Introduction
Lipids constitute about 50% of energy intake (EI) in exclusively breast-fed infants. Exclusive breastfeeding for the first 6 months of life has a strong consensus for recommendation(1,2) , and human milk is used as a model to define fat and fatty acid (FA) intakes in early life for healthy infants.
Despite breastfeeding promotion campaigns, the rate of breastfeeding initiation remains low in France, which has one of the lowest rates in Europe (around 66%), and has even tended to decline in recent years(Reference Guajardo-Villar, Demiguel, Smaïli, Boudet-Berquier, Pilkington and Blondel3). In addition, to enhance breastfeeding support, it is therefore also important to define as accurately as possible the nutritional framework for infant formulas offered in the absence of breastfeeding.
Recent clinical studies have underlined the essential contribution of fats consumed by children to metabolic programming, and their involvement in the development of neuronal and immune functions in children under 3 years old(Reference Nicklaus4). Thus, infant formulas currently available have a FA profile that more or less faithfully reproduces the natural composition of human milk(Reference Boué-Vaysse, Billeaud, Guesnet, Couëdelo, Alessandri and Putet5). However, regulations for infant formulas only define a limited number of parameters for lipids(6,Reference Motarjemi7) in terms of quantity, while leaving other criteria (such as the addition of arachidonic acid (ARA), and the source of fat, etc.) to the discretion of the manufacturer.
Finally, public health messages may appear contradictory to the parents. Indeed, while prevention messages aimed at the lay public advise to reduce fat intake, young children need significantly higher proportions of EI from fat than adults (three to five times higher for children younger than 3 years v. adults)(Reference Martin8,Reference ANSES9) . It is therefore essential to establish clear recommendations to guide parents during the first 1000 days of life. Providing such recommendations is challenging because clinical trials are often not ethically feasible in nutrition, particularly in children, and in the absence of a strong meta-analysis, child nutrition remains a subject of debate among experts.
This literature review aims to provide a comprehensive analysis of current lipid intake recommendations for infants from birth to 3 years of age. By highlighting inconsistencies, contradictions and gaps in existing guidelines, this study seeks to propose evidence-based pathways for enhancing nutritional strategies, particularly in the formulation of infant formulas, to better align with the physiological needs of young children during this critical developmental period.
Methods
This article is a narrative literature review on the topic of lipid recommendations for young children from birth to 3 years old. This review is based on literature identified through searching PubMed, Embase and Cochrane databases (2001–22) conducted using the keywords: ‘alpha-linolenic acid (ALA), arachidonic acid (ARA), children, cholesterol, docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA), guidelines, infant, LC PUFA, linoleic acid (LA), lipids and dietary intakes, newborn, palmitic acid and toddler’.
The selection of relevant literature included articles in English and French and identified 861 articles in the following search engines: Pubmed (n = 620), Embase (n = 60) and Cochrane (n = 181). Articles were excluded if they did not provide the information sought, if they had an insufficient number of cited references (<2), if they presented redundant content or were deemed irrelevant or published in a journal of insufficient importance (classified as a ‘predatory journal’, without peer review and low classification rank in the specialty (less than D)).
After deduplication, there were 422 articles left and 104 were withdrawn after reviewing citation eligibility. After removing redundant citations, non-accessible and/or missing sought information, 133 different articles were included (Fig. 1).
Flow chart showing the methodology used to carry out the narrative literature review. The search period was between 2001 and 2022. The selection of relevant literature included articles in English and French and identified 861 articles in the following search engines: PubMed (n = 620), Embase (n = 60) and Cochrane (n = 181). Articles were excluded if they did not provide the information sought, had an insufficient number of cited references (<2), presented redundant content or were deemed irrelevant. After deduplication, there were 422 articles left, and 104 were withdrawn after reviewing citation eligibility. After removing redundant citations, non-accessible articles and/or those missing the information sought, 133 different articles were included.
Total lipids
Definition
Lipids are water-insoluble substances found in plant and animal tissues and are essential for structural and energy needs(10). They exist in various forms, including triacylglycerols, which are the primary nutritional form; fatty acids are an important constituent of many of these lipids including triacylglycerols.
Energetic function
The body’s physiological needs change with age and require specific nutrient intakes. Indeed, from birth to 3 years, the energy needs derived from total lipids are particularly important, as energy expenditure is high due to rapid growth during this stage of life. In the first months, lipids represent about 35% of weight gain, or 80–90% of the energy value of new tissues(Reference Fomon, Haschke, Ziegler and Nelson11).
Current recommendations
Current recommendations on the proportion of EI derived from lipids are not consistent and have varied over the years and between different scientific organisations, with an upward trend in recent years (Table 1).
Evolution of the recommendations on the proportion of energy intake derived from lipids in children aged 0–3 years
EI, energy intake; ANSES, French Agency for Food, Environmental and Occupational Health and Safety; EFSA, European Food Safety Authority; ESPGHAN, European Society of Pediatric Gastroenterology, Hepatology and Nutrition; FAO, Food and Agriculture Organization; m, month; SFP, French Paediatric Society; WHO, World Health Organization.
Currently, the French Agency for Food, Environmental and Occupational Health and Safety (ANSES) recommends 50–55% of EI from lipids in infants under 6 months, which is similar to the lipid content of human breast milk, and then 45–50% for children from 6 months to 3 years of age(12). In fact, epidemiological studies suggest adequate growth as long as lipids are above 30% of the EI(Reference Briend, Legrand, Bocquet, Girardet, Bresson and Chouraqui13); below this there is a risk of inadequate intake of energy and fat-soluble vitamins. However, the ANSES recommendation derives from the fact that lipid consumption lower than 50% of total EI (TEI) does not ensure the minimum requirements for certain essential FAs, such as omega 3, and fat-soluble vitamins for infants under 6 months. The 2014 recommendation from the European Food Safety Authority (EFSA) – around 40% for infants aged 6–12 months(14) – is very similar to that of ANSES (45%).
Long-chain polyunsaturated fatty acids (PUFA)
Generalities
Definition
Polyunsaturated fatty acids (PUFA) are fatty acids with at least two double bonds, classified as long-chain when containing more than twelve carbon atoms. Essential PUFA, such as linoleic acid (LA) C18 : 2 (n-6) and alpha-linolenic acid (ALA) C18 : 3 (n-3), cannot be synthesised by the body and must be obtained through diet(Reference Carlson, Schipper, Brenna, Agostoni, Calder and Forsyth15,Reference Einerhand, Mi, Haandrikman, Sheng and Calder16) . Omega-6 (LA, ARA) and omega-3 (ALA, DHA, EPA) families (Table 2) play critical roles in structural functions, inflammation modulation and neurodevelopment.
Main roles of lipids in child development
Metabolism
Endogenous synthesis of DHA and ARA is influenced by the genetic polymorphism of the fatty acids desaturases (FADS) FADS1 and FADS2 genes (which encode desaturases 1 and 2, respectively, and limit their synthesis) and by elongases 2/5. Furthermore, the activities of these enzymes depend on the organ (e.g. liver, brain, testicle, kidney) in relation to their protein concentration levels and substrate availability(Reference Valenzuela, Metherel, Cisbani, Smith, Chouinard-Watkins and Klievik17). The synthesis of PUFA is dependent on various factors, such as diet, nutrients, age, nutritional status and certain pathologies, particularly those associated with insulin resistance, as insulin induces the enzymatic activity involved in PUFA synthesis(Reference Videla, Hernandez-Rodas, Metherel and Valenzuela18). Nutritional intake of DHA and ARA during infancy is important to ensure their availability for growth and development. There is a competition phenomenon between these PUFA since the essential fatty acids (EFA) conversion pathways use the same set of enzymes (desaturases and elongases) for synthesis from their precursors ALA and LA (Fig. 2) (Reference Sprecher19). A high intake of LA induces excessive synthesis of ARA, which in turn limits DHA and EPA synthesis from ALA, resulting in reduced local availability of long-chain (LC) n-3 PUFA (notably DHA) into membranes. An excessively high LA:ALA ratio in early in life is likely to have negative short- and long-term health effects by reducing cerebral DHA availability and uptake through competition with ALA in the conversion stages(Reference Simopoulos20). In addition, high levels of circulating ARA compete with DHA for incorporation into neuronal membranes. Conversely, high levels of ALA improve the infant’s DHA status(Reference Simopoulos20,Reference Brenna21) . Thus, dietary intake of LA and ALA, and the balance between them, has the potential to affect LC PUFA status, as well as affecting numerous physiological functions such as neuro-cognitive development(Reference Uauy, Mena and Rojas22). Moreover, circulating DHA levels depend not only on the ratio of LA to ALA but also on the amount of dietary preformed PUFA supplied(Reference Lewis, Steenson, Haslam, McDonald, Sharman and Traka23); conversely, DHA plays a fundamental role in PUFA synthesis because it downregulates its own liver synthesis by inhibiting EPA elongation(Reference Metherel, Valenzuela, Klievik, Cisbani, Rotarescu and Gonzalez-Soto24,Reference Valenzuela, Metherel, Cisbani, Smith, Chouinard-Watkins and Klievik25) .
