ARCHIVE - FOOD SURVEILLANCE INFORMATION SHEETS

	\| Help \| Site map \| Search \| Text links \|
	Archive - MAFF

Home

MAFF UK - PLANT OESTROGENS IN SOYA-BASED INFANT FORMULAE

Index to MAFF UK Food Surveillance Information Sheets, 1998

Summary

A survey of phytoestrogens (plant oestrogens) in soya-based infant formulae has been completed. This is the first such survey carried out in the UK and is a snapshot of the position at the time these soya-based infant formulae samples were bought in the UK between March 1996 and July 1997. Six brands of soya-based infant formulae (5 powdered and 1 concentrated liquid) and three major brands of cows' milk-based infant formulae were analysed for their isoflavone contents. Total isoflavone phytoestrogen concentrations ranged from 18-41 mg isoflavone (expressed as the aglucone)/litre made up formulae. To obtain a measure of batch to batch variation, eight batches of a single brand of soya-based infant formulae were also analysed for isoflavones. Estimated average intakes were 5 mg isoflavone /kg bw /day for 1 to 2 month old infants and 4.5 mg isoflavone/kg bw/day for 4 to 6 month old infants. These intakes are comparable with the estimate of around 4 mg/kg bw/day during the first 4 months of life which was used as the basis for the Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment (COT)'s advice on soya-based infant formulae given in 1996 and further endorsed in 1998, that breast milk, or if necessary cow's milk formulae, are the preferred sources of nutrition for infants, but where parents have been advised by their doctor or health professional to feed their baby soya-based infant formulae, they should continue to do so.

Background

Phytoestrogens or plant oestrogens have oestrogenic properties but are much less potent than the human female oestrogen 17-beta oestradiol.¹ Soya and its products are particularly rich sources of a group of phytoestrogens known as isoflavones. Consequently soya-based infant formulae also contain these substances. The main types of isoflavone found in soya are genistein, daidzein and to a lesser extent glycitein (which is not considered to be oestrogenic). These compounds or aglucones are usually linked (conjugated) to a sugar molecule to form a glycoside. The conjugated forms of the isoflavones are not oestrogenically active but are converted by the gut microflora to the active aglucones. To make meaningful exposure calculations, it is important to obtain information on the concentrations of the individual isoflavones.

The Department of Health recommends breast feeding and if necessary cows' milk formulae as the best way of feeding infants.² Infant formulae provide an alternative source of nutrition when mothers cannot breast feed or choose not to do so. Cows' milk formulae are suitable for the majority of bottle fed infants. However, a small number of babies cannot tolerate cows' milk and they can be given soya-based infant formulae as an alternative. This is usually done on the advice of a health professional. Approximately 2 per cent of infants are reportedly³ fed with soya-based infant formulae.

In 1996, the COT reviewed the possible effects on infants of consuming phytoestrogens in soya-based infant formulae. At this time the Committee stated that the potential for isoflavones to adversely affect infants is of particular concern since it is possible that a hormonal imbalance in early life can permanently affect sexual development and fertility. However it acknowledged that there have been no reports which suggest that populations which habitually ingest large quantities of soya (e.g. Chinese, Japanese) have impaired fertility or altered sexual development. At the time of the 1996 COT review limited data indicated that the estimated intake of isoflavones by infants fed soya-based infant formulae was in the region of 4 mg/kg bw/day. This is larger (on a mg/kg bw/day basis) than the intake reported to cause hormonal effects in pre-menopausal women.⁴ Following its review, the COT issued a statement.⁵ The COT endorsed current DH advice that children whose parents have been advised to use soya-based infant formulae by their doctor or health professional should continue with its use. However the Committee recommended that research be undertaken as a matter of high priority to determine whether ingestion of soya-based infant formulae carries any risk for infants.

The survey

The purpose of this survey was to obtain data on the concentrations of isoflavone phytoestrogens (isoflavones) in soya-based infant formulae to determine whether any more specific advice was required on the consumption of soya-based infant formulae.

Concentrations of isoflavones were determined in three brands of cows' milk-based formula; Cow and Gate Premium, Farley's Firstmilk and SMA Gold, and in five brands of soya-based powdered infant formulae; Cow & Gate Infasoy, Farley's Soya Formula, SMA Wysoy, Isomil, ProSobee Soya Formula and in one brand of liquid formula; ProSobee Concentrated Liquid. Isoflavone levels were also measured in another 7 different batches of the market leader, SMA Wysoy, with best before dates ranging from December 1997 to November 1999 in order to establish whether isoflavone concentrations varied between batches.

