REVIEW URRENT C OPINION
Challenges in breast milk fortification for preterm infants Niels Rochow, Erin Landau-Crangle, and Christoph Fusch
Purpose of review To outline new evidence published from 2013 to 2014 about breast milk fortification in preterm infants. Recent findings Breast milk is the feeding choice for preterm infants because of its immunoprotective properties. However, breast milk’s nutrient content is not sufficient for preterm infants, and interindividual variation is high. The variation challenges standard fortification, which assumes a standard breast milk composition. Two new fortification strategies (adjustable fortification and target fortification) optimize macronutrient intake and improve growth. Adjustable fortification uses blood urea nitrogen levels to adjust fortifier strength. Target fortification analyzes breast milk and fortifies macronutrients individually to achieve targeted intake. Its feasibility is shown in clinical routine. Current breast milk analyzers used for target fortification achieve acceptable precision for protein and fat but not for lactose and energy. Evidence of benefits for postdischarge breast milk fortification is lacking. Eliminating cow’s milk products and feeding exclusively breast milk may decrease the occurrence of feeding intolerance and necrotizing enterocolitis. To facilitate exclusively breast milk diets, a collaboration of prenatal, nutrition and lactation stakeholders is key. Fortification increases osmolality; however, safety cutoffs to prevent necrotizing enterocolitis are unclear. There is also new evidence that composition and structure of various macronutrients and micronutrients affect growth and development, and might play a role in future fortification concepts. Summary Recent research focuses on the variability of breast milk composition, its impact on postnatal growth patterns and the usefulness of target fortification. As well, diets exclusively composed of human milk are a promising approach to improve feeding tolerance. For safe fortification, osmolality cutoff levels are needed. Keywords breast milk analysis, growth, lactation, nutrition, target fortification
INTRODUCTION Progress in neonatal care has improved survival and reduced morbidity of very low birth weight (VLBW, birth weight < 1500 g) infants. However, between 43 and 97% of these infants still experience postnatal growth restriction and do not follow in-utero weight gain trajectories after birth [1,2]. This is of concern as poor growth is associated with impaired neurological development and risk of early onset of adult metabolic diseases [3]. To improve growth [4–8], a bundle of strategies such as higher parenteral and enteral nutritional intake, advancing enteral feeds earlier and faster, adjusting enteral feeding according to daily weight gain, tablet-based nutritional monitors for fluid, energy and macronutrient intake and daily plotting of individual growth trajectories on reference curves [9 ,10,11] have been used during daily &
www.co-clinicalnutrition.com
patient management rounds [5]. Since 2013, 12 reviews have summarized feeding strategies for preterm infants [1,2,12–21]. All recommend early, ample supply of protein and energy, preferably provided enterally. Breast milk is the preferred choice because of digestibility and immunoprotective effects. The high growth rates of VLBW infants increases nutritional needs above what can be provided by native breast milk [17]. Therefore, fortification is required [13,18,22 ,23]. The widely used &&
Division of Neonatology, Department of Pediatrics, McMaster University, Hamilton, Ontario, Canada Correspondence to Niels Rochow, MD, Division of Neonatology, Department of Pediatrics, McMaster University, 1280 Main Street West, HSC4F5, Hamilton, ON L8S 4K1, Canada. Tel: +1 905 521 2100x75721; fax: +1 905 521 5007; e-mail: [email protected] Curr Opin Clin Nutr Metab Care 2015, 18:276–284 DOI:10.1097/MCO.0000000000000167 Volume 18 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Challenges in breast milk fortification Rochow et al.
KEY POINTS
APPROACHES TO BREAST MILK FORTIFICATION
Target fortification using a fortifier and modular products for protein, fat and carbohydrates to adjust breast milk content to the recommended levels is feasible in clinical routine, although the effects on growth and neurodevelopment need to be studied.
To compensate for the low macronutrient levels of breast milk and provide sufficient intake for VLBW infants, three fortification approaches are currently used in neonatal intensive care units (NICUs) [22 ,23].
Exclusively breast milk diets for preterm infants seem to be able to reduce high NEC rates in some centers; however, the effect on centers with low NEC rates needs to be studied.
Standard (‘blind’ or fixed-dose) fortification
Preterm infants have immature buffer systems and are at risk for metabolic acidosis when acid load of fortifiers exceed thresholds, which may impair growth and development. The structure of specific macronutrients (e.g., types of fat) and micronutrients varies in breast milk and may affect growth and development. Fortification leads to increase of osmolality, in particular in combination with additional supplementation; however, the impact on NEC rates is not proven, and the science behind the safety cutoff is limited.
&&
Standard (‘blind’ or fixed-dose) fortification is widely established and assumes a standard composition of breast milk. Fortifiers containing protein, carbohydrates, fat, electrolytes, calcium, phosphate, vitamins and trace elements are added in fixed doses. These products are in powder or liquid form, and contain hydrolyzed or intact protein derived from cow’s milk. One commercially available fortifier product is derived from breast milk. The energy contents of fortifiers are similar whereas their composition differs. At recommended doses, they provide extra protein (1–1.1 g/dl), but the amount and composition of nonprotein calories varies considerably by manufacturer (from 0 to 1 g/dl for fat, 0.4– 3.4 g/dl for carbohydrates) [22 ]. In case of insufficient growth, it is a common practice to increase fortifier strength and/or to arbitrarily add modular protein, fat and/or carbohydrates. In a retrospective study (n ¼ 206), infants fed more than 75% fortified breast milk showed less growth on day 28 and at discharge when compared with infants fed less than 25% fortified breast milk. Standardized fortification of breast milk at current recommended doses may not adequately support early postnatal growth [32]. It could be hypothesized that the lower growth associated with fortified breast milk is due to underachieved target macronutrient levels. In another three-arm randomized trial (n ¼ 84), infants received fortification at different strengths with a protein intake of 3, 3.3 and 3.6 g/kg/day and found better head growth with higher protein intake [33]. Renal glomerular and tubular functions were unaffected by protein intake up to 3.6 g/kg/day [34]. In a multicenter randomized control trial (RCT), (n ¼ 150) using an ultraconcentrated liquid breast milk fortifier or powdered breast milk fortifier, protein intake was a significant predictor of length but not weight [35]. Addition of 0.5–2.3 g protein/kg/day to standard fortified breast milk led to appropriate growth (18 g/kg/day, head 7 mm/week, length 1.0 cm/week, n ¼ 43) [36]. The standardized fortification approach was recently applied in a study on breast milk with human donor milk-derived fortifier (n ¼ 104; 130 kcal/kg/day, 3.6 g protein/kg/day). Fortification was increased to 140 (maximum 150) kcal/kg/day, &&
concept of breast milk fortification is calculated with the assumption of a standard composition of macronutrients and micronutrients. Challenges arising from this approach include risk to develop postnatal growth restriction in infants fed nutrientdeficient breast milk; adverse effects of fortifiers including reaction to cow’s milk [allergy, inflammation and necrotizing enterocolitis (NEC)], metabolic acidosis and feeding intolerance; establishing an adequate fortifier composition (fat:carbohydrate ratio and micronutrients); optimizing on-site milk handling/fortification infrastructure.