Conversion pathways for linoleic and alpha-linolenic acids. The synthesis of omega-6 and omega-3 FA from their respective precursors, the EFA linoleic acid (LA) and alpha-linolenic acid (ALA), is illustrated here. The successive action of desaturases and elongases, which are common to both metabolic pathways, explains the competitive phenomenon that can occur with an excessive intake of one of the EFA.
Main functions
EFA, but especially their LC PUFA derivatives, are esterified mainly in the phospholipids of cell membranes, where they have a structural and functional role from fetal life onwards. While PUFA can be used as an energy source, they have numerous physiological roles such as inflammation modulation, immunity and regulation of lipid metabolism, as well as playing an essential role in the functioning of the brain and vision. EFA deficiency impairs lipid and energetic metabolism, cell membrane structures and intracellular signalling pathways. A deep and prolonged deficiency can be detrimental, hence there is dependence on a diet containing both LA and ALA(Reference Uauy, Mena and Rojas22).
Structural and modulation of gene expression
In the form of phospholipids, these FA are universal constituents of biological membranes: modulating their fluidity and the activity of the proteins they contain (enzymes, receptors and transporters, etc.). They are important for lipid synthesis in physiological barriers such as the epidermis. They are also involved in the regulation of inflammation through the activation of transcription factors(Reference Bazan26,Reference Serhan, Arita, Hong and Gotlinger27) ; it is well documented that n-6 PUFA demonstrate a higher pro-inflammatory effect than the n-3 PUFA.
Immunity
FA are precursors of oxygenated mediators that specifically modulate a wide range of cellular functions and can produce multiple effects: haemostasis, platelet aggregation, immune system activity, etc.(Reference Bazan26).
A number of clinical studies have shown that LC PUFA can modulate a child’s immune status, thereby potentially preventing allergy development. These studies show a positive association between n-6 PUFA levels in breast milk and the risk of developing asthma, while high levels of total n-3 FA are associated with a reduced risk of atopy(Reference Soto-Ramírez, Karmaus, Zhang, Liu, Billings and Gangur28).
The impact of high ARA intake in the evolution or triggering of common childhood inflammatory diseases, such as asthma, eczema, atopic dermatitis and food allergies, is a topic of interest, with ARA intake often inversely related to these allergic diseases(Reference Miyake, Tanaka, Sasaki and Arakawa29,Reference Decsi, Marosvölgyi, Muszil, Bódy and Szabó30) .
Neurological and ophthalmic
Furthermore, some LC PUFA are involved from the fetal stage onwards in body growth, central nervous system development and, consequently, cognitive development and retinal function in children(Reference Simopoulos20,Reference Spector31) .
Specifically, DHA is central to fetal and infant growth, as well as retinal and visual development. It is a major lipid constituent of photoreceptor membranes, where it plays a crucial role in maintaining their structural and functional integrity(Reference Hadley, Ryan, Forsyth, Gautier and Salem32). It has been shown, for example, that the intake of DHA from algae in infant formulas improves visual acuity in children as young as 12 months(Reference Birch, Carlson, Hoffman, Fitzgerald-Gustafson, Fu and Drover33). This beneficial effect could be explained by the fact that DHA increases mitochondrial activity and has antioxidant, anti-inflammatory, anti-apoptotic and anti-angiogenic effects at the retinal level. Since the continuous renewal of retinal membranes requires a constant supply of n-3 PUFA, DHA-rich diets may even improve retinal function, particularly where damage has already occurred. In particular, DHA is a precursor of oxygenated derivatives, giving it specific properties in the brain such as anti-inflammatory effects and involvement in the apoptosis process(34).
Importance of PUFA during pregnancy and breastfeeding
Polyunsaturated fatty acids (PUFA), particularly DHA and ARA, are essential during pregnancy and breastfeeding owing to their critical roles in fetal and infant development(Reference Muñoz, Mercado, Farias, Beyer, Alvear and Echeverría35). DHA supports neurodevelopment, visual function and cognitive growth, while ARA contributes to immune function and cellular signalling. Maternal diet significantly influences the PUFA composition of breast milk, as dietary intake of n-3 and n-6 fatty acids directly affects their levels in milk(Reference Barrera, Valenzuela, Chamorro, Bascuñán, Sandoval and Sabag36). For example, diets rich in DHA, such as those including fatty fish or DHA supplements, enhance DHA levels in breast milk, promoting optimal infant brain and retinal development(Reference Barrera, Valenzuela, Chamorro, Bascuñán, Sandoval and Sabag36). Conversely, high intake of LA can increase ARA levels but may reduce DHA synthesis due to competition for shared enzymatic pathways(Reference Einerhand, Mi, Haandrikman, Sheng and Calder16). Gestational obesity further alters the lipid profile, often increasing n-6 PUFA levels while reducing n-3 PUFA concentrations, which may impact infant health(Reference Gulecoglu Onem, Coker, Baysal, Altunyurt and Keskinoglu37). Therefore, balanced maternal dietary intake of n-3 and n-6 PUFA during pregnancy and lactation is crucial for ensuring adequate PUFA transfer to the infant, supporting growth, immune function and long-term health outcomes(Reference Newberry, Chung, Booth, Maglione, Tang and O’Hanlon38).
Long-term effect of PUFA intake in children
Preclinical evidence underlines the deleterious short- and long-term role of excess LA, and it has been shown that an increase in dietary intake of LA in Western societies over the last few decades coincides with higher incidences of obesity(Reference Innis39) and immune diseases(Reference Prescott and Dunstan40) at the population level.
Projections for lipid intake from birth to 3 years emphasise the importance of maintaining a balanced lipid profile to support optimal growth and development while mitigating obesity risks. Lipid availability, particularly essential fatty acids such as DHA and ARA, is critical during infancy, as these contribute to neurodevelopment, immune function and metabolic programming(Reference Bazan26). Maternal diet during pregnancy and lactation directly influences the lipid composition of breast milk, with high n-6 PUFA intake potentially reducing DHA levels owing to enzymatic competition(Reference Barrera, Valenzuela, Chamorro, Bascuñán, Sandoval and Sabag36). This imbalance may predispose infants to inflammatory profiles and metabolic dysregulation. Maternal obesity further exacerbates these effects, altering breast milk lipid profiles by increasing n-6 PUFA and saturated fatty acids while reducing n-3 PUFA levels(Reference Álvarez, Muñoz, Ortiz, Maliqueo, Chouinard-Watkins and Valenzuela41). In childhood, excessive lipid intake, particularly from diets high in saturated fatty acids (SFA) and trans fats, can contribute to adiposity and long-term obesity risks(Reference Malin Igra, Ekström, Andersson, Ljungman, Melén and Kull42). Therefore, ensuring adequate intake of balanced lipids, particularly n-3 PUFA, during early life is essential for preventing obesity and promoting lifelong health.
