VYH Yu

Abstract

An increasing number of infant formulas are being marketed with nucleotide supplementation, because the composition of human milk is considered the gold standard for infant nutrition, and human milk has been shown to have a higher concentration of nucleotides than bovine milk which is the source of most infant formulas. This paper provides an outline of the biology of human milk nucleotides, and reviews the evidence for beneficial effects of feeding infants with nucleotide-supplemented formulas in clinical studies. Published randomised controlled trials have to date not convincingly demonstrated clinical benefits of feeding healthy term infants nucleotide-supplemented formula compared with non-supplemented formula. However, one study reported that term infants with severe intrauterine growth retardation have better catch-up growth in the first six months with nucleotide supplementation. Further studies are required to define the role of human milk nucleotides, and to evaluate the potential clinical benefits and appropriate level of nucleotide supplementation of infant formula, before a recommendation can be made for nucleotide supplementation of infant formulas for feeding either healthy or at-risk infants.

What are Nucleotides?

Nucleotides are non-protein nitrogen compounds in human milk and milk of other species. Thirteen acid-soluble nucleotides have been reported since nucleotides were first identified in human milk in 1960. They consist of a nitrogenous base, a five-carbon sugar (ribose or deoxyribose) and one to three phosphate groups. The nitrogen-containing bases are derivatives of two parent heterocyclic compounds, purines (adenine and guanine, mainly) and pyrimidines (cytosine, thymine and uracil, mainly). The ribo-nucleotides and deoxyribo-nucleotides serve as the monomeric precursor units of RNA and DNA respectively. Because nucleotides are the structural units of nucleic acids, RNA and DNA, and are essential compounds in the energy transfer systems (that is, in ATP and GTP), they have been assumed to play an important role in carbohydrate, lipid, protein and nucleic acid metabolism, and as modulators of many neonatal physiological functions.¹

Proteases and nucleases degrade dietary nucleoproteins and nucleic acids into nucleotides. Intestinal alkaline phosphatases and nucleotidases cleave the phosphate groups from nucleotides to form nucleosides which are absorbed in the small intestine. The absorbed nucleosides are mainly degraded to uric acid and allantoin, but some are reconverted to nucleotides. Nucleotides can also be synthesised de novo via a metabolically costly process using elemental components such as amino acids and glucose. Intestinal mucosa and bone marrow haematopoietic cells, which have a limited capacity for this de novo synthesis, depend more on the salvage pathway that produces nucleotides from either exogenous dietary nucleosides or endogenous purine and pyrimidine bases derived from the degradation of RNA and DNA.

It remains uncertain what the relative contribution of dietary nucleotides is to the overall pool of nucleotides in individual organs or in the entire body.² Theoretically, dietary nucleotides might become a vital source when the infant's metabolic demand exceeds the capacity for de novo synthesis and the endogenous salvage pathway. They could then potentially become a semi-essential or conditionally essential nutrient which, by definition, is one which becomes essential only in specific situations when the endogenous supply is inadequate for normal function, but which under normal circumstances, its absence from the diet does not result in a clinically evident syndrome consequent to its deficiency. The three scenarios in which a semi-essential nutrient might become essential are periods of insufficient intake, high rate of physical growth, and the presence of disease. It is therefore possible in conditions such as prematurity, intrauterine growth retardation (IUGR), and intestinal injury from necrotising enterocolitis or diarrhoeal states, an adequate dietary intake of nucleotides might spare infants the metabolic cost of de novo synthesis or salvage, and optimise their physiological function.