BREAST MILK COMPOSITION A recent systematic review of 41 studies, including 843 mothers who delivered preterm and 2299 mothers who delivered at term, found highly variable levels of energy, protein, lactose, oligosaccharides, fat, phosphorus and calcium in breast milk between individuals and across postnatal and gestational age [24 ]. This variability was further confirmed for preterm, term and donor milk by seven recent studies [25,26 ,27–30,31 ]. Interestingly, macronutrient levels do not correlate to one another and cannot be predicted from each other. Prior to individualized or target fortification, all three macronutrients in breast milk need to be analyzed to guide fortification and avoid unexpected nutrient content [26 ,28]. &&
&
&
&
1363-1950 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-clinicalnutrition.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
277
Paediatrics
4.4 (maximum 5.25) g protein/kg/day when growth was less than 15 g/kg/day. Infants achieved a high weight gain of 24.8 5.4 g/kg/day from birth (gestational age 27.6 2.0 weeks) to discharge (gestational age 40.8 6.0 weeks), starting at a birth weight of 913 182 g [37]. However, there was a discrepancy reported between weight gain and the weight at discharge of 2795 g [37]. A weight gain of almost 25 g/kg/day may be inappropriate to achieve a normal body composition. Current standard fortification of breast milk does not ensure appropriate weight gain for all preterm infants. This may be because of interindividual and intraindividual variation in the macronutrient content of breast milk, including some low protein and/or fat levels [22 ,24 ,26 ,38 ]. Two approaches are proposed to account for varying breast milk composition: ‘adjustable’ and ‘target’ fortification [22 ,23]. &&
&&
&
&
&&
Adjustable fortification Adjustable fortification is a modified approach to standard fortification using blood urea nitrogen (BUN) as a surrogate marker for metabolic response to determine appropriate protein intake. The fortification strength is adjusted according to BUN values [39,40]. In an observational study (gestational age 32 weeks or less), higher protein intake of 4 g/kg/day via adjustable protein fortification (standard fortifier plus additional protein, n ¼ 29) compared with standard fortification (2.8 g protein/kg/day, n ¼ 29) led to higher weight gain, length and head circumference. Protein was increased by 0.55 g/dl breast milk when BUN was less than 9 mg/dl, decreased by 0.55 g/dl if BUN was 14–20 mg/dl. Addition of protein was withheld for 1 week if BUN was 20 mg/dl or more [39]. In a retrospective study, infants received a protein product (n ¼ 33) in addition to standard fortification when BUN was less than 5 mg/dl and/or prealbumin was 8 mg/dl or less. Supplementation was decreased gradually once BUN 9 mg/dl or more and prealbumin reached 8 mg/dl or more, and was discontinued at 40 weeks gestational age. Infants with additional protein fortification received 0.5–1.5 g/kg/day more protein than the standard fortification group (n ¼ 32). Weight gain and head growth improved, and Bayley scores at 18 months were higher in the group with extra protein [40].
Target (individualized or customized) fortification Target (individualized or customized) fortification accounts for variability of protein, carbohydrate 278
www.co-clinicalnutrition.com
and fat through regular breast milk analysis and fortification to provide the recommended intake of macronutrients [38 ,41,42] and to optimize growth. In a study comparing data from standard and individualized fortification approaches, daily breast milk composition was measured with a mid-infrared milk analyzer (n ¼ 24 infants). For their individualized approach, breast milk fat content was first adjusted to 4 g/dl by adding modular fat. Next, a fortifier was added to reach a protein intake of 4.3 g/kg/day (increased protein by 1.1 g and increased fat by 1 g). As a result, the variability of macronutrients in the individualized approach was significantly decreased, but the average fat intake was 8.6 g/kg/day [41], which exceeded recommendations [21]. In a two-center RCT (n ¼ 78), breast milk energy density was measured with a near-infrared analyzer. Infants received cream in addition to donor milkderived fortifier if energy density was less than 20 kcal/oz (intervention group n ¼ 39) and showed superior weight (14.0 2.5 vs. 12.4 3.0 g/kg/day) and length (1.0 0.3 vs. 0.8 0.4 cm/week) gain vs. the control group without cream [42]. However, in recent articles the infrared analyzers have been extensively validated and have determined that the measurement of calories is not precise because of the inability to accurately measure lactose with these devices [31 ,43]. A recent pilot study (gestational age 32 weeks or less) was the first to show the feasibility of target fortification of all macronutrients through daily breast milk analysis (near-infrared), using modular products to adjust levels of fat to 4.4 g/dl, protein to 3 g/dl and carbohydrates to 8.8 g/dl. Infants with target fortification but not standard fortification, showed growth rates with a high correlation to volume of breast milk fed. The lack of correlation between volume of breast milk fed and growth rates for the standard fortification approach can be explained by the 30–40% of infants who experience poor growth as their breast milk falls below the average composition of breast milk used to calculate fortification with the standard fortification approach, and thus are fed nutrient deficient breast milk [38 ]. To ensure appropriate fortification for each infant, standard, adjustable and target fortification utilize different approaches. Standard fortification is adjusted when infants show failure to grow. Adjustable fortification relies on monitoring metabolic response via markers such as BUN. Standard and adjustable fortification risk feeding macronutrientdeficient breast milk as they are outcome-based approaches. Target fortification analyzes breast milk &
&
&
Volume 18 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Challenges in breast milk fortification Rochow et al.
and provides individual fortification to achieve target macronutrient levels.
