886
Essentiality of Dietary Fatty Acids for Premature Infants: Plasma and Red Blood Cell Fatty Acid Composition 1 Dennis R. Hoffman and Ricardo Uauy* The University of Texas Southwestern Medicat Center, Department of Pediatrics and The Human Nutrition Center, Dallas, Texas 75235-9063
Pre~term infants, that are not breast-fed, are deprived of vital intrauterine fat accretion during late pregnancy and must rely on formula to obtain fatty acids essential for normal development, particularly of the visual system. Preterm infants (30 wk postconception) receiving human milk were compared to infants given one of the following formulae: Formula A was a commercial preterm formula with predominantly 18:2c06 (24.2%) and low (0.5%) 18:3~3; Formula B was based on soy oil and contained similar 18:2w6 levels {20%) and high 18:3w3 {2.7%); Formula C was also a soy oil-based formula {20% 18:2, 1.4% 18:3) but was supplemented with marine oil to provide w3 long-chain polyunsaturated fatty acids {LCP) at a level {docosahexaenoie acid, DHA, 0.35%) equivalent to human milk. At entry (I0 days of age), the fatty acid composition of plasma and red blood cell (RBC) membrane lipids of the formula groups were identical. By 36 wk postconception, the DHA content in lipids of group A was significantly reduced compared to that in the human milk and marine oil formula groups. Omegas3 LCP results were further amplified by 57 wk with compensatory increases in 22:5c~6 in both plasma and RBC lipids. Provision of 2.7% a-linolenic acid in formula group B was sufficient to maintain 22:6w3 levels equivalent to those in human milk-fed infants at 36 wk but not at 57 wk. Effects on the production of thiobarbituric acid reactive substances and fragility of RBC attributable to the marine oil supplementation were negligible. The results support the essentiality of co3 fatty acids for preterm infants to obtain fatty acid profiles comparable to infants receiving human milk. Formuia for preterm infants should be supplemented with c03 fatty acids including LCP. Lipids 27, 886-895 (1992}.
Since animals lack the desaturating enzymes (1,2) necessary to form the co6 and co3 series of fatty acids (FA), they must rely on plant sources to obtain the parent essential fatty acids (EFA). Animal tissues are capable of further elongating and desaturating the parent EFA, linoleic acid (18:2co6) and a-linolenic acid (18:3co3), generating a family of long-chain polyunsaturated fatty acids (LCP) of each respective series. Alternatively, dietary LCP may be provided to infants directly from sources such as breast milk, marine oils or egg phospholipids.
The functional and structural roles of specific long-chain f a t t y acids are being increasingly recognized. Synthesis of eicosapentaenoic acid (EPA, 20:5co3) from a-linolenic acid in animals provides a precursor for series-3 prostaglandins which may suppress some effects of the series-2 prostaglandins originating from co6 FA by limiting their production. Although the actions of EPA may be beneficial in pathological conditions, concern persists as to the safety of limiting the action of arachidonate-derived eicosanoids in infants. Docosahexaenoic acid (DHA, 22:6co3} may provide a structural micro-environment in the membrane that influences important membrane functions such as ion or solute transport, receptor activity or enzyme action (3,4}. The enrichment of DHA in retina and neural tissues (5,6} may reflect a unique and critical role for this co3 LCP in visual function. Although the essentiality of co6 FA in the diet has long been recognized, only recently have studies focused on the more subtle biochemical and functional effects of dietary ~3 FA deficiency. The highly unsaturated nature of DHA may be responsible for the observed effects of co3 FA deficiency on the function of developing brain and retina (3,7}. Wheeler e t al. (8} demonstrated dramatic enhancement of both a- and b-waves of electroretinograms from rats fed dietary supplements of co3 FA compared to fat-free diets or supplementation with only co6 FA. Neuringer e t al. (9) reported visual acuity impairments of up to 50% in newborn rhesus monkeys receiving safflowe~based diets deficient in r FA pre- and postnataUy compared to controls given a soybean oilbased diet. These primates have depressed levels of DHA in plasma phospholipids at birth and nearly undetectable amounts at 12 wk of age (10). These visual and neural derangements may also occur in human infants; however, evidence of FA status in the brain presently must rely on the direct correlations of brain and red blood cell (RBC) lipid FA distribution found in rat and primate studies (9,11).
