Long-chain polyunsaturated fatty acidsupplementationin infant formulasdoes not influencecalcium and magnesium bioavailability in rats

Running tittle: LCPUFA-supplemented infant formula, Ca and Mgbioavailability

Dr. M Pilar Vaquero, Dr. Beatriz Sarriá

Departamento de Metabolismo y Nutrición, Instituto del Frío, CSIC,

Madrid, Spain

Acknowledgment for research support: this work was supported by the Spanish Commission of Science and Technology (project ALI 96-0465). The authors thank Dr. Ana MariaPérez-Granadosfor scientific advice andMarcel Veldhuizen for technicalassistance.

Author for correspondence:

Dr. M. Pilar Vaquero

Instituto del Frío, CSIC

C/José Antonio Nováis 10, 28040 Madrid, Spain.

Phone number: 00 34 91 549 00 38 ext. 295

Fax number: 00 34 91 549 36 27

Email:

Key words: long-chain polyunsaturated fatty acids; infant formula; calcium; magnesium; mineral bioavailability

ABSTRACT

The influence of infant formula supplementation with long-chain-polyunsaturatedfatty acids(LCPUFA)on calcium and magnesium bioavailability was assessed in rats. Two test dietscontaining the unsupplemented (F) or supplemented (SF) infant formula as the fat source and a control diet(C) were administered to weaning rats for 28days. Food intake and body weight were monitored, during the last week faeces and urine were collected to calculate apparentcalcium and magnesium absorption and retention.Food intake and body weight evolution showed no significant differences between F and SF but were lower in both groups compared to C. Calcium and magnesium intake did not differ between F and SF, but both parameters weresignificantly lower in SF compared toC. Calcium absorption efficiency in Fand SF was significantly higher than in C.Formula-fed groups showed higher urinary calcium excretion and thus,no differences were observed in calcium retention.Magnesium absorption was significantly higher in F compared to SF and C but there were no significant differences between test groups for the absorption efficiency. Magnesium apparent retention was similar in the three groups. The consumption of a diet containing an infant formula supplemented with LCPUFA compared to the unsupplemented formula does not affect calcium and magnesium bioavailability in rats.

INTRODUCTION

Human milk is the optimal source of nutrients for healthy young infants. As an alternative to breastfeeding, cow’s milk based infant formulas are consumed. Modifications to infant formulas are continually being made as the components of human milk are characterised and as the nutrient needs of diverse groups of infants are identified. Regarding the fat content, although there is evidence that infants are able to synthesize arachidonic acid (AA, 20:4, n-6) and docosahexanoic acid (DHA, 22:6, n-3) from linoleic (18:2, n-6) acid and linolenic acid (18:3, n-3), respectively, the increased requirements for the longer chain metabolites during development has raised concern that infants may also require preformed long-chain polyunsaturated fatty acids (LCPUFA). Therefore, formulas with LCPUFA in amounts similar to those in human milk have recently become available1.Beneficial effects of supplementing infant formulas with LCPUFA, of n-3 and n-6 series, on visual and neural development in rodent models2and preterm infants have been reported3,4 whereas in term infants neutral or positive outcomes have been described5,6. Controversy exists concerning the critical period during which the dietary supply of LCPUFA may influence the maturation of cortical function in term infants. The potential long-term benefits of LCPUFA supplementation are still being explored, biochemical data indicate that breast-fed infants accumulate DHA in the brain until >12 months of age and at a greater rate than do infants fed formula without DHA7. Das8 recently suggested supplementation of LCPUFA from the second trimester of pregnancy to the age of 5 years in order to prevent coronary heart disease in adult life.

The repercussions of LCPUFA supplementationon mineral bioavailability have been studied to a limited extent. Our research group evaluated the effects of consuming an infant formula supplemented with LCPUFA on iron absorption using a rat model, this parameter was found to be unaffected, haemoglobin did not change but the erythrocytic concentrations of iron9, as well as copper and zinc10 increased, probably due to increased AA and DHA in erythrocytic membranes and changes in membrane permeability.Few studies have investigated the potential for polyunsaturated fatty acids to influence calcium and bone metabolism during early growth stages. Calcium absorption is enhanced by linoleic acid compared to oleic acid11, whereas magnesium absorption is not affected, although thermally treated fats did alter magnesium bioavailability12. Research in adult rodents fed diets containing varied amounts of fish oils indicates that a total n-6/n-3 ratio of 3:1 or 1:1 results in higher amounts of calcium in bone compared with the control, but the ratio 1:3 had no effect. Feeding the diet with an n-6/n-3 ratio of 3:1 also resulted in significantly greater absorption of dietary calcium compared with the other diets13. However, a higher excess in n-3 fatty acids intake (n-6/n-3 ratio of 1:5) induced growth retardation and reduced calcium absorption and retention efficiencies in young rats14. Martinez et al15 assessed in very low weight preterm infants the influence of LCPUFA supplementation in infant formulas, versus a non-supplemented formula, on mineral bioavailability and obtained comparable intake and net retention of calcium and magnesium through a three-day balance. However, data on term infants are lacking although first age formulas containing LCPUFA are available in the market.