Each of the brands were analysed for fourteen individual isoflavones and one coumestan (coumestrol). The isoflavones can be divided into four groups:

(a) aglucones: daidzein, genistein, glycitein, formononetin and biochanin A

(b) glycosides: daidzin, genistin and glycitin

(d) malonylglycosides: 6"-O-malonyldaidzin, 6"-O-malonylgenistin and 6"-O-malonylglycitin

Except where explicitly stated the isoflavone content is expressed as mg aglucone equivalents/litre made up formula i.e. the individual isoflavones are normalised to the aglucone to enable direct comparison between brands. (For example, in order to express the concentration of malonylglycoside as mg aglucone equivalents/litre or kg the following calculation should be made: mg measured x (molecular weight of the aglycone/molecular weight of the malonylglycoside) .

Sampling

This MAFF survey was commissioned prior to the introduction of a new policy on reporting of survey results in 1997. The policy involves the release, where relevant, of brand names of samples analysed in a survey. More information on these changes is in the MAFF Press Release dated 21 July 1997^6, and the September '97 edition of the Food Safety Information Bulletin (No. 88).⁷

In the interim period, the results from some MAFF surveys, including this one, will also be reported under the new procedures even though the survey was not designed with the new policy in mind. As far as we are aware, the survey included samples of all of the retail brands of soya-based infant formulae available when the samples were bought (between March 1996 and July 1997).

The infant formulae analysed for their isoflavone contents were all obtained from retail outlets, such as supermarkets or chemist's shops. For an estimate of interbatch within brand variation, a further 7 different batches of a single brand, (SMA Wysoy, the known market leader⁸) with best before dates ranging from December 1997 to November 1999 were analysed and compared to establish whether isoflavone concentrations varied between different batches.

Method of Analysis

The analysis was carried out at the Department of Food Science and Technology, University of Reading. With the exception of ProSobee Concentrated Liquid which was freeze dried, all samples were analysed as purchased. The samples were extracted with 80 per cent aqueous methanol at room temperature, using a procedure based on that of Barnes et al. (1994).⁹ Analysis of the isoflavones in the methanolic extracts was performed by reversed-phase high performance liquid chromatography (HPLC) with gradient elution and diode array detection, using a modified procedure based on Wang and Murphy (1994).¹⁰ Each lot of infant formula was analysed in triplicate. Liquid Chromatography-Mass Spectrometry(LC-MS), with atmospheric pressure chemical ionisation in positive and negative mode monitoring selected ions, was used to provide confirmation of the identity of isoflavonoid peaks in the chromatogram for each brand analysed. Flavone and flavanone were used as internal standards.

Recoveries were determined using isoflavone standards in Arcon F, a virtually isoflavone free soya protein isolate. Recoveries ranged from 94-120 per cent (corrected for recovery of flavone). The limit of detection as defined was 0.25-0.5 mg/litre and the limit of quantification is 0.5-1 mg/litre. Inter-laboratory repeatability studies were also carried out by analysing the same batch of infant formula on three different days using the same conditions. Results corrected for percentage recovery of flavone showed that RSD for "within days" repeatability for the two sets of duplicate determinations of the same soya-based formula ranged from 2-5 per cent for isoflavones present above 5 mg/kg. RSD for isoflavones present below 5 mg/kg (e.g. malonylglycitin which represents up to 2 per cent total) was about 19 per cent. These results were deemed acceptable in the context of this survey particularly since there are as yet no established standards, validated methods or established quality assurance schemes for isoflavone analysis.

Results

Isoflavone concentrations in the three samples of cows' milk-based formulae analysed were below the limit of detection (LOD = 0.25-0.5 mg/litre). Isoflavones were detected in all brands of soya-based infant formulae analysed with total concentrations ranging from 18-41 mg isoflavone (expressed as the aglucone)/litre made up formulae. Table 1 gives the total mean isoflavone concentration found in each brand of formulae expressed as mg aglucone equivalents/litre made up formula, together with date of purchase, "best before date", lot number, size of sample purchased and the manufacturer's name and address.