Current recommendations
LC PUFA requirements vary from one child to another even within a population of healthy term infants, so the LC PUFA composition of infant formulas is described as a range rather than as a single precise value.
Essential fatty acids
Table 3 summarises the main official recommendations for EFA intake and their evolution over the past three decades.
ALA and LA recommendations for children aged 0–3 years
ALA, alpha-linolenic acid; d, day; EI, energy intake; AFSSA, French Agency for Food Safety; LA, linoleic acid; ANSES, French Agency for Food, Environmental and Occupational Health and Safety; FAO, Food and Agriculture Organization; m, month; WHO, World Health Organization; y, year.
The current recommendations’ evolution reflects the difficulty of setting a maximum limit for LA intake besides the experimental data available, showing possible adverse effects of high LA intakes as described earlier. Unlike the recommended dietary allowance (RDA) proposed by the French Agency for Food Safety (AFSSA) in 2001(Reference Martin8), the ANSES 2011 report does not give a maximum value for LA and ALA(Reference ANSES9). The 2010 the Food and Agriculture Organization and World Health Organization (FAO-WHO) report gives a range for ALA intake between 0.2% and 0.3% EI(43), but no recommendation is given for LA.
Scientific societies and organisations worldwide have mainly specified the minimum physiological requirement for LA in order to limit the imbalance between the two PUFA families when n-3 PUFA consumption is low. In the 1970s, supplementation studies proposed a recommended minimum intake of LA representing 2·7% EI on the basis of the levels observed at that time in human milk. As for ALA, a minimum intake of 0·45% EI is recommended to obtain an optimal DHA status for nerve and visual function(Reference Motarjemi7). These values are currently recommended by ANSES(12), whereas FAO-WHO recommend a slightly lower ALA intake of 0·3% EI(43). It is of importance to note that only the EFSA recommends both lower and upper limits for LA and ALA. Subsequently, the 2016 European regulations (based on EFSA’s recommendations) set LA levels at between 500 and 1200 mg/100 kcal and ALA levels at between 50 and 100 mg/100 kcal(44).
EFSA’s most recent evolution of EFA RDAs raised the lower limit of LA from 300 to 500 mg/100 kcal and lowered the upper limit of ALA to 100 mg/100 kcal owing to the mandatory addition of DHA to infant formula, thus lowering the required level of its precursor, ALA, in the presence of the preformed dietary DHA provided to adequately cover the physiological needs of infants(14).
Regarding the LA:ALA ratio, the 2006 European directive on infant formulas and follow-on formulas states that it should be between 5 and 15(44). As seen before, this ratio is important for the composition of cell membranes, brain and neurosensory development, and the child’s overall health considering the competition phenomenon in EFA conversion pathways. Thus, any imbalance between these two PUFA families could have harmful consequences. However, the rate of conversion of ALA to DHA depends on the absolute amounts of intake of LA and ALA, and not only on their ratio(Reference Briend, Legrand, Bocquet, Girardet, Bresson and Chouraqui13). Therefore, this ratio seems to be of less interest today, and the ANSES and FAO-WHO reports in 2010, unlike their predecessors, no longer mention limits for this LA:ALA ratio(45). Nevertheless, this ratio can remain a practical benchmark in the context of overall diet, and remains important in the event of excessive intakes of LA or inadequate intakes of DHA and EPA.
Long-chain polyunsaturated fatty acids
FAO-WHO and ANSES currently recommend providing preformed ARA and DHA for all children aged 0–6 months, i.e. 0·5% and 0·32% of total FA for ARA and DHA, respectively (Table 4). These values are based on average values from global surveys of human breast milk(Reference Brenna, Varamini, Jensen, Diersen-Schade, Boettcher and Arterburn46). For DHA, EFSA’s recommendations focus, in particular, on the effect of DHA on children’s visual functions: a level of DHA equal to 0·3% of total FA in infant formulas is recommended to ensure good visual development in children(Reference Carlson and Colombo47).
ARA and DHA recommendations for children aged 0–3 years
AFSSA, French Agency for Food Safety; LC PUFA, long-chain polyunsaturated acids; TFA, total fatty acids; ANSES, French Agency for Food, Environmental and Occupational Health and Safety; ARA, arachidonic acid; d, day; DHA, docosahexaenoic acid; EFSA, European Food Safety Authority; EI, energy intake; EPA, eicosapentaenoic acid; FAO, Food and Agriculture Organization; m, month; WHO, World Health Organization; y, year.
For EFSA, ARA is optional even before 6 months old. As for DHA derived from ALA(Reference Brenna21), ARA indispensability, linked to its low formation by conversion of LA, has led to the definition of a minimum physiological requirement of 70 mg/d, ensuring sufficient accumulation of this PUFA in cerebral membranes(Reference ANSES9). For infant formulas, this DHA intake should be between 20 and 50 mg/100 kcal, with no specific recommendation for ARA.(44)
For EPA, only an EPA/DHA ratio <1 was previously recommended for newborns and infants up to 6 months of age(Reference ANSES9), but the current data available is insufficient to define a physiological requirement and a RDA for EPA. Table 4 shows the main changes in LC PUFA recommendations over time for children aged 0–3 years according to various scientific societies and organisations.
Monounsaturated fatty acids
Monounsaturated fatty acids (MUFA) are FA that have a single unsaturated double bond in their carbon chain. They typically have between sixteen and eighteen carbon atoms in their carbon chain, although variations can exist. In quantitative terms, oleic acid is the major component of MUFA, is actively synthesised by cells and is very abundant in all plant and animal foods. It therefore accounts for almost all the MUFA in human nutrition.
Oleic acid C18 : 1 (n-9) is used as a source of energy and is a constituent of all types of lipids, particularly reserve triacylglycerols (adipose tissue) (1). Increasing the percentage of intake of MUFA, mainly oleic acid, at the expense of saturated fatty acids (SFAs), leads to a reduction in total cholesterol and LDL-C, without reducing HDL-C in adults(Reference Kris-Etherton, Pearson, Wan, Hargrove, Moriarty and Fishell48).
According to the latest ANSES recommendations, the MUFA RDAs are the same for children, adolescents and adults(Reference ANSES9). Oleic acid intakes recommended by AFSSA(Reference ANSES9) are between 15% and 20% TEI. This intake was set at 20% TEI in the agency’s official report in 2011(Reference ANSES9).
MUFA are constituents of human milk lipids (1·7 g/100 g). To date, there are no specific recommendations on MUFA for infants (0–3 years). The only exception is that European directives recommend a maximum level of erucic acid C22 : 1 (n-9), equal to 1% of total fats for the preparation of infant formulas because of potential adverse effects of this specific minor MUFA(49,50) . As MUFA, mainly oleic acid, are very well represented in both the plant and animal world, a varied and balanced diet provides adequate quantities.
Saturated fatty acids
Generalities
SFAs are fully hydrogen-saturated carbon chains, classified by chain length. Different SFA families have very different origins, metabolism and functions, and it is therefore essential to distinguish between them (Table 2). Short-chain SFAs, such as butyric acid, are synthesised by the body and support energy regulation. Medium-chain SFAs, found in coconut oil, are rapidly absorbed independently of chylomicrons and the lymphatic system and are used for energy. Long-chain SFAs, such as palmitic acid, are abundant in the diet and essential for cell membranes and protein acylation (Table 2).
Human milk contains about 2 g of SFA per 100 g milk, and they account for almost half of the total FA content. Palmitic acid is the main SFA in human milk comprising around 25%(Reference Koletzko51). SFA have several functions in the body. Beyond being a source of energy, they are important constituents of membranes(Reference Legrand52) and needed for the FA acylation of proteins. FA acylation regulates intracellular trafficking, protein–protein and protein–lipid interactions, and each FA confers different biochemical properties. SFAs can also regulate gene transcription via the recruitment of transcription factors(Reference Legrand52,Reference Legrand and Rioux53) .
