The optimal quantity of dietary protein for preterm infant formulas is still a matter of debate. Dietary protein needs for preterm infants were calculated through two different methods [15]. The first, the empirical approach, measures biochemical or physiological responses to graded intakes. The second method, the factorial approach, considers the requirements as the sum of the essential losses (e.g., urine, feces, skin), plus the amount incorporated into newly formed tissues accounting theoretical fetal accretion at the same PMA. The empirical approach evaluates physiological and biochemical variables to determine minimum and maximum protein needs in the growing preterm infant.

In the factorial approach, compositional analysis of fetal tissues has been a valuable source of data for our understanding of the nutrient needs of the fetus and, by extension, those of the growing preterm infant. Fetal accretion rates have been obtained from compositional analyses of aborted fetuses or stillborn infants. From these data, the protein increment for growth has been estimated as approximately 2.5 g/kg/day [5]. Protein requirement is thus calculated by adding to this value the obligatory losses, and the amount needed for the additional catch-up growth. In a large population of preterm infants receiving a controlled energy intake, the minimal protein supply necessary to obtain a zero nitrogen balance (to cover all nitrogen losses) was calculated to be 2 g/kg/day. A similar value was estimated to cover the need for catch-up growth. Thus, the protein requirements of premature infants are between 3.5 and 4.0 g/kg/day (Fig. 6.2) [5].

Protein intake must be sufficient to achieve normal growth without negative effects such as acidosis, uremia, and elevated levels of circulating amino acids [1, 15, 17]. A recent Cochrane analysis aimed to determine whether higher (≥ 3.0 g/kg/day) versus lower (< 3.0 g/kg/day) protein intake during the initial hospital stay of formula-fed preterm infants or LBW infants (< 2500 g) results in improved growth and neurodevelopmental outcomes without evidence of short- and long-term morbidity [17]. Five studies comparing low versus high protein intake were included. Improved weight gain and higher nitrogen accretion were demonstrated in infants receiving formula with higher protein content while other nutrients were kept constant. No significant differences were seen in rates of NEC, sepsis, or diarrhea. One study compared high (3 g/kg/day) versus very high (> 4 g/kg/day) protein intake during and after an initial hospital stay [18]. Very high protein intake improved gain in length at term, but differences did not remain significant at 12 weeks corrected age. Three of the 24 infants receiving very high protein intake developed uremia, defined as serum urea nitrogen level greater than 6 mM/ml. Formula with higher protein content improves all growth parameters and neurodevelopment [19, 20]. However, available evidence is not adequate to permit specific recommendations regarding the provision of very high protein intake (≥ 4.5 g/kg/day) from formula during the initial hospital stay or after discharge.

Nitrogen absorption rate (absorbed/intake) differs significantly according to the feeding regimen (Fig. 6.3) [21] . It is higher with powder whey-predominant protein preterm formulas than with human milk supplemented with fortifiers powder protein hydrolyzed formulas or ready-to-use liquid whey-predominant preterm formulas. These differences result from the nature of the various types of protein supply. Also net protein utilization is higher for preterm formulas compared with human milk, with or without protein supplementation. Technical processes (i.e., hydrolysis) may alter macronutrient concentrations of formula and the pasteurization of human milk before delivery to newborn also cause a reduction in the protein concentration of about 4 % [22].

The whey/casein ratio significantly influences individual amino acid intakes and plasma amino acid concentrations. Plasma threonine increases and tryptophan relatively decreases in infants fed a whey-predominant formula, whereas methionine and aromatic amino acids increase in those fed a casein-predominant formula [23]. The high plasma threonine concentration observed in preterm infants fed whey-predominant formulas is related to the glycomacropeptide obtained from casein by enzymatic casein precipitation of cow’s milk proteins . Acidic precipitation, by contrast, removes the glycomacropeptide rich in threonine from the soluble phase [23]. In preterm infants receiving either an enzymatic or an acidic whey-predominant formula, a significant reduction in plasma threonine concentration was observed in acidic whey protein formula-fed infants, compared to those receiving conventional enzymatic whey protein formulas. All other plasma amino acid concentrations were similar, with the exception of valine, which was slightly reduced in acidic whey protein formula-fed infants [15, 24].

The relative percentage of α-lactalbumin depends on the cow’s milk protein content [25]. This protein fraction is naturally rich in tryptophan and helps normalize the low levels of this amino acid frequently observed in whey-predominant formula-fed infants. The use of a higher percentage of whey in protein-hydrolyzed formulas may worsen the plasma amino acid pattern by increasing threonine and decreasing aromatic amino acid concentrations [25]. Moreover, a significant decrease of plasma histidine and tryptophan concentrations, probably due to a relative reduction in amino acid bioavailability, could be also observed. Using a more appropriate technology, these formulas have been corrected for the threonine content and supplemented with histidine and tryptophan.