BREAST MILK ANALYSIS IN CLINICAL ROUTINE Macronutrient analysis of breast milk is needed for target fortification. To make analysis feasible in clinical routine, a breast milk analyzer should fulfill certain requirements: easy-to-use bedside devices, low sample volume (1 ml), low maintenance and rapid analysis. Few milk analyzers using mid-infrared or near-infrared spectroscopy are commercially available [44 ] and were originally developed for use in the dairy industry. Analyzers require recalibration as breast milk has a different matrix and optical characteristics than cow’s milk. Standard chemical methods require 30–60 ml of milk, and cannot be used to measure or validate protein, fat and carbohydrate levels in clinical settings. A set of modified micromethods (1.5 ml) have been developed using Mojonnier ether extraction (fat), elemental analysis (protein) and ultraperformance liquid chromatography-tandem mass spectrometry (lactose) [45]. To overcome the drawbacks of the first validation attempts (limited sample size [46] and inappropriate secondary conclusions [47]), two low sample volume breast milk analyzers were validated in a recent study (n ¼ 1188 breast milk samples, 63 mothers of preterm and term infants). These analyzers have the potential to be introduced in clinical routine to measure fat and protein content using correction algorithms, but require major adjustments for lactose and calories. It is speculated that the lactose measurements are confounded by oligosaccharides, which are highly abundant in breast milk but scarcer in cow’s milk. Consequently, as a precise lactose measurement is not available, calorie content cannot be calculated. Preanalytical sample handling has also been assessed: the evidence for the influence of aliquoting, storage time, temperature, container wall absorption, on stability and availability of macronutrients in frozen breast milk was reviewed [31 ]. Use of laboratory analyzers in clinical routine requires quality checks. Owing to the lack of a commercially available breast milk standard, the authors suggest creating one by pooling 500 ml of random leftover breast milk samples and dividing them into 1 ml aliquots to use as daily quality controls before measurement of patient samples [45]. A promising method using Fourier transform mid-infrared spectroscopy was validated using (n ¼ 116) breast milk samples with chemical methods (Kjeldahl, Mojonnier and high-performance liquid chromatography). Agreement between the breast milk analyzer and chemical methods was &
high for protein and fat, low for lactose and energy. These authors also speculate the interference of oligosaccharides in the accurate measurement of lactose [43]. Can milk analysis be simplified? Current commercial breast milk analyzers cost US$ 17 000– 59 000 [44 ]. Creamatocrit analysis, measuring the cream layer of breast milk, could be a low-cost alternative (for capital/one-time costs). A recent study used 51 leftover breast milk aliquots to compare creamatocrit analysis with mid-infrared spectroscopy. The analyzer was validated with reference methods (Kjeldahl, high-performance liquid chromatography and Mojonnier) and a correction equation was developed [48]. The data about energy content need to be interpreted with caution because of limitations to measure lactose with current infrared methods [31 ,43]. Even if fat content could be determined with creamatocrit analysis, other components cannot be predicted from a measured fat level. It has been shown that macronutrient levels do not correlate but need to be measured separately [26 ]. Therefore, the creamatocrit analyzer currently cannot be recommended as a low-cost alternative. With respect to work hours, daily breast milk analysis for target fortification would increase NICU workload by 10–15 min/patient/day and may not be feasible in all nurseries. An observational study (n ¼ 210 pooled native 24 h breast milk samples from 10 infants) varied the number of measurement days and analyzed the variation of macronutrient intake. Day-to-day variation of macronutrients (protein 20%, carbohydrate 13%, fat 17%, and energy 10%) decreased with increased frequency of analysis. Two measurements per week led to a mean macronutrient intake within a range 5% of target levels. The extent of the effects of day-to-day variation of nutritional intake on growth are unclear [49]. In summary, infrared analysis seems to be a promising tool for fat and protein with calibration, but lactose and energy cannot be assessed with the current state of technologies. &
&
&
&
BREAST MILK MANAGEMENT CENTER To make breast milk available in neonatal intensive care requires an adapted infrastructure. Key prerequisites are collaboration between healthcare providers and nutrition experts, counseling mothers on nutritional needs of preterm infants, donor milk availability, breast milk analysis and handling and storage of breast milk. Recent articles described challenges and suggestions to improve breast milk availability [44 ,50,51,52 ,53]. One study reported a ‘mobile’ milk cart for preparation/fortification of breast milk for NICUs with limited space [50].
1363-1950 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
&
&
www.co-clinicalnutrition.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
279
Paediatrics
Another study presented a centralized milk management center for breast milk measurement, target fortification and safe preparation [44 ]. Measuring breast milk composition in lab and target fortifying at bedside has also been described [38 ]. A ‘Supporting Premature Infant Nutrition’ program includes the collaboration of the hospital nutrition and lactation stakeholders and 10 key points [52 ]. Another challenge is to provide fortification at home. Postdischarge preterm infants could benefit from breast milk fortification; however, two recent reviews demonstrated limited evidence of the impact of postdischarge breast milk fortification on growth and neurodevelopment [54,55]. &
&
&
EXCLUSIVELY FEEDING BREAST MILK Seven recent reviews suggest that breast milk decreases the risk of NEC and late-onset sepsis, and should be the primary enteral diet for premature infants. Donor milk is a resource for premature infants whose mothers are unable to provide adequate milk supply. Further, donor milk-derived fortifier can be used to feed infants exclusively breast milk. However, donor milk and derivatives require pasteurization, which decreases the concentrations of most cytokines and lactoferrin, and is limited in supply. Transforming growth factor-b2 is a protective intestinal cytokine and surprisingly showed comparable concentrations in breast milk and donor milk-derived fortifier, despite pasteurization [56]. In general, donor milk differs nutritionally from breast milk, and neonatologists should be aware that these differences might affect the need for nutrient fortification [57–63]. In the last 2 years, further evidence favoring an exclusively breast milk diet has shown that intake of cow’s milk protein may lead to a proinflammatory reaction on the neonatal gut. Cytokine secretion of interferon-g, interleukin-4, interleukin-10 and transforming growth factor-b1 was observed in response to b-lactoglobulin and casein [64]. The clinical impact of feeding exclusively breast milk on NEC rates has been shown in four recent clinical studies. A multicenter study demonstrated a significantly different NEC rate for cow’s milk protein-based fortifiers vs. breast milk protein-based fortifiers with 17 and 5%, respectively, (n ¼ 93 vs. n ¼ 167) [65]. Another multicenter RCT (n ¼ 53) reported similar findings with a significantly lower NEC rate: 3% (exclusively breast milk) vs. 21% (preterm formula) [66]. A retrospective before–after study (n ¼ 642) compared a 2.5-year period of exclusively breast milk diets with the preceding 6.5 years when cow’s milk products were given and found a decrease in NEC rate from 3.4% (15/443) to 1% 280
www.co-clinicalnutrition.com
(1/199) [67]. Another observational study (n ¼ 104) analyzed data from infants who received a breast milk-derived fortifier and found only three medical and one surgical NEC case [37]. Data from these studies show a promising trend of decreased NEC rates with an exclusively breast milk diet. However, the reported NEC rate in infants receiving cow’s milk-derived products was 17–21%, unexpectedly high compared with average NEC rates reported by international networks of 1.6–6% [68]. A recent study reviewed enteral feeding type (breast milk, formula and nutrient fortification) for effects on microbiota. Altered colonization of the microbiota has been associated with increased morbidity in preterm infants. There is a lack of evidence demonstrating the impact of breast milk fortification on gut microbiota [69]. An observational study (n ¼ 27) investigated the addition of prebiotic oligosaccharides to stimulate commensal bacterial growth and to bind pathogens within the intestinal lumen, as a safer method of NEC prevention. Formula-fed infants (n ¼ 12) received galactooligosaccharide or concentrated donor milk with human oligosaccharides. In (n ¼ 15) infants, breast milk was fortified with a donor milk-derived fortifier or cow’s milk-derived fortifier. None of the prebiotic interventions resulted in significant change of microbiota. Limitations of the study were lower-than-proposed oligosaccharide dose for treatment group and antibiotic use [70]. Feeding exclusively breast milk appears to decrease feeding intolerance and NEC rates. Future clinical studies are required in which the initial NEC rate for cow’s milk-based fortifiers are closer to the international average. The microbiota or addition of specific oligosaccharides may be a further target to improve feeding tolerance.