Short- and long-term effects of a low SFA diet in children
Questions have been raised about the appropriateness of recommending a reduction in SFA intake for children, as excessive restriction could lead to inadequate nutritional intake and subsequently have a negative impact on the child’s normal growth and development. However, there was no evidence of adverse effects of reduced SFA intake on anthropometric measures of growth, cognitive development or micronutrient intake in children(Reference Lifshitz and Moses54,Reference Kaplan and Toshima55) . In addition, clinical studies show that diets low in SFA are associated with statistically significant reductions in total and LDL cholesterol and diastolic blood pressure in children and adolescents aged between 2 and 19 years(Reference Te Morenga and Montez56).
The Finnish STRIP study assessed the long-term effects of a low-SFA diet started early in infancy. Over a thousand healthy infants were included and randomised at the age of 7 months. At the end of the follow-up, at the age of 10 years, it was shown that current dietary recommendations targeting SFA restriction could be initiated in young children without deleterious consequences for growth, with beneficial effects on atherosclerosis risk factors and arterial function in boys(Reference Hakanen, Lagström, Kaitosaari, Niinikoski, Näntö-Salonen and Jokinen57).
To our knowledge, there is no specific paediatric study about the deleterious effect of long-chain SFA on cholesterolaemia.
Current recommendations
The ANSES and FAO-WHO reports do not include any recommendations on SFA for infants under 6 months of age, even though human milk provides 20–25% of energy in the form of SFA(Reference Lifshitz and Moses54). However, for infants and young children aged between 6 months and 3 years, ANSES recommends a total SFA intake of less than 12% TEI and 8% for lauric, myristic and palmitic acids(Reference ANSES9). There are no recommended minimum intakes for short- and medium-chain SFA, which are present in breast milk. Furthermore, the FAO-WHO report recommends a reduction in SFA, without a reduction in lipids, in children aged over 2 years from families with hypercholesterolaemia. Regarding the composition of infant formulas, European regulations impose a maximum level of lauric and myristic acid of 20% of total fats, owing to a possible atherogenic effect, while imposing no restrictions on other SFA such as palmitic acid(44).
Trans fatty acids
Generalities
Trans fatty acids (TFAs) have a trans configuration of hydrogen atoms around their double bonds. They can be naturally produced by ruminants or industrially synthesised through hydrogenation. Industrial TFA, found in processed foods, are linked to cardiovascular risks, while natural TFA in dairy and meat are less harmful (Table 2)(Reference Tardy, Lambert-Porcheron, Malpuech-Brugère, Giraudet, Rigaudière and Laillet58,Reference Tardy, Morio, Chardigny and Malpuech-Brugère59) .
Long-term effects of trans fatty acid intake in children
The harmful effects of industrial TFA include an increase in LDL cholesterol and a drop in HDL cholesterol(Reference Brouwer, Wanders and Katan60). Clinical studies and meta-analyses in adults have shown that an excessive intake of TFA is associated with an increased risk of cardiovascular disease (CVD) in a dose–response manner(Reference Zhu, Bo and Liu61). One study on Spanish children aged 4–5 years old found that the highest quartile of industrial, but not natural, TFA intake was associated with overweight. Another study carried out by Greek researchers on a hundred or so mothers and their 3-month-old infants showed that mothers who consumed at least 4·5 g of TFA per day were almost six times more likely to have a body fat percentage 30% higher. In this case, the risk of the infant having a fat mass greater than 24% is more than doubled. This study showed an association between high consumption of TFA and an increase in body fat in both mother and baby, raising the hypothesis that TFA are an early determinant of obesity(Reference Anderson, McDougald and Steiner-Asiedu62).
Current recommendations
In 2005, the ANSES set a maximum intake threshold for total TFA at 2% TEI(63) and at 1·5% TEI for TFA of technological industrial origin, whereas the WHO currently recommends limiting their consumption to less than 1% TEI(64), an indication of the increase in the negative effects linked to CVD of this class of lipids in the diet of the general population. For infant formulas and follow-on formulas, the European legislation authorises a maximum level of TFA of 3% of the total fat content to allow milk fat to be used in infant formulas(49).
Cholesterol
Generalities
Cholesterol is crucial for neurological development, cell membrane formation and hormone synthesis. It is abundant in human milk ranging between 90 and 150 mg/l (ten times higher than that of vegetable-oil-based formulas and six times higher than that of cow’s-milk-based formulas) (Reference Wu, Zhang, Zhang, Shi, Tan and Zheng65), regulating endogenous cholesterol synthesis in infants. Breast-fed infants have higher cholesterol levels than formula-fed infants, which may positively influence long-term cholesterol metabolism and cardiovascular health (Table 2) (Reference Nakada, Ho, Celis-Morales and Pell66).
Long-term effects of a high cholesterol diet in children
Lipid profile in adulthood
The hypothesis of the long-term impact of a diet rich in cholesterol during childhood emerged in the 1980s with the epigenetic notion that a diet rich in cholesterol early in life could favourably regulate the metabolism of cholesterol and lipoproteins in adulthood(Reference Reiser and Sidelman67,Reference O’Brien, Pullarkat, Darsie and Reiser68) .
Animal studies in baboons suggest that the amount of cholesterol in the diet of the first few months of life (formula or breastfeeding) can affect adult cholesterol metabolism, but in a nonlinear way(Reference Mott, Jackson, McMahan and McGill69). The potential impact of early life nutrition in CVD incidence has been calculated by the following model: if 30% of infants were exclusively breast-fed, the reduction in endogenous cholesterol synthesis would result in a 0.15 mM drop in adult cholesterol levels, and therefore a reduction in the prevalence of CVD of up to 5% in the population(Reference Fall, Barker, Osmond, Winter, Clark and Hales70).
Cardiovascular risk
The role of dietary cholesterol in cardiovascular risk is still a matter of debate, and epidemiological or long-term follow-up studies do not allow us to conclude with certainty about the role of early exposure to cholesterol in childhood in relation to cardiovascular risk in adulthood, apart from pathologies such as familial hypercholesterolaemia(Reference Kratz71,Reference Spence, Jenkins and Davignon72,Reference Lecerf and De Lorgeril73) . In fact, there is no evidence to support its role or lack of effect on cardiovascular risk, as it is not possible to separate its role from that of other nutrients such as TFA, with which it is co-ingested.
Recent epidemiological studies comparing breastfeeding with milk formula show a higher risk of CVD in adulthood in infants fed with infant formulas containing no or low amounts of cholesterol. In a population of 109 young adults who died prematurely and were autopsied, atherosclerosis occurred in 60% of cases in individuals initially fed on artificial milk, compared with 25% of subjects fed on breast milk(Reference Martin and Davey Smith74). In another study, it was shown that mortality rates from ischaemic heart disease were higher in a formula-fed population than in a breast-fed population, and that these mortality rates were associated with higher adult total and LDL cholesterol values in the formula-fed population(Reference Fall, Barker, Osmond, Winter, Clark and Hales70). Of note, a major bias in human studies is the comparison of breast-fed newborns versus those fed artificial formula without considering the cholesterol content of the formula and the proportion of breastfeeding, if any, and its duration.
Current recommendations
Besides the studies published on early exposure of cholesterol impact on long-term cardiovascular risk, to date, there is no recommendation for the addition of cholesterol to infant formula.
Molecular species carrying FA: glycerolipids, glycerophospholipids, sphingolipids and glycosphingolipids
Generalities
Even if triacylglycerols (TG) are the main lipids providing dietary FA, breast milk and dairy products also contain many other bioactive lipids such as polar lipids. Their properties depend on the quality of the FA(Reference Osawa, Fujikawa and Shimamoto75).
