|Home | About | Journals | Submit | Contact Us | Français|
Dietary DHA (22:6n-3) is a long-chain PUFA that has provocative effects on inflammatory signal events that could potentially affect preterm infant health. It is well known that the essential fatty acid of the (n-3) series; α-linolenic acid (18:3n:3) can be desaturated and elongated in the liver endoplasmic reticulum and peroxisome to produce the 22-carbon DHA. Nevertheless, concern exists as to the efficiency of this mechanism in providing the preterm infant with adequate DHA. Activity of the δ-6-desaturase and the δ-5-desaturase necessary for DHA synthesis is decreased by protein deprivation. The combined effects of suboptimal intake of both DHA and protein in the preterm infants could have substantial clinical consequences.
Perinatal investigations of DHA have focused on neurodevelopmental outcomes or pregnancy duration. However, fatty acids have a much broader role in perinatal health. They are an integral component of membrane phospholipids, which impart changes to cell membrane fluidity, prostaglandin synthesis, cytokine expression, and bioactive molecules such as N-acyl ethanolamines that act as messengers in promoting anti-inflammatory pathways. Preterm infant diseases such as bronchopulmonary dysplasia (BPD),4 necrotizing enterocolitis (NEC), and retinopathy of prematurity (ROP) create significant morbidity and mortality and are largely due to an unchecked inflammatory pathophysiology that may improve with dietary DHA. The clinical challenges in ensuring optimal intake for the preterm infant are determining an appropriate target dose of DHA and whether to supplement the infant directly or via the maternal diet. Based on the most logical approach, the target intake of DHA for the preterm infant should be the intrauterine accretion value of the last trimester of pregnancy at 52 mg/d (1). A maternal dietary intake of DHA at 1 g/d therefore to meet this requirement is suggested for lactating mothers providing breast milk to preterm infants (2).
Preterm infants miss the large portion of DHA (1, 3) nutrient accretion that occurs during the last trimester of pregnancy, making it essential that they receive this nutrient in their dietary management in the neonatal intensive care unit (NICU).
All mammals require the fatty acids linoleic acid (C18:2n-6) and α-linolenic acid (C18:3n-3) in the diet because they are incapable of inserting the cis double bond at the (n-6) and (n-3) positions from the carboxyl terminus (according to International Union of Pure and Applied Chemistry terminology) (4). The liver is the predominant organ that receives the triglyceride from fat in the diet and, using a series of lipases, extracts the nonesterified fatty acid (5) and in the endoplasmic reticulum will desaturate and elongate the acyl chain of the parent fatty acid to its longer chain and more unsaturated compounds: arachidonic acid (ARA) (C20:4n-6) and DHA (C22:6n-3) (6) (Fig. 1). These long-chain PUFA (LCPUFA) not only produce different prostaglandins but DHA metabolites become resolvins, which can promote the resolution of inflammation (7). DHA is also the predominant fatty acid in the cerebral cortex, retinal rods, and cones (8) that significantly influence their function (9). Historically, our dietary intake of (n-6) and (n-3) fatty acids provided a homeostasis between the 2, but as dietary habits had changed during the 20th century (10), there has been a dramatic increase in (n-6) fatty acid consumption. In addition, the biosynthesis dependent on the desaturase and elongase enzyme activities may not be able to produce enough DHA because of the competition that the fatty acids have for the same enzymes (11), thus down-regulating the (n-3) pathway and altering the balance between pro- and anti-inflammatory events (12). Activity of the specific desaturases occurs in humans even as early as 26 wk of gestation (13), but preterm infants have been shown to convert variable percentages of the α-linolenic acid to DHA (14–16).
Maternal dietary sources for linoleic acid and α-linolenic acid are found primarily in plant oils such as corn, olive, poppy seed, palm, soybean, rapeseed, safflower, sunflower, and wheat germ (17). Dietary sources of preformed ARA are found in animal products and those of DHA are found in oily fish sources such as salmon, mackerel, tuna, and herring (18). Eggs that are the product of (n-3) fed hens are also good sources of DHA (19). The advisable intakes during lactation are 13 g/d of linoleic acid and 1.3 g/d of ALA (18). Recommended intakes during infancy are 4.4 g/d linoleic acid and 0.5 g/d α-linolenic acid to prevent essential fatty acid deficiency (18). Recommendations have not been officially made for DHA, but based on randomized, controlled trials, it is suggested that pregnant and lactating women receive a minimum of 200 mg/d of DHA in the diet (20). For the preterm infant, recommended enteral intake of DHA is 12–30 mg/kg/d according to the Nutrition Committee of the European Society for Pediatric Gastroenterology, and Hepatology (21).