Due to the facts that LCPUFA supplemented infant formulas are widely consumed and to the difficulties in performing studies in term infants, we investigated the effects of LCPUFA supplementation of an infant formula on calcium and magnesium bioavailability in weanling rats. Suckling16-18 and weanling rat models19 have been extensively used and are well validated to carry out mineral bioavailability experiments.

METHODS AND MATERIALS

Diets

Two term commercially available infant formulas were obtained from the same manufacturer (Milupa,Spain). Both were provided in powder and the only difference was the fat composition. The standard infant formula was devoid of LCPUFA while the infant formula supplemented with LCPUFA contained egg-lipid-derived LCPUFA. Details about the nutrient composition of these infant formulas have been published20.

Three isoenergetic diets were prepared, two contained the unsupplemented (F) or supplemented (SF) infant formula exclusively as the fat source (150g/kg) and the third was a control diet (C) based on the American Institute of Nutrition recommendations (AIN-93) except for the fat level, which was also 150g/kg of corn oil. In order to elaborate F and SF diets, the appropriate amounts of all dietary components were combined to make up with the control diet. The resulting mixture was divided into two parts and the test infant formulas were added to make F and SF diets. The control diet (modified AIN diet, Dyets Inc., Bethlehem, PA, USA) was purchased. The composition of the experimental diets has been previously detailed9 andthe fatty acid content is shown in table 1. The three diets were kept at 4ºC until use. The calcium content of the diets (g/kg) was (mean  standard error of 5 determinations per diet) 4.790.04, 4.730.05 and 4.590.02 for F, SF and the control, respectively. Magnesium contents were 0.470.02, 0.440.03 and 0.460.03 for F, SF, and the control, respectively.

The diets were given to weanling rats for 28d. During this period, food intake and body weight were monitored and in the last week faeces and urine were collected to calculate apparent calcium and magnesium absorption and retention.

Biological assay

Thirty-four weanling Wistar rats, initial body weight 40.30.1g (mean  standard error), were housed individually in metabolic cages in an environmentally controlled room maintained at 20-22ºC, with a 12h light-dark cycle and 55-70% humidity. F, SF and the control group contained 10, 12 and 12 animals, respectively. An equal number of males and females in each group were randomly assigned to the dietary treatments. Animals had free access to food and demineralised water (Milli-Q plus, Ultrapure Water System, Millipore Corporation, Bedford, U.S.A.) for 28 days. Body weight and food intake were monitored and faeces and urine were collected and pooled separately from day 21 until day 27. Faeces were dried, weighed and homogenised. Urine was collected in 0.5% v/v HCl solution, filtered (nº41, ashless Whatman Filter Papers, Whatman Ltd., England) and diluted.

On day 28, after an overnight fast, animals were anesthetised using sodium pentobarbital (Abbott Laboratories, S.A., Madrid, Spain) and blood was drawn from the carotid artery into acid-washed (HNO3 10 mol/L) plastic vials and allowed to clot. Throughout the study, rats were handled following the European Science Foundation Statement on the Use of Animals in Research (

Analytical Techniques

Diets and faeces were dry-ashed in a muffle furnace at 450ºC. Ashes were dissolved in a HCl/HNO3/H2O solution (1:1:2) (Suprapur, E.Merck, Darmstadt, Germany).

Calcium and magnesium analysis in the diets, faeces, urine and serum was performed in duplicate by flame atomic absorption spectrometry (Perkin-Elmer 1100B, Norwalk, CT., U.S.A.). A stock standard solution of calcium (1gl-1) and magnesium (1gl-1) was prepared from Titrisol (E. Merck). Lanthanum chloride (Merck) was added to the calcium samples and standards (final concentration: (5 gl-1 lanthanum).