An interesting observation is the relatively lower absorption rate of ready-to-use liquid preterm formulas or protein hydrolyzed preterm formulas, where the technical process seems to impair nitrogen absorption in relation to heat treatment—inducing some Maillard reaction—or to preliminary hydrolysis, altering the physiological absorption process in the lumen or at the border of the GI tract. Formulas based on hydrolyzed proteins have been recently proposed for the feeding of preterm infants to reduce GI problems such as delayed gastric emptying, abdominal distension, hard stools, and feeding intolerance [26]. The technological processes of hydrolysis may modify the amino acid content and/or amino acid bioavailability [27]. The efficiency of protein differs also according to non-protein nitrogen contentof different feeding regimens. The highest values were obtained in preterm infants fed powder and liquid preterm formulas . It was significantly lower in those fed protein-hydrolyzed formulas and fortified human milk. The lower value obtained with fortified human milk may be related to nonprotein nitrogen content, which represents 20–25 % of total nitrogen content of human milk, but still 13.5–17 % of total nitrogen content of fortified human milk. The contribution of nonprotein nitrogen fraction of fortified human milk to protein gain is lower than that of the α-lactalbumin or the casein content of human milk [1, 15, 23, 24]. Due to its biological activities, the nonprotein nitrogen of human milk may not be included in protein intakes when human milk is compared to formulas. This may explain why we still frequently observe differences between infants fed human milk or PTF with theoretical similar nitrogen content .

Formulas do not contain any of the biologically active immune proteins nor enzymes, hormones, or growth factors are found in human milk [2]. The long-term implications of this deficiency have not been completely determined. Further long-term randomized studies with higher number of patients, and standardized supplemental amounts and experimental conditions are needed to establish the efficacy of nonnutritional compounds, such as nucleotides, polyamines, and growth factors supplementation in preterm formula [1, 5].

Fat provides the major source of energy in growing preterm infants. Recommended fat intake is 4.8–6.6 g/kg/day [5]. Although some infants with restricted fluid may need high fat content in their feeds to meet energy needs, 4.4–6.0 g/100 kcal is a reasonable range for most preterm infants [16]. Lipid availability by enteral route depends more on digestion and absorption capability than on enteral nutrition content and composition. Digestion of lipids is not fully developed in premature infants [2]. Lipid malabsorption is the result of low levels of pancreatic lipase and bile salts. The bile acid pool is only half the size of that of a full-term infant. Human milk provides additional lipases: lipoprotein lipase and bile salt-stimulated lipase , which contribute to intestinal lipolysis. However, additional fat should be also provided to human milk because of the possible reduction of fat content in expressed human milk, especially when various processes are applied to human milk prior to delivery to neonates [19, 20].

Fats represent approximately 50 % of the nonprotein energy content of human or formula milk. Fat could be provided in different form with different nutritional power .

Recent studies have repoKrted outcome data in preterm infants fed milk with a DHA content two to three times higher than the current concentration in infant formulas. Overall, these studies show that providing larger amounts of DHA supplements, especially to the smallest infants, is associated with better neurologic outcomes in early life [34]. DHA intakes of 55–60 mg/kg/day from the time of preterm birth to expected term have been tested and appear to be safe, to promote normal DHA status, and to improve visual and neurocognitive functions. These values are likely to represent an adequate intake for very preterm infants, but further research is needed to confirm that they represent adequate intake for other infants subgroups (i.e., extremely, very, and moderately preterm infants, with or without intrauterine growth restriction , and males and females). Because the maximum DHA content of human milk is 84 mg/kg/day (equivalent of 1.5 % of fatty acids), and because no studies with such a high intake in preterm infants have been reported, no upper limit can be set with certainty.

Synthesis of long-chain LCPUFAs is limited in preterm infants, thus, these molecules are considered as semi-essential or essential. Several studies have suggested that the LCPUFAs supply in preterm infants has a beneficial effect on growth, visual, and cognitive function, as well as on the immune system . Although some methodological issues in these studies do not allow definitive conclusions to be drawn, most of the formulas in Europe and the USA are currently supplemented with LCPUFAs [32–34] .