Polar lipids include mainly glycerophospholipids and sphingolipids. These molecules are amphiphilic, which means they have surface-active properties and can be used as emulsifiers of lipids (TG) in a water medium (micelles), and can auto assemble to form liposomes or be part of mixed lipid micelles. They have two different parts: a hydrophilic ‘head’ (phosphate group) and a hydrophobic ‘tail’ constituted by two FA in glycerophospholipids; the junction between these two parts is a glycerol molecule. They constitute the main part of the cell membranes (cell, mitochondria, endoplasmic reticulum, nucleus, etc.) forming lipid bilayers, and they influence the membrane’s biophysical properties depending on the composition of their different components (FA unsaturation level, organic molecules attached to the phosphate group, etc.). This large family of lipids has also many other roles in cellular signalling, inflammation pathways, neurological transmission and intestinal mucosal integrity among others(Reference Feingold, Feingold, Anawalt, Blackman, Boyce, Chrousos and Corpas76).
Saccharolipids are complex molecules containing FA or a sphingosine branched on a sugar molecule. They are amphipathic and incorporated in cell membranes as the phospholipids. They are involved in a large panel of physiological pathways such as immunological response, cell interaction (apoptosis, tissue repair), brain function and cholesterol regulation(Reference Park, Choi, Kim, Lee, Lee and Lee77).
Current recommendations
To our knowledge, it is of interest to note that there is no formal recommendation about the level of supplementation of glycerophospholipids and sphingolipids in infant formula.
Discussion: gaps, controversies and paths of improvement for the recommendations on the lipid profile of infant formulas
The evolution of the current recommendations we have just detailed aptly demonstrates the progress in our understanding of the lipid requirements of infants and young children. Maintaining significant calorie intake while achieving an optimal omega 6:omega 3 ratio is at the core of all recommendations put forth by various scientific organisations. The composition of breast milk stands as the gold standard for determining the infant’s nutritional intake. Our literature analysis identifies several potential avenues for improvement based on the disparities between the composition of breast milk and the current recommendations regarding the composition of infant and follow-on formulas. These differences can be attributed either to technical difficulties in modifying infant formula or to the lack of studies providing answers to the questions raised.
Quantitative characterisation of infant formulas’ lipid profile
Quantitative differences in the lipid profiles of breast milk and infant formulas
Unlike infant formulas, breast milk mainly provides MUFA (45–50%), such as oleic acid (n-9) and palmitoleic acid (n-7), in higher quantities than PUFA (15%), as well as a large proportion of SFA (35–40%) such as palmitic acid and myristic acid (Table 5)(Reference Grote, Verduci, Scaglioni, Vecchi, Contarini and Giovannini78). However, the precise quantity of MUFA and SFA reported in literature may vary depending on the studies. Indeed, the composition of these FA in milk depends mainly on the nature of the mother’s lipid reserves and her dietary intake during pregnancy and lactation. These FA come from both endogenous synthesis and diet(Reference ANSES9).
Comparison of the nutritional lipid composition of breast milk, cow’s milk and infant formulas based on vegetable oils(Reference Brenna, Varamini, Jensen, Diersen-Schade, Boettcher and Arterburn46,Reference Ballard and Morrow130,Reference Delplanque, Gibson, Koletzko, Lapillonne and Strandvik131,Reference Hageman, Danielsen, Nieuwenhuizen, Feitsma and Dalsgaard108)
MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; SFA, saturated fatty acids; N/A: Not Available; Nq: Not Added.
SFA supplementation
Palm oil
The composition of infant formulas aims to be as close as possible to that of breast milk, in which palmitic acid represents about 25% of total FA(Reference Koletzko51). Palm oil constitutes a widely available source of palmitic acid, and consumption of formulas with palm oil is associated with an increase in weight-for-length (WLZ) in infants under 6 months, but a decrease in weight-for-age (WAZ) and WLZ in infants aged 6–12 months(Reference Young79). Because of eco-responsibility, the use of palm oil is a concern for many parents, but currently there is no evidence suggesting that palmitic acid at the levels observed in infant formulas, or the presence of palm oil, have a deleterious effect on child health. Still, current research suggests that other SFA sources, namely, dairy lipids, can be more optimal than palm oil regarding global FA profile and TG structure (mentioned later). There are now infant milks made from other vegetable oils, such as coconut oil(80), but their intensive production may also have an ecological impact and their FA profile is also quite different from human milk, notably the high proportion of C12 in coconut fat.
Dairy lipids
Dairy fats are rich in SFA (60–65% of FA including 30% palmitic acid) and low in EFA(Reference Legrand52). The varied profile of cow’s milk (>400 different FA species) is a richness that vegetable oils lack, including palm oil. This means that infant formula with milk fat provides a greater variety of SFA, and this complex profile is closer to breast milk than can be achieved with vegetable oils alone (Table 5).
Dairy lipids are often considered to be potentially deleterious for CV health in case of excessive intake, as they are rich in SFAs. However, many studies suggest that, when present in a balanced diet, because of the short- and medium-chain SFA and the presence of myristic acid in particular, dairy milk fat optimises cellular and tissue availability of the EFA and LC PUFA. In other words, dairy milk fat’s SFA could help the body to synthesise and/or maintain very-long-chain, highly unsaturated FA, particularly those of the omega-3 family such as eicosapentaenoic acid (EPA) and DHA, provided that there is a sufficient intake of the precursor ALA. For example, in primates, compared with a diet low in myristic acid (0·6%), a richer diet containing twice as much myristic acid (1·2%) is associated with an increase in EPA and DHA levels in phospholipids and DHA levels in cholesterol esters(Reference Dabadie, Peuchant, Bernard, LeRuyet and Mendy81). This raise the hypothesis that, in humans, a diet with moderate myristic acid intake (1·2% of EI), compared with a lower diet (0·6%), increases DHA and EPA levels in some plasma lipid classes, despite equivalent intake of ALA.
These results suggest that higher dietary SFA of dairy origin contributes at least in part to the maintenance of long-chain n-3 PUFA(Reference Rioux and Legrand82).
At present, regulations do not specify the fat source to be used in infant formulas. It is up to the manufacturer’s either to use only vegetable fats, such as palm oil or coconut oil, or to incorporate dairy lipids in the form of milk cream or milk fat.
LC PUFA supplementation
Long-chain polyunsaturated fatty acids
Intake of LC PUFA in children has been correlated with neurological and immune benefits.
Numerous studies have examined the effect of LC PUFA supplementation on psychomotor and neurosensory development in childhood, particularly in premature infants owing to the increased needs associated with their rapid growth. Better cognitive performance was demonstrated when infant formulas enriched with LC PUFA and milk fat globule membrane (MFGM) were consumed for more than 6 months(Reference Salas Lorenzo, Chisaguano Tonato, De La Garza Puentes, Nieto, Herrmann and Dieguez83). Paediatric nutritionists agree that it is important for children to receive sufficient and balanced quantities of LC PUFA from the fetal stage and for at least the first 6 months of their post-natal life, and that any serious deficiency in EFA and LC PUFA could have deleterious and irreversible consequences for the brain, hence the importance of supplementing infant milks and ensuring that the diet of pregnant or breast-feeding women is adequately covered in these FA(Reference Crawford84).
Moreover, the antenatal and neonatal periods should be favoured to influence the maturation of the immune system. Supplementing hydrolysates with n-3 PUFA could prevent allergies in at-risk populations(Reference Vandenplas, Meyer, Chouraqui, Dupont, Fiocchi and Salvatore85). Fish oil supplementation during pregnancy leads to an increase in n-3 LC PUFA in breast milk and a decrease in n-6, more markedly than in control mothers. Thus, n-3 PUFA levels are associated with IgA, IL10, IL6 and CD14 levels in breast milk on day 3, which could have a preventive effect on allergy risk(Reference Denburg, Hatfield, Cyr, Hayes, Holt and Sehmi86). The number of studies is still limited and the evidence inconclusive, as such, n-3 PUFA supplementation is not the subject of official recommendations for allergy prevention. Further studies are needed to determine whether n-3 PUFA have an anti-allergy protective effect, and to determine the optimal dose and type of supplementation to ensure this putative anti-allergy effect.