A pool of faeces was used as an internal control for both mineral determinations. The interassay relative standard deviation was 2.26% and 1.13% for calcium and magnesium, respectively. Certified reference material (milk powder, CRM 63, Community Bureau of Reference, Brussels) was used to assess accuracy, values of 12.8±0.4 and 1.11±0.03mgg-1 (mean±standard deviation of five determinations) were obtained for calcium and magnesium, respectively (certified value: 12.6±0.3 mgg-1 and 1,12±0.03 mgg-1 for calcium and magnesium respectively). Distilled-deionised water (Milli Q plus, Millipore) was used for the preparation of dilutions and calcium and magnesium standards.

Indices

The following indices were calculated from data on calcium and magnesium, intake and faecal and urinary excretion:

Apparent absorption(A)= Intake (I) - Faecal excretion

%A/I = Apparent absorption/ Intake x 100

Apparent retention(R)= Apparent absorption – Urinary excretion

%R/A = Apparent retention/Apparent absorption x 100

%R/I= Apparent retention/Intake x 100

Statistical Analysis
Food intake and body weight data were analysed by analysis of variance (ANOVA) of repeated measures. The rest of data were studied by means of one-way ANOVA. Two sample comparisons were made using the Bonferroni test. Data were processed with the SPSS Statistical Package. Significance of the results was established at p<0.05.

RESULTS

Food intake and body weight evolution

Throughout the assay, rats fed the C diet showed a higher food intake compared to those administered the diets containing F and SF (table 3; ANOVA repeated measures p=0.002). The differences between C and both F and SF were very significant in the periods 3-7 (p=0.002) and 22-27 days(p=0.001) but not in the period 7-14 days. Animals fed SF showed significantly lower food intake compared to C in the stage 14-22 days (p=0.014). The differences between F and SF were not significant, although SF exhibited the lowest values of food intake.

Accordingly, there were significant differences in the body weight evolution (table 3; ANOVA repeated measures p=0.002). This parameter was significantly lower in SF than C (p=0.001). On days 3 and 7, body weight was significantly lower in both F and SF than C (p=0.007 and p<0.001, respectively), the differences were not significant on day 14. However, on days 22 and 27 SF rats presented significantly lower body weights than C animals (p=0.015 and p=0.004, respectively).

Calcium

Calcium concentration in serum was similar in the three groups (mg/dl) 11.33±0.11, 10.24±0.97 and 10.83±0.66 (mean ± standard error mean) for F, SF and C, respectively.

Calcium intake was significantly lower in SF compared to C (p<0.001). In the test groups, calcium faecal excretion was 3-4 times lower (p<0.001) and therefore the apparent absorption was elevated, although the differences were only significant between F and C. However, urinary calcium was also 3-4 times higher in SF and F compared to C (p<0.001), therefore the differences between groups in calcium apparent retention were not significant. The ratios %A/I and %R/I were significantly higher in both SF and F (p<0.001 and p=0.001, respectively) respect to C, while %R/A was significantly lower in F and SF (p<0.001) than in C.

Magnesium

Magnesium concentration in serum was similar in the three groups (mg/dl) 2.15±0.14, 1.9±0.24 and 2.22±0.15 (mean ± standard error mean) for F, SF and C, respectively.

Magnesium intake was significantly lower in SF compared to C (p<0.001) but there were not significant differences between F and SF. The experimental groups eliminated significantly less of this mineral in faeces (p<0.001) and F showed significantly higher magnesium absorption than SF and C (p=0.016). However, in F the urinary excretion of magnesium was slightly higher compared to SF and C, therefore the apparent retention was similar to the other groups and no significant differences were obtained for this parameter.The digestive ratio, %A/I, was significantly higher in the test groups compared to the control (p<0.001). The percentages %R/A and %R/I were within the same range in the three groups, not showing any statistical differences.

DISCUSSION

The standard infant formula used in the present study (F) contains a mixture of polyunsaturated and saturated fatty acids, which is also present, although in slightly different quantities, in the supplemented infant formula, together with LCPUFA from egg-lipid, which is one of the richest and most commonly used source of these fatty acids21. In contrast, diet C contained approximately 5 times more linoleic acid than F and SF, very low amounts of saturated fatty acids, particularly myristic and palmitic acids, and was devoid of LCPUFA. Considering that diet C is designed for rats, it is not surprising that food intake and body weight were the highest in the group that consumed it. [Similar values were obtained for these parameters in F and SF, although lower in SF, through the experimental period].