During fetal growth in the human, accretion of C18EFA is not linear but is approximately 400 mg of co6 and 50 mg of co3 EFA per kg of body weight deposited daily during the last trimester of gestation (12) and continues postnataUy in brain tissue (13). However, a preterm infant born at less than 34 wk will be deprived of this critical period of intrauterine lipid accretion and thus, is dependent on diet to fulfill its EFA requirements. Although formula supplementation with co6 and co3 EFA (i.e., 18:2co6 and 18:3c03) may be beneficial (14}, evidence for a limited 1Based on a paper presented at the Symposium on Milk Lipids held FA desaturase activity (t5) indicates that LCP such as at the A O C S Annual Meeting, Baltimore, M D , April 1990. arachidonic acid (16) and DHA (17,18) may also be re*To w h o m correspondence should be addressed at the Retina Founquired by the low-birth-weight infant. dation of the Southwest, 9900 N. Central Expressway, Suite 400, Preterm infants fed their own mother's milk receive a Dallas, T X 75231-3303. typical intake of 140 mg co3 FA/kg/d including 45 mg co3 Abbreviations: A N O V A , analysis of variance, D H A , docosahexa- LCP/kg/d (19}, thus meeting the requirements for neural enoic acid (22:6co3);EFA, essentialfatty acid(s);EPA, eicosapentaenoic acid (20:5co3);FA, fatty acid(s);H M , human milk; LCP, long- tissue and whole body EFA accretion. Preterm infants chain polyunsaturated fatty acid(s);MCT, medium-chain trig|ycer- given commercial formula receive less a-linolenic acid and ides; RBC, red blood cell(s);TBARS, thiobarbituric acid reactive elevated linoleic acid compared to breast-fed infants. This substances. relative excess of 18:2co6 may limit the formation of co3 LIPIDS, Vol. 27, no. 11 (1992)
887
co3 FATTY ACIDS IN PRETERM FORMULA LCP in ceil membranes (20-22) and blood lipids (11,17,18) because the A6 desaturase shared by b o t h series of FA is subject to s u b s t r a t e inhibition (1). Typical p r e m a t u r e infant formulas available up to 1987 (when the s t u d y commenced) had a low content of a-linolenic acid and no r LCP. Thus, we considered the p r e t e r m infant at risk for r LCP deficiency because biosynthesis is limited and formula diet supply was insufficient to meet the needs of the developing visual and neural s y s t e m s (23). The purpose of the present s t u d y was to evaluate the potential effects of oJ3 FA deficiency in very-low-birthweight h u m a n infants by s t u d y i n g functional and biochemical indicators of r FA status. The FA composition of p l a s m a and RBC lipids of h u m a n milk-fed infants and three formula-fed groups receiving different EFA and LCP supplementation are reported herein. A secondary objective included the evaluation of RBC m e m b r a n e peroxidation and resistance to hemolysis in the four diet groups. Interim visual function results for the infants under study have been recently published (18). MATERIALS AND METHODS
Subjects. Ninety-five newborns with body weights of 1000-1500 g, who received enteral feedings b u t were free of major neonatal m o r b i d i t y by day 10 of life, were eligible for the study. Twelve infants receiving breast milk served as controls for the study. The h u m a n milk (HM) fed group (supplemented with Enfamil Fortifier, MeadJohnson, Evansville, IN) received i>75% of their intake as H M up to 36 wk postconceptional age (usual discharge age); if mothers were unable to fully provide H M for a brief period ( Formula A. An important enrichment was observed in the D H A content of the plasma and RBC lipids of the infants receiving the diets containing r LCP, HM and Formula C. Of additional interest were alterations in the end product/precursor ratios for both co6 and co3 FA desaturation pathways at this time point. Since the 18:2r 18:3r and 22:6co3 contents vary in the four diets, a true comparison of metabolic conversion is confounded; however, similar 20:4co6/18:2Go6ratios for all diets in both plasma and RBC lipids suggests t h a t metabolism of the co6 series was unaffected by diet FA composition. I n contrast, the co3 FA