LC PUFA also appear to have an epigenetic programming effect during the period of early ante- and neonatal development in animals and humans(Reference Lepping, Honea, Martin, Liao, Choi and Lee87). Thus, recent studies show that LC PUFA (DHA such as ARA) could have a long-term effect on children’s body growth as they regulate the expression of genes responsible for the development of adipocytes(Reference Uauy, Mena and Rojas22).
ARA and DHA
To date, the European Commission has made it compulsory to add DHA to infant formula manufactured from February 2020 onwards, but the addition of ARA remains optional(Reference Koletzko, Bergmann, Brenna, Calder, Campoy and Clandinin88).
This decision comes from several arguments, mainly that no additional benefit for infants has been demonstrated after the addition of ARA and DHA to formulas. In terms of growth, compared with supplemented formulas, the absence of added ARA and DHA does not alter the growth of infants (weight, height, etc.) (Reference Jasani, Simmer, Patole and Rao89). In neurological terms, since a large proportion of DHA is found in the cerebral cortex, optimal cerebral development seems to be possible with DHA alone(Reference Birch, Hoffman, Uauy, Birch and Prestidge90,Reference Tounian, Bellaïche and Legrand91) . It should be remembered that DHA levels increase with age and depend mainly on diet, unlike ARA whose levels depend on age and very little on diet(Reference Makrides, Neumann, Byard, Simmer and Gibson92). The beneficial effect on visual acuity is only associated with DHA. In fact, supplementation during the first 4 months of life with DHA alone, without ARA, resulted in visual maturation at the age of 4 years, which was identical to that of breast-fed children and superior to that of infants receiving a formula not enriched with DHA. From this study including sixteen and seventeen children in DHA alone and DHA + ARA supplemented groups, respectively, the use of milk enriched with ARA and DHA seemed not to provide any benefit compared with formula enriched only with DHA.(Reference Birch, Garfield, Castañeda, Hughbanks-Wheaton, Uauy and Hoffman93) It should be noted that according to the Diamond study, DHA intakes higher than the DHA levels measured in breast milk are not associated with additional visual improvement.(Reference Birch, Carlson, Hoffman, Fitzgerald-Gustafson, Fu and Drover33)
However, the latest official recommendations from the European Commission and most learned organisations making the addition of ARA to infant formulas optional are the subject of real debate in the field of paediatric nutrition. Several groups of experts suggest that these two LC PUFA should be added simultaneously to infant formulas(Reference Koletzko, Bergmann, Brenna, Calder, Campoy and Clandinin88). They support this position because human milk, considered the gold standard in infant nutrition, contains more ARA than DHA(Reference Fu, Liu, Zhou, Jiang and Chai94). In addition, in the absence of adequate dietary intake, endogenous synthesis of ARA is insufficient to ensure the non-neurological biological functions in which it is involved(Reference Lien, Richard and Hoffman95). Specifically, genetic variants of desaturases present in 30% of children reduce the endogenous synthesis of ARA(Reference Lattka, Klopp, Demmelmair, Klingler, Heinrich and Koletzko96), so a higher dietary intake of ARA is necessary in carriers of these variants. In fact, the 30% reduction in ARA and EPA is observed in the case of minor alleles of FADS, whereas DHA levels are less affected by this genetic variant than by dietary intake. FADS is a gene that appears to modulate the effect of nutrition on cognition and immune development(Reference Salas Lorenzo, Chisaguano Tonato, De La Garza Puentes, Nieto, Herrmann and Dieguez83). Miklavcic et al. suggest that increasing the supplementation of infant formulas to 34 mg/100 kcal of ARA and 17 mg/100 kcal of DHA prevents the reduction in ARA due to minor alleles(Reference Miklavcic, Larsen, Mazurak, Scalabrin, MacDonald and Shoemaker97). Without the addition of ARA, anti-inflammatory and immunosuppressive effects can be obtained and are undesirable in the post-natal period when there is a significant risk of infection.
For all of these reasons, the addition of ARA to infant formulas, in addition to the mandatory addition of DHA, seems desirable. According to the Diamond study, the ARA:DHA ratio in infant formula could influence neurological development, and the respective proportions of ARA and DHA should also be taken into consideration, highlighting an important and relevant effect of the ARA:DHA ratio(Reference Colombo, Carlson, Cheatham, Shaddy, Kerling and Thodosoff98). Similarly, it has been shown that when infants receive both ARA and DHA supplementation, they perform better in terms of cognitive performance than when they receive DHA alone(Reference Liao, McCandliss, Carlson, Colombo, Shaddy and Kerling99). The controversy mainly comes from the heterogeneous results of neurodevelopmental studies on DHA and ARA supplementation. It can be explained by several confounding factors: heterogeneity between studies; a wide variety of judgement criteria and methodological approaches; the impact of genetic variability modulating the rate of endogenous LC PUFA synthesis; the interaction of breastfeeding, which provides preformed LC PUFA; lifestyle; smoking; and socio-economic status(1).
Therefore, the need for high-quality clinical trials is of the utmost importance to answer these questions. The size of the samples chosen is a central parameter for assessing complex intellectual performance, identifying sex differences and the effect of different polymorphisms known to influence FA metabolism. The results of these trials will be important in forming the basis of evidence-based guidelines for LC PUFA formula supplementation for infants and young children.
Cholesterol supplementation
Human milk contains a higher cholesterol content compared with commercial infant formulas containing only vegetable fats (Table 5). While conclusive evidence from randomised clinical trials are still lacking, the arguments supporting the addition of cholesterol to infant formula are based on its physiological functions; cholesterol is crucial for neurological development, cell membrane formation and hormone synthesis. Because cholesterol is a key component of the developing brain, its addition to infant formulas may support the growth and functioning of the central nervous system.
A potential impact on lipid metabolism has been suggested by epidemiological studies. Breast-fed infants have higher plasma concentrations of total and LDL cholesterol (LDL-C) than formula-fed infants; however, in the long term, studies have shown lower LDL-cholesterol in individuals who were breast-fed. This difference is attributed to cholesterol in breast milk (15 mg/100 ml), which is absent in infant formula. Oral ingestion of cholesterol early in life may play a part in the hepatic development of lipid degradation enzymes or hepatic receptors for LDL-cholesterol. However, other studies have reported conflicting data regarding the assumed protective effect on cardiovascular level, with similar cholesterol levels after the age of 1 year(Reference Friedman and Goldberg100,Reference Huttunen, Saarinen, Kostiainen and Siimes101,Reference Bayley, Alasmi, Thorkelson, Jones, Corcoran and Krug-Wispe102) .
Of note, the addition of cholesterol to infant formulas presents technological and economic challenges that currently limit its widespread implementation. However, the incorporation of milk fat globule membrane (naturally containing cholesterol) in infant formula is a possibility to achieve higher cholesterol content (mentioned later). Although the presence of significant amounts of cholesterol in human breast milk and epidemiological data suggesting its nutritional value, the current lack of recent and rigorous studies makes it challenging to formulate a recommendation on adding cholesterol to infant preparations.
Qualitative characterization of infant formulas’ lipid profiles
Triacylglycerol structures
Triacylglycerols (TG) are esters composed of one glycerol and three FA. The main sources of TG are dairy products, vegetable oil and animal fats. Their primary role is energy-related, serving as reserves in adipose tissue and contributing to ATP production.