SPECIFIC MACRONUTRIENTS AND MICRONUTRIENTS Breast milk macronutrient and micronutrient content varies in composition within and between individuals and over time. Additionally, there is variation of composition amongst diverse commercial fortifiers in regard to quality of macronutrients and micronutrients. Some fortifiers contain v-3 fatty acids or iron whereas others do not. In order to provide infants with the recommended nutrition, the composition of fortifiers (not only breast milk) must be considered. In this paragraph, we will list all the new studies that have either tested additional components in fortifiers or that provide evidence that a component has the potential to be a useful supplement in fortifiers or needs to be tested. Volume 18 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Challenges in breast milk fortification Rochow et al.
A recent review (seven studies, n ¼ 1914) analyzed neurodevelopment at less than 3 years (three studies) and at 5–11 years (four studies) and speculated that breast milk contains neurotrophic factors that benefit brain development. The authors concluded that increasing use of donor milk, new preterm formulas and formulation of new fortifiers requires reassessment of the effects on neurodevelopment [71]. One potential candidate for a neurotrophic factor could be v-3 fatty acids that are needed for visual and cognitive development in early life. Two reviews showed that v-3 content of breast milk is highly variable following maternal diet. Nutritional counseling during lactation was suggested [72,73]. In a multicenter RCT (14 sites, n ¼ 150), infants received powdered fortifier or liquid fortifier with additional docosahexaenoic acid (DHA) 12 mg/dl and arachidonic acid (ARA) 20 mg/dl. Infants on liquid fortifier had higher phospholipid DHA/ARA levels at day 28 and were related to maternal DHA/ARA intake [74 ]. Sphingomyelin plays a role in brain cell membrane structure. Feeding sphingomyelin to rats led to increased levels of DHA in red blood cell membranes and accelerated myelination of cortical areas. Infants in an RCT (n ¼ 24) received either 13 or 20% sphingomyelin-fortified milk. At 18 months, infants with 20% sphingomyelin showed a trend toward better neurobehavioral development [75]. Increased phenylalanine levels may have unfavorable effects on brain development. A recent study (n ¼ 24) found an increase of plasma phenylalanine levels from 12 1 to 30 1 mmol/l in infants fed breast milk fortified with cow’s milk protein. This was not found in infants fed exclusively breast milk. The long-term effects of this increase are unclear [76]. Vitamin D, calcium and phosphorus are essential for bone development in preterm infants. A recent clinical report suggested a Vitamin D intake of 200–400 IU/day [77]. For phosphate, an intake of 184–230 mg/kg/day is recommended. A retrospective study (n ¼ 106) showed that under standard fortification, additional phosphate supplementation was necessary to achieve phosphate levels between 1.8 and 2.6 mmol/l [78], this was confirmed by a prospective study (n ¼ 79) [79]. The importance of maintaining adequate phosphate levels was shown in an RCT (n ¼ 50) in which a high sepsis rate was related to low phosphate levels and electrolyte imbalances [8]. Folate is essential for rapidly growing preterm infants. Heat treatment of donor milk decreases folate by 14–25%. Nonfortified, heated breast milk supplied only 25% of the recommended daily intake to preterm infants [80]. Another study (n ¼ 162) &
showed that providing parenteral nutrition and fortified breast milk with sufficient folate content reduces the risk for folate deficiency during the first 2 postnatal months [81]. Zinc is needed to achieve sufficient growth rates. In a retrospective study (n ¼ 52), supplementation of 1.7 mg/kg/day was associated with an increase in weight gain from 11 to 20 g/kg/day, and linear growth from 0.7 to 1.1 cm/week [82]. Elemental iron supplementation of 2–4 mg/kg/ day for breast milk-fed preterm infants during the first 12 postnatal months is recommended. An observational study (n ¼ 152) measured ferritin levels at term, 3 months and 6 months and showed a higher occurrence of low ferritin less than 12 mg/l in breast milk-fed infants until 6 months. Iron-fortified formula with 0.8–1.0 mg iron/dl led to normal ferritin levels [83]. Carnitine plays a role in energy metabolism via long-chain fatty acids. A recent observational study (n ¼ 38) showed higher carnitine levels of formulafed compared with breast milk-fed infants during the first 28 days of life [84]. These studies provide evidence that micronutrients’ composition and the structure of proteins and fat may affect preterm infants’ metabolism. Depending on fortifier composition, micronutrients or other specific nutrients might need to be added.
(POTENTIAL) SIDE-EFFECTS OF FORTIFICATION It is obvious that fortification of breast milk has improved growth rates and neonatal care. However, like with any other intervention, it has some sideeffects, though the aim is to minimize these. The American Academy of Pediatrics has recommended that osmolality of enteral feeds should not exceed 450 mOsm/kg (400 mOsm/l). This recommendation is based on a series of case reports and a small RCT from 1973–1974 using a formula with an osmolarity of 650 mOsm/l. Powdered fortifiers increase osmolality of breast milk to 360–414 mOsm/kg depending on the manufacturer. Osmolality is further increased by adding mineral and vitamin supplements [85 ]. Osmolality of standard fortified breast milk with additional hydrolyzed protein has been studied. Native breast milk (n ¼ 84) had 297 mOsm/kg, standard fortified breast milk (hydrolyzed) reached 436 mOsm/kg and adding hydrolyzed protein supplements increased osmolality by 24 mOsm/kg/0.5 g. The authors reported that additional micronutrients increased osmolality up to 868 mOsm/kg [86]. This finding is not generalizable as the osmolality increase is dependent on the quality of protein (intact or hydrolyzed) and
1363-1950 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
&
www.co-clinicalnutrition.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
281
Paediatrics
carbohydrate used (lactose, maltodextrin or glucose polymer). Osmolality data from a trial testing target fortification (n ¼ 650 samples) found that osmolality increased from 298 7 mOsmol/kg in native breast to 436 13 mOsmol/kg (maximum 477 mOsmol/kg) in target fortified breast milk [38 ]. Although breast milk fortification leads to an increase of osmolality, a recent review found a lack of evidence of feeding intolerance caused by fortification [87]. To provide further insight into this important area of gut functionality, an ultrasound method to monitor gastric volume and emptying was developed [88]. Gastric emptying is influenced by infant positioning and feeding frequency. Fortified breast milk showed slower emptying than unfortified breast milk. Higher casein concentration led to faster gastric emptying [89,90]. A study reported frequent sedimentation with liquid fortifiers, and these infants (n ¼ 82) had lower weight z scores at discharge and BUN levels on days 21 and 28 compared with infants with powdered fortifier (n ¼ 99). Protein absorption may be reduced with liquid fortifiers [91]. An in-vitro study showed that acidification caused decreased white cells, lipase activity and total protein but increased creamatocrit. The authors concluded that changes in the milk’s cellular and nutritional components may decrease absorption [92]. Lactobezoar refers to a coagulation of milk curds leading to intestinal obstruction, presented in eight infants (gestational age 25–27 weeks, birth weight
Challenges in breast milk fortification for preterm infants Niels Rochow, Erin Landau-Crangle, and Christoph Fusch
