The effects of early diets on later obesity may be subtle and hard to measure in observational studies given residual confounding by earlier milk feeding practices and the sociodemographic characteristics of caretakers. Nonetheless, substantial evidence suggests that the early feeding environment can shape the physiological, metabolic, and behavioral pathways surrounding eating and can contribute to the development of child and adolescent obesity. Prospective, longitudinal examination of early feeding behaviors and their physiological consequences is needed to better assess the causal relationship between feeding and obesity.

While further research is warranted, complementary feeding practices, including the provisioning of healthy foods in appropriate quantities with appropriate responsiveness to infant hunger and satiety cues, are nonetheless important for shaping health infant growth and preventing the development of obesity. In addition to their nutritive effects, early feeding practices shape long-term eating and health behaviors by teaching children how, what, when, and how much to eat and transmitting cultural and familial beliefs, attitudes, and practices surrounding food and eating. Food habits formed over this critical period are likely to persist through childhood into adulthood, indicating that interventions to reduce obesity focusing on improving diet quality, food acceptance, and eating behaviors need to begin early. Differences seen in the timing and progression of complementary foods across settings highlight the importance of examining “typical” infant diets in diverse contexts and the need for context-specific measures of “healthy” infant diets. Further, the finding that social disparities in diet quality and obesity emerge early in life signifies that interventions should focus on children and families who are most vulnerable. A focus on the impact of early complementary feeding on the physiological and behavioral regulation of eating behaviors and energy metabolism is critical for intervening early to prevent the development of obesity.

 

Supporting Information

The following Supporting Information is available through the online version of this article at the publisher’s website.

Table S1  Studies examining the association between infant diet patterns and concurrent and later size, body composition and overweight