The isoflavones were found to be predominantly present as glycosides but their acetyl and malonyl forms were also present as were the aglucones although in much smaller amounts. ProSobee Concentrated liquid contained isoflavones almost exclusively in the form of glycosides with no 6''-O-substituted conjugates present.

Isoflavones detected in quantifiable amounts in the brands analysed were:

(a) Diadzein and related compounds: daidzein, diadzin, 6"-O- malonyldaidzin and 6"- O-acetyldaidzin

(b) Glycetein and related compounds: glycitin and 6"-O-malonylglycitin

The following isoflavones were not detected in any the brands of soya-based infant formulae analysed: glycitein, 6''-O-acetylglycitin, formononetin and biochanin A; in addition the coumestan; coumestrol was not found to be present.

In order to look at variation between batches of soya-based infant formulae, a further seven batches of SMA Wysoy, the market leader¹¹ , purchased at different times and in different locations were analysed. Concentrations in the 8 batches ranged from 18-33 mg total isoflavone/litre formulae as made up and are shown in Table 2.

Discussion

The results showed that genistin was present at the highest concentration in all brands ranging from 5-13 mg isoflavone/litre formulae as made up. Acetyl and malonyl genistin also contributed significantly to the isoflavone concentrations in all brands except in the case of ProSobee Concentrated Liquid where the acetyl and malonyl derivatives were not detected. ProSobee Concentrated liquid contained isoflavones almost exclusively in the form of glycosides with no 6''-O-substituted conjugates present (confirmed by analysis of a second batch). This is in contrast to the powdered formulae where isoflavones were present as aglucones, beta-glycosides and their conjugated forms. This can be seen from Figure 1 which shows the five isoflavones present in the highest concentrations in each brand analysed, and for clarity also shows levels for the batches of SMA Wysoy with the highest and lowest concentration of isoflavones ( batches B and F respectively). Figure 2 shows the percentage distribution of the three main types of isoflavone (expressed as total aglucone). Genistein and related compounds represented on average about 58 per cent, daidzein and related compounds about 36 per cent, glycitein and related compounds 6 per cent of the total isoflavonoid on a molar basis. For the eight batches of SMA Wysoy, the ratio was 60:34:6. Figure 3 shows a large degree of consistency in the percentage distribution between aglucones and glycosides for the six brands and plus batches B and F of SMA Wysoy. Conjugation patterns for the additional six batches of Wysoy were also similar to those of batches B and F.

In the soya bean^12,13 isoflavones normally occur as glycosides while in processed foods such as soy sauce, tempeh, tofu and miso^10,14 they may be partially or totally broken down to the aglucone which, unlike the glycoside, does not need to be broken down by gut bacteria before becoming oestrogenically active. However, since (especially in man) the detailed metabolism of the isoflavones is still poorly understood, it is not known in which form (aglucone or glycoside) these compounds are most readily absorbed so it is not possible to comment on the physiological relevance of this observation.

Within brand batch variation

Eight different batches of SMA Wysoy were analysed to establish variations in composition between batches of soya-based infant formulae. Concentrations in the 8 batches ranged from 18-33 mg total isoflavone/litre formulae as made up (see Table 2). Comparison of total isoflavone concentrations in the 8 batches showed that levels did vary though not enough to have any significant impact on exposure. Highest concentrations were found in batch B with a best before date of April 1998 (33 mg isoflavone/litre formulae as made up), and the lowest in batch F with a best before date of December 1999 (18 mg isoflavone/litre formulae as made up). These concentrations were similar to those found in the other brands.

Statistical analysis

Statistical analysis of the results was carried out to consider whether there was significant variation in the levels of isoflavones between the 6 brands analysed and between the 8 batches of SMA Wysoy analysed. An analysis of variance was performed in which between-brand variation in total isoflavone levels was compared with within-brand variation (as judged from 8 batches of SMA Wysoy), using an F-test. All calculations were made on concentrations on an "as made up" basis.

Statistical analysis provided evidence of between batch variation equivalent to significance at well beyond the 0.1 per cent level. Differences between brands can only be gauged in relation to the within-brand variation (i.e. differences between batches of the same brand). As this information is only available for SMA Wysoy it has been assumed that for the purpose of statistical analysis, variation within this brand accurately represents the variation within all brands. Statistical analysis at the 5 per cent level was unable to provide evidence of any real differences among brands.