Human milk contains 4 g fat/L; thus, the exclusively breast-fed term infant consumes 750 mL of mother’s milk and receives ample essential fatty acids (22). Commercial formulas are designed to provide a minimum of a 5:1 ratio of linoleic acid to α-linolenic acid to be comparable to human milk samples. Infants fed formulas with a lower 4:1 ratio had significantly decreased fatty acid erythrocyte profile compared with a reference breast-fed group (23). Dietary fatty acid intake is reflected in plasma and red blood cell profiles (24). Red blood cell phospholipids significantly increase as dietary DHA intake increases (24, 25). Biochemical documentation of essential fatty acid deficiency is determined by the ratio of the nonessential, eicosatrienoic (triene) fatty acid or Mead acid of the (n-9) family compared with the ARA of the (n-6) family (tetraene). A triene:tetraene ratio >0.4 is considered diagnostic of essential fatty acid deficiency (26). Preterm infants exhibit evidence of fatty acid deficiency by 5 d of life when they are not fed an essential fatty acid source (27). Clinical signs of deficiency of the essential fatty acids include scaly dermatitis and platelet dysfunction (28). Signs of LCPUFA deficiency symptoms have not been established, but preterm infants have demonstrated slow growth when their ratio of n-6:n-3 fatty acids was altered in earlier fish oil supplement studies (29).
Preterm infants, unlike term infants, are not ingesting 750 mL/d of human milk or formula for months. Rather, they are often on small amounts of enteral milk (20 mL/kg) for days and are supplemented with intravenous nutrition as feeding is advanced (30). Current intravenous sources of lipid emulsions in the United States provide adequate essential fatty acids but only trace amounts of ARA and DHA (Intralipid). Compassionate therapy with an intravenous emulsion that contains fish oil (as a source of DHA) has been prescribed in the NICU but is not available in the United States, so it has primarily been reserved for the infant with significant hepatic cholestasis (31). Reliance on enteral sources of DHA is therefore needed. The biosynthesis of LCPUFA from the dietary ingestion of the precursor fatty acids can be of special concern, however, for the preterm infant, unlike the term infant, because of additional confounders. For instance, Mayes et al. (32) determined that preterm infants have enzymes available for conversion, but the total quantity of DHA produced may be small. Additionally, the δ-6-desaturase has been documented to be rate limiting in the rat model if protein is marginal (33) and under conditions of glucocorticoid and mineralocorticoid steroid inhibition (34, 35). Preterm infants are known to have a marginal dietary intake of protein throughout hospitalization (36) and have steroid exposure from maternal supplementation before delivery (37) As clinicians, we need to re-examine and question the assumption that desaturases will be fully able to provide the mechanism for preterm infants to receive adequate DHA. Figure 1 illustrates the effects that insufficient dietary protein and steroids may have on DHA metabolism for the preterm infant. The significance of inadequate DHA intake may affect the ability of the phospholipid membrane to activate pathways that would attenuate inflammation-related events.
Premature birth accounts for 75% of perinatal death (38). Inflammatory conditions such as BPD, NEC, and ROP in the preterm infant are a tremendous source of morbidity and mortality. The inflammatory response mediated through both the innate immune system (granulocytes, monocytes, macrophages, dendritic cells, and natural killer cells) and adaptive immune systems (antibodies) are a necessary protection against pathogens, viruses, dying cells, and tumors (39). Without robust proinflammatory expression, patients are at risk of death from multiorgan system failure (40). TNF-α and IL-1α effectively trigger natural killer cells to attack locally. They also mediate dendritic cell maturation and subsequent capacity to prime T helper cells in lymphoid tissue (41). The complex orchestration of events balances inflammation and then the attenuation of the inflammatory response (42). The fetus and neonate typically have a down-regulated Th1 function (43) and an inefficient monocyte response to protect the maternal-fetal dyad from autoimmunity. Circulating monocytes appear in the fetal blood by 18–20 wk of gestation; however, interferon-γ production and response are reduced (44). Pattern recognition receptors via the Toll-like receptors (TLR) recognize a wide range of pathogens and mediate transmembrane and downstream signal transduction for the innate immune system (45). For the preterm infant, homeostasis in inflammation is often elusive, resulting in an increased risk of sepsis (46) and, paradoxically, pathological damage to vital tissues. For instance, in the baboon model of BPD, the T cells are autoreactive (47). Postnatal brain and lung injury is significantly increased when proinflammatory cytokines such as IL-6, IL-1β, and TNF-α are present in the amniotic fluid (48). In addition, expression in preterm infants of the concentrations of the cytokines IL-6, IL-10, and TNF-α is significantly increased during disseminated coagulopathy (49). After 48 h, the IL-10:TNF-α ratio decreases significantly whereas the IL-6:IL-10 ratio does not (49). In preterm infants with chronic lung disease, the proinflammatory cytokines IL-1, IL-6, IL-8, and TNF-α predominate in their plasma and bronchoalveolar lavage (50). In very low birth weight infants, a significant risk of BPD is evident with higher concentrations of IL-8 expression (51). TLR-4 also plays a crucial role in the pathogenesis of NEC in the murine model (52, 53). Supplementing animal models with DHA changes these specific inflammatory markers and outcomes.