Purpose of review To outline new evidence published from 2013 to 2014 about breast milk fortification in preterm infants. Recent findings Breast milk is the feeding choice for preterm infants because of its immunoprotective properties. However, breast milk’s nutrient content is not sufficient for preterm infants, and interindividual variation is high. The variation challenges standard fortification, which assumes a standard breast milk composition. Two new fortification strategies (adjustable fortification and target fortification) optimize macronutrient intake and improve growth. Adjustable fortification uses blood urea nitrogen levels to adjust fortifier strength. Target fortification analyzes breast milk and fortifies macronutrients individually to achieve targeted intake. Its feasibility is shown in clinical routine. Current breast milk analyzers used for target fortification achieve acceptable precision for protein and fat but not for lactose and energy. Evidence of benefits for postdischarge breast milk fortification is lacking. Eliminating cow’s milk products and feeding exclusively breast milk may decrease the occurrence of feeding intolerance and necrotizing enterocolitis. To facilitate exclusively breast milk diets, a collaboration of prenatal, nutrition and lactation stakeholders is key. Fortification increases osmolality; however, safety cutoffs to prevent necrotizing enterocolitis are unclear. There is also new evidence that composition and structure of various macronutrients and micronutrients affect growth and development, and might play a role in future fortification concepts. Summary Recent research focuses on the variability of breast milk composition, its impact on postnatal growth patterns and the usefulness of target fortification. As well, diets exclusively composed of human milk are a promising approach to improve feeding tolerance. For safe fortification, osmolality cutoff levels are needed. Keywords breast milk analysis, growth, lactation, nutrition, target fortification
INTRODUCTION Progress in neonatal care has improved survival and reduced morbidity of very low birth weight (VLBW, birth weight < 1500 g) infants. However, between 43 and 97% of these infants still experience postnatal growth restriction and do not follow in-utero weight gain trajectories after birth [1,2]. This is of concern as poor growth is associated with impaired neurological development and risk of early onset of adult metabolic diseases [3]. To improve growth [4–8], a bundle of strategies such as higher parenteral and enteral nutritional intake, advancing enteral feeds earlier and faster, adjusting enteral feeding according to daily weight gain, tablet-based nutritional monitors for fluid, energy and macronutrient intake and daily plotting of individual growth trajectories on reference curves [9 ,10,11] have been used during daily &
www.co-clinicalnutrition.com
patient management rounds [5]. Since 2013, 12 reviews have summarized feeding strategies for preterm infants [1,2,12–21]. All recommend early, ample supply of protein and energy, preferably provided enterally. Breast milk is the preferred choice because of digestibility and immunoprotective effects. The high growth rates of VLBW infants increases nutritional needs above what can be provided by native breast milk [17]. Therefore, fortification is required [13,18,22 ,23]. The widely used &&
Division of Neonatology, Department of Pediatrics, McMaster University, Hamilton, Ontario, Canada Correspondence to Niels Rochow, MD, Division of Neonatology, Department of Pediatrics, McMaster University, 1280 Main Street West, HSC4F5, Hamilton, ON L8S 4K1, Canada. Tel: +1 905 521 2100x75721; fax: +1 905 521 5007; e-mail: [email protected] Curr Opin Clin Nutr Metab Care 2015, 18:276–284 DOI:10.1097/MCO.0000000000000167 Volume 18 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Challenges in breast milk fortification Rochow et al.
KEY POINTS
APPROACHES TO BREAST MILK FORTIFICATION
Target fortification using a fortifier and modular products for protein, fat and carbohydrates to adjust breast milk content to the recommended levels is feasible in clinical routine, although the effects on growth and neurodevelopment need to be studied.
To compensate for the low macronutrient levels of breast milk and provide sufficient intake for VLBW infants, three fortification approaches are currently used in neonatal intensive care units (NICUs) [22 ,23].
Exclusively breast milk diets for preterm infants seem to be able to reduce high NEC rates in some centers; however, the effect on centers with low NEC rates needs to be studied.
Standard (‘blind’ or fixed-dose) fortification
Preterm infants have immature buffer systems and are at risk for metabolic acidosis when acid load of fortifiers exceed thresholds, which may impair growth and development. The structure of specific macronutrients (e.g., types of fat) and micronutrients varies in breast milk and may affect growth and development. Fortification leads to increase of osmolality, in particular in combination with additional supplementation; however, the impact on NEC rates is not proven, and the science behind the safety cutoff is limited.
&&
Standard (‘blind’ or fixed-dose) fortification is widely established and assumes a standard composition of breast milk. Fortifiers containing protein, carbohydrates, fat, electrolytes, calcium, phosphate, vitamins and trace elements are added in fixed doses. These products are in powder or liquid form, and contain hydrolyzed or intact protein derived from cow’s milk. One commercially available fortifier product is derived from breast milk. The energy contents of fortifiers are similar whereas their composition differs. At recommended doses, they provide extra protein (1–1.1 g/dl), but the amount and composition of nonprotein calories varies considerably by manufacturer (from 0 to 1 g/dl for fat, 0.4– 3.4 g/dl for carbohydrates) [22 ]. In case of insufficient growth, it is a common practice to increase fortifier strength and/or to arbitrarily add modular protein, fat and/or carbohydrates. In a retrospective study (n ¼ 206), infants fed more than 75% fortified breast milk showed less growth on day 28 and at discharge when compared with infants fed less than 25% fortified breast milk. Standardized fortification of breast milk at current recommended doses may not adequately support early postnatal growth [32]. It could be hypothesized that the lower growth associated with fortified breast milk is due to underachieved target macronutrient levels. In another three-arm randomized trial (n ¼ 84), infants received fortification at different strengths with a protein intake of 3, 3.3 and 3.6 g/kg/day and found better head growth with higher protein intake [33]. Renal glomerular and tubular functions were unaffected by protein intake up to 3.6 g/kg/day [34]. In a multicenter randomized control trial (RCT), (n ¼ 150) using an ultraconcentrated liquid breast milk fortifier or powdered breast milk fortifier, protein intake was a significant predictor of length but not weight [35]. Addition of 0.5–2.3 g protein/kg/day to standard fortified breast milk led to appropriate growth (18 g/kg/day, head 7 mm/week, length 1.0 cm/week, n ¼ 43) [36]. The standardized fortification approach was recently applied in a study on breast milk with human donor milk-derived fortifier (n ¼ 104; 130 kcal/kg/day, 3.6 g protein/kg/day). Fortification was increased to 140 (maximum 150) kcal/kg/day, &&
concept of breast milk fortification is calculated with the assumption of a standard composition of macronutrients and micronutrients. Challenges arising from this approach include risk to develop postnatal growth restriction in infants fed nutrientdeficient breast milk; adverse effects of fortifiers including reaction to cow’s milk [allergy, inflammation and necrotizing enterocolitis (NEC)], metabolic acidosis and feeding intolerance; establishing an adequate fortifier composition (fat:carbohydrate ratio and micronutrients); optimizing on-site milk handling/fortification infrastructure.