conversion ratio (22:6co3/18:3r was significantly reduced in plasma and RBC lipids of infants fed Formula B compared to HM and Formulas A and C. However, interpretation of this ratio is complicated by the high proportion of 18:3co3 in this formula. Results of FA analysis from infants at 57 wk postconceptional age, receiving diets for 5-6 mon, presented very little difference in the m o n o u n s a t u r a t e d FA of either plasma or RBC lipids (Tables 7 and 8). Despite the fact t h a t all formulas contained a greater percentage of saturated FA (62-67%) than in HM (45%), both plasma and RBC lipids from HM-fed infants were enriched in saturates compared to formula groups, particularly the corn oil-based Formula A group. Of additional interest, the plasma and RBC lipid FA distributions of the HM- and Formula C-fed groups were remarkably similar, yet the r E F A (18:2co6) compositions of these two diets varied markedly, 12.7% vs. 20.4%. In Formula A and B groups, the 18:2co6, 22:5co6 and 22:4co6 content in plasma and RBC fractions were comparable but differed from t h a t of the two groups receiving co3 LCP (Formula C and HM). Studies of ~3 FA dietary enrichment in rats specifically implicate an inhibition of A6 desaturase as leading to a reduction of r LCP in serum lipids (33). These observations are confirmed in long-term h u m a n infant studies (34). The r FA composition of plasma and RBC lipids were lowest in the Formula A-fed group, intermediate in Formula B infants and essentially equivalent for the HM and marine oil (Formula C) groups as found at 36 wk. Similarly, a reduction in the r product/precursor ratio (22:6r162 was found for LIPIDS, VoI. 27, no. 11 (1992)
890 D.R. H O F F M A N
A N D R. U A U Y
TABLE 3 F a t t y A c i d C o m p o s i t i o n of T o t a l P l a s m a L i p i d s a t E n t r y a Human milk
Formula A
Formula B
Formula C
(n = 12)
(n = 22)
(n = 26)
(n = 29)
ANOVA F
P
Saturated
32.5 +. 3.1
32.4 +_ 3.5
34.9 +- 3.5
34.5 +. 2.8
2.3
n.s.
Monounsaturated
26.7 +_ 5.8 b
19.9 +. 2.6
19.9 +. 2.4
19.9 +. 2.8
15.8
Essentiality of Dietary Fatty Acids for Premature Infants: Plasma and Red Blood Cell Fatty Acid Composition 1 Dennis R. Hoffman and Ricardo Uauy* The University of Texas Southwestern Medicat Center, Department of Pediatrics and The Human Nutrition Center, Dallas, Texas 75235-9063
Pre~term infants, that are not breast-fed, are deprived of vital intrauterine fat accretion during late pregnancy and must rely on formula to obtain fatty acids essential for normal development, particularly of the visual system. Preterm infants (30 wk postconception) receiving human milk were compared to infants given one of the following formulae: Formula A was a commercial preterm formula with predominantly 18:2c06 (24.2%) and low (0.5%) 18:3~3; Formula B was based on soy oil and contained similar 18:2w6 levels {20%) and high 18:3w3 {2.7%); Formula C was also a soy oil-based formula {20% 18:2, 1.4% 18:3) but was supplemented with marine oil to provide w3 long-chain polyunsaturated fatty acids {LCP) at a level {docosahexaenoie acid, DHA, 0.35%) equivalent to human milk. At entry (I0 days of age), the fatty acid composition of plasma and red blood cell (RBC) membrane lipids of the formula groups were identical. By 36 wk postconception, the DHA content in lipids of group A was significantly reduced compared to that in the human milk and marine oil formula groups. Omegas3 LCP results were further amplified by 57 wk with compensatory increases in 22:5c~6 in both plasma and RBC lipids. Provision of 2.7% a-linolenic acid in formula group B was sufficient to maintain 22:6w3 levels equivalent to those in human milk-fed infants at 36 wk but not at 57 wk. Effects on the production of thiobarbituric acid reactive substances and fragility of RBC attributable to the marine oil supplementation were negligible. The results support the essentiality of co3 fatty acids for preterm infants to obtain fatty acid profiles comparable to infants receiving human milk. Formuia for preterm infants should be supplemented with c03 fatty acids including LCP. Lipids 27, 886-895 (1992}.