Dietary intakes

Dietary intakes for 1-2 month old infants and 4-6 month old infants were calculated on the basis of the reconstitution of the feed and using average figures for body weights and consumption data extrapolated from the data indicated on the packs of the 5 powdered formulae under study (no information was provided on the liquid formula). Mean estimated average intakes as given in Table 3 ranged from 5 mg isoflavone /kg bw/day for 1 to 2 month old infants to 4.5 mg isoflavone/kg bw/day for 4 to 6 month old infants. These intakes are comparable with the approximate intake figure of 4 mg/kg bw/day during the first 4 months of life which the COT used as the basis of their 1996 advice on phytoestrogens and which was subsequently endorsed in 1998. Actual intakes will vary depending on the actual consumption and weight of individual babies.

Other recent studies

Details of two studies measuring the levels of phytoestrogens in soya-based formulae sold in the US were published by Setchell et al¹⁵ in July 1997 and Murphy et al¹⁶ in November 1997 respectively. Work in New Zealand by Irvine and co workers¹⁷ and in Australia by Knight et al¹⁸ also investigated levels in of phytoestrogens in soy based infant foods. The results from all four studies are compared in Table 4. The results obtained in this present survey indicate that levels in UK formulae are similar to those reported elsewhere. Differences are likely to be due to batch to batch variation in the isoflavone content of the soya isolate used, slight variations in formulation of soya infant formula in different countries or minor differences in the analytical methodology used.

Interpretation

These data were considered by the COT in September 1998. It welcomed the publication of information on the concentrations of phytoestrogens in soya-based infant formulae sold in the UK. The Committee was reassured that the survey results indicated that levels of phytoestrogens in soya infant formulae were comparable to those considered by the Committee in 1996 and to data published from other countries.

The COT therefore advised that there was no need to amend the advice issued in July 1996 that: "it endorsed the advice of the Department of Health that breast milk and cows' milk formulae are the preferred sources of nutrition for infants. However, parents who have been advised by their doctor or other health professionals to feed their baby soya-based infant formulae should continue to do so."

Isoflavone concentrations, and the distribution of the individual isoflavonoids can vary both within and between varieties of soya. Concentrations can also be dependent on factors such as climatic conditions, growth locations, seasonal variation, time of harvest and the degree of "environmental stress" to which the soyabean plants are subjected.^12,19,20 Soya protein isolate²¹ used in the manufacture of soya-based infant formulae, is manufactured to a constant protein composition from a natural product (the soya bean), thus minor differences in the isoflavone content of individual extracts are to be expected.

The results from this survey indicate that levels of phytoestrogens in the six brands of soya-based infant formula analysed are between 18-41 mg/isoflavone/litre formula as fed. Levels are comparable to those used by the COT in its previous consideration of this issue: that is for infants in the first four months intakes are around 4 mg/kg bw/day of total phytoestrogens. It is reassuring that the levels of isoflavones measured in UK soya-based infant formulae are similar to those reported by other workers in the field in the US, Australia and New Zealand and any small differences seen are likely to be due to differences in the individual batches of soya isolate used, the formulation or minor differences in analytical methods.

A copy of the full report of this survey has been placed in the MAFF Library, Nobel House, London, SW1P 3JR Tel. No. +44 (0) 171 238 6575. If you wish to consult a copy please contact the library for an appointment giving at least 24 hours notice or alternatively copies can be obtained from the library: a charge will be made to cover photocopying and postage.

Summary of units

kilogram (kg): one thousand grams
milligram (mg) : one thousandth of a gram
litre (L) : one thousand millilitres
mg/kg bw/day: milligrams per kilogram of body weight per day.

Note: soya and soy are synonymous.