BREAST MILK COMPOSITION A recent systematic review of 41 studies, including 843 mothers who delivered preterm and 2299 mothers who delivered at term, found highly variable levels of energy, protein, lactose, oligosaccharides, fat, phosphorus and calcium in breast milk between individuals and across postnatal and gestational age [24 ]. This variability was further confirmed for preterm, term and donor milk by seven recent studies [25,26 ,27–30,31 ]. Interestingly, macronutrient levels do not correlate to one another and cannot be predicted from each other. Prior to individualized or target fortification, all three macronutrients in breast milk need to be analyzed to guide fortification and avoid unexpected nutrient content [26 ,28]. &&
&
&
&
1363-1950 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-clinicalnutrition.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
277
Paediatrics
4.4 (maximum 5.25) g protein/kg/day when growth was less than 15 g/kg/day. Infants achieved a high weight gain of 24.8 5.4 g/kg/day from birth (gestational age 27.6 2.0 weeks) to discharge (gestational age 40.8 6.0 weeks), starting at a birth weight of 913 182 g [37]. However, there was a discrepancy reported between weight gain and the weight at discharge of 2795 g [37]. A weight gain of almost 25 g/kg/day may be inappropriate to achieve a normal body composition. Current standard fortification of breast milk does not ensure appropriate weight gain for all preterm infants. This may be because of interindividual and intraindividual variation in the macronutrient content of breast milk, including some low protein and/or fat levels [22 ,24 ,26 ,38 ]. Two approaches are proposed to account for varying breast milk composition: ‘adjustable’ and ‘target’ fortification [22 ,23]. &&
&&
&
&
&&
Adjustable fortification Adjustable fortification is a modified approach to standard fortification using blood urea nitrogen (BUN) as a surrogate marker for metabolic response to determine appropriate protein intake. The fortification strength is adjusted according to BUN values [39,40]. In an observational study (gestational age 32 weeks or less), higher protein intake of 4 g/kg/day via adjustable protein fortification (standard fortifier plus additional protein, n ¼ 29) compared with standard fortification (2.8 g protein/kg/day, n ¼ 29) led to higher weight gain, length and head circumference. Protein was increased by 0.55 g/dl breast milk when BUN was less than 9 mg/dl, decreased by 0.55 g/dl if BUN was 14–20 mg/dl. Addition of protein was withheld for 1 week if BUN was 20 mg/dl or more [39]. In a retrospective study, infants received a protein product (n ¼ 33) in addition to standard fortification when BUN was less than 5 mg/dl and/or prealbumin was 8 mg/dl or less. Supplementation was decreased gradually once BUN 9 mg/dl or more and prealbumin reached 8 mg/dl or more, and was discontinued at 40 weeks gestational age. Infants with additional protein fortification received 0.5–1.5 g/kg/day more protein than the standard fortification group (n ¼ 32). Weight gain and head growth improved, and Bayley scores at 18 months were higher in the group with extra protein [40].
Target (individualized or customized) fortification Target (individualized or customized) fortification accounts for variability of protein, carbohydrate 278
www.co-clinicalnutrition.com
and fat through regular breast milk analysis and fortification to provide the recommended intake of macronutrients [38 ,41,42] and to optimize growth. In a study comparing data from standard and individualized fortification approaches, daily breast milk composition was measured with a mid-infrared milk analyzer (n ¼ 24 infants). For their individualized approach, breast milk fat content was first adjusted to 4 g/dl by adding modular fat. Next, a fortifier was added to reach a protein intake of 4.3 g/kg/day (increased protein by 1.1 g and increased fat by 1 g). As a result, the variability of macronutrients in the individualized approach was significantly decreased, but the average fat intake was 8.6 g/kg/day [41], which exceeded recommendations [21]. In a two-center RCT (n ¼ 78), breast milk energy density was measured with a near-infrared analyzer. Infants received cream in addition to donor milkderived fortifier if energy density was less than 20 kcal/oz (intervention group n ¼ 39) and showed superior weight (14.0 2.5 vs. 12.4 3.0 g/kg/day) and length (1.0 0.3 vs. 0.8 0.4 cm/week) gain vs. the control group without cream [42]. However, in recent articles the infrared analyzers have been extensively validated and have determined that the measurement of calories is not precise because of the inability to accurately measure lactose with these devices [31 ,43]. A recent pilot study (gestational age 32 weeks or less) was the first to show the feasibility of target fortification of all macronutrients through daily breast milk analysis (near-infrared), using modular products to adjust levels of fat to 4.4 g/dl, protein to 3 g/dl and carbohydrates to 8.8 g/dl. Infants with target fortification but not standard fortification, showed growth rates with a high correlation to volume of breast milk fed. The lack of correlation between volume of breast milk fed and growth rates for the standard fortification approach can be explained by the 30–40% of infants who experience poor growth as their breast milk falls below the average composition of breast milk used to calculate fortification with the standard fortification approach, and thus are fed nutrient deficient breast milk [38 ]. To ensure appropriate fortification for each infant, standard, adjustable and target fortification utilize different approaches. Standard fortification is adjusted when infants show failure to grow. Adjustable fortification relies on monitoring metabolic response via markers such as BUN. Standard and adjustable fortification risk feeding macronutrientdeficient breast milk as they are outcome-based approaches. Target fortification analyzes breast milk &
&
&
Volume 18 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Challenges in breast milk fortification Rochow et al.
and provides individual fortification to achieve target macronutrient levels.