Since animals lack the desaturating enzymes (1,2) necessary to form the co6 and co3 series of fatty acids (FA), they must rely on plant sources to obtain the parent essential fatty acids (EFA). Animal tissues are capable of further elongating and desaturating the parent EFA, linoleic acid (18:2co6) and a-linolenic acid (18:3co3), generating a family of long-chain polyunsaturated fatty acids (LCP) of each respective series. Alternatively, dietary LCP may be provided to infants directly from sources such as breast milk, marine oils or egg phospholipids.
The functional and structural roles of specific long-chain f a t t y acids are being increasingly recognized. Synthesis of eicosapentaenoic acid (EPA, 20:5co3) from a-linolenic acid in animals provides a precursor for series-3 prostaglandins which may suppress some effects of the series-2 prostaglandins originating from co6 FA by limiting their production. Although the actions of EPA may be beneficial in pathological conditions, concern persists as to the safety of limiting the action of arachidonate-derived eicosanoids in infants. Docosahexaenoic acid (DHA, 22:6co3} may provide a structural micro-environment in the membrane that influences important membrane functions such as ion or solute transport, receptor activity or enzyme action (3,4}. The enrichment of DHA in retina and neural tissues (5,6} may reflect a unique and critical role for this co3 LCP in visual function. Although the essentiality of co6 FA in the diet has long been recognized, only recently have studies focused on the more subtle biochemical and functional effects of dietary ~3 FA deficiency. The highly unsaturated nature of DHA may be responsible for the observed effects of co3 FA deficiency on the function of developing brain and retina (3,7}. Wheeler e t al. (8} demonstrated dramatic enhancement of both a- and b-waves of electroretinograms from rats fed dietary supplements of co3 FA compared to fat-free diets or supplementation with only co6 FA. Neuringer e t al. (9) reported visual acuity impairments of up to 50% in newborn rhesus monkeys receiving safflowe~based diets deficient in r FA pre- and postnataUy compared to controls given a soybean oilbased diet. These primates have depressed levels of DHA in plasma phospholipids at birth and nearly undetectable amounts at 12 wk of age (10). These visual and neural derangements may also occur in human infants; however, evidence of FA status in the brain presently must rely on the direct correlations of brain and red blood cell (RBC) lipid FA distribution found in rat and primate studies (9,11).
During fetal growth in the human, accretion of C18EFA is not linear but is approximately 400 mg of co6 and 50 mg of co3 EFA per kg of body weight deposited daily during the last trimester of gestation (12) and continues postnataUy in brain tissue (13). However, a preterm infant born at less than 34 wk will be deprived of this critical period of intrauterine lipid accretion and thus, is dependent on diet to fulfill its EFA requirements. Although formula supplementation with co6 and co3 EFA (i.e., 18:2co6 and 18:3c03) may be beneficial (14}, evidence for a limited 1Based on a paper presented at the Symposium on Milk Lipids held FA desaturase activity (t5) indicates that LCP such as at the A O C S Annual Meeting, Baltimore, M D , April 1990. arachidonic acid (16) and DHA (17,18) may also be re*To w h o m correspondence should be addressed at the Retina Founquired by the low-birth-weight infant. dation of the Southwest, 9900 N. Central Expressway, Suite 400, Preterm infants fed their own mother's milk receive a Dallas, T X 75231-3303. typical intake of 140 mg co3 FA/kg/d including 45 mg co3 Abbreviations: A N O V A , analysis of variance, D H A , docosahexa- LCP/kg/d (19}, thus meeting the requirements for neural enoic acid (22:6co3);EFA, essentialfatty acid(s);EPA, eicosapentaenoic acid (20:5co3);FA, fatty acid(s);H M , human milk; LCP, long- tissue and whole body EFA accretion. Preterm infants chain polyunsaturated fatty acid(s);MCT, medium-chain trig|ycer- given commercial formula receive less a-linolenic acid and ides; RBC, red blood cell(s);TBARS, thiobarbituric acid reactive elevated linoleic acid compared to breast-fed infants. This substances. relative excess of 18:2co6 may limit the formation of co3 LIPIDS, Vol. 27, no. 11 (1992)
887