In human milk, up to 30 per cent of its total nitrogen content is non-protein nitrogen, of which free nucleotides account for 2 to 5 per cent, and have a concentration of 50 to 150 mmol/L, or 2 to 6 mg/100 kcal.³ By three months of lactation, the free nucleotide content goes down to 75 per cent of that in human colostrum.⁴ In addition to the free nucleotides, there are free nucleosides and polymeric and cellular nucleotides present in human milk, derived from the structural units of nucleic acids RNA and DNA. The concentration of total potential available nucleosides (TPAN) is 189±70 mmol/L (mean±SD), with a range of 82 to 402 mmol/L⁵ (84±25 mmol/L present as nucleotides 68±55 mmol/L as nucleic acids, and 10±2 mol/L as nucleosides).⁶

In bovine milk, non-protein nitrogen accounts for only 2 to 5 per cent of the total nitrogen, resulting in a nucleotide content in infant formulas which is significantly lower than that of human milk. In addition, cytidine and adenosine derivatives are present in relatively lower proportions in bovine milk than in human milk. Nevertheless, no report of clinical deficiency due to the lower nucleotide content in infant formulas has been published. Because the goal of infant formula development is to mimic human milk as closely as possible, infant formulas supplemented with nucleotides have been available in Japan from 1965, in Spain from 1983, in the USA from 1989, and in some countries in South-East Asia from 1990, and no deleterious effects have been reported. The European Commission's Scientific Committee for Food has published guidelines in 1991 and in 1996 on nucleotide supplementation of infant formulas.^7,8 The Committee has approved infant formula supplementation for five nucleotides with the following maximum limits: cytidine 5'-monophosphate 2.5 mg/100 kcal, uridine 5'-monophosphate 1.75 mg/100 kcal, adenosine 5'-monophosphate 1.5 mg/100 kcal, guanosine 5'-monophosphate 0.5 mg/100 kcal, and inosine 5'-monophosphate 1 mg/100 kcal. The Committee has authorised the use of the sodium salts of these nucleotides which are readily water soluble and hydrolysed in the intestine and absorbed as nucleosides. The Committee stated that the total nucleotide concentration should be less than 5 mg/100 kcal, that is, in the same order of magnitude as the free nucleotides in human milk. It specifically rejected a submission for a two-fold to threefold increase in nucleotide supplementation, which would result in a level equivalent to the TPAN content of human milk.⁸

Effect on Intestinal Microflora

Bifidobacteria which reduce the pH of intestinal contents and impair the growth of pathogenic bacteria, are found in the stools of breast-fed infants. Since nucleic acids have been shown to enhance the growth of bifidobacteria in vitro, a study on intestinal microflora was conducted in healthy term infants to compare the effects of feeding human milk (HM group), formula milk supplemented with nucleotides (NFM group), and non-supplemented formula milk (FM group).⁹ The numbers of infants in the groups were only 10, 11 and 12 respectively, and the study was not conducted as a randomised controlled trial (RCT). Bacterial culture was performed on faecal samples taken at 1 and 4 weeks of age. Only data from healthy children were analysed; those with diarrhoea or had antibiotic therapy were excluded. The number of bacteria per gram of dry faeces reported for each of eight organisms (aerobes, anaerobes, lactobacilli, bifidobacteria, enterobacteria, enterococci, clostridia, staphylococci) was not significantly different among the three groups at both 1 and 4 weeks of age. Significant differences were however found between groups when the bacterial counts for bifidobacteria, lactobacilli and enterobacteria were expressed as percentages of the total bacterial count. Although the percentage of bifidobacteria in the NFM group was significantly higher than that in the FM group at 4 weeks, both FM/NFM groups had a significantly lower percentage of bifidobacteria compared with the HM group at both 1 and 4 weeks. Both FM/NFM groups had significantly higher percentages of lactobacilli and enterobacteria compared with the HM group.