Dietary DHA attenuates nuclear factor-κB activity (54) and resultant inflammatory mediators such as IL-6, IL-8, and TNF-α (55). The attenuation of TLR-4 by PUFA reduces the activation of the nuclear factor-κB inhibitory protein, which therefore decreases the inflammatory cascade of IL-6, IL-8, TNF-α and the adhesion molecules (56). In the preterm baboon model, lung surfactant was found to be higher in concentration in mothers who had received an LCPUFA supplement (57). In our murine model, macrophage density and neutrophil concentration markers of inflammation in the alveoli were significantly lower in pups whose mothers had received DHA (51, 58). In an experimental design to investigate NEC in the rat model, PUFA supplementation inhibited intestinal platelet-activating factor receptor and TLR-4 expression (59) and reduced the incidence of NEC and death in the neonatal rat model (60). DHA also produces a potent anti-inflammatory N-acyl ethanolamine that has been described to reduce adipocyte expression of IL-6 and monocyte chemotactic protein 1, thus improving metabolic health (61). The developing brain may be protected from oxidant stress by the hydroxyl radical scavenging activity shown to be enhanced in rat fetuses supplemented with DHA (62). Based on this evidence, dietary DHA has a fundamental part of cellular biology and in animal models of preterm disease; thus, providing adequate DHA in the preterm diet could translate to reducing disease for the preterm infant, as illustrated in Figure 2.
Providing an enteral source of DHA may attenuate the inflammatory condition of the preterm infant. Most U.S. NICU have a policy of providing human milk to preterm infants because human milk–fed preterm infants have a decreased risk of BPD (63), NEC (64–66), and ROP (67). Pasteurized donor milk is often used to augment mother’s milk in the NICU if her milk volume is not sufficient. A meta-analysis (68) and a multicenter trial (66) have both demonstrated that preterm infants receiving pasteurized donor milk have a decreased likelihood of NEC. Human milk is therefore life-saving, but nutrients vary in composition from mother to mother (69) and over the course of lactation. Lactational stage is widely variable among milk banks ranging from early milk obtained from bereaved mothers to milk obtained at 12 mo postpartum (70). Milk obtained at later lactational stages can be very low in DHA compared with the nutritional needs of the preterm infant (71). Maternal dietary intake of fatty acids and the mother’s body stores influence her milk composition (19). The mammary gland can synthesize fatty acids (72), but the LCPUFA are primarily dependent on maternal stores and diet (15). Approximately 30% of the LCPUFA in human milk are from the mother’s diet, and depending on the population, milk concentrations of DHA range from 0.1 to 2 mol % (73). Geographic differences exist in the maternal diet, and thus the DHA concentrations in milk vary (74) and are decreasing in populations previously described as having higher concentrations of DHA in breast milk (10, 15). In a study examining the nutritional contents of pasteurized donor milk, we found that despite the donor women eating a diet aligned with U.S. Dietary Guidelines, their milk DHA concentration was 0.1 mol weight %, which translated to a mere 13 mg/kg/d of DHA for the preterm infant (71), which is only 25% of the daily DHA accretion during the last trimester of pregnancy (1).
The term infant may be less affected by low breast milk DHA concentrations by consuming a larger volume of milk (750 mL/d) compared with the preterm infant (150 mL/d). However, formula-fed term infants may be vulnerable to consuming decreased dietary DHA (75, 76). More specifically, Birch et al. (77) group demonstrated that decreased LCPUFA intake was significantly associated with lower visual acuity scores at 17, 26, and 52 wk of age among term infants weaned from breast milk to unsupplemented formula versus DHA-supplemented formulas. Hoffman et al. (78) demonstrated increased red blood cell DHA and visual-evoked acuity and stereoacuity in term infants breast-fed for 6 mo and then randomized to supplemented or unsupplemented formula. More recently, Jensen et al. (79) reported significantly improved sustained attention scales at 5 y of age in breast-fed children whose mothers had modest DHA supplementation (200 mg/d) for the first 4 mo of breastfeeding compared with a placebo control group. Cardiovascular health may also be improved with DHA supplementation. Children who were previously breast-fed or given supplemented DHA formula have significantly lower diastolic blood pressure at age 6 y compared with controls (80). These findings suggest that modest maternal and/or infant intake of DHA has sustained effects into childhood. This type of nutritional programming demonstrates the need to continue long-term examinations of health effects by early nutrition interventions (81).