The position of some FA in the triacylglycerol structure (sn-1 and sn-3 for the outer positions and sn-2 for the central position) can influence their digestion and absorption in the intestine. In this respect, TG structure is different in human, bovine and vegetable fat. Human and bovine milk fat contain a wide variety of FA, whereas vegetable fat variety is poorer and depends on the vegetable species used (blends are used to improve the variety). In human milk, the main TG structures are palmitic acid (C16) at the sn-2 position (see next paragraph) and oleic acid (18 : 1) at the sn-1 or sn-3 position. In bovine milk fat, palmitic acid (C16) at the sn-2 position is also the main TG structure (40–45% of the total amount of palmitic acid is branched at sn-2) (Reference Mehrotra, Sehgal and Bangale103), with oleic acid (18 : 1) at the sn-1 or sn-3 position(Reference Bourlieu and Michalski104). Of note, butyrate branches at the sn-3 position and stearic (18 : 0) acid at the sn-1 position. In vegetable fat used for infant formula, palmitic acid at the sn-2 position is less frequent (10–20% only) (Reference Mehrotra, Sehgal and Bangale103), whereas higher levels of SFA are present at the sn-1 and sn-3 position(Reference Miles and Calder105) compared with human and bovine milk(Reference Tu, Ma, Bai and Du106).
sn-2 palmitate level optimisation
Palmitic acid (C16) is the most abundant SFA in breast milk, and between 60–70% is esterified at the sn-2 position. During digestion, the action of lipases produces two Non-Esterified Fatty Acids (NEFA) (positions sn-1 and sn-3) and one monoacylglycerol. If palmitic acid is released as NEFA because it was on the sn-1/3 position, it can bind calcium and form insoluble calcium soaps that are not absorbed and are associated with increased constipation and infant colic(Reference Innis107). Studies have shown a positive correlation between the sn-2 central position of palmitic acid on triacylglycerols and improved digestive comfort in children. This beneficial effect is linked to an increase in the number of Lactobacillus and bifidobacteria in the intestine(Reference Miles and Calder105,Reference Hageman, Danielsen, Nieuwenhuizen, Feitsma and Dalsgaard108) , and secondly with the lower amount of calcium soaps in faeces, which favours the formation of softer stools(Reference Bar-Yoseph, Lifshitz and Cohen109,Reference Mehrotra, Sehgal and Bangale103) .
Recent studies also show that increasing the level of sn-2 palmitate in infant formula, using milk fat instead of palm oil, is associated with better development of motor skills in children at the age of 16 months, and that the beneficial effects on neuronal development in children are associated with an increase in the level of bifidobacteria in the intestinal microbiota(Reference Wu, Zhao, Liu, Ye, Su and Li110).
The optimisation of sn-2 palmitate level in infant milk formulas can be achieved either by industrial restructuring of vegetable fats (palm oil) or the addition of milk fat instead of regular palm oil. This is because in dairy lipids, palmitic acid (C16 : 0) accounts for approximately 25% of total FA and is mostly found in the sn-2 position (Fig. 3). However, despite the importance of the positioning of FA on the TG molecule for FAs bioavailability and the associated benefits on digestion and neuronal development, no recommendation has been formulated so far. Therefore, specific consideration on this point appears necessary for future recommendations.
Comparative FA composition and regio-distribution in human milk fat (A) and bovine milk fat (B) (% mol of main FA >0·5 % total). This figure compares human versus bovine milk fat in terms of FA composition (expressed as % mol of main FA >0·5% of total) and also the positional distribution of FA within the triacylglycerol molecule (regio-distribution in the triacylglycerol (TG) core between the external positions sn-1 and sn-3, or the middle position sn-2).
Fat emulsion ultrastructure
Differences in fat emulsion ultrastructure between breast milk and infant formulas
In terms of quality, the fat emulsion in human milk and infant formula differs significantly in both structural and biochemical terms.
The fat globules in human breast milk are distinguished from those in infant formula by their larger size and the presence of a phospholipid membrane that modifies intestinal absorption and may confer beneficial properties in terms of body composition and subsequent cardio-metabolic prevention(Reference Gallier, Vocking, Post, Van De Heijning, Acton and Van Der Beek111). Phospholipids are also essential components of the neurological system and supplementation in animal diet improves brain myelinisation(Reference Oshida, Shimizu, Takase, Tamura, Shimizu and Yamashiro112). In human clinical studies, supplementation of milk with phospholipids, including sphingomyelin, is associated with improvement of neurological development, mainly cognitive efficiency(Reference Tanaka, Hosozawa, Kudo, Yoshikawa, Hisata and Shoji113).
In its native form, women’s milk fat is organised into dispersed globules enveloped by a triple phospholipid membrane known as the milk fat globule membrane (MFGM), which originates from the fat-secreting epithelial cells of the mammary gland(Reference Bourlieu, Deglaire, Oliveira, Ménard, Gouar and Carrière114). These MFGM comprise proteins inserted into the phospholipid membrane (mainly at the inner layer), an intermediate layer composed of the phospholipid triple membrane and an outer layer consisting mainly of high-molecular-weight glycoproteins and sphingomyelin/cholesterol complexes (Fig. 4). Milk fat globules have an average diameter of around 3–5 µm, but show a wide size distribution (from 0·1 to 15 µm) in both women’s and cow’s milk(Reference Michalski, Briard, Michel, Tasson and Poulain115). In infant formulas, however, the fat is dispersed as a result of the homogenization of vegetable oils in the presence of milk proteins. This process produces a stable micro-emulsion of plant lipids, mainly stabilised by caseins and potentially added stabilisers such as lecithins, in the form of small lipid droplets, with an average diameter of less than 0·5 µm and no membrane coating (Fig. 4).
Structural differences in lipid droplets between breast milk containing native milk fat globules (A) and a microemulsion of plant lipids in infant formula (B). The structural differences between fat in mammalian milk and standard infant formula are presented. Panel A illustrates the structure of the milk fat globule with detailed milk fat globule membrane structure. Panel B, depicting the fat droplet in standard infant formula, highlights differences in terms of size (smaller diameter: 0·3–1 µm v. a few nm) and structure (aggregates of protein and emulsifier v. trilayers).
Size of milk fat globule optimisation
Recent experimental data suggest that the size of milk fat globules is a key parameter in the long-term beneficial effects of breastfeeding, particularly in protecting against metabolic syndrome and obesity(Reference Brink and Lönnerdal116). During the developmental period in mice, from weaning to young adulthood, consumption of a milk diet composed of large (10 µm) fat globules surrounded by a phospholipid membrane reduced total adipose tissue mass and circulating leptin levels by 25% in adulthood, compared with a diet with a droplet structure characteristic of infant milks(Reference Oosting, Kegler, Wopereis, Teller, van de Heijning and Verkade117). This beneficial effect on body composition is also accompanied by an improvement in the animal’s metabolic status and, more specifically, insulin sensitivity, with a reduction in the homeostatic model assessment (HOMA) index, fasting glycemia and circulating resistin levels. The authors hypothesise that changes in lipid absorption and digestion kinetics may in turn alter their metabolic utilisation – the balance between catabolism by beta-oxidation and lipogenesis. In rats, the absorption of small fat globules (0·4 µm) formed with adsorbed casein causes a marked reduction in postprandial beta-oxidation of FA compared with large native globules. This difference is thought to result from a delay in the kinetics of triacylglycerol appearance in plasma, linked to slower gastric emptying and lipolysis when small fat globules are ingested(Reference Michalski, Briard, Michel, Tasson and Poulain115,Reference Borel, Armand, Pasquier, Senft, Dutot and Melin118) .
Recent studies show that the interfacial coating of breast milk fat globules with MFGM facilitates lipolysis by digestive lipases and bile salts naturally present in breast milk(Reference Bourlieu, Deglaire, Oliveira, Ménard, Gouar and Carrière114). Lipolysis being an interfacial process, gastric and pancreatic lipases hydrolyse fats at the oil–water interface. Thus, the composition and structure of the oil–water interface is likely to affect lipolysis in the gastrointestinal tract. In infants, levels of pancreatic lipase and bile salts are low compared with adults, so the products of gastric lipolysis play an important role in the digestion of milk lipids, compensating for the low levels of pancreatic lipase and emulsifying lipids, respectively. Today’s infant formulas contain much smaller thick protein-coated fat droplets than the MFGM-coated fat globules of breast milk. Thus, infant formulas containing bigger fat droplets resulting from MFGM enable a digestion process closer to that of human milk(Reference Gallier, Vocking, Post, Van De Heijning, Acton and Van Der Beek111).