The chance of biases in this study was increased by failing to randomise subjects, the small numbers in each group, and excluding data from ill children. Infants fed a nucleotide-supplemented formula did not develop an intestinal microflora pattern which is quantitatively different from those fed a non-supplemented formula, as there was a lack of significant differences in the bacterial counts of the eight bacteria groups obtained from the three feeding regimes at both 1 and 4 weeks. Even when the data was expressed as a percentage of the sum of the bacterial counts, both NFM/FM groups were significantly different from that in the HM group, suggesting that nucleotide supplementation did not change the intestinal microflora pattern to that found with breastfeeding. The difference in the percentage of bifidobacteria between the NFM/FM groups was statistically significant only at 4 weeks and only with a p-value of <0.05. The magnitude of this difference was much smaller than that reported between the HM group and NFM/FM groups, and the probability of finding a false-positive result due to a type I error is high, because of small sample size and comparison of multiple variables within multiple groups (a higher level of statistical significance, for example p<0.01, should be used). This study had failed to show that nucleotide supplementation of cow's milk-based formula above the quantity already present, changes significantly the number and type of intestinal bacteria in normal healthy term infants to that found in those who are breast-fed. A more recent study had reported similar findings.¹⁰

Effect on Diarrhoea and Sepsis

Since dietary nucleotides can enhance mucosal regeneration rats with chronic diarrhoea,¹¹ a clinical study was conducted to investigate the effect on the pattern of diarrhoea of feeding healthy infants nucleotide-supplemented (n=194) and non-supplemented (n=198) formula milk.¹² Over a three-month period, assessments were made on physical growth, diarrhoeal episodes, presence of enteropathogens, incidence of infectious illnesses, and hospitalisation rate, in infants from the periurban slums of Santiago in Chile. There were no significant differences in body weight or length, total number or duration of diarrhoeal episodes, and total number of days with diarrhoea (all episodes). Although the number of first episodes of diarrhoea was significantly lower in the nucleotide-supplemented group compared to the non-supplemented group, no significant difference was found in the duration of first episodes nor in the number of children with more than one episode. No significant differences were found in the presence or type of enteropathogens, incidence of infections of the upper/lower respiratory tract, skin, urinary tract, eye, or other sites, and in the rate of hospital admission.

It is uncertain how applicable the findings are to an urban infant population within a developed country or an industrialised nation, since this study was conducted among infants from a low socioeconomic stratum living within a contaminated environment. The chance of biases in this study was increased by it not being a RCT and by the exclusion of 26 per cent of enrolled infants for a variety of reasons. Of the multiple comparisons used to describe the incidence or severity of diarrhoea, all except one showed no significant differences between the two groups. Therefore, nucleotide- supplemented formula feeding was not shown to reduce the incidence or severity of diarrhoeal disease compared to non-supplemented formula feeding. Data from this study also supports the findings of two previous studies^9,10 which reported no difference in intestinal microflora between the nucleotide supplemented and non-supplemented groups. Furthermore, the incidence of infectious illnesse and the need for hospitalisation were not different between the two groups.

Effect on Immune Function and Sepsis

A number of in vivo animal studies had demonstrated significant effects of nucleotides on tests of cellular and humoral immunity. These include increased mortality of graft versus host disease, increased rejection of allogenic grafts, improved delay cutaneous hypersensitivity and alloantigen-induced lymphoproliferation, reversed malnutrition-induced immunosuppression, improved resistance to challenge with staphylococcus aureus and candida albicans, and enhanced T-cell maturation and function. Therefore two studies have been conducted on the immune function of healthy term infants who were breast-fed (HM group) and fed with two versions of formula milk, one supplemented (NFM group) and one not supplemented (FM group) with nucleotides.

The first study measured natural killer cell activity and interleukin-2 production in peripheral blood mononuclear cells at 2 and 4 months of age, and reported the incidence and severity of infections in the four-month period. ¹³ The three groups had only 9, 13 and 15 infants respectively, and random allocation was performed for the two formula-fed groups. Physical growth, haematological indices and plasma biochemistry profiles among the three groups were not significantly different. Both the natural killer cell activity (at the 50:1 and 25:1 effector-to-target cell ratios but not at the 12.5:1 ratio) and interleukin-2 activity were significantly higher at 2 months of age in the NFM group compared to the FM group, but no significant difference was found at 4 months of age. No significant differences in these two immune function tests were found between the HM group and the NFM/FM groups. Neither were there significant differences found in the incidence or severity of infections among the infants in the three groups. The findings of this study cannot be extrapolated to infants beyond the age of 4 months. The probability of finding a false-positive result due to a type I error is high, because of small sample size and comparison of multiple variables within multiple groups. The lack of effect on the incidence or severity of infections among the infants in the three groups suggests that the transient differences (they were statistically non-significant by 4 months) reported for these two peripheral blood mononuclear cell functions do not have clinical relevance. It is important to realise that, in order to support a particular viewpoint, investigators have been known to play the substitution game in which an intermediate outcome (in this case, an immune function test) was substituted for a clinical relevant outcome of prime importance (in this case, the infection rate).¹⁴