Recommended intakes are 12–30 mg/kg/day or 11–27 mg/100 kcal for DHA and 18–42 mg/kg/day or 16–39 mg/100 kcal for AA [5]. The ratio of AA to DHA should be in the range of 1.0–2.0 to 1 (wt/wt), and EPA supply should not exceed 30 % of the DHA supply [5, 34]. Recently, an experts panel has recommended (18-) 55–60 mg/kg/day or (16.4-) 50–55 mg/100 kcal for DHA and (18-) 35–45 mg/kg/day or (16.4-) 32–41 mg/100 kcal for AA [16].

Pasteurization and storage of banked human milk induces lipolysis, inactivates bile salt-stimulated lipase and lipoprotein lipase, reduces fats, and increases the absolute amount of free fatty acids in pooled samples. Holder pasteurization involves a modification of the biological activity of a large number of enzymes, including lipases [19, 20]. These effects alter the integrity of human milk and may contribute to slower growth in preterm infants fed banked milk versus their mothers’ own milk . On the basis of these evidences, a further supplementation with specific fatty acids could be considered for these infants [34–36].

Carbohydrates are essential to provide adequate energy supply and they represent a major source of dietary energy. Based on carbohydrate equivalents of total energy expenditure, the experts have recommended a minimum of 10.5 g/100 kcal and a maximum of 12.0 g/100 kcal [5, 16]. Carbohydrates are essential to provide adequate energy supply. Additionally, carbohydrates promote epithelial cell proliferation, insulin secretion, and calcium absorption [5, 16]. Thus, they are also essential to the overall health of the GI tract other than the fulfillment of energy requirements.

The predominant carbohydrate in mammalian milk and in term infant formula is lactose. After digestion by lactase, lactose is absorbed as glucose and galactose, which utilize the same carrier mechanism. In the human fetus, intestinal lactase activity is measurable by 10–12 weeks of gestation [37] . There is a gradual increase in lactase activity with advancing gestation, although the activity remains low until about 36 weeks of gestation, when it reaches the levels seen in full-term neonates [37]. Based on the low lactase activity in early gestation and the estimated length of the bowel, it was calculated that a preterm infant weighing 1300–1400 g might be expected to absorb only 30–50 % of the ingested lactose. The lactose that is not fully digested serves as a source of nutrition for bacterial flora in the colon, where it is transformed into short fatty acids and then absorbed [2]. Through the process of fermentation, this not only facilitates colonic water and electrolytes absorption but also stimulates cell turnover of both the colon and the small intestine. After lactose, oligosaccharides are the largest other carbohydrates component in mature human milk. They represent the third largest solute load of the diet after lactose and fat [37]. Oligosaccharides are hydrolyzed by salivary, pancreatic, and intestinal amylase and maltase to free glucose, which is rapidly absorbed. Many of the protective effects of human milk (i.e., against diarrheal diseases, respiratory, and ear infections) have been attributed to oligosaccharides. A further advantage of oligosaccharides for preterm infant formulas includes the improved gastric emptying [38, 39]. Oligosaccharides may be considered as prebiotics. There are some reports that such prebiotics have beneficial effects on various markers of health. For example, primary prevention trials in infants have provided promising data on prevention of infections and atopic dermatitis. However, additional well-designed prospective clinical trials and mechanistic studies are needed to advance knowledge further in this promising field .

Several nutrients cause a reduction of calcium bioavailability [45–47]. Various factors affect calcium absorption: vitamin D status, solubility of calcium salts, and the quality and quantity of fat intake. In preterm infants, the vitamin D body stores at birth depend mainly on maternal vitamin D status. In the USA, where dairy products are routinely supplemented with vitamin D, a daily additional intake of 400 IU in the formula may sometimes be sufficient to maintain an adequate plasma concentration of 25-OH and 1–25(OH)2 vitamin D. By contrast, in most parts of Europe, cord blood concentration of 25-OH vitamin D of premature infants is frequently less than 10 μg/ml; in this case, up to 1000 IU/day of vitamin D are recommended [5]. The increased needs of premature infants could be partly due to a relative malabsorption of vitamin D, resulting from a low secretion of bile acids [48]. A calcium retention close to 90 mg/kg/day could currently be expected in preterm infants fed preterm formula with a highly soluble calcium content. These values are still lower than the estimated fetal accretion during the last trimester of gestation (100–120 mg/kg/day). However, there are dramatic physiological changes in bone metabolism resulting from various factors after birth: disruption in maternal mineral supply, stimulation of calciotropic hormone secretion, change in hormonal environment, and relative reduction in mechanical stress. These events stimulate the remodeling process leading to an increase in endosteal bone resorption and a decrease in bone density. In preterm infants, these adaptation processes modify the mineral requirement, since, by itself, the increased remodeling and bone turnover provides a part of the mineral requirement necessary for postnatal bone growth [46].