co3 FATTY ACIDS IN PRETERM FORMULA LCP in ceil membranes (20-22) and blood lipids (11,17,18) because the A6 desaturase shared by b o t h series of FA is subject to s u b s t r a t e inhibition (1). Typical p r e m a t u r e infant formulas available up to 1987 (when the s t u d y commenced) had a low content of a-linolenic acid and no r LCP. Thus, we considered the p r e t e r m infant at risk for r LCP deficiency because biosynthesis is limited and formula diet supply was insufficient to meet the needs of the developing visual and neural s y s t e m s (23). The purpose of the present s t u d y was to evaluate the potential effects of oJ3 FA deficiency in very-low-birthweight h u m a n infants by s t u d y i n g functional and biochemical indicators of r FA status. The FA composition of p l a s m a and RBC lipids of h u m a n milk-fed infants and three formula-fed groups receiving different EFA and LCP supplementation are reported herein. A secondary objective included the evaluation of RBC m e m b r a n e peroxidation and resistance to hemolysis in the four diet groups. Interim visual function results for the infants under study have been recently published (18). MATERIALS AND METHODS
Subjects. Ninety-five newborns with body weights of 1000-1500 g, who received enteral feedings b u t were free of major neonatal m o r b i d i t y by day 10 of life, were eligible for the study. Twelve infants receiving breast milk served as controls for the study. The h u m a n milk (HM) fed group (supplemented with Enfamil Fortifier, MeadJohnson, Evansville, IN) received i>75% of their intake as H M up to 36 wk postconceptional age (usual discharge age); if mothers were unable to fully provide H M for a brief period ( Formula A. An important enrichment was observed in the D H A content of the plasma and RBC lipids of the infants receiving the diets containing r LCP, HM and Formula C. Of additional interest were alterations in the end product/precursor ratios for both co6 and co3 FA desaturation pathways at this time point. Since the 18:2r 18:3r and 22:6co3 contents vary in the four diets, a true comparison of metabolic conversion is confounded; however, similar 20:4co6/18:2Go6ratios for all diets in both plasma and RBC lipids suggests t h a t metabolism of the co6 series was unaffected by diet FA composition. I n contrast, the co3 FA conversion ratio (22:6co3/18:3r was significantly reduced in plasma and RBC lipids of infants fed Formula B compared to HM and Formulas A and C. However, interpretation of this ratio is complicated by the high proportion of 18:3co3 in this formula. Results of FA analysis from infants at 57 wk postconceptional age, receiving diets for 5-6 mon, presented very little difference in the m o n o u n s a t u r a t e d FA of either plasma or RBC lipids (Tables 7 and 8). Despite the fact t h a t all formulas contained a greater percentage of saturated FA (62-67%) than in HM (45%), both plasma and RBC lipids from HM-fed infants were enriched in saturates compared to formula groups, particularly the corn oil-based Formula A group. Of additional interest, the plasma and RBC lipid FA distributions of the HM- and Formula C-fed groups were remarkably similar, yet the r E F A (18:2co6) compositions of these two diets varied markedly, 12.7% vs. 20.4%. In Formula A and B groups, the 18:2co6, 22:5co6 and 22:4co6 content in plasma and RBC fractions were comparable but differed from t h a t of the two groups receiving co3 LCP (Formula C and HM). Studies of ~3 FA dietary enrichment in rats specifically implicate an inhibition of A6 desaturase as leading to a reduction of r LCP in serum lipids (33). These observations are confirmed in long-term h u m a n infant studies (34). The r FA composition of plasma and RBC lipids were lowest in the Formula A-fed group, intermediate in Formula B infants and essentially equivalent for the HM and marine oil (Formula C) groups as found at 36 wk. Similarly, a reduction in the r product/precursor ratio (22:6r162 was found for LIPIDS, VoI. 27, no. 11 (1992)
890 D.R. H O F F M A N
A N D R. U A U Y
TABLE 3 F a t t y A c i d C o m p o s i t i o n of T o t a l P l a s m a L i p i d s a t E n t r y a Human milk
Formula A
Formula B
Formula C
(n = 12)
(n = 22)
(n = 26)
(n = 29)
ANOVA F
P
Saturated
32.5 +. 3.1
32.4 +_ 3.5
34.9 +- 3.5
34.5 +. 2.8
2.3
n.s.
Monounsaturated
26.7 +_ 5.8 b
19.9 +. 2.6
19.9 +. 2.4
19.9 +. 2.8
15.8