The second study measured the antibody responses to haemophilus influenzae type b polysaccharide (Hib), diphtheria, tetanus toxoids, and oral polio virus (OPV) immunisation, at 6, 7 and 12 months of age, as well as physical growth and serum immunoglobulin G and A concentrations.¹⁵ The three groups had 124, 121 and 125 infants respectively, and random allocation was performed for the two formula-fed groups. However, 16 per cent of enrolled infants did not complete the study. The antibody responses to QPV and tetanus were not significantly different between the NFM/FM groups at any age, and the antibody response to QPV was significantly lower both NFM/FM groups compared with the HM group at 6 months of age. The NFM group had a significantly higher Hib antibody concentration compared with the FM group at 7 and 12 months of age, and a significantly higher diphtheria antibody concentration at 7 months but not at 12 months. No significant differences were found in physical growth or serum immunoglobulins levels.

Effect on Catch-up Growth in Severe IUGR Infants

At least 13 studies had shown that in healthy term infants who are appropriately grown at birth, nucleotide supplementation has no effect on physical growth.¹ However, a RCT conducted in term infants whose birthweight was below the 5th percentile, had shown that nucleotide supplementation of formula milk was associated with better catch-up growth.¹⁶ This RCT was conducted in 39 infants fed a nucleotide-supplemented formula and 35 infants fed a non-supplemented formula over a six-month period. Unfortunately no breast-fed group was included for comparison, and 16 per cent of the enrolled infants did not complete the study. The nucleotide-supplemented group has significantly higher growth rate (weight, length and head circumference) compared to the non-supplemented group. The pattern of illnesses, which might have influenced their physical growth, was not different between the two groups. Therefore, the improved catch-up growth during the first 6 months after birth in the nucleotide-supplemented group was postulated to be due to trophic effects of nucleotides on the intestinal mucosa compromised by intrauterine malnutrition. Whether the improved early growth results in long-term benefit which persists into late infancy and beyond is unknown.

Discussion

Nucleotides are not essential nutrients as they are synthesised endogenously. The hypothesis which researchers have been testing is that nucleotides are semi-essential nutrients, that is, they might become essential under certain clinical conditions, when supplementation might confer beneficial effects upon the gastrointestinal tract, the immune system and physical growth. However, to answer such questions, more basic research is required to study the absorption and metabolism of nucleic acids, nucleotides, nucleosides, bases, and related metabolic products in newborn infants, and the impact of conditions such as prematurity, IUGR, intestinal injury and limited nutrient intake.

While the biology of nucleotides is being scientifically clarified, an increasing number of infant formulas are already becoming available which are supplemented with nucleotides. This is partly because the composition of human milk has traditionally been considered as optimal for infant nutrition, and has served as the gold standard in research and development of infant formulas. As no adverse effects had been reported with their use in healthy term infants, such products are currently considered safe, at least within the range of nucleotide concentrations approved by the Scientific Committee for Food of the European Commission.^7-8 However, scientific proof to show that nucleotide supplementation is beneficial can only be obtained from clinical evidence based on medical studies performed in human infants. RCTs offer maximum protection against selection bias which can invalidate comparisons between groups because of confounding variables. In conducting such RCTs, the choice of outcome measures is important to avoid the substitution game, in which a risk or surrogate factor (for example, immune function test) is substituted for an event of prime clinical importance (that is, gastrointestinal or systemic infection). The limited number of quality clinical studies to date have yet to provide definitive medical evidence that nucleotide-supplemented formulas confer the benefits of human milk, and that in formula-fed healthy term infants, nucleotide supplementation results in improved physical growth or gastrointestinal function, and lower rates of infectious diarrhoea or other forms of sepsis. Consequently, a review article concluded that "the role of human milk nucleotides for breast-fed infants is not known, and the issue of nucleotide supplementation of infant formula remains controversial."¹⁷ This present review of the clinical evidence also comes to the same conclusion that, although the use of nucleotide-supplemented cow's milk-based formulas appear to have physiological rationale, its theoretical beneficial effects have not yet been demonstrated in RCTs. Until the clinical benefits are proven, it is prudent to remain cautious in the interpretation or acceptance of exaggerated or misleading nutritional performance claims found in some commercial promotional material associated with nucleotide-supplemented formulas.