In preterm infants, the effects of DHA on developmental indices is even more striking (82), particularly in our most immature infants (83). Henriksen et al. (84) directly supplemented preterm infants by adding 32 mg/d of DHA and 31 mg/d of ARA to their human milk feedings in the NICU by a sonification process and found that the supplemented group had higher problem-solving scores at 6 mo. More recently, increased whole-blood concentrations of DHA in preterm infants receiving standard of care were retrospectively found to correlate with decreased BPD outcome in the NICU (85).
Based on the clinical evidence, it has been suggested that milk sources for the premature infant contain 1–1.5% fatty acid as DHA (2). Preterm formulas contain both DHA and ARA but not always at the suggested concentrations. Exogenous commercially available powdered human milk fortifiers contain the DHA precursor lipid-linolenic (18:3, n-3) and ARA precursor linoleic (18:2, n-6) acids, but they do not contain a source of DHA or ARA. A new sterile liquid commercial fortifier is now available with these essential fatty acids as well as DHA and ARA (Mead Johnson Nutritionals - Enfamil Human Milk Fortifier Acidified Liquid) that can be added to the human milk feeding. The practice of direct supplementation of the neonate’s milk with an exogenous supplement is compelling but can be difficult in the NICU. A more practical approach is the addition of >3 dietary sources of high DHA fish, egg source, or direct supplementation of the lactating mother to produce an adequate balance of fatty acids, while ensuring the mother’s optimal health as well. Next, the optimal dose to give a lactating mother needs to be considered.
Makrides et al. (86) demonstrated that the DHA concentration in human milk increases in a dose-dependent fashion when measuring DHA milk concentrations in a placebo-supplemented group (0.21 mol %) to 0.86 mol % in a 0.9-g supplemented group and 1.13 mol % in a 1.3-g DHA–supplemented group, respectively (P < 0.05). Milk DHA concentrations of 0.9 mol % would be an adequate concentration to achieve intrauterine goals. In addition, epidemiologic studies have correlated an adult average intake of both eicosapentaenoic acid and DHA at 1.8 ± 1.2 g/d with a significant decrease in cardiovascular mortality (87) and inflammatory diseases (35). Moderate fish intake (up to 3 meals per week) has also been associated with a reduction in a repeat preterm delivery for women at risk (88). A systematic review and meta-analysis concluded that (n-3) fatty acid dietary intake in pregnancy increases gestational age and weight in the offspring (89). This provides further evidence that maternal practice of supplementation should be the paradigm for providing adequate DHA for the infant-maternal dyad. It appears that a maternal dietary intake of 1 g/d DHA meets the preterm infant’s dietary needs and improves maternal health as well.
Finally, current clinical evidence suggests that DHA whole-blood status in the preterm infant is related to BPD and late-onset sepsis (85), thus making it imperative that NICU staff examine the perinatal diet and recommend optimal dietary DHA sources for mothers and high-risk neonates. Future prospective randomized, controlled trials should evaluate the maternal dietary DHA dose of 1 g/d and relationships to maternal/infant health.
The author thanks the mothers who donated human milk; the Mother’s Milk Bank of Ohio; Georgia Morrow, RN, IBCLC; Lynette K. Rogers, PhD; and Ardythe Morrow, PhD, as mentors during these investigations; Donna Wuest for her expert technical assistance; and Laurie Nommsen-Rivers, PhD, RD, for her excellent review of this manuscript. The sole author had responsibility for all parts of the manuscript.
1Published as a supplement to Advances in Nutrition. Presented as part of the symposium entitled “Impact of Maternal Status on Breast Milk Quality and Infant Outcomes: An Update on Key Nutrients,” given at the Experimental Biology 2011 meeting, April 12, 2011, in Washington, DC. The symposium was sponsored by the American Society for Nutrition and supported by an unrestricted educational grant from Medela. The symposium was chaired by Laurie Nommsen-Rivers and Donna J. Chapman. Guest Editors for this symposium publication were Donna J. Chapman and Shelley McGuire. Guest Editor disclosure: Donna J. Chapman received travel support and compensation for editorial services provided for this symposium from the International Society for Research on Human Milk and Lactation. Shelley McGuire had no conflicts to disclose.
2Supported by The Research Institute at Nationwide Children’s Hospital.
3 Author disclosure: C. J. Valentine, no conflicts of interest.
4Abbreviations used: ARA, arachidonic acid; BPD, bronchopulmonary dysplasia; LCPUFA, long-chain PUFA; NEC, necrotizing enterocolitis; NICU, neonatal intensive care unit; ROP, retinopathy of prematurity; TLR, Toll-like receptor.