BREAST MILK ANALYSIS IN CLINICAL ROUTINE Macronutrient analysis of breast milk is needed for target fortification. To make analysis feasible in clinical routine, a breast milk analyzer should fulfill certain requirements: easy-to-use bedside devices, low sample volume (1 ml), low maintenance and rapid analysis. Few milk analyzers using mid-infrared or near-infrared spectroscopy are commercially available [44 ] and were originally developed for use in the dairy industry. Analyzers require recalibration as breast milk has a different matrix and optical characteristics than cow’s milk. Standard chemical methods require 30–60 ml of milk, and cannot be used to measure or validate protein, fat and carbohydrate levels in clinical settings. A set of modified micromethods (1.5 ml) have been developed using Mojonnier ether extraction (fat), elemental analysis (protein) and ultraperformance liquid chromatography-tandem mass spectrometry (lactose) [45]. To overcome the drawbacks of the first validation attempts (limited sample size [46] and inappropriate secondary conclusions [47]), two low sample volume breast milk analyzers were validated in a recent study (n ¼ 1188 breast milk samples, 63 mothers of preterm and term infants). These analyzers have the potential to be introduced in clinical routine to measure fat and protein content using correction algorithms, but require major adjustments for lactose and calories. It is speculated that the lactose measurements are confounded by oligosaccharides, which are highly abundant in breast milk but scarcer in cow’s milk. Consequently, as a precise lactose measurement is not available, calorie content cannot be calculated. Preanalytical sample handling has also been assessed: the evidence for the influence of aliquoting, storage time, temperature, container wall absorption, on stability and availability of macronutrients in frozen breast milk was reviewed [31 ]. Use of laboratory analyzers in clinical routine requires quality checks. Owing to the lack of a commercially available breast milk standard, the authors suggest creating one by pooling 500 ml of random leftover breast milk samples and dividing them into 1 ml aliquots to use as daily quality controls before measurement of patient samples [45]. A promising method using Fourier transform mid-infrared spectroscopy was validated using (n ¼ 116) breast milk samples with chemical methods (Kjeldahl, Mojonnier and high-performance liquid chromatography). Agreement between the breast milk analyzer and chemical methods was &
high for protein and fat, low for lactose and energy. These authors also speculate the interference of oligosaccharides in the accurate measurement of lactose [43]. Can milk analysis be simplified? Current commercial breast milk analyzers cost US$ 17 000– 59 000 [44 ]. Creamatocrit analysis, measuring the cream layer of breast milk, could be a low-cost alternative (for capital/one-time costs). A recent study used 51 leftover breast milk aliquots to compare creamatocrit analysis with mid-infrared spectroscopy. The analyzer was validated with reference methods (Kjeldahl, high-performance liquid chromatography and Mojonnier) and a correction equation was developed [48]. The data about energy content need to be interpreted with caution because of limitations to measure lactose with current infrared methods [31 ,43]. Even if fat content could be determined with creamatocrit analysis, other components cannot be predicted from a measured fat level. It has been shown that macronutrient levels do not correlate but need to be measured separately [26 ]. Therefore, the creamatocrit analyzer currently cannot be recommended as a low-cost alternative. With respect to work hours, daily breast milk analysis for target fortification would increase NICU workload by 10–15 min/patient/day and may not be feasible in all nurseries. An observational study (n ¼ 210 pooled native 24 h breast milk samples from 10 infants) varied the number of measurement days and analyzed the variation of macronutrient intake. Day-to-day variation of macronutrients (protein 20%, carbohydrate 13%, fat 17%, and energy 10%) decreased with increased frequency of analysis. Two measurements per week led to a mean macronutrient intake within a range 5% of target levels. The extent of the effects of day-to-day variation of nutritional intake on growth are unclear [49]. In summary, infrared analysis seems to be a promising tool for fat and protein with calibration, but lactose and energy cannot be assessed with the current state of technologies. &
&
&
&
BREAST MILK MANAGEMENT CENTER To make breast milk available in neonatal intensive care requires an adapted infrastructure. Key prerequisites are collaboration between healthcare providers and nutrition experts, counseling mothers on nutritional needs of preterm infants, donor milk availability, breast milk analysis and handling and storage of breast milk. Recent articles described challenges and suggestions to improve breast milk availability [44 ,50,51,52 ,53]. One study reported a ‘mobile’ milk cart for preparation/fortification of breast milk for NICUs with limited space [50].
1363-1950 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
&
&
www.co-clinicalnutrition.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
279
Paediatrics
Another study presented a centralized milk management center for breast milk measurement, target fortification and safe preparation [44 ]. Measuring breast milk composition in lab and target fortifying at bedside has also been described [38 ]. A ‘Supporting Premature Infant Nutrition’ program includes the collaboration of the hospital nutrition and lactation stakeholders and 10 key points [52 ]. Another challenge is to provide fortification at home. Postdischarge preterm infants could benefit from breast milk fortification; however, two recent reviews demonstrated limited evidence of the impact of postdischarge breast milk fortification on growth and neurodevelopment [54,55]. &
&
&
EXCLUSIVELY FEEDING BREAST MILK Seven recent reviews suggest that breast milk decreases the risk of NEC and late-onset sepsis, and should be the primary enteral diet for premature infants. Donor milk is a resource for premature infants whose mothers are unable to provide adequate milk supply. Further, donor milk-derived fortifier can be used to feed infants exclusively breast milk. However, donor milk and derivatives require pasteurization, which decreases the concentrations of most cytokines and lactoferrin, and is limited in supply. Transforming growth factor-b2 is a protective intestinal cytokine and surprisingly showed comparable concentrations in breast milk and donor milk-derived fortifier, despite pasteurization [56]. In general, donor milk differs nutritionally from breast milk, and neonatologists should be aware that these differences might affect the need for nutrient fortification [57–63]. In the last 2 years, further evidence favoring an exclusively breast milk diet has shown that intake of cow’s milk protein may lead to a proinflammatory reaction on the neonatal gut. Cytokine secretion of interferon-g, interleukin-4, interleukin-10 and transforming growth factor-b1 was observed in response to b-lactoglobulin and casein [64]. The clinical impact of feeding exclusively breast milk on NEC rates has been shown in four recent clinical studies. A multicenter study demonstrated a significantly different NEC rate for cow’s milk protein-based fortifiers vs. breast milk protein-based fortifiers with 17 and 5%, respectively, (n ¼ 93 vs. n ¼ 167) [65]. Another multicenter RCT (n ¼ 53) reported similar findings with a significantly lower NEC rate: 3% (exclusively breast milk) vs. 21% (preterm formula) [66]. A retrospective before–after study (n ¼ 642) compared a 2.5-year period of exclusively breast milk diets with the preceding 6.5 years when cow’s milk products were given and found a decrease in NEC rate from 3.4% (15/443) to 1% 280
www.co-clinicalnutrition.com
(1/199) [67]. Another observational study (n ¼ 104) analyzed data from infants who received a breast milk-derived fortifier and found only three medical and one surgical NEC case [37]. Data from these studies show a promising trend of decreased NEC rates with an exclusively breast milk diet. However, the reported NEC rate in infants receiving cow’s milk-derived products was 17–21%, unexpectedly high compared with average NEC rates reported by international networks of 1.6–6% [68]. A recent study reviewed enteral feeding type (breast milk, formula and nutrient fortification) for effects on microbiota. Altered colonization of the microbiota has been associated with increased morbidity in preterm infants. There is a lack of evidence demonstrating the impact of breast milk fortification on gut microbiota [69]. An observational study (n ¼ 27) investigated the addition of prebiotic oligosaccharides to stimulate commensal bacterial growth and to bind pathogens within the intestinal lumen, as a safer method of NEC prevention. Formula-fed infants (n ¼ 12) received galactooligosaccharide or concentrated donor milk with human oligosaccharides. In (n ¼ 15) infants, breast milk was fortified with a donor milk-derived fortifier or cow’s milk-derived fortifier. None of the prebiotic interventions resulted in significant change of microbiota. Limitations of the study were lower-than-proposed oligosaccharide dose for treatment group and antibiotic use [70]. Feeding exclusively breast milk appears to decrease feeding intolerance and NEC rates. Future clinical studies are required in which the initial NEC rate for cow’s milk-based fortifiers are closer to the international average. The microbiota or addition of specific oligosaccharides may be a further target to improve feeding tolerance.