References

Markiewicz, L., Garey Coldham N.G., Dave M., Sivapathasundaram S., McDonnell, D.P., Connor, C. and Sauer, M.J., (1997), Evaluation of a Recombinant Yeast Cell Estrogen Screening Assay; Environmental Health Perspectives; 105, no 7, p 735 - 742.
Department of Health. Weaning and the Weaning Diet. London: HMSO, (1994), Report of Health and Social Subjects: 45 Report of the Working Group on the Weaning Diet of the Committe on Medical Aspects of Food Policy.
Foster K., Lader, D. and Cheesbrough, S., (1997), Infant Feeding 1995, p68; The Stationary Office, London.
Cassidy, A., Bingham, S. and Setchell, K., (1994), Biological effects of a diet of soy protein rich in isoflavones on the menstrual cycle of premenopausal women. Am. J. Clin. Nutr 60, 333-340.
MAFF (1996), Statement by the COT on phytoestrogens. Food Surveillance Paper No. 57, p 59-81.
MAFF Press Release 21 July 1997, Product Brand Names to be Published in New Openess on Food Surveys.
MAFF/DH, 1997, Food Safety Information Bulletin No. 88
Corporate Intelligence Report on baby foods, Retail Business No.481, March 1998.
Barnes, S., Kirk, M., Coward, L., (1994), Isoflavones and their conjugates in soy foods: extraction conditions and analysis by HPLC-mass spectrometry. J. Agric. Food Chem., 42, 2466-2474.
Wang, H-J. and Murphy, P., (1994a), Isoflavone content in commercial soybean foods; J. Agric. Food Chem., 42, 1666-1673.
Corporate Intelligence Report on baby foods, Retail Business No.481, March 1998.
Wang, H-J., and Murphy, P., (1994b), Isoflavone composition of American and Japanese soyabeans in Iowa: effects of variety, crop year, and location. J. Agric. Food Chem., 42, 1674-1677.
Tsukamoto, C., Shimada ,S., Igita, K., Kudou, S., Kokubun, M., Okubo, K. and Kitamura, K., (1995), Factors affecting isoflavone content in soybean seeds: changes in isoflavones, saponins and composition of fatty acids at different temperatures during seed development. J. Agric. Food Chem., 43, 1184-1192.
Coward, L., Barnes, N.C., Setchell, K.D.R. and Barnes, S., (1993), Genistein, daidzein and their beta-glycoside conjugates: antitumor isoflavones in soybean foods from American and Asian diets. J. Agric. Food Chem., 41, 1961-1967.
Setchell, K.D.R, Zimmer-Nechemias, L., Cai, J. and Heubi, J.E., (1997), Exposure of infants to phytoestrogens from soy-based infant formula. The Lancet, Vol. 350, pp 23-27.
Murphy, P.A., (1997), Isoflavones in soy-based infant formulas J. Agric. Food Chem., 45. 4635- 4638.
Irvine, C.H.G, Fitzpatrick, M.G. and Alexander, S.L., (1998), Phytoestrogens in soy-based infant foods: concentrations, daily intake, and possible biological effects. Proceedings of the Society for Experimental Biology and Medicine, Vol. 217, pp 247-253.
Knight, D. C., Eden, J. A., Huang, J. L. and Waring, M. A., (1998), Isoflavone content of infant foods and formulas. J. Paediatr. Child Health, Vol 34, pp 135 - 138.
Eldridge, A.C.and Kwolek, W.F. (1983) Soybean isoflavones: effect of environment and variety on composition. J. Agric. Food Chem., 31, 394-396.
Tsukamoto, C., Shimada,S., Igita, K., Kudou, S., Kokubun, M., Okubo, K. and Kitamura, K., (1995) Factors affecting isoflavone content in soybean seeds: changes in isoflavones, saponins and composition of fatty acids at different temperatures during seed development. J. Agric. Food Chem., 43, 1184-1192.
Messina, M J.,Persky, V., Setchell,K.D.R., and Barnes, S., (1994), Soy intakes and cancer risk: A review of the in vitro and in vivo data. Nutrition and Cancer, 113-131.

Contact point:

Dr Catherine Boyle
Food Standards Agency,
Chemical Safety and Toxicology Division
Room 507C, Aviation House,
125 Kingsway, Holborn
London WC2B 6NH
Tel: +44 (0)20 7276 8509
Fax: +44 (0)20 7276 8513

Spreadsheet Tables

Table 1: Details of Samples of Cows' Milk (C) and Soya-Based Infant (Sb) Formulae Analysed with Levels Expressed

Click here to download the Excel 5.0 version of Table 1
Shift-Click here to download the .csv version of Table 1 (if you have any other spreadsheet package)

Table 2: Details of Samples of Sma Wysoy Analysed for Within Brand (Interbatch) Variation: Levels Expressed as mg Aglucone per litre Infant Formulae as Made Up

Click here to download the Excel 5.0 version of Table 2
Shift-Click here to download the .csv version of Table 2 (if you have any other spreadsheet package)

Return to Index to Surveillance Information Sheets, 1998 page
Go to top of page

These pages were last updated on 30th October 1998