Milk fat globule membrane supplementation
In humans, clinical studies have reported the beneficial effects of MFGM supplementation of infant formulas on neurocognitive development and protection against infectious agents, with no deleterious impact on growth(Reference Timby, Domellöf, Lönnerdal and Hernell119,Reference Hernell, Timby, Domellöf and Lönnerdal120) .
The COGNIS study is a prospective, randomized, double-blind, nutritional intervention study. This study compared seventy children, during their first 18 months of life, fed with either a standard infant formula (SF, n = 29) or a bioactive-compound-enriched infant formula (EF, n = 41) and thirty-three breast-fed (BF) children (reference group). The results suggest that enriched-infant-formula-fed infants seem to show fewer behavioural problems up to 2·5 years compared with standard-infant-formula-fed infants(Reference Nieto-Ruiz, Diéguez, Sepúlveda-Valbuena, Herrmann, Cerdó and López-Torrecillas121).
In Indonesian infants aged under 8 weeks to 6 months supplemented with MFGM complex lipids, a higher general intelligence quotient (IQ) and better hand–eye coordination and Griffiths scale performance IQ were found at 24 weeks of intervention compared with those fed without MFGM(Reference Gurnida, Rowan, Idjradinata, Muchtadi and Sekarwana122). Another trial in Swedish infants aged 2–6 months supplemented with MFGM (Lacprodan MFGM-10) showed a higher Bayley-III cognitive score at 12 months compared with the unsupplemented group(Reference Timby, Domellöf, Hernell, Lönnerdal and Domellöf123,Reference Timby, Lönnerdal, Hernell and Domellöf124) .
In a Swedish, double-blind, randomized controlled trial including 160 formula-fed healthy term infants and 80 breast-fed reference children, formula supplemented with a protein-rich MFGM concentrate given between 2 and 6 months decreased infectious morbidity until 6 months of age (significantly lower incidence of acute otitis, lower antipyretic use and lower serum concentrations of IgG against pneumococci after vaccination)(Reference Timby, Domellöf, Lönnerdal and Hernell119,Reference Timby, Hernell, Vaarala, Melin, Lönnerdal and Domellöf125) .
A Franco–Italian study in infants aged 14 d to 4 months, comparing supplementation with either a lipid-rich MFGM fraction (MFGM-L) or MFGM-P (Lacprodan MFGM-10: the first MFGM ingredient to enter the global infant formula market, rich in phospholipids and gangliosides), showed no difference in weight gain for the two groups(Reference Billeaud, Puccio, Saliba, Guillois, Vaysse and Pecquet126).
The use of MFGM in child nutrition can be provided by specific added products such as buttermilk-based ingredients or directly from milk cream.
These results call into question the historical choice of not including milk fat as part of infant formula and may serve as a basis for new strategies to evolve the nutritional formulations of infant milks. Recently, some formulas have begun to add dairy lipids, whereas the majority of current infant formulas contain lipids of exclusively vegetable origin. This addition of dairy lipids positively impacts both triacylglycerol structure (with increased sn-2 palmitate levels) and fat emulsion composition/ultrastructure with MFGM supplementation. By playing on the complementarity of plant and dairy lipids, rather than eliminating the latter, it seems possible to optimise the composition of infant formulas to make them more similar to breast milk. Preserving dairy lipids would thus have beneficial effects on child development, while retaining interesting taste properties.
Conclusions
This narrative review highlights the consensus and changes in current recommendations for the lipid intake of young children under 3 years of age. The contribution of lipids to EI appears to be the most consensual recommendation at the international level. At present, the benefits of a diet rich in omega 3 and 6 are widely recognised in lipid recommendations for young children aged 0–3 years.
However, this work raises persistent questions about current recommendations such as the absence of regulations on the addition of ARA in infant formulas despite clinical evidence showing that concomitant supplementation of ARA and DHA can mimic the effects of breastfeeding and ensure good brain and retinal development in children. It also points out that the evolution of official recommendations mainly concerns the quantitative aspect of LC PUFA while other lipids present in breast milk, such as SFA, MUFA, cholesterol, EPA and the structure of lipids for infants under 6 months, remain without recommendations.
The introduction of milk fat, including sn-2 palmitate and MFGM, seems to be a relevant way to improve the quality of infant formula. However, further randomised controlled trials evaluating the nature and function of lipid matrices are needed to increase scientific knowledge, specifically studies assessing the impact of the structure of lipids ingested during the neonatal period in metabolism programming in adulthood. Recent animal studies have also reported promising results on the supplementation of cholesterol in infant formulas, but the underlying mechanisms still need to be better understood and could help in designing improved nutritional strategies. We are confident that these aspects will lead to future improvements in infant formula and, possibly, to better preventive nutrition for future adults who cannot be initially breast-fed.
Acknowledgments
We thank Mrs Claire Bourlieu-Lacanal for providing Fig. 4 and for meaningful dialogue about the article.
Financial support
The salary of N. Nazek was funded by Biostime Laboratory during the year of the Master-2.
Competing interests
Najdi Nazek has received financial support from Biostime Laboratory for her Master-2 studies. Jung Camille has received grants from: Nestlé, Danone, Menarini and Dairy Goat Co-operative, which are not related to the present literature review. Castañeda-Gutiérrez Eurídice is an employee of Health and Happiness Group. Michalski Marie-Caroline received research funding from CNIEL, Sodiaal-Candia R&I and Danone Nutricia Research; congress travel funding from CNIEL; and symposium honorarium from IMGC, which are not related to the present literature review. Belaïche Marc has received financial support from Danone, Nestlé, Modilac, Reckitt and Biostim, which are not related to the present literature review. Bouziane-Nedjadi Karim declares none. Clouzeau Haude declares none. Coopman Stéphanie received grants from Sanofi, Cell Trion, Mirum and Danone, which are not related to the present literature review. De L’Hermuzière Clémentine declares none. Degas Vanessa declares none. Fabre Alexandre declares none. Garcette Karine has received financial support from Danone, Nutricia, Nestlé, Novalac, Sodilac, Reckitt and Lactalis, which are not related to the present literature review. Lalanne Arnaud declares none. Ley Delphine has received lecture fees from AbbVie and Sandoz, as well as travel grants from Nestlé, all of which are not related to the present literature review. Martinez-Vinson Christine declares none. Piloquet Hugues has received grants from Nestlé, Danone Hipp and Menarini laboratory, which are not related to the present literature review. Scheers Isabelle declares none. Beraud Virginie was an employee of Health and Happiness Group at the moment of her contribution to this review. Peretti Noël has received grants from Lactalis, Nestlé NHS and Nutricia-Danone, which are not related to the present literature review.
Authorship
Najdi Nazek: bibliographic research and writing of the first draft. Jung Camille: conception of the article and writing of the article. Castañeda-Gutiérrez Eurídice: supervision and correction of the article. Michalski Marie-Caroline: correction and significant improvement of the article and elaboration of the figures. Bourlieu-Lacanal Claire: correction and significant improvement of the article and elaboration of the figures. Belaïche Marc: correction of the article. Bouziane-Nedjadi Karim: correction of the article. Clouzeau Haude: correction of the article. Coopman Stéphanie: correction of the article. De L’Hermuzière Clémentine: correction of the article. Degas Vanessa: correction of the article. Fabre Alexandre: correction of the article. Garcette Karine: correction of the article. Lalanne Arnaud: correction of the article. Ley Delphine: correction of the article. Martinez-Vinson Christine: correction of the article. Piloquet Hugues: correction of the article. Scheers Isabelle: correction of the article. Beraud Virginie: conception of the article and writing of the article. Peretti Noël: conception, supervision of N. Najdi, writing and significant improvement of the article.
Open access