SPECIFIC MACRONUTRIENTS AND MICRONUTRIENTS Breast milk macronutrient and micronutrient content varies in composition within and between individuals and over time. Additionally, there is variation of composition amongst diverse commercial fortifiers in regard to quality of macronutrients and micronutrients. Some fortifiers contain v-3 fatty acids or iron whereas others do not. In order to provide infants with the recommended nutrition, the composition of fortifiers (not only breast milk) must be considered. In this paragraph, we will list all the new studies that have either tested additional components in fortifiers or that provide evidence that a component has the potential to be a useful supplement in fortifiers or needs to be tested. Volume 18 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Challenges in breast milk fortification Rochow et al.
A recent review (seven studies, n ¼ 1914) analyzed neurodevelopment at less than 3 years (three studies) and at 5–11 years (four studies) and speculated that breast milk contains neurotrophic factors that benefit brain development. The authors concluded that increasing use of donor milk, new preterm formulas and formulation of new fortifiers requires reassessment of the effects on neurodevelopment [71]. One potential candidate for a neurotrophic factor could be v-3 fatty acids that are needed for visual and cognitive development in early life. Two reviews showed that v-3 content of breast milk is highly variable following maternal diet. Nutritional counseling during lactation was suggested [72,73]. In a multicenter RCT (14 sites, n ¼ 150), infants received powdered fortifier or liquid fortifier with additional docosahexaenoic acid (DHA) 12 mg/dl and arachidonic acid (ARA) 20 mg/dl. Infants on liquid fortifier had higher phospholipid DHA/ARA levels at day 28 and were related to maternal DHA/ARA intake [74 ]. Sphingomyelin plays a role in brain cell membrane structure. Feeding sphingomyelin to rats led to increased levels of DHA in red blood cell membranes and accelerated myelination of cortical areas. Infants in an RCT (n ¼ 24) received either 13 or 20% sphingomyelin-fortified milk. At 18 months, infants with 20% sphingomyelin showed a trend toward better neurobehavioral development [75]. Increased phenylalanine levels may have unfavorable effects on brain development. A recent study (n ¼ 24) found an increase of plasma phenylalanine levels from 12 1 to 30 1 mmol/l in infants fed breast milk fortified with cow’s milk protein. This was not found in infants fed exclusively breast milk. The long-term effects of this increase are unclear [76]. Vitamin D, calcium and phosphorus are essential for bone development in preterm infants. A recent clinical report suggested a Vitamin D intake of 200–400 IU/day [77]. For phosphate, an intake of 184–230 mg/kg/day is recommended. A retrospective study (n ¼ 106) showed that under standard fortification, additional phosphate supplementation was necessary to achieve phosphate levels between 1.8 and 2.6 mmol/l [78], this was confirmed by a prospective study (n ¼ 79) [79]. The importance of maintaining adequate phosphate levels was shown in an RCT (n ¼ 50) in which a high sepsis rate was related to low phosphate levels and electrolyte imbalances [8]. Folate is essential for rapidly growing preterm infants. Heat treatment of donor milk decreases folate by 14–25%. Nonfortified, heated breast milk supplied only 25% of the recommended daily intake to preterm infants [80]. Another study (n ¼ 162) &
showed that providing parenteral nutrition and fortified breast milk with sufficient folate content reduces the risk for folate deficiency during the first 2 postnatal months [81]. Zinc is needed to achieve sufficient growth rates. In a retrospective study (n ¼ 52), supplementation of 1.7 mg/kg/day was associated with an increase in weight gain from 11 to 20 g/kg/day, and linear growth from 0.7 to 1.1 cm/week [82]. Elemental iron supplementation of 2–4 mg/kg/ day for breast milk-fed preterm infants during the first 12 postnatal months is recommended. An observational study (n ¼ 152) measured ferritin levels at term, 3 months and 6 months and showed a higher occurrence of low ferritin less than 12 mg/l in breast milk-fed infants until 6 months. Iron-fortified formula with 0.8–1.0 mg iron/dl led to normal ferritin levels [83]. Carnitine plays a role in energy metabolism via long-chain fatty acids. A recent observational study (n ¼ 38) showed higher carnitine levels of formulafed compared with breast milk-fed infants during the first 28 days of life [84]. These studies provide evidence that micronutrients’ composition and the structure of proteins and fat may affect preterm infants’ metabolism. Depending on fortifier composition, micronutrients or other specific nutrients might need to be added.
(POTENTIAL) SIDE-EFFECTS OF FORTIFICATION It is obvious that fortification of breast milk has improved growth rates and neonatal care. However, like with any other intervention, it has some sideeffects, though the aim is to minimize these. The American Academy of Pediatrics has recommended that osmolality of enteral feeds should not exceed 450 mOsm/kg (400 mOsm/l). This recommendation is based on a series of case reports and a small RCT from 1973–1974 using a formula with an osmolarity of 650 mOsm/l. Powdered fortifiers increase osmolality of breast milk to 360–414 mOsm/kg depending on the manufacturer. Osmolality is further increased by adding mineral and vitamin supplements [85 ]. Osmolality of standard fortified breast milk with additional hydrolyzed protein has been studied. Native breast milk (n ¼ 84) had 297 mOsm/kg, standard fortified breast milk (hydrolyzed) reached 436 mOsm/kg and adding hydrolyzed protein supplements increased osmolality by 24 mOsm/kg/0.5 g. The authors reported that additional micronutrients increased osmolality up to 868 mOsm/kg [86]. This finding is not generalizable as the osmolality increase is dependent on the quality of protein (intact or hydrolyzed) and
1363-1950 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
&
www.co-clinicalnutrition.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
281
Paediatrics
carbohydrate used (lactose, maltodextrin or glucose polymer). Osmolality data from a trial testing target fortification (n ¼ 650 samples) found that osmolality increased from 298 7 mOsmol/kg in native breast to 436 13 mOsmol/kg (maximum 477 mOsmol/kg) in target fortified breast milk [38 ]. Although breast milk fortification leads to an increase of osmolality, a recent review found a lack of evidence of feeding intolerance caused by fortification [87]. To provide further insight into this important area of gut functionality, an ultrasound method to monitor gastric volume and emptying was developed [88]. Gastric emptying is influenced by infant positioning and feeding frequency. Fortified breast milk showed slower emptying than unfortified breast milk. Higher casein concentration led to faster gastric emptying [89,90]. A study reported frequent sedimentation with liquid fortifiers, and these infants (n ¼ 82) had lower weight z scores at discharge and BUN levels on days 21 and 28 compared with infants with powdered fortifier (n ¼ 99). Protein absorption may be reduced with liquid fortifiers [91]. An in-vitro study showed that acidification caused decreased white cells, lipase activity and total protein but increased creamatocrit. The authors concluded that changes in the milk’s cellular and nutritional components may decrease absorption [92]. Lactobezoar refers to a coagulation of milk curds leading to intestinal obstruction, presented in eight infants (gestational age 25–27 weeks, birth weight