Supplementation with LCPUFA did not affect food intake and growth significantly, althoughlower values were observed in FS which tended to ingest less calcium and magnesium.In other animal models, such as mouse pups, reduced growth was observed after dietary supplementation with up to 9% DHA (wt/wt), regardless of dietary AA or total n-6:n-3 ratios22. On the contrary, feeding piglets infant formula supplemented with DHA and AA (0.1% and 0.5% wt/wt total fat, respectively) did not compromise growth but furthermore, greater whole body weight, bone mineral content and density were observed compared to feeding an isoenergetic standard formula23. The suggested mechanism for the stimulated growth and bone mass is that a greater amount of precursor, AA, is available for synthesis of prostaglandin E2, a potent stimulator of bone formation, in response to the calciotropic hormones during growth24.

Results in human infants are very limited and not conclusive, although no influence of consuming infant formulas with LCPUFA supplements on standard growth indexes were reported25,26,27, specific negative associations were observed in few studies. An infant formula supplemented with DHA and eicosapentaenoic acid (EPA, C20:5)(both n:3) reduced concentrations of AA in plasma phospholipids that were significantly associated with reduced weight and length growth in preterm babies26. In a different study using a supplement with DHA and AA, infant growth was not affected, although a small association between DHA status at 16 weeks of age and weight at 1 and 2 years was observed27.As indicated by Innis28, inconsistencies in the findings of n-6 and n-3 requirements of infants may be attributed to the different necessities of preterm/term infants, differences in the types of oils used to provide DHA and AA, differences in the AA:DHA ratios, and even to the home environment. Expert committees and advisory panels recommend that the ratio linoleic acid: α linolenic acid not be more than 16:1 nor less than 6:1 which points to the possible adverse effects of excess n-3 fatty acids (EPA, AA and DHA) on growth29. According to these recommendations, the supplemented infant formula used in the present study included fatty acids in the adequate amounts and ratios, and accordingly food intake and body weight were not affected compared to the unsupplemented formula.

It is well known that a high polyunsaturated/saturated ratio of dietary fat favours calcium and magnesium absorption30. Pérez-Granados et al11observed that linoleic acid favours calcium absorption more than oleic acid. In contrast, long chain saturated fatty acids and erucic acid (C22:1) reduced calcium bioavailability31, the former forming more insoluble soaps32. Moreover, the addition of oleic acid to stearic acid improves the absorption of this saturated fatty acid33, the mechanism involves increasing the solubility of stearic acid in the bile salt-monoglyceride micelle. Concerning palmitic acid, if this fatty acid is predominantly esterified in sn-2 position (beta-position), higher fat digestibility and enhanced calcium absorption has been shown34. Lopez-Lopez et al35 evaluated the influence of the position of the long chain fatty acid in triglycerides on the fatty acid, calcium and magnesium content of term newborn faeces and found that those infants that consumed human milk and infant formula with palmitic acid predominantly esterified in sn-2 position (beta-position) showed reduced contents of calcium and total fatty acids in faeces, without yielding any differences on magnesium.

Inthe present experiment, although the differences between the two infant formulas were small, SF contained more stearic acid, less palmitic acid, oleic and linoleic acids but more LCPUFA. To these differences the slightly lower bioavailability of calcium and magnesium may be attributed.

Magnesium is affected by dietary fat to a lower extent than calcium12. It is known that, long chain saturated fatty acids also form insoluble soaps with this mineral33,and that magnesium soaps produced with oleic and linoleic fatty acids are 10 to 20 times more soluble than the corresponding ones formed by saturated fatty acids36.The fact that magnesium absorption was higher in F compared to SF may be associated to the higher food intake and to the differences in fatty acid composition between the two infant formulas: F contained lower levels of long chain fatty acids, either saturated or unsaturated, and higher levels of palmitic acid.

The main differences obtained in the present study are between the C group and the infant formula groups. The reference group showed 3 to 4 times higher faecal excretion of the two studied mineralsinducing lower absorption in C particularly respect to F. This may be explained becauseinfant formulas contain 16 times more palmitic acid, slightly more oleic acid but only 5 times less linoleic acid. In S and SF although more calcium and magnesium was absorbed,at later stages higher proportion of these minerals was excreted by urine, possibly in response to animal’s physiological needs. Consequently, the calculated retention of both minerals were similar in the three groups.