References

1. Carver JD, Walker WA. The role of nucleotides in human nutrition. J Nutr Biochem 1995;6:58-72.

2. Quan R, Barness LA, Uauy R. Do infants need nucleotide supplemented formula for optimal nutrition? J Pediatr Gastroenterol Nutr 1990;11:429-37.

3. Janas LM, Picciano MF. The nucleotide profile of human milk. Pediatr Res 1982;16:659-62.

4. Gil A, Sanchez-Medina F. Acid-soluble nucleotides of human milk at different stages of lactation. J Dairy Res 1982;49:301-7.

5. Leach JL, Baxter JH, Molitor BE, et al. Total potentially available nucleosides of human milk by stage of lactation. Am J Clin Nutr 1995;61:1224-30.

6. Thorell L, Sjöberg L, Hernell O. Nucleotides in human milk: sources and metabolism by the newborn infant. Pediatr Res 1996;40:845-52.

7. Scientific Committee on Food. Second Addendum to the Report concerning the essential requirements of infant formulae and follow-up milks based on cow's milk proteins. European Commission of the European Communities, Luxembourg (Directive 91/321/EEC). 28th Series, pgs. 30-5.

8. Scientific Committee for Food. Minutes of the 100th Plenary Session, March 7-8, 1996, in Brussels. European Commission of the European Communities, Luxembourg (Directive 96/4/EC). pg. 7.

9. Gil A, Corral E, Martinez A, et al. Effects of the addition of nucleotides to an adapted milk formula on the microbial pattern of feces in at term newborn infants. J Clin Nutr Gastroenterol 1986;1:127-32.

10. Balmer SE, Hanvery LS, Wharton BA. Diet and faecal flora in the newborn: nucleotides. Arch Dis Child 1994;70:F137-40.

11. Nunez MC, Ayudarte MV, Morales D, et al. Effect of dietary nucleotides on intestinal repair in rats with experimental chronic diarrhea. J Parenter Ent Nutr 1990;14:598-604.

12. Brunser O, Espinoza J, Araya M, et al. Effect of dietary nucleotide supplementation on diarrhoeal disease in infants. Acta Paediatr 1994;83:188-91.

13. Carver JD, Pimentel B, Cox WI, et al. Dietary nucleotide effects upon immune function in infants. Pediatrics 1991 ;88:359-63.

14. Sinclair JC. Assessing evidence concerning treatment and prevention of diseases of the newborn. In: Sinclair JC. Bracken MB (eds). Effective Care of the Newborn Infant. Oxford University Press, Oxford. 1992:3-12.

15. Pickering LK, Granoff DM, Erickson JR, et al. Modulation of the immune system by human milk and infant formula containing nucleotides. Pediatrics 1998;101:242-9.

16. Cosgrove M, Davies DP, Jenkins HR. Nucleotide supplementation and the growth of term small for gestational age infants. Arch Dis Child 1996;74:F122-5.

17. Rudloff S, Kunz C. Protein and nonprotein nitrogen components in human milk, bovine milk, and infant formula: Quantitative and qualitative aspects in infant nutrition. J Pediatr Gastroenterol Nutr 1997;